COMMON_Proceedings_of_the_San_Fransisco_Meeting_196712 COMMON Proceedings Of The San Fransisco Meeting 196712
COMMON_Proceedings_of_the_San_Fransisco_Meeting_196712 COMMON_Proceedings_of_the_San_Fransisco_Meeting_196712
User Manual: COMMON_Proceedings_of_the_San_Fransisco_Meeting_196712
Open the PDF directly: View PDF 
.
Page Count: 624
| Download | |
| Open PDF In Browser | View PDF |
COMMON
Proceedings of the San Francisco Meeting
December 11-13, 1967
C
·,:·':t
'",
o
o
d)
t·
·hH{ .... H' .. ···'j· .. ·w ..·rt5f· .. 'ihi.dwM+E"Biift"Ft8ijpifj".. w·wlHW·
r··
WFmma.
GENERAL MEETING INFORMATION
c'
Meeting'Rooms
All meeting rooms will be in the Sheraton Palace Hotel, San Francisco.
Specific room assignments are noted in the agenda.
Room Reservations
Reservations must be made by submitting the reservation card directly
to the hotel.
Registration
Registration fees for COMMON cover all publication and operating
costs and represent the only source of funds for the organization. Registration for this meeting has been set at $30 for the entire meeting) which
includes the cost of the luncheon on Tuesday. The registration desk at the
hotel will be open during the following hours:
Sunday,
Monday,
Tuesday,
December 10, 1967
December 11) 1967
December 12) 1967
5:00 p.m. - 9:30 p.m.
7:45 a.m •. - 5:00 p.m.
8:00 a.m. - 3:00 p.m.
Birds - of - a - Feather Sessions
()
Attendees may schedule additional informal sessions by posting a notice
on the bulletin board.
Systems Reference Library Publications (SRL)
A selection of 1620, 1130, 1800, and S/360 publications are on display
in the registration area. You may order an SRL publication by filling out
an order form- provided o You should expect to receive the items ordered
in about 3 weeks.
Computer Equipment Displays
Arrangements have been made for delegates to attend demonstrations
and receive time on the San Francisco IBM Data center machines including the 1130, 360/20, 360/30, and 360/40.
Meeting Headquarters
Typing and reproduction services will be available in the headquarters
room: Regency
Agenda Notes
The agenda listed herein represents the program at the time the booklet
went to press. Modifications and/or additions will be noted on the information board in the coffee area. Abstracts for the papers listed in the program
appear at the end of this bOGklet.
- 1 -
¥¥.E .i¥4;;;;S;4.# ....
444 ..... #4Z¥J4Al4¥IlWi¢&EIi\4?Z¥4&#.
iiIii,i&)l!¥iiit¥¥n ,•.f & ....u,,,
.JJ2t!!iitiiSiiih'lf.ii&lii,.•,LA" ,,:# /.
I
-
=----~--=--~------~-----=------~--==--
---'--~--"---'-'--------~--------.~---
-----------------
.=~:..:..::.:
LETTER TO MEMBERSHIP
FROM PRESIDENT OF COMMON
Members of COMMON:
With this letter is the preliminary agenda for the December meeting
of COMMON, in San Francisco. The September meeting in Cincinnati
was a good one; we expect the San Francisco meeting to be better. With
your help and cooperation, we believe that each meeting can continue to
be better than its predecessors.
There are several ways in which COMMON can affect IBM and the world
in general; the majority of them, branch office, project liaison, etc., are
for today's problems, but two are specifically for the future. The first
is our comments and suggestions to the U.S.A. Standards Institute (USASI) ..
Lynn Yarbrough, of USASI's X3.4 Committee on Programming Languages,
wIll ha ve a session on the Standards Institute: its organization, how it
affects us, and what we can do to influence it. There will be time for
questions and comments.
We will have a brief discussion of the Systems Objectives and Requirements Committee (SORC). This is a committee of user and IBM representatives, whose purpose is to provide IBM with the benefit of informed
user opinion in determining the requirements for the next cycle. There
will not be a formal session on SORCi their purpose is to acquire information, and not to dispense it. There will be an explanation of it at the
general assembly, and there will be opportunities to discuss your ideas
with SORC representatives. There is a serious lack of representation
in SORC of the small computer user and process control installation.
Remember that SORC is concerned almost exclusively with the problems
of the future. We have vociferously exercised our right to criticize the products of
both IBM and USA!; we are morally obligated to do what we can to alleviate
the situation for the next generation. And our most important contribution
is information about our requirements.
See you in San Francisco.
James C. Stansbury
o
- 2-
ALL ABOUT COMMON
o
During the first half of the 1960's while IBM built and marketed the
"1620" Computer, COMMON was known as "The 1620 Users Group."
. It was the "SHARE" of the small scientific computer user. All "1620"
installations becarne members by applying and sending a representative
to anyone of a dozen or so regional or national meetings occurring during
any two year period.
As
that
was
was
the third generation computers began to appear and it became apparent
the days of the IBM "1620" were numbered, the name of organization
changed to "COMMON" and a myriad of regional and national meetings
reduced to three national meetings per year.
The COMMON Executive Board met in Chicago on September 15 and 16,
1966, to discuss ways of improving the effectiveness of the organization.
The board members felt that an organizational change was necessary to
cope with the imminent growth of COMMON, the integration into the organization . of multiple machine types (1130, 1800, 1620, 360/30, 360/40 and
360/44), and the development of increasing numbers of interest areas by
the members.
o
It was decided to break the organization down into four nl....!n divisions
along the lines of tre other two IBMUser Groups "SHARE" and "GUIDE".
Each division would be divided into projects and the projects would be
further subdivided into working committees as needed.
The divisions established were as follows:
1. Systems Division. This division has broad responsibility for coordination
of user activity .and opinions in the areas of software and hardware regardless of application.
2. Adm.inistrative Division. This division has responsibility for the various
activities which are necessary tooperate COM.MON as an organization, both
internally and in relation to other organizations. These include the newsletter, CAST, bylaws, membership list, program library, meeting plans, etc.
3. Installation Management Division .. This division has responsibility for
coordinating those user activities which are concerned with the management
of a computer installation. These include personnel training, standards,
computing center operation, etc.
4. Applications Division. This division has responsibility for coordinating
user activities which are application or industry orientl~d ..
c
Each division is headed by a Division Manager who supervises the operation of his division and reports the activities and requirements of his
division to the Executive Board through the Executive Vice President.
- 3-
Each division is composed of a number of projects. Each project is
headed by a Project Manager who is responsible for specific areas of
interest within the division and who reports to the Dvision Manager.
Each project is divided into committees each of which is headed by a
Committee Chairnlan. The committee chairmen have responsiblity for
seeing that their committees carry out the duties within their areas of
interest and for seeing that the consensus of opinion of the users at the
committee meetings are transmitted to the Project Managers.
Although in some areas committees may be divided into special working
subcommittees dealing with specific areas, it is intended that the participation of users in the activities of COMMON be carried out at the committee
level. I.B.M. supplies liaison people to the established committees so that
user's opinions and requests are transmitted to I.B.M.
The committees meet at every COMMON meeting and these sessions are
open to any user who has an interest in the topics covered by the committee.
Some meetings on the agenda are listed as project meetings. Most of these
are intended to be general discussions by groups of users having the same
machine types and are open to all users. These sessions replace the general
hardware meetings and problem-solving sessions that occupied much of
the 1620 Users Group meetings.
There are also opportunities at every COMMON meeting for users with
special interests for which no formal project or committee exist to get
together and discuss their problems and techniques. These are called
"birds-of-a-feather" sessions and are organized by groups of users
posting a notice on th~ bulletin board and holding a meeting. If enough
interest persists, a committee may be formally created and meetings
scheduled at future Common meetings.
o
-4-
------.....
-.--~---~.-.-
...- . - - - - -
COMMON OFFICERS
President
James Co Stansbury
Halcon International
Two Park Avenue
New York, New York 10016
Secretary - Treasurer
Charles Eo Maudlin, Jr.
Computer Center
Texas Woman's University
Denton, Texas 76204
Eastern Region President
Norman Goldman
Boston University
111 Cummington st.
Boston, Massachusetts 02215
()
European Region President
(Acting)
Dr. Hans Tompa
European Research
Associates S.A.
95 Rue Gatti de Gamond
Brussels 18, Belgium
Western Region Presiden~
William G. Lane
Director of Computer
Sciences
Chico State College
Chico, Californid. 95926
Mid-West Region President
Mrs. Laura B. Austin
General Supervisor
Computer Services
General Motors Institute
1700 W. Third A venue
Flint, Michigan 48502
Canadian Region President
stuart Do Baxter
National Research Council of Canada
Computation Centre M23A
Montreal Road
Ottawa, Ontario, Canada
Board Members at Large
Frank H. Maskiell
McGraw-Edison Power Systems
Division
Box 330
Cannonsburg, Pa. 15317
Richard L. Pratt
Data Corporation
7500 Old Xenia Pike
Dayton, Ohio 45432
Paul A. Bickford
Computer Center
DePauw University
Greencastle, Indiana
- 5-
COMMON COMMITTEE OFFICERS
ADMINISTRATIVE DIVISION
Manager - Laura B. Austin (3070)
General Motors
General
Reference Manual
Chairman - Dr. Raymond E. Roth (1715)
State University College
Contributed Program Library
Chairman - Miss B. Baber (1454)National Education Assn.
(Prep Forms)
CPL Shipment Analysis (1303)
Chairman - J. H. Keith
Miami - Dade Junior College North
Jug Inter-Library Exchange
Chairman - W. A. DeLagall (1582)
Schering Corp.
APPLICATIONS DIVISION
Manager - F. H. Maskiell (3081)
McGraw-Edison Co.
Techniques Project
. Chairman - W. Pease, Jr. (1516)
West Virginia Pulp & Paper CQ.
Utilities
Chairman - G. Haralampu (1041)
New England Electric· System
Civil Mechanical Engineering
Chairman .. L. W.. Brown (3301)
International Paper Co.
Chemical Engineering
Chairman - G. Hertel
U. S. Rubber Co •
.. 6--~
o
J'"
"ri'ri'
'.* . -.
f"5'it·· .. , .. ···tt-iF . -
t . ···tf£"·· rt'iifi'--' HT·-"w·ri'i9i8tiih"BflE iiiiiihHifbH·-S·.·t±H! hHfi&8 t&W- t ii'iBW-wu··,mij··W"[]"'·2'J[ w·w,""[rf·!iiiFi£riJFF'P
llI
yr"'irrm-g'vWO¥*'UZWf
Education
Chairman - M. Goldberg (1416)
Fordham University
INSTALLATION MANAGEMENT, DIVISION
Manager - C. Baker
Pioneer Hi-Bred Corn Company
Personnel
Chairman - Ro B. Thomas (3038)
Federal Reserve Bank
Operations
Chairman - R. J Snailer (1495)
Metropolitan Life Ins. Co.
c>
Standards
(Inactive)
SYSTEMS DIVISION
c
Manager - J. S. Taylor (3121)
Data Corporation
System/360
Project Manager - R.L. Pratt (3121)
Data Corporation
DOS
Committee Chairman - D. R. McIlvain (1100)
Catalytic Construction Co.
OS
Committee Chairman - Wade Norton
Southern Services, Inc.
S/360-44
Committee Chairman - A" G. WigdahL(3134)
Allen Brady Company
Hardware
Committee Chairman - J. L. Tunney) Jr. (3076)
James R. Ahart & Associates
- 7-
360 Commercial Users
Chairman .. Mr. F. Hatfield (3008)
Line Materials Industries
0
1800
Project Manager - C. R. Pearson (1497)
J. P. stevens & Co., Inc.
(Non-control)
Software
Chairman - R. Cox (1132)
Humble Oil Company
Applications
Chairman - J. C. Deck (3237)
Inland Steel Co.
Hardware & RPQs
Chairman - W. Barnes (531)
Pacific Gas & Electric
(Control)
Process Systems
Chairman - L. Jones (5046)
Bonneville Power Adm.
Software
Chairman - J. K. Tanatra (3462)
GMC Truck & Coach Division
Hardware
---Chairman - D. L. Kraatz (5255)
Corn Products Company
Applications
Chairman- Dr. W. Carlson (3009)
Champion Papers, Inc.
1130 Project
Chairman - D. Dunsmore (3428)
Ohio River Valley Commission
1620 Project
Chairman - Richard Ross
University of Mississippi
- 8-
-
------------------------------
0
lip w"'t"-j\,,j,
tti'·j·.·,} ·f ·k
itt&·±·¥·*%ttMiiiF,F· . Tbp . Jj····h.jiHibii!Rii#tE,jifrJIU·} ·.. ··r-.. -·WjijHiJJ"iAfIiiIi·u[·
wW-iuw
nf w
ryirnwH0D
SESSION M. O.
Monday 8:00-8:45 a.m.
M.O.l
New Members Session
Chairman: James Stansbury, President
M. o. 2
Room:
Comstock
Subject:
Topics of interest and concern for those attending
COMMON for the first time.
Session and Panel Chairmen Briefing
Chairman: Robert Forstrom
Room:
English
Subject:
General information and arrangements briefing.
for session and panel chairmeR.
c
o
- 9-
--.----.--.-- ...
---.--~~~--
SESSION M. 1
Monday 9:00-10:00 a.m.
o
M. 1. 1. General Session
Cha.irrnan: James stansbury, President
Room:
Rose
Subject,s:. a.
Introduction of COMMON Officers and IBM
Representatives.
b. Reports from Divisional Chairmen
c. Secretary-Treasurer's Report
d. Arrangements Chairman's Announcements
e. - Program Chairman's Announcements
f. Announcements by Members
g. Announcement of Chicago Meeting
COFFEE BREAK 10:00 - 10:30
o
.. 10-
SESSION I'rt2
Monday 10:30 - 12:00 Noon
M.2.1
360 Project
Chairman:
Richard Pratt
.Room:
Rose
a. COBOL to PL/l Conversion:
Subject
b.
IBM
FORTRAN to PLll Conver sian: IBM
M.2.2 ;
1130 Project
J
Chairman:
Dave Dunsmore
Room:
California
Subject:
~ a.
: b.
M.2.3
Dynamic Model Simulator for the IBIvl
1130: George Polyzoides
Three-and-Four-Bar Linkage System with
Plotted Output: Robert Cushman
1620 Project
Chairman:
Richard D. Ross
Room:
Golden Gate
Subject:
Tutorial Sessions Non-IBM Compilers
a. Data SPS: Richard Ross
b. One Dimensional Blast Wave Theory For
Explosions: J . Goresh and J. Caslin
c.
M.2.4
M.2.5
o
Discussion of Related Topics
1800 Project
Chairman:
Robert Forstrom
Roonl:
Comstock
Subject:
Multiprogramruing Executive, MPX: D. Schade (IBM)
Numerical Control Project
Chairman:
J .. Moschetti
Room:
Royal Suite (262)
Subject:
Discussion of Problems Relating to Numerical
Control.
- 11-
M.2.6
/
Installation Management and Administration
Chairman:
Room:
Subject:
o
Laurence Baker
Forty ... Niner .
,., a. Systems Reference Library SRL:
G. Goesch
(IBM)
v'b. Program Library Discussion: L. Austin
PREP forms for 1130, 1800
and 360 libraries
(;)
o
SESSION M" 3
V/
C
M.3.1
Monday 1:30 - 3:00 p.m.
360 Project
Chairman:
Richard Pratt
Room:
Rose
Subject:
a.
b.
M.3.2
J
M.3.3
C:i
as/DOS Data Management:
(IBM)
as/DOS Linkage Editor:
(IBM)
William
William
Post
Post
1130 Project
Chairman:
Dive Dunsmore
Room:
California
Subject:
Monitor Version II:
G. Lester (IBM)
1620 Project
Chairman:
Tony Ross
Room:
Golden Gate
Subject:
Tutorial Session (Continued)
a.
'/b.
Kingston and Forgo FORTRAN:
F. Windham and G. Smith
A Batch Processing FORTRAN system for a
minimal configuration 1620: Gaylord Henry
c. Discussion of Related Topics
M.3.4
1800 Project
Chairman:
Robert Forstrom
Room:
Comstock
Subject:
1800 MPX: B. Lanek (IBM)
o
-lj-
M.3.6
M.3.7
M.3.8
Numerical Control Project
Chairman:
J. Moschetti
Room:
Royal suite (262)
Subje~t:
Discussion of Problems Relating to Numerical
Control
o
Installation Management
Chairman:
Laurence Baker
Room:
Forty-Niner
Subject:
360 Operator Training: H. Cad ow (IBM)
Administration
Chairman:
L. D. Yarbrough
Room:
Bonanza
Panel Discussions:
USASI Standards
Panelists:
T. B. steel, Jr., Chairman, USASI X3.4
L. D. Yarbrough, X3.4 and COMMON
(Others to be announced)
SUbject:·
a. Introduction to USASI:
Mr. Steel
b. Current USASI Activities: Mr. Yarbrough
c. Fortran: USASI vs 1620 vs 360
o
COFFEE BREAK 3:00 - 3:30 p.m.
o
-14~-
~-~-~-~~~.~~~
- -
SESSION M.4
c'
Monday 3: 30 - 5:00 p.m.
M.4.1
360 DOS Committee
Chairman:
Don Mcllvain
Room:
Rose
Subject:
SOUND-OFF
Members sound off to IBM: regarding suggestions,
problems and needs for DOS.
M.4.2
360 OS Committee
Chairman:
o. a. ABe9F~on
Room:
Royal Suite (264)
SUbject:
SOUND-OFF
a.
C,;'
b.
M.4.3
W tltl>.
NOrzrdAl
Members sound off to IBM regarding suggestions,
problems and needs for OS.
Organization of OS Committee
360-44 Committee
Chairman:
Allen B. Wigdahl
Room:
Royal Suite (260)
Subject:
SOUND-OFF
Members sound off to IBM regarding suggestions,
problems and needs for System 360-44.
M.4.4
o
1130 Project
Chairman:
Dave Dunsmore
Room:
California
SUbject:
Continuation of monitor
Ver II: G. Lester (IBM)
SOUND - OFF For System 1130 will
be Held 7:30 to 9:00 p.m.
-15-
·MAy
1620 Project
Chairman:
Frank Windham
Room:
Golden Gate
SUbject:
Monitor I Papers
o
a. A Large 1620 DOS, Reasonably 7094
Compatible: Lanny Hoffman
b. Basic Problems of Information Retrieval
and a Solution on the 1620: Lanny Hoffrnan
c. Discussion of Monitor I
M.~ 1800 Project
M.4.
J
Chairman:
Robert Eorstrom
Room:
Comstock
Subject:
Teleprocessing Support on 1800/1070/1050/2740/
1030: R" Smith (IBM)
Techniques Project
Chairman:
W. Pease
Room:
Parlor A (214)
Subject:
M.4~8V;
!,il
Fast Fourier Transforms with Applications
for the IBM 1800: Joe Howard Smith
'b. NIMS ... The Aerospace Scientific Data Reduction Monitor System: Charles R. Aumann
Installation Management
Chairman:
Laurence Baker
Room:
Forty-Niner
SUbject:
a.
standards in Operation: D. Fuiz (IBM)
be
Investigation of Abnormal Operation Conditions:
K. Anderson (IBM)
V
M.4.9
Administration
Chairman:
L. D. Yarbrough
Room:
Bonanza
: Subject:
-16-
a.
Current Research Concerned with the Standardization and Formal Definitions ofPL/l: Dr.J.A .. N.
Lee
c
SESSION T. 1
Tuesday 8: 30 - 10:00 a.m.
. T .1.1
T .1.3
o
T.1.4
T.1.5
360 Project
Chairman:
Richard Pratt
Room:
Concert
Subject:
a. A Small OS System: Wa_Norton
b. Panel Discussion:
Does OS Belong in
COMMON?
800 Project
Chairman:
Robert Forstrom
Room:
Comstock
Subject:
PL-1 Language Development COP: C. Burwick
(IBM)
Electric Utility project - 1130 Working Group
Chairman:
S. A. Clark
Room:
Golden Gate
Subject:
a.
b.
1130 Load Flow
1130 Short Circuit Calculations
Electric Utility Project - 360 Working Group
Chairman:
O. B. Anderson, Jr.
Room:
Royal Suite (260)
Subject:
a.
b.
360 Load Flow
360 Short Circuit Calculations
·0.
-17-
T .1.6
Electric Utility Project . . 1800 Wor.king Group
Chairman:
R. W. Page
Room:
English
Subject:
a.
b.
1800 Load Flow
1800 Short Circuit Calculations
T.1. 7 /Techniq~es Project
Chairman:
W. Pease
Room:
Parlor A (214)
Subject:
a.
~i b.
Data Collection: G. A. Gallaway
Large Matrix Inversion on a Small Computer
T. E. Bridge
/
T.1.8 ':;Education Project
Chairman:
Jack Underwood
Room:
Bonanza
Subject:
T .1.9
J a.
Chico state Col1~ge Registration/Scheauling
Analysis: Neil McIntyre
student Information System of Christian
Brothers College Employing the IBM 1130:
Brother Jerome We~ener
o
Installation Management
Chairman:
Laurence Baker
Room:
Forty-Niner
Subject:
~!
a. Programmer Expectation: D. Mayer (IBM)
b. Programmer Selection and Testing:
D. Mayer (IBM)
COFFEE BREAK 10:00 - 10:30
c"
-18-
SESSION T.2
Tuesday 10:30 - 12:00 Noon
c'~
T.2.1
T .2.2
T .2.3
c
360 Project
Chairman:
Richard Pratt
Room:
Concert
Subject:
44PS and its differences from OS: D. Rumney (IBM)
1130 Project
Chairman:
Da ve Dunsmore
Room:
California
Subject:
1130 as a Terminal for S/360: K. Gabbert (IBM)
1800 Project
Robert Forstrom
Room:
Comstock
Subject:
a.
~W"'7
T .2.4
T.2.5
-
Chairman:
-9
Installation Descriptions
Bonneville Power: B. Hoffman
Colorado -Public Service Corp: E. McLaughlin
b. -aa1Q 8&11. ~u "gar..
Electric Utility Project - 1130 Working Group
Chairman:
S. A. Clark
Room:
Golden Gate
Subject:
a.
b.
c.
1130 Hardware Difficulties
Sound Off
New business
Electric Utility Project - 360 Working Group
ChairIl\an:
O.Bo Anderson, Jr.
Room:
Royal Suite (260)
Subject:
a. 360 Hardware difficulties
b. Sound off
c. New business
-19-
.~~~--
...... ...
~
-.----.~
..-.....•....-- ...-...-.-
T .2.6
T
.2,j
Electric utility Project - 1800 Working Group
Chairman:
R. W. Page.
Room:
English
Subject:
a. 1800 Hardware Difficulties
b. Sound off
c. New business
Techniques Project
Chairman:
Room:
Subject:
W·. Pease
, Parlor A (214)
J
J
T
.2.J
T.2.9
o
a. Non-linear Regression Analysis with three
Independent Variables: T.E. Bridge
b.
User Experience with 1130 LP-Moss: S. A.
Lynch
Petro Chern Engineering Project
Chairman:
Gene Hertel
Room:
Forty-Niner
Subject:
Application of Simulation in Control, Design and
Optimization of Chenlical Processes: M.J. Shah
(IBM)
Education Project
Chairman:
Jack Underwood
Room:
Bonanza
Subj ect:
Panel Discussion on Software Requirements in
an Instructional Program
o
- 20._--_
.. _ - - - _ . _ ....
DELEGATES LUNCHEON
Tuesday 12:00 Noon
Rose Room
()
Speaker:
Dr. Robert E. Hill
President, Chico State College
Dr. Hill is a recognized authority in the
field of international education, finance, investment, and international economics. He
has written numerous articles, monographs
and short papers and has risen from Assistant Professor of Finance (1957) at the
University of Illinois to Professor of Business and President of Chico state College
(1966).
'0'"
,.
)/:..
.
Dr. Hill will speak on the subject: "Technology and Adrninistration; A Paradigm."
- 21-
SESSION T.3
Tuesday 1:30 - 3:00 p.m.
T.3.1
T .3.2
T.3.3
T.3.4
T.3.5
Special Session:
Time ... Share Computing
Chairman:
W. G. Lane
Room:
Concert
Subject:
a. Conversational Computing: James Babcock
b. RUSH, Conversational PL-l: Paul DesJardine
c. Computer Assisted Instruction at Stanford
University: Max Jerman
360-44 Committee
Chairman:
Allen B. Wigdahl
Room:
Royal Suite (260)
Subject:
Division of Problems related to Systern 360-44
1800 Project
Chairman:
Robert Forstronl
Room:
Comstock
Subject:
PROSPRO: O. Merklinghouse (IBM)
Electric Utilities Project
Chairman:
G. S. Haralampu
Room:
California
Subject:
a.
b.
Reports of Working Groups
Progress reports on fault calculations,
transient stability engineering operating
systems
Petro Chem Engineering Project
Chairman:
Gene Hartel
Room:
Forty-Niner
Subject:
Laboratory Automation: RQ A. Edwards (IBM)
T.3.6. 1130 Project
Chairman:
Room:
Larry Armbruster
')"'"
\
Subject:
LP-MOSS: IBM
COFFEE BREAK 3:00 - 3:30 p.m.
- 22-
o
SESSION T.4
Tuesday 3:30 - 5:00 p.m.
T.4.1
Open Board Meeting
Chairman:
James stansbury, President
Room:
Rose
Subject:
a o Members Sound Off to Board
b.. Discussion of IBM-COMMON Relations
c. COMMON Publications
T.4.2, 1800 Project
.,../
Chairman:
Robert Forstrom
Room:
Comstock
Subject:
C;!
T .4.3
a. PROSPRO (continued): O. Merklinghouse (IBM)
U/b. TSX Modification for 6 Disk Drives:
E.H. Spencer
Electric Utilities Project
Chairman:
G. S. Haralampu
Room:
California
Subject:
a.
b.
c.
Engineering Data Banks
New Business
Plans for next meeting
o
- 23-
SESSION W.1
Wednesday 8:30 - 10:00 a.m.
W.l.l
W.l.2
360-0S Committee
Chairman:
"s i. Apderiia,.#r.
Room:
Royal Suite (264)
Subject:
a.
W.l.4
,W.l.5
OS Reread in FORTRAN: WaN Norton
a60-DOS Committee
Chairman:
W.l.3
W"bf' JJol:rO'-J
Don McIlvain
Room:
Ralston
Subject:
a. Discussion of problems
under DOS
relat~d. to... FfORTR}N
360-Commercial Committee
Chairman:
To be announced
Room:
Royal Suite (262)
Subject:
a.
Discussion of problems related to
COBOL, RPG, COS
360 ...44 Committee
Chairman:
Allen B. Wigdahl
Room:
Royal Suite (260)
Subject:
Discussion of problems relating to System 360-44
1130 And Techniques Projects
Chairman:
Da ve Dunsmore
Room:
Bonanza
Subject:
a. Overlapped printing for IBM 1130 Commercial
~pplications USing FORTRAN Write Statement:
Brian Swain
b .. 1130 Commercial Subroutines
- 24·,
o
W.1.6
1620 Project
Chairman:
Richard Karpinski
Room:
Golden Gate
General information on CAl with special
"emphasis on 1620 version of COM PUT EST
SUbject:
w.~ 1800 Project
Chairman:
Robert Forstrom
,I
Roorn:
California
SUbject:S·A comparison between 2310 and 2311 Disk
storage Systems: Salomon Saroussi
'2.
Process Control in Natural Gas Transmission:
,
• ;;
A. A. Douloff
-rZ
W.l.8
Admini straHon Proj ect
Chairman:
Laura Austin
Room:
English
Subject:
a. Discussion of Reference Manual Project
h. Call for help
J1.9
Installation Management
Chairman:
Laurence Baker
Room:
Forty-niner
Subject:
Systems and Programming Project Management:
Ralph Sackman,Jr.
COFFEE BREAK 10:00 - 10:30 a.m.
- 25-
--------.....-...
-~--
SESSION W.2
o
Wednesday 10:30 .. 12:00 Noon
W.2.1
W.2.2
W.2.3
W.2.4
W.2.5
360-0S Committee
It, Anderson
Jr... W Ai>&
Chairman:
Q,
Room:
Royal Suite (264)
Subject:
RAX: D Madden aBM)
~aT4"-'
360-00S Committee
Chairman:
Don Mcnvain
Room:
Ralston
Subject:
Discussion of problems relating to PL-l
360-44 Committee
Chairman:
Allen B. Wigdahl
Room:
Royal Suite (260)
Subject:
Discussion of problems relating to System 3
r.f~
V
1130 Project
Chairman:
Da ve Dunsmore
Room:
Bonanza
Subject:
Discussion of problems relating to 1130
1620
proi~1Jo
eSM' ,d....
Chairman:
Richard Ross
Room:
Golden Gate
Subject:
a. Continuation of CAl discussion
b. Suggestions and Wind up
c. Plans for next meeting
o
.. 26 ..
- - - - --_._---
W.2.6
1800 Project
J
Chairman:
Robert Forstrom
Room:
California
Subject:
a.
b.
W:2.7
W.2.8
(
Techniques Project
Chairman:
W. Pease
Room:
English
Subject:
a.
b.
W.2.9
Summation and Review
Plans for future meetings
Petro Chern Engineering Project
Chairman:
Gene Hertel
Room:
Royal Suite (262)
Subject:
a.
b.
"'"
)
User Experience with. Teletype Terminals
on the IBM 1800: W. M. Schonlau
Plans for next meeting
Instal1at~on
Summation and Review
Plans for future meetings
Management
Chairman:
Laurence Baker
Room:
FottY-Niner
Subject:
A Review of the GUIDE Installation
Managem:ent Project: Arthur Nich()ls
.........
----.--~
.... -.-.-
~-
..
_--_.__.__.._...-
..•__._•.....
".,r-') ,
(~-J~
SPECIAL ATTRACTION
VVednesday Afternoon
IBM San Jose Plant Tour
a.
b.
c.
1500 and 1800 Production line
Direct access storage devices
Demonstration of 1800 MPX
Complementary Transportation by IBM Charter Bus
Leave
San Francisco
1:15 PM
Rtn
San Francisco Airport 4:45 PM
.Rtn
San Francisco
(Sheraton Palace Hotel) 5:30 PM
Reservations must be made at the registration desk by 12:00
Noon Tuesday.
Delegates may choose to tour the facilities
( a and b) or to attend the demonstration
(c). We regret that time does not perrnit
both.
c
- 28-
"FpprWT filt"W7Wr
W
SESSION W.3
C:
Wednesday 1:30 - 3:00 PM
W.3.1
360 Project
Chairman:
Richard Pratt
Room:
Ralston
Subject:
a.
b.
c.
~
W.3.2
Reports of 360 committees
Recommendations to IBM
Plans for future meetings
1130 Project
Chairman:
Da ve Dunsmore
Room:
Bonanza
Subject:
a.
b.
Suggestions and Improvements
Plans for future meetings
C'
C,
',c,
,
.
- 29-
."
---...
......•
~-~~-
--....... .... ..
---,---..............
,.''''
,,,.,
,-,.~-~
..
~~~---------
Abstracts of Paper s
Session M.O
M.O.1
o
New Members Session
At past meetings the orientation session for new members of COMMON
has always taken place at the end of the first day. This arrangement
has the result of permitting many new members to walk around in a
fog for one-thirq of the meeting! In an attempt to remedy this situation
we have scheduled a short session before the meeting proper begins to
familiarize new members with the workings of COMMON and to make them
feel more at ease at the sessions that follow.
Session M.l
M.l.l
General Session
This meeting will be chaired by the president of COMMON, Mr.
James stansbury and will be in the nature of a formal welcome to the
attendees from the Executive Board. There are no concurrent sessions,
so that all registrants may be present. In addition to the usual introductions of CO~1MON Officers and special guests there will be time reserved
for last minute changes to be announced by the Arrangements Chairman
and the Program Chairman. If possible, procedural questions from the
floor will be invited.
o
Session M.2
DYNAMIC MODEL SIMULATOR FOR THE IBM 1130
The Dynamic Model Simulator is a chain of FORTRAN source programs
that permits the IBM 1130 user to investigate in depth the dynamic behavior
of physical systems that can be modeled into linear or nonlinear ordinary
differential equations.
It can be used for the investigation of a wide range of engineering
and mathematical problems from reaction kinetics and reactor responses
to the design of electrical networks and structural assemblies.
The system does not :r:equire familiarity with analog computers although in a way it fo.rces the IBM 1130 hardw~re to perform the function
Of an analog computer while at the same time it offers the digital computer
advantages of random memory access and ten digit (extended) precision.
The input consists of English command words (START , WAIT, RESET,
etc.) and numeric data which are read in free format.
The system output includes tabulated values of the variables at specified
intervals, a table of the 'variables involved and detailed error and information messages. Plotting of the variables is also available as an option.
-30-
o
j"."j"'H'i" ,; "" "f
"""
"E" t1j ".
""'j""
"&¥'""Tf,( HiT"" "j&i!fditffWEtiWii£#E¥&""" 'T"" ""'"WnWYUWillTillHW'NP
System Source Language:
o
FORTRAN IV, Level E
System Hardware Requirements:
1131 CPU-SK, one 2315 dis.k cartridge
. 1442 Card Reader
1132 Printer
Benson-Lehner 305 Digital Incremental
Plotter
(optional)
PROGRAM LIBRARY
Discussion of the proposed PREP Forms for the 1130, lS00 and 360
Library. There will also be a discussion of an extension of the minimal
standards for this Library. It is requested that there be representatives
from each of the machine system projects in the Systems Division present at this meeting.
THREE-BAR AND FOUR-BAR LINKAGE SYSTEM WITH
PLOTTED OUTPUT
A mathematical model of a linkage system is controlled by an array
of inputs to plot out the path of the linkage intersection point. In addition,
the system simUlates turning the entire mechanism about a major axis.
The output of the system has yielded some very interesting designs.
In conjunction with this report, a list-oriented free field floating point
input program has been developed. This routine enables the control
program to modify only specified items of the system input. Under break
character control, an automatic interactive procedure can be set up.
Session M.4
'A LARGE 1620 DISK OPERATING SYSTEM, REASONABLY
7094 COMPATIBLE.'
o
A new monitor for a 40K or 60K 1620 has been developed at Princeton
for use as a debugging aid for the 7094 and some 360/0S programs.
FN II I/O and FN IV I/O· are included as well as private storage of programs. Many control card options exist. Speed can be 2 to 4 times faster
than MON-I (if one has hardware FLT. PT.)both in compilation and execution. A users manual does exist. Many students are presently using the
system on a completely open-shop basis.
- 31-
'BASIC PROBLEMS OF INFORMATION RETRIEVAL AND A
SOLUTION ON THE 1620.'
Some of the basic problems of information retrieval and implementation
are discussed. The techniques of fi1e-organiza~ion are discussed. These
techniques are discussed for the most common file devices (disk and tape).
A solution is shown for the 1620 with disk to demonstrate disk utility.
FAST FOURIER TRANSFORM
WITH APPLICATIONS FOR THE IBM 1800
The Fast Fourier Transform technique as developed by Cooley and
Tukey has had a widespread effect in the field of time series analysis.
However, some difficulty has been encountered. by potential users in
determining exactly how the technique works. An effort to explain the derivation in detail will be made
Also, a description will be given of an analysis package program in
which the Fast Fourier Transform technique is used to yield amplitude/
phase spectra, power spectra, cross power spectra, and auto correlations.
NIMS- THE AEROSPACE SCIENTIFIC DATA REDUCTION
MONITOR SYSTEM
The IBM-TSX System was designed for an IBM - 1800 operating
in a real time environment as a process controller. The Aerospace
IBM-1800 is used primarily as a post flight data reduction system interacting with telemetry ground station equipment. This demands the
fuIl interrupt facilities of the system but in a batch processing mode of
operation such as offered by the TASK stand alone type system. The
, resulting NIMS Monitor System combines the features of the process
and non-process modes of operation into a general mode of operation
in which the full interrupt and peripheral facilites of the machine are
available to the active' application program under execution.
The system basically is a TASK OFF -LINE system with the system
components modified to allow Process Subroutines and Interrupt Service
Subroutines to be included in the core load generation. Other features
include a one card cold' start, ability to assemble or compile from magnetic
tape input, and a magnetic tape backup system. In addition, the TRACE
and CORE DUMP routines of TASK have been modified to provide and
extended mnemonic output listing.
--32---------.-.-.~--
. . .-
---------_._----_ ........
_
.. -.
-
----
- .... - ..---------..
-.~---.-----
-_.- --------..
...
- - - - - - - -.. -
.. _ - - - - - - - - - - - _ . . -
o
;- '"1
o
.
t·l· ..
··t" ...... ··\ ...... if .... ili·
j
--s .
CURRENT RESEARCH CONCERNED WITH THE STANDARDIZATION
AND FORMAL DEFINITION OF PL/I
In April 1966, a subcommittee of USASI Task Group X3.4.2 was
. established to investigate the standardization of PL/I. Since that time,
two working committees have established with exnlicit charters:
Group I: The resolution of the language PL/I
and
Group II: Formal Definition of PL/I
This paper deals specifically with the work of the committee on Formal
Definition and its relations with the definitional task forces within IBM.
This paper will discuss the techniques of definition as proposed by the
cOlnmittee, the uses to which the definition is to be put and the implications
of the work of this committee on the standardization of other computer
languages.
A familiarity with PL/I is not presumed.
Session T. 1
c
THE 1620 AS A DATA COLLECTOR OR
SOFTWARE, THE CRUTCH HARDWARE BOYS LEAN ON
The paper will cover the date collection methods used at the Sacramento
Peak Observatory. The principle reasons for designing a computerized
data collection system were:
1) Improve existing methods which used sumrnary punches.
2) Gain experience and insight into problem~ associated with data
collection in preparation for third generation equipment.
The paper will also cover the roles played by the programming staff
in this type of operation. Namely, software as a diagnostic tool for the
hardware boys in developing data reduction/ collection instruments, and
software as a useful and flexible tool for the research scientist.
Three methods for data collection and reduction by computer have
been tried.
1) On line, one point at a time, testing each point for parity and
structural errors and storing it on the disk before the next
point is generated.
2) On line, a record at a time, locking the source device out until
the data has been tested and stored.
o
3) Off line, using 7 track incremental tape recorders, with testing
and reduction being done during slack times.
- 33-
t·b····'i'ri'FfMi*ii'j'j·-t- .. a-·· u tlF#ii±i8-%it .. ,.
CHICO STATE COLLEGE REGISTRATION AND
. SCHEDULING ANALYSIS
"Salient features of Chico .state College's computer approach to
student scheduling and registration are discussed; applicability to other
colleges and other computer configurations is emphasized."
o
STUDENT INFORMATION SYSTEM OF CHRISTIAN BROTHERS
COLLEGE EMPLOYING THE IBM 1130
We received our 1130 around July 4 and immediately began programming
it for our Student Information Svstem. This is a great improvement over
the 1620 system since that was chiefly card oriented and we can now keep
all the information on the disk. All of the programs are written in FORTRAN
with the exception of two or three adapted with the Commercial Subroutine
Package. After the students are registered, we are able to produce class
lists, student schedules. mailing labels, all types of statistics, statements,
report cards, and permanent record labels. The system works very well
and will become more efficient when we add an additional two disks and
receive our 1403 printer.
o
G
- 34-
Session T. 2
o
USER EXPERIENCE WITH 1130 LP-MOSS
U. S. Reduction Co. has used the 1130 LP-MOSS application package
since its first release. Various problems have been encountered during
its, use; however, useful answers have been obtained. Due to the size of
the aluminuln alloy blending problem being studied, a second 1130 had to
be leased for full dedication to this problem. Early results indicate a
reasonable payback may be obtained from this hardware and software
combination.
An Outline
1.
2.
3.
4.
5.
Company background
Early uses of linear programming
Problems encountered in implementing 1130 LP-MOSS
Results to date and future planning
Specific application modifications desired
APPLICA TION OF SIMULATION IN CONTROL, DESIGN AND
OPTIMIZATION OF CHEMICAL PROCESSES
The problem of supervisory control and optimization of a chemical
process requires that the control as well as the optimization programs
be provided with a mathematical relationship between the dependent and
independent variables in the process.
In this presentation we will discuss methods of arriving at this relationship based on the fundamental laws of physics and chemistry. We will
make the proper approximations valid for control purposes to obtain
the solution of the differential equations, describing this relationship.
Use of simulation languages helps in reducing the programming time as
well as the number of trials required to attairl successful solutions.
An example of methanol plant will be discussed in some detail to illustrate the mathematical and programming technique's to achieve the desired
relationships for control and optimization. It will be shown how these
differential equations with modification of the objective function when
used in the optimization program can lead to optimum design of the plant.
o
- 35-
~1i!iiiiiiiiiIBi;:;;;:;'HI1I1M31",;;;,r\l...,"a;K,6;;:'=P"d;::;b,,,IA!;1J[
·A" .. rutL''''·'',·j@.'''JO·'i.·f!#I''f..liF·'':;'''·'\'I··' II····· "" .. ".;.. """.(;Z;"""""""I'! ... ~ '~"""" ,.
44..1.,';'.:
......- .... __ ._ ..................
~~---
_-------------_.__..........
..
"".•...•-"~
....
,~~~"'-"""
Session T.3
LA BORA TORY AUTOMATION
The session on Laboratory Automation will include a general description
of program'ming and equipment aspects of the application of the 1800 and
1130 to this newly emerging and fast growing field. Following this
discussion, an application will be discussed in detail. This will include
a description of the type of research to be accomplished, the incentives
of the on-line computer, the programming system to run the instrument
in a closed loop fashion, and other experiences to date with the system.
o
o
-36-
--"
......
-~
..
~-~-
..-
- - -....
--rrK-W"T "Wl?Tvr---zgprwn
Session W.1
o
OVERLAPPED PRINTING FOR IBM 1130 COMMERCIAL
APPLICATIONS USING FORTRAN WRITE STATEMENT.
A method is described for incorporating overlapped printing into
IBM 1130 Commercial FORTRAN programs. Communication to the suosubprograrn which performs a printing operatIon IS acnlevea tnrough
the FORTRAN WRITE statement rather than through the CALL statement.
The advantage of this method is that limited use can be made of the
formatting ability of the FORTRAN language. Headings can readily be
incorporated, and the layout of the printed page specified by use of FORMAT
statempnts"
A BA TCH PROCESSING FORTRAN SYSTEM
FOR A MINIMAL CONFIGURATION IBM 1620
o
This paper describes a software package designed to increase throughput on a 20K card system IBM 1620 computer in an environment where
a large number of fairly simple FORTRAN programs must be processed.
The increase in throughput is accomplished in two ways; (1) programs
are batch executed under control of a loader-monitor routine, and (2)
a powerful precompiler reduces the number of execution errors and de':
creases the requirement for on-line debuggingo
The cOlnpiler is a version of the PDQ FORTRAN compiler which
has been modified to handle the batch processing features of the system.
Batch execution is made possible by keeping the entire subroutine library
resident in core rather than reloading it with each object deck. Object
programs, separated by control cards, are stacked for input. The reading
of a control card by the FORTRAN card read subroutine or the execution
of a CALL EXIT statement will cause the next object deck to be automatically loaded and executed. The monitor will also ternlinate a job if the
output line count exceeds a control card specification.
The precompiler detects more than seventy distinct errors based on
PDQ FORTRAN specifications. Many of these errors, such as undefined
synlbols, are undetected by the compiler and cause execution checkstops. Recognition of multiple errors in a single statement is possible,
thus eliminating multiple debugging passes.
o
- 37-
-----..---------.---.-"..-...... ......-"""'-..
"
-.-.~~~---------------------------.
A COMPARISON BETWEEN 2310 AND 2311 DISK STORAGE SYSTEMS
A series of programs has been written to allow 1800 users to fully
utilize the capabilities of the IBM' 2841-2311 disk storage system within
the frame work of the Time-Sharing Executive -- TSX, Version 3, Mod L
o
A comparison is made between the 2311 and the 2310 disks in programming techniques, timing and cost effectiveness.
PROCESS CONTROL IN NATURAL GAS TRANSMISSION
WITH AN IBM. 1800
This paper will present the progress made by Trans-Canada Pipe
Lines Limited, Toronto in the field of process control computers. A
brief description is given of the feasibility study prior to ordering the
computer, the organization of the implementation team and the methods
of implementation.
The purpose of installing the process control computer at TransCanada is to save fuel by more efficient operation of compressor stations,
and to guide the dispatcher into better control of the line, thus allowing
more throughput at the same or better operating cost. The computer is
not closed loop, but accepts telemetered data from all compressor
stations on a priority interrupt basis, optimizes the line by means of an
on-line simulation program and then informs the dispatcher by typewriter output what change, if any, to make to achieve optimal operation.
o
A description is given of the telemetering system that feeds the
computer and of the Simulation/optimization program that is used to
control the line. The computer, the tB"M. 1800, was installed in June
1967.
SYSTEMS AND PROGRAMMING PROJECT MANAGEMENT
The allocation of limited systems and programrning resources to the
highest payoff areas is becoming more· important as data processing installation costs continue to rise. Long and short-range plans properly
approved by top management and formalized systems project managell1ent
are vital tools to assure that systems capability is focused on the major
areas of the enterprise and adequately controlled to attain the stated objectives. To work effectively in this environment, additional demands
are placed onto the systems group to develop systems plans in terms
easily understood by top management and onto data processing man . .
agement to bring about fulfillment of the approved plans on time and as
economically as possible.
- 38- .
-..
---.----~-.--
..-.-
--~
o
Session W.2
0.:\
USER EXPERIENCE WITH TELETYPE
TERMINALS ON THE IBM 1800
'.)
The
the IBM
teletype
to solve
following describes and addition to the programming system for
1800 computer. The expanded system will support up to 16 remote
terminals being used in a time-sharing environment primarily
realtime inforrnation processing problems.
1800 Hardware
The 1800 CPU Inakes extensive use of hardware interrupt levels and
data channels to operate its standard data processing I/O equipment. In
addition the 1800 can have analog and digital I/O capable of communicating
directly with almost any kind of equipment. The terminai system uses
2 words of digital input and 1- word of digital output (16 bits/word) to
control all 16 terminals simultaneously. No data channels or additional
hardware are employed.
TSX Programming System
o
The 1800 program.ming system provides many conveniences. Programs
are of two types, process and non-process.
Process programs can be initiated by externally generated interrupts
or can be queued by any program for execution. They can be a part of the
system skeleton which is in core at all times or they can be kept on disk
in core-image form. Process programs have highest priority and generally
use SaIne analog or digital 1/00
Non-process programs ordinarily do not use the process (analog and
digital) I/O. They are usually stacked jobs of a conventional type to be
run under the non-process monitor.
When time-sharing, the systenl does the following:
(1).
(2)..
(3)..
Runs non-procE::ss programs for background (job shop).
Periodically checks the queue for process programs.
Permits externally or internally generated interrupts to
load and execute process programs at any time.
Both (2) and (3) require that the non-process job be saved and restored
when the process work is finished ..
o
The Terminal System
The ternlinals extend the tiIne-sharing capabilities of the system by
allowing up to 16 users to communicate with the system simultaneously.
- 39·'·J",-" .••. ,4,
,.,·."·,;,44;.·,·",··,,
".,,,:+.8f'iiifiT\'I$fff,
'"#'''''
,.. "4,...., ..'.I>'.. r'.'.Q..'li£#%2W,,.,.:
....
_-_...._........_...__.._----_._._.._...__..... ................... ....."
,
..
~.-~--~~--.~~~~~~~~~--~~-~------------
Terminal activities include:
(1).
(2).
(3).
Queing process programs for execution.
Communicating with programs during execution.
Programming in NUtran (conversational fortran).
o
Many other activities are planned.
Most of the terminal system capabilities are achieved by simply
making the standard TSX functions more accessible. The only programming
efforts unique to' the terminal system are the communications controller
(simulated by an in-skeleton program), and the time-slicing of terminal
service programs.
o
o
-40I
I.
I
__
~S
D
ALE X AND E f<
.
C
'Of:. 1"1 f-
~j~.t~. t= I':,l.< C_tL s.JJ.~!l[;j:L__.__ ..__.--1i.()F E.LI 1._ F JtJ.".JJ_.~.1:\.b_Lf_. ___..J,.J:Ll~.o.. _.
C I 0 L~f\ LTC; Y
C Ci'.~ P
H () L L J S
hi
Y
0
3 60
0
__ ~.h L_H_R_'!.I.I_SJ_i,,____ . ___ E .. _..~Q_l1J'J DJJ:
_f_t:;.!::.~ i~~Q_. __ .___1 b 0 ~_~.b {\J.i..~~:~yJJ.:c'= .L.H!L_______1 H D.S!
ALLEN
R F IF~t,·;
Ct: I Cl'-,(~U ILL
360
_.~b.y.,t~:,. __.__.__ .__ .__._. ____ F_r~!1J~~..:.... ___ ._._. ___
...:,_:~
_.:'?o!~,.I'·1 __ .JJl~if_5:;.(L~. JE __._____ .._______ .__ ._
I __
I
ANDERSOI\!
.... __1\ ~.!<
_ _ •• _ , . _ _ _ _ ....__...
SJH! Fr·__ f.h.!.!~.·F_i~l I C t·l ___. ____.___ .____.____1.:9.2 0 .
AYE R S
f~ [) HO \,-':i·: c: T C UI": I~
CHI C 1\ (;U I L 1_
.... ,,3'.6 (}
I NS
HADFN
-,---.. _- ..
:~---
8ft,.QEN
___ t_6J..b E Y
S
F ~;: A ~~~ S 1-1 /\ L L
._______ .J:.~_.C.!Jj_L~__ .__ .. ___ .____________._____._ ___~~P/\ T CJ G !\_~J~LJ£_
1l-:; (;
360 ..
__. ______ ~_1:JiJj)J:Lt~5.T._JLU= __(JJ£~ ________ Dr:: i',~\Lc;.,~L~J.!...l:_.!L. _____ ._________.__ 36 Ci ._
t\1
JUr'~IV
OF ~~EhJTUCIJD :'~lCh
IBOO
2 l;. 00 COL U;.: [:3 US If'! D I A f·.! A
-1 1 30
B CHi/~ Y ,_ _ _ _ _ _ _ _ _ _ J PC U j\'1 i't! i'l SoC I\! G If',1 E en
V.J
ST
S -PI;~~~ t:: ~~-------'-BR-O
HAvE·~J'~,:l:i-s-s----- ··~-~·-1·80(-;::BOl'JD
V L I l-~.~:: ____._.
. ._______.__<; /~~1?b E L L C 1\ L I F
___ ~ 6 D__
_ ~f:3 q~? a N.:......._ _
I iv'iA
-F -RE'"GI
BOUC
HEF~
H
f3 0 Vj E "._f_-________
.J::{f..~J
-.---"': ----.
ILLli. ~
CD
NE:\'!
HAVEN
1130
3:6'0
COf\IN
_._~
---.-----..----- ,-
DGE/vi AI',)
K
-~
.HR
UNITF'::D
I
R VJUS
Gt::CJL
.. CJGI
~U!~VEY
CAL IF
__M ...
_____
_ _ _ _C/\L
_ _ .. _ _ _
.. _ _ _ _ _ _ _ _ _jviENLO
_ _ _ _ _PAf-~K
.. _._. ___ _
V T I b r,,'j \'! T C
SAt·..) J 0 S l-:: C ;.\ L 1 F
t,j XCENTF~!\L LI-\ ~::LECT~ -1 _
J
OE~!ISON
PA
YO~~I(TU1:i~,1
1620
360
1130
f'j Y
:16 0
F O:·0
F R I r=: [) t J-\ N D
1\1 J U ~_LU~}1__L(__.t1.~;:~~=.I..rJ.__;;., L:".UV ____ L 1\ ~ V ~~i£\..~ f~lE V /\[! p'-_ _ _ _ _ .-lL~CL-_
FULLAN
0 JI8M CORP
WHITE PLAINS NoYo
360
.___ 0~-,'l6.JRA
H
.J
G A NAT R A
GARD
D
GAR [) 1\1 E R
_
.D
_____.___ DE_T:~. L L J\ i·,;
A F F Hd:: STU i',j f-:: T I ion:
J-\ K R (l!' I C1 H I G
3 60
H/\ LL 11'1 G
[-3
I-~ OLL s-~ Cl YC E LTD
DE U<3 Y Et,1 GL !\f\D
__ 1800__ _
HJ-\ f-i I L
Vi Rlbf·i-5DD
1 1 30
H A ~~ A L J\fvl P U G S I'~ E ~',; E: r\1 GL II. f\j D l:: LEe S Y S 1 () 11 1 b D S T (I i·) r~': t\ s s
1130
- - - - - - - - - _ . ._-_ ._---_._-------_..
.---------.--HATFIELD
F Id,iCGf~A\".! EDISl)l,1 CC!
::;Otk;) ZI-\r'Il:.<'VILLf OHIO
360
HA YE S
~'1 E I E::j'v! DE PT
?06
S/\ r'! J os r: C,I\L 1 F
360
-H-t-\-Y·,.~-/\--'~-D
- - - - - . A P"[)'UOUEs-r:~[,~ ___._ _ _ _ _ _ _3 6
D_.I_V_I_S l_l_-.ii'_'J_________. __
-.---.---------------~----~.-----.--.-.----
.
.3 (:) 0.__ _
(~_._
vn~~_ ~_f.~_~~_J\_I_i·_,1~:_ .• ~_~ _______. __1_1_3_(~___
---------_..._._._---------"-_._------ _-------
-------.._._-------------_ .. ._-_._-------_._-_._-_..
..
--~-------------.----.-.--.-.-.-------
___.___--.------.--._.________. . __________.__.______________#:B.L______.
llSLEZLWLW CWJ';:G£J;g;b;;;:;.g,;;:;b"'k.;:;;:;;;;;;;:t"JI".~R..".,'" .
..~ .. , ., .... ,.
-_ _--
~.
•.
I
..
1
•
~
~
PAGE
HILL
HI-P-S-'K-r-N--O------H0
H 0 F F f"j AN
\fJ
J
1tJ
_t!. C1 VJ E~.hL_
HUGHES
Hl3ENDIX cor~r:~
G C BESTOf< "j'
161';:' TETI'R"OPO tjJ
~
. ':~-CARr~,~~: C'-AL~;F-'--"----'-.-..---~ 1 ~VJ--'
-
A.SSOC
_._-.::....._ _ _ _~~i:.l
IBM
'I)
E HER C U L F S " I_tJS_f) R ~,______ ._.__ __.JJ_!..bli.!.~~J.JJ;U:i._p Eb______
OMINN PO~ER + LIGHT
DULUTH ~INN
R
v: A SA r<. I
S
U I'\J I V E 1.( SIT Y
c
1-\ L T U . UN r-= [)~_~ H
P t\ L 0
FY
117 0
._____J~!;B.I5£l..J::;)/_~ !-\L. 1£-=-__..
It-)25
f)
R RD M ~EATHE~LY CO
S r':!.l.Ul:i....
_________ r_~
JOHNSON
._~QJ;.L
1 H 00
1 62 0
L P 1-,( INC E T
ij \\' AN G _. _ _ _ _ _ _ _ _ _.-=C~~E-::..:.:\E:: .:.8!.~E
·
L E '(__ ~_~I_E;j~.LLF I
INDIVERr
r~ LSI1\!CLAlk OIL e()~~p
I
J 0 ~_E _C AbJ_F
N. J •
P l~ INC E T (l N
L
(J f\I
4
CHE'=L.~LL~,~i£~_1:.£~I,.._J£_._ .. ___
360
360
1 1 30
360
1 62'0
360
. __.~_____L(ULQ___.
JONAS
C RoeHl CHEi"ICi\L ell
3£:!·17 l'.'lIDLAND ,'iICI-'j
leon
JONES
C E lJ:: N 1\1 (\ + 1_._5 T /-\ Il~'___VtLl_V___ .__-.!..i'..:.:.~A~S H_Y_LLL..t';"_l E:: f\!!j
1620
JUL IAN
F c} C 0 U NT 0 F SA j\J 0 I t: (~O
S /\ N DIE (~() C IX L I F
JS.[iJ =Z_ _--,-.--'--_ _ _ _-=-_ _E==-:_--'L=~~._.l!.I-\ T E f-<-±-.!-::1~i t.:} f~EL _________ IJ_~) i:.f i'if~y__(:.!:~~.lJ=- _____.____ ~.9~_._
t< END RIC K
N H 1\ E C H E S BUT I~\ j,! E P F~ (j U
P () ~~ T NEe HE S T E X fA S I b () 0
JSJ':: Nf\! Y
.. ___._ _ .,, _ _ _ _ .•. _... _~.bJl~
. t!__~....Y..QJ,iL~. __ C.JJ:...'L_b~f!..Y...!.________U:i'O_Q_ _
A
D IJ; j'i
D
Fiv-;ICHIGi\N
\'1
J I !·~~~_. _______ •. _._ ..._.___.______...• __ ._~!:\l:LfJ..~:::!~~~J .?_~. ~·L.~.f.:\...~.J F. _ _ ._._1.J...::.2.L__
EMARATHGN OIL CO
5317 T~XAS CITY TEXAS
11~G
ST/\Tt~
KIEL
KING
KIRKHAiVl
K LEE S
K NO ELL
LAI,ISING j"iICH
1130
J LAF~Ko P()\·.rf,J~~_cH:L__.. _.1t}'.~(_E.U\;E .£lLUFF /-\Hf(
11,.:,~:;',,-,)C==-J__
G KId/vI
S/\N ..JCJSt:: C/~\LIF
1800
G Nt-, U 'I~Q ,,!f:':: TIC S _ _ _. _. _________.-E:.lJ_'= _~: E f'~JJ2:~~ I~ L IE __ . ______l~_~)JL___.
D L U /\ SIC V E G L l' i\ U L E P h: f 1[) !':-) I~ C () V /-\ C A V ILL E C t\ L I F
1 1 3 (;
K 0 CH
KOSTER
D
K R lEG E i-~
L ,i\ j·1 A !-~
L/\NDFCK
3(,25
E=./\ST
__ £'~.~!i' r-~< E: E_ PSI t: ---.t.'!....~-.Y_~__
3 6 ~~____
K H C /-\ L P U L Y C U L L l::: (; t:
p () j. i 0 I~ /-\ C i\ L I F '
1
N CANOGA ELECTRONJC~
CANOGA PARK CALIF
11
---------... --- .. _----------J E \.' EST U~ 1'1 C U Vi T I'~ C [) I-~ P S I 0 U x CIT Y I () ~'i A
36 0
J
K F( lEG E
KUREK
LAC 0 U R S I ERE
J.:: A I-ific_R S
lJi-'IIV
fVl It·.:) j",
10"
-~-.----.----
F L v./-\ H 0 S P I
J R I hi'"
_ _ _ _---=-B ~'/ I 6 I,,;
T~.'=___.__ .. __~ ______ ;~~~~~__~~J.:Le:..f::!~
__!.::Lf_ ~~ ~~ ;.\ s ~SA ___________L2..?Q._,._
C t'l I C A G ()
1S
I l. L
°
0
LAI'·~D\'jEHR
S /\ ~,l J 0 SEC AL I F
12.00
LANE
S E OK L i\H 0:1,; I~ U I,d V t.:-. PSI I..~ ______ .~'J (Ji3.Ll!:~':LJ2!~l:_A _______________3 C?_o___ _
LANE
\"j GCHICD ST.!:I..TE CULLEC;c=
CHICO C/\L IF
1620
__L /'. . F< U E
H H ldJ::L!.Y.._~~E-.-:?_.~!.._l) A 1<' CJ IJ:..~ ______.. __._____ 'y C::J3.i JJ..b.kJJ-l 1,1 oS t:"l..l.L~.~.LLIlL._____~ 6 ('L-r~
LA\'!REr'-lCE, Jr~
\'} \fIltH;;
POUGHKEEPSIE I':Y
1(300
LA Y_N..c.iG-=-_ _ _ _ _ _ _ _ _---.:.:..V'--=L:;,...S_-Ut'-!DS.TP I-\h!~~. I I~ -!:'.l..C1N __ .__~~Q~~.~J2kD__ l~ _______...__.__ ..11 3(2 __
E S U t\J I V D F T (1 f< UN T (l
-I CJ 1 T 0 F< U t,JT U () 0! T C f:, j',! /lD A
1 6:2 (;
LEE
LEE
J
A
Ut,!
I
V
DF
HI.
SSACt-IUSr:: TT S 1 1 07 A,\IHE ~.(S T Ii; ~\SS
1620
- - - --_._-------
K 1-\ 1',1 S A SCI T Y tv1 n
L MI.Ji~-.l'!...!..r_.____ _
18Q.. .;,; O_ _
SAN JOSE CALIF
1130
BeLL I NGH/\;.. l Vl ASH
1800
POUGHKEEPS IE-·t\-!-o-Y-c-------
L E HR
F< V P /\ N H ft. N D L E l:-: .L\ S T E ~ ~ i'J
L E:..:..N.:...T~_ _ _ _ _ _ _ _ _---:F< SIb i·1
LESTER
. G
I~~
LEVY
l. NI'/iO!:3 I L 0 I L CU
-CF\:vf·~s--------·--~ N -,J I lJj:~·-·----------
E~
LEWIS
E
SYLVANIA
BUFFALO NY·
360
LIN I C K E F S 0 F TVI ARE --R' E S() U ~~ C1::: S
LOS· ,i\ N C; E CE s C .£\ L. I F
36 0
LIPSOr,J
A LVIRGINIJ-\ ELECT + P~'!f< 10£:~t~ RICHf\101,11) VA.
360
-L I T T r · , l A N - - - - - - - - - - I - £:30 I SE CASC·/\L.>G:-c·Cf~~-·------Ji:)Tsf~I-D;~;i"(J-·~---·-··-··----1i3--O-~~'"
LOMAS
W AlUM
P---I-Hfvl
LONE~~GAN
SAN JOSE CALIF
\In'lITE PLAINS N.Y.
----------.-----------
------------.---'1,
- - - -..--1\--.
--_._----_._--- ._--
"--..-/
. - - - - - - - - - - - - - - - ._------_.----_._--------_._-_. .._-------_._-_ -----_ _-_.•_----...
...
-----.----~.-
.......- ...
- - - _........._.._ . _ - -
.
..
-----0-
· t""
..:..-..
j
t . ·.. jt#"S "#· . ff.'ffJW······· W2Mif'itiiWRWWiWSiiitw -r-' Wf"-,"R[
n
~-
..
-.-"----------_._---------'-----------------_...------_.---------------------_._--------_
-------------------------_._------------------------------_._---------.....,-"PAGE
~s
I S~A.J2l.Q__~J~(:J _._~J_)___
.---'-'--=-.-'--_ _ _ _ _ _ _ _LYLJ.l~,JG.
LOVE
1120 H 1 Ci<..~_lLL_';:
tvi A I Ht·i
L UF3 Y
\'.1 K C U I~ 1'1
LUI< 1 N S
J C UN I V
,. . :L:;:-. . ;.Y..:C:...;,I-_i;..E:;:;.:=_ _ _ _ _ _ _ _,_ _..::.D::...-::.:.\'i.';:S;.;,:.UfA
P P. U r~~..I~~S:~~~~l~.s2g
() F I<. E ~'J T U C K Y
1 ~) 5 3
Ny _________ .-_~16iL.
CLEVELAND OH 10
11:30
DO if.' hI E: H :2__ G ~< 0 V Ll L L
1 b QJL
L ~ X 11"; G T c) N K Y
1 t3 0 0
'=-______,__.__ ._._____.____r_..-f;.l-_.._~_s._(...vJ~.P cL._~ 1\!,.._1E.
1 f.HliL:
LYNCH
S AUS REDUCT I LJi'-l CD
EAST CH ICAGO IND
1130
___f'Ji ~C A
SA I:J J D S E ( A L I F'
J
ES I
P
R
R
K S U I\J 0 I L C (l
V' I H jvi
SSUNDSTRAND
j\!
AVIATION
DENVE~~
1 H 00
1 a 00
1 1 :-5 ()
360
COLD
__~:i.AYER
MC GUIRE
S
(·i e l L V A 1 I'.]
I?
Me KINNON
G
rviC LAUGHL IN
E
jv; CPA H I L .
R
r·./;c au AR F~ I E _ £ .
j\':CCALL
E
--OE A C HE:' f.( ~!
,\ 1C I [\J T Y R E,
R ~HITE CO
DENVER COLO
1130
P F~ () [) LJ_~ T S +~tj E :"./ _ _ _ _ _A_~I-=-~ N T f.lJL N P J\
~i?.Q __
JIBM
ENDICOTT NoYo
360
EPutLh) C SJz.£:..~.J CI~_. CG __._ _ _ _ [).f:::I\!VE r.:~ CCJl,-_~ __________~§_(~..Q__
e c; [.: ~,! F.::r~ A L D Y !,·jfd·: I C 5
G R () TeN C (J N j\j
3 60
p.E; ~-.&~~_'LI..__ _ _ _~ __.___ . . __ .Y~~In.f:li.t·~_D C_.~.t:~t! !\[) 0.
. 3 (2i!_
Hlbrv; cu
12HJ<;i ST Pt,UL j.'IIi'.li-J
360
WKEN
f~ A
I k
N_-'Y..QI: LEU \'~_ C I~.l_~~.f.~/ C A..?~. ~..!) J~--.J (l fj
JR
tv; F H L
J
C CH I C (j
\-.1 I ['.;, i
fJ K Y
1 1 30
o
L [) E ~I
L R (; E r~ D Y h! cur ~ v t\ I I-<
~) 2 7 b SA f\! D I f. (~O C J\ L I F
1e 0 0
------------------------------_.
-------- --OLSON
R wIbM
CHICAGO ILL
369
_.t:0 C H ______ ..::>_?.2._
__
O_R.~0_F_F_. ____::... ________t\_1_J_L_F.._:I_:.._F<.__~IJ:: Ci.'- t~~ R _1. hi ~_. _____ ~ ____.__ G 1-< I:~:I.!:_.?:.!\ p I !~ s
ORTdf\LS
~/; H\'iESTERN FA~,b,,.-_ _ • _ _ _ _ _ _ _ _ _ .•_ _ •_ _ _ _ _ _ _ _ _· -•...- - - - - - - - - - - - - - -
----_._-----------_._-------_._--_.
__._-----------_._.--------
..
~-
· ..
PAGE
PINFIELD
E
PODOLSI< Y
J
R ~L!~} V_.Ql:~_~_~]-':=-() - fil ~12~ F: N T __
L F A I R CHI L D S E: [.'j I C U f\J D C T
A ___I 8/'!i F 8P. N C;.S,______
P_QLLE
POLYZOIDES
.. P 0 FZZ I'J.!S. _____
G
J
PO\'1ELL
PF~
J
DH0
(J I<
E. F<
C H E 1\: I C /\ L
"If EST j'-'U~I S I~.g C ()~ . _ __
jvj T
V I E Vi CAL I F
6.
IV-
P J\ F< IS. F 1:<"6 j-'·l C E ____.__________--.l1:i..illL__ _
C U F< P
N I A G A f-.1 A F t\ L L S
N. Y •
1 1 .3 0
_!?.LAF ~~.Y£.I.IX;__~J:.Lb:J _~_..__________J2..GIEQJ:~_~U~G.l:L:."_.~ ___ ~-_______Lb-OJ1____
EUI\1IV
OF
SCJ
D,b,i I
U-1'1-I-V-E_
~~ SIT Y "/1 ISS ISS I P P I
1
. _ - - - - _ . _ - - - - - - - - - -62
- -0 - - -
R 0 S S T 1\ U 1\1 I V 0 F
f.ti ISS ISS I P P I
UN
r V E ~~ SIT Y
(-"\ ISS ISS I P P I
___8_~~f'-_!________,____ ,__ ,.J~__!_9J..E::vr~ ~~ R F:..~.t~J:\F< C/-~_\J2. __ . _____£.<...1.~J~lL!Jl~ID~. 6'=. ,!J:_
CH I Ct,
SAADAT
:3 (") C
1 t~()_(L. __ _
ILL
(~()
360
____.b!t~:}'__YJlL~~c.;~IL.N y_o________.____
SI-\ fv1UFLS
CI
SAf\jDEFU F<
SAUNDERS
SAN
J~J=.._I!~t\y_L~ ~ t=: R
SAUTER
_;; C H /.'I~-,,[).:..:.F=-=._ _ _ _ _ _ _ _ _ _~[):......E
~ARINO
S.___~s.S r.:.t\ I-< ~...!i.. ____.~~.~:.._ H t\ F~T f
SA t'J
J
CALIF
1130
() RJL~.nr-1 01,'_ _ _ _ _ _ _ _ ~ 6 Q _____
(J SEC 1\ L
::,60
IF
lilll_________ ~____..,. _____._____.sA:A __ .!>LC,L5..Lc;__ ~..(J).,LLL_ _______~_l..f..i.Q.Q___~
K C L II ~ I< S () j',j
C U L L F:: (; f:'-:
S CHI F T N E R
S
S C H CJ D ITS C H
(~ _E.L~I~)JJ.;;_j\l\iJ~~L _ _ _ _ _ _ _ _ _
SCHORO\:.j
tvi
Cc.I',j F()f-~ F~ES r,;r::D Ue I .
EVA~"ST(JN
ILL
l~?
H L C ?_CJ:!.=t!j~,.~~··~ J\f-<"Q__.___________ ~!-\ YJ:L~\ gD C .0J- I F _ _ _ _ _ _ _ _ _.L~__
J
f:: A S T iv": {I, ,',1 K Cl D ,t\ K
1 03 S ~~ n C HE S T E ~;.: i\i Y
113 0
P E t\F-~ TH ur:< c; rviC K El::
SAr,,! F FO\ f'.IC I SC tJ C 1\ L I t=
1 1 30
-=-____,_
__________ .
F Ofl.r:RCJ CUf.;;·I/-\!'JDt::f-!
N()Pj.l/\~\! OKL;,-\
1f)(J()
S C H R AD E R
S CUD D E R
SC ULL Y
SE~UJUSS
I
S
J
FlbV
SHOEtJIAI 60
0'-"'_·
__________2____~2.~D~ __ ._______________ .___________S!.~~I_~~ ___ ~~_~.~~~_} C!2.._c;/\L. ~~ ______ ~?i:.!J___ ~
H-iJ:------····-·--'-·--··'-'---·-·--------'--0"~:.j"I-:ti~--~p L
S TEE L E L L I H 1\'1
-
0_0_.__.
PA
.•
.
d t t'W
i····t-\·ttf···'···t··#itf'&·h»'i%fHi6d-f1.)"¥f!'iffbifid*8¥ijij"·····irp·OO·p·,.·wIT WUry"f·r·w·-··'
.
.
l'i!iiAiRilNlljljp'ymflfWR
..
"""
•.
----_._-------_._-_._.•
--------------------
-.-~.- ... - '-" ,---
7
---
SDgl.!_._CH~!::;J~~.~ _____ .
T
WHALCON
_ _ _R_' F I 8;'·1
--------------
~-.------
---~-----.--.--~---.
________ D
STEIN.
STEINHART
.
PAGE
-_..__ ._-------_~~~~ERG
__
S TEPHEf~SDN
J
PIT '(~!2~}l~£:; __ ~L -=.I.:...F_ __
11_:;tD__ _
INTERNATIONAL
360
NEIl! YOf~K CITY No Y.
. ______________\"JI-u.I[L_E1-Al_~_~::..._'_'N...:"......;Y:......:::...e_ _ _ _ _____"36_s T f:,:.NFg~HYSICS II'.:Tt'::KI,!ATICH,IAL
SAN Lf:ANDf~O CALIF
360
R P \'-i H I ~~ l.~ p ®._~lr·! ~ _____;; 3 ~1 1 S -r J [i S L;J:>J::~I C H_ _ _ _ _ _ _-=-1.=8Jl.Q._
G WIR~
WHITE ~LAINS NnYo
H i·I,l t3~.. \ CQF~~ ______.__ .._. ______ S !LVEP SPf-( H~G r--·'i.:-..)_ _ __
1 1 0°_
V
V!ARREN
~Iif.:::
--
AF I _~t!_.. R ..I::~Jii? __<;'f.~ti /\ I) A_._•.':L~) -(.~....!:~.ti~J ~·:·0:!_~:.~_5J~.!.~?A,_._______1_L~i~ __.
L D 0 ~,; C H E (.j I C J\ L
PIT T SHU I-! (~ C /\ L I F
3 60
f::f,~OCl<
IA,
ISLAND
r=_()r_
__._ _ _ _ ~ FU I\! I
;.,1 I
Cf'h
o
59.
.,
."""-:-,
PL/1HAS
MORE CONCISE
•
EXPRESSION
.. PROBLEM I NIT I AL I Z[ ARRA Y A AND COMPUTE
o
FOR ALL I. J
FORTRAN
oI MENS I ON
.5
10
A
(i 00 • 100) 100 )
DO 5
1=1,100
DO 5
J :: 1 ~ 100
DO 5
K:1J100·
A (I, J ~ K); 0 .0
DO 10
1::: 1 100
DO 10
J:; 1 100
o
A (I,J~2)::B (IJJ)+ M (I}MS(6),J)
Del A 600,100,100);
A:.O;
A (~~~) 2): B+Ml*~ MS( 6)~
*);
o
o
PL/l HAS BETTER I NTERr~UPT CONTROL
PROBLEM
"
ALLOW FLOATING POINT UNDERFLOW FIRST
100 T I. [vIES THEN KILL JOB.
FORTRA.N REQUIRES ASSElvlBL YlANGUAGE ROUTINE.
ON UNDERFLOW
BEGIN~
DCL COUNT FIXED~3)
IF
PUT
COUNT
INIT(O)j
= 99
THEN D0)
LIST (1100 U~JDE~FLOWSI)
SiGNAL FINISH)
SKI
pel) .,
END)
COUNT:::: COUNT + 1 )
RETURN)
END)
6/.
-----,,-,,----------,,--""'---"'-,--,,-""""''',,''''''''''''''''
PL/l HA$ MORE EXTENSIVE DATA EDITING
SOURCE
TAR(;FT
00100
'l ~ 100
"'0203
1 2 3
1234.56
1.234.56
12
1
001 .23
$1 .23
-123
$1 .23CR
123
123
-123
123
"2
+"
-
~-
o
PL/l HAS SUPERIOR ARRAY HANDLING
EXAfviPLES
DECLARE A(-5~-25, 17~18);
DO 1=-5 TO -25 BY -5 )
- 25 TO - 30 BY - 1 )
WH I LE
( X~19. 6) )
A=B+C/Ej
o.
WHERE A)B)C)AND E ARE
ALL N-DIMENSIONAL ARRAYS
,...-
..----.-..... -.--------.. -----.-....
-~-.--.----.~.--.--_.
__._._._----
·c·
PL/l HAS ~RE BUILT IN FUNCTIONS
EXA~iIPLES
.',
TIME
RETURNS CHARACTER STRING YYMMDD
RETURNS CHARACTER STRING HHMMSSTTT
SUM(X)
PROO(X)
RETURNS SUM OF ALL ELEMENTS ·OF X
RETURNS PRODUCT·OF ALL ELEMENTS. OF X
. DATE
o
c
. PL/'
HAS CHARACTER.STRING AND BIT STRING PROCESSING
. EXA>iPLES
DCL SYi,iPTOMS BH( 64) P.ACKED)
HEADACHE
BIT(l) DEFINED SYivIPTO[vIS
POS I T I ON ( 25) )
FE VER BI T( 1) DEF I NED SYiviPTOiv1S
POS I T I or~ (35) ;
IF HEADACHEf1FEVER THEN GO TO ASPIRIN;
o
ELSE GO TO
PENECILLIN~
8IT(64) REQUIRES 8 BYTES OF STORAGE
X :'XYCOlviA8C!\!iON' j
Y-= I NDEX( X, 1 COrvi I);
Z ;::; I NDEX( )( J I r' 10 N~ ;
A·~ SUBSTR(X) Y ,3) II SU3sm(x, Z)3) j
6$.
·-.-......---.--~"-.-.-"~-~"."---.~----- --------..- ... ---"-..... ..... ~."..".--..............-........ ,,''' ..................... ,.
"
I
I
"
.,I
•
I
~
,.
~.
('.!~
,
.
o
IN ADDITION PL/l OFFERS
CQt-/iPILE TliliE CAPABILITY
(i"iACROS)
LIST PROCESSING
jviUL T I ~ TASK I NG
DYNAMIC ALLOCATION AND RELEASE OF
EXTENSIVE DEBUGGING
EXTENSIVE I/O CAPABILITY
STORAG~
.0
66.
fl.
I· hooker
DYNAMIC MODEL SIMULATOR FOR THE IBM
1. .
l.- (o.!
1130
by
George D. Polyzoides
Hooker Chemical Corporation, Niagara Falls, New York
The Dynamic Model Simulator is a chain of FORTRAN source programs
that permits the IBM 1130 user to investigate in depth the dynamic behavior
of physical systems that can be modeled into linear or nonlinear ordinary
differential equations.
It can be used for the investigation of a wide range of engineering
and mathematical problems from reaction kinetics and reactor responses to the
design of electrical networks and struct~ral assemblies.
The system does not require familiarity with analog computers although in a way it forces the IBM 1130 hardware to perform the function of
an analog computer while at the same time it offers the digital computer advantages of random memory access and ten digi t (extended) precision.
c
The input consists of Engli sh conrnand words (START, WAIT, RESET, etc.)
and numeric data which are rC0d in free tormal.
The system output includes tabulated values of the variables at
specified intervals, a table of the variables involved and detai led error and
information messages. Plotting of the variables is also available as an option.
System Source Language:
FORTRAN IV, Level E
System HardvJare Requi rements:
1131
CPU-8K, one 2315 disk cartridge
Card ;{eclder
1132 Printer
Ben son· L e h ne r 305 0 i 9 ita 1 Inc reme n tal P lot t e r
(op t i ana 1)
1l~4~
fa ?
DYNAMIC MODEL SIMULATOR
FOR
THE IBM 1130 SYSTEM
George D. Po1yzoides
Hooker Chemical Corporation
December ·1967
o
t
\
II hooker
NIAGARA
FALLS,
NEW
YORK
14302,
PHONE
(716)
285-6655
ABSTRAC!
The Dynamic Model Simulator (DYNAMO) consists of a series of
FORTRAN IV programs that allow the IBM 1130 user to investigate with
detail and accuracy the behavior of physical systems that can be modeled
into linear or nonlinear ordinary differential equations.
The
DYNAMO Simulator can be of great help to engineers and
scientists who consider such time consuming problems as the investigation
of the transient behavior of chemical reaction systems, electrical networks,
process variables and control systems.
c
The input to the DYNAMO Simulator ia in free format and in the
form of distinct key words and numbers.
The digital computer setup
prohibits real time operation but it offers the additional advantages of
accuracy, data storage and detailed plotting.
Besides the input-output
and calculation mainlines DYNAMD utilizes the 1130 Plotting System, a
plotting software package created by Hooker's Systems Engineering Group.
The minimum equ'ipment requirements for program execution arel
1131-cPU-SK-2B
1442 Card Reader and 1132 Printer
The Plots (optional) can be obtained through an on line digital
plotter (.OOS"/step).
o
II hooker
o
TABLE OF CONTENTS
Page
INTRODUCTION
PART ONE
GENERAL INFORMATION
A)
PROGRAM CHARACTERISTICS
5-7
B)
PROGRAMMITNG TECHNIQUES
8-13
PARTTW)
EXAMPLES OF APPLICATIONS
CASE ONE
A CONTINUOUS STIRRED REACTOR BATTERY
14
CASE
DESIGN OF AN RLC CIRCUIT
15a-
T~
APPENDIX A DYNAMO INPl1r LANGUAGE
Q- ~
p
17
APPENDIX B PLOTTING INSTRUCTIONS
21 a..-b
APPENDIX C FORTRAN-ALGEBRAIC EQUIVALENCE OF THE MODEL
22- 24
APPENDIX D CONTENTS OF THE STANDARDS FILE
24 -
0
o
II hooker
3
INTRODUCTION
The expansion of the range of the applicati.ons of digital
computers into fields that were the traditional domain of analog
computers is a rather recent trend that has produced such interesting
results as the GPSS, the CONSIM, the CSMP, the PACTOLUS and other
simulation systems.
The DYNAMO Simulator does not claim "a place in
the sun" among these systems.
and less sophisticated.
Its mode of operations is more restricted
The object of DYNAMO is the solution of ordinary
differential equations and the emphasis is on the mathematics rather than
the block diagrams.
Since differential equations and block diagrams are
frequently equivalent in describing a system, it follows that in many
o
instances DYNAMO can be used in obtaining results similar to the ones
obtained from larger and more complicated systems.
The point
ot
decision when programming for analog to digital
equivalences is whether analog procedures and nomenclatures should be
carried over to the digital system's specifications.
For example, should
gain be labeled as such, or should it be implied by a multiplication?
One
can argue about the advantages and the nuisances of both types of approach.
The DYNAMO Simulator approaches the problem from the point of
view of the person who has more experience in digital than in analog
computation.
This implies that analog computer nomenclature, wiring
o
71.
- - - - - - - - - - - - - - - - , - - - -_ _ _ _ _ _ _ _ _,_ _w.......... •
II hooker
4
diagrams, and scaling are not necessary to deaeribe and set up the
problem.
The analog concepts, although still
_i ..ent
afid helpful,
tend to fall in the background ~ile ~amii!arlty withdiglta1 concepts
becomes essential.
o
gm
t"iH8U'i't··'#·¥ss·_···· T·..P--····-y··
II hooker
s
PART ONE
o
PROGRAM CHARACTERISTICS
A)
INPUT REQUIREMENTS
The input to the DYNAMO Simulator consists of some key words
and numeric data.
The programs have been
~itten
in such a way as to
permit experimentation with the differential equations that describe the
model.
The execution of the programs can be interrupted at five distinct
points and restarted at some later time without loss of continuity.
1.
Input Language
The key input words or commands constitute a very elementary input
language •. The term "language" is applicable only as far as it is understood as the substitution of a numeric code with English words.
The
input commands can be divided into six categories:
o
i)
Integration Information commands:
They define the step length
and interval of the integration, value of the equation coefficients, etc.)
CONTROL, COEF
ii)
iii)
iv)
v)
vi)
Identification commands:
TITLE, TABLE
Initial conditions commands: START, RESET, REPEAT, RESTART
Utility commands.t SETUP, LOG, NOLOG
End Indicator commands: WAIT, END, HALT
Output commands: PLOT (8ee Appendix B), LIST, FILE
For those interested in more details the DYNAMO input language is presented
in more length in Appendix A.
c·
...........•-.--.---..- .....-.~~----..
_._-_._ .............................._.............._......•--.. ..
'~
-~
.....
~-~-----------------
II hooker
6
2.• ' Input Procedures
The DYNAMO commands and the required numeric data can be entered through
punched cards or the 1131 console.
during program execution.
o
The two input modes can be alternated
All commands are listed on. the 1131 console,
but listing can be stopped upon use of the NOLOG command.
The SETUP command
can be used to reset the system standard files (see Appendix D).
All commands and data are entered on a free format basis.
special plotting instructions are only sequence dependent.
The
More will be
said about free format in subsequent sections.
Upon detection of F- type input errors in the numeric data or an
erroneous conmand the user has the option of causing a "PAUSE" and correcting
the error.
MOre serious logic errors (i.e. no initial conditions specification)
cause exits to the monitor.
The files containing the standards for the system
are not closed and most of the times the error can be corrected and the run
can be restarted without loss of continuity.
3.
o
The Mathematical Model
All the first order ordinary differential equations describing the model
must be previously stored as a function subprogram.
The present dimensioning
allocations allow a maximum of twenty variables (equations).
These ordinary differential equations may be linear or nonlinear.
Second or higher order equations can be expressed as two or more equivalent
first order equations.
The coefficients of the equation terms can be
()
II hoolier
o
7
variable and their values must be present during the input step.
The DYNAMO Simulator will accept any first order differential
which can be arranged in the formr*
- iwhere Yi is the i~h dependent variable and x is the independent variable.
EXAMPLE:
Algebraie expression<
~'I:
'+- •
C~ Y2 = S IN X'h
5 Yi
DYNAMO FORTRAN:
The value of L is determined in the calling mainline.
()
The
function subprogram must contain a computed GO TO statement so that the
appropriate derivative value for each variable is selected.
Appendix C
contains sample subprograms in both algebraic and FORTRAN notations.
* Other equation forms can be expressed in the form of equation -1after some mathematical manipulation.
o
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _''''''''''''''''_ _ _ _ -c
II hooker
B)
8
o
PROGRAMMING TECHNIQUES
The present version of the program consists of two input mainlines,
an error checking mainline, the mainline that solves the differential equations
and the mainline that generates the plotting specifications.
plotting is accomplished through a separate software package.
The actual
Due to the
number of CALL LINK'$ that is involved the majority of the programs have been
stored in core image to allow for fast loading.
1.
System Standards
A seventy (70) digit integer vector is initialized and then continuously
updated and maintained through each execution of the DYNAMO programs.
Appendix D contains a list of the information contained in the vector.
The
standards can be reset by using a special "standards generation" program or
by using the SETUP command during the program execution.
2.
The predictor-Corrector method for numerical integration.
The solution of the differential equations describing the model is the
most vital part of DYNAMO.
The Predictor-Corrector method of Hamming(l)
has been found to fulfill the important considerations of stability,
accuracy and relative calculation speed.
No attempt will be made here to
d~scuss
the method in detail.
Any such effort would be a reproduction of the excellent article by
Ralston(2) descrfbing the details.
The limited information to be presented
o
76.
"u,a"
:
-
II hOOker
9
o
in this paper is here just for clarifying the overall approach.
Given the values of M dependent variables (Y) and an independent
variable (X) at four equispaced points in the X domain (0, 1, 2, 3) the
predictor corrector method (fifth order accuracy) is a means for obtaining
the M values of (Y) at point (4).
This pr.ocedure can be generalized by
setting the four known points of the domain as n-3, n-2, n-l, and n and by
considering the points to be obtained as the x domain value becomes Xntl.
A
set of four equations is used for each ith variable (i=l,M) and this
-t
procedure will repeated until xn+l reaches a present value:
i)
ii)
The predictor equation gives a rough estimate of Y(i,n+l)
r" •. ::
Y(L
J
n.
~) .. ( 4~/:3") ~ ( 2 ti J( ~.)"tt) _ ~ '( ~) n _I) .. 2 ~ ~
I (
) "-
2))
- 2-
The modifier and corrector equations eliminate the need for iterative
convergencez
o
- 4-
where~t\-('Y\) is the trunc'a.tion error from the last step, ~' is ~~e. den".~.~
c.\'I"4
"I:' ~he
Inh1'\lr.\
~,"""~~n
'IC."'J~,,-,,1"_1.,"ft_l _.... 4
c.f d
Yo" ••
III) The final equation takes the form:
Y( i.)
(lTC)
-= Cn~'
-4-
...9..
(-p C','
12.1 V,,+I- ",""J
wher~" ...-(1\+:)iB the present truncation step.
From the point of view of calculations the only additional
requirement is that the values for
from the initial conditions
•
y( ~-,1)
y(
t
,K), where (K=2,4) must be calculated
before equations (2-5) can be executed.
This initial step, conunonly called the STARTER consists of a first order
o
Newtonian extrapolation and an iterative convergence technique described in (2).
77
_ _ _ _ _ _ _ _ _ _ _ _,_ _ _ _ _ _ _ _ _ _ _ _ _ .~ _ _ "~~O"···
II hooker
10
Stability checks have been incorporated in the integration routines.
The user can check the stability of the calculations by using DATA SWiTCH 4
o
during the execution of the program.
3.
Data
co~version
from extended to standard precision.
In order to minimize the accumulation of the inherent roundoff errors, the
DYNAMO mainlines (with the exception of the plot generation mainline) have
been compiled in extended precision.
of standard precision mainlines.
The 1130 Plotting System is
compos~d
The DYNAMO data must be converted into
standard precision before they are stored in the files from where the plotting
mainlines are to pick them up.
Conversion is accomplished by equivalencing the extended precision
floating point values of X and Y(i,k) to a three digit vector 3 and by creating
the new integers IY(20,2) and rx(20).
The technique can be outlined as
fo11owsr
i)
Equivalence,
o
--
S
, -.~I\A. NTI5~A
J(2)
j(~)
ii)
3
2
/~> 'NV.l NI V·.lN3\tJdlflJ -=IJ
I
f\[]I.l:J~ ~
I
II hooker
15
CASE No. 2
-0
Design of an RLC circuit (5)
The investigation of the transient behavior of electrical networks
is another potential area of applications.
The signle loop RLC circuit pictured
below is a textbook example of differential equation solving.
The diff.
equations describing the transient characteristics of the circuit arel
Currents
L £ (~') +
d+ z
R.dt ... .!. L =0
d+
C
Voltage (across the capacitor):
wherel
JV!Jt = t~
R
i = current amp
L = inductance
henry
C = capacitance
farad
R = resistance
ohm
1----------.
L
..
c
ClOh
E
= potential
v= voltage
--
volts
'~t 5~1.'" .t
t~
to
tt
.. ."
The behavior of such a circuit will be investigated for three casess
a)
over damping
b)
critical damping
c)
underdamping.
This will be
accomplished by using the COEF command to set appropriate values for R, L,
and C.
The REPEAT command will cause continuous execution with the original
initial conditions.
The next pages show the input requirements and the
resulting output (the plotting instructions have been omitted for the sake
of brevity).
o
TITLE
RLC CIRCUIT CALCULATIONS-OVEkUAMPED
TABL.E
1
1
1
1
1
1
1
1
1
1
1
1
1
1
TH~
,CAS~
INITIAL CONDITIONS FUR THE RUN ARE
A)
B)
I(O)=U
V(U)=Q
(D(l)/OT).O = (E-VCO)/l = 4
E) THE CIRCUIT POTtNTIAL IS 4 VOLTS
VARIABLE COEFFICIENTS
NUMBER 1
RE~lSTANCt
NUMBER 2
CAPACITANCE
NUMBER 3
INDUCTANCE
C)
OUTPUT
TIME IN SECONDS - FIRST COLUMN
CURRtNT IN CIRCUIT - THIkD CULUMN
VOLTAGE A'ROSS CAPACITOR - FUURTH
COEF
1
3.0
1.0'
1.U
2
3
Ci
CONTROL
6
1
2
,
COLUM~
3
10.
.001
3
6
4
1
0
0
START
o.
o.
o.
'4.
, LIST 1000
WAIT
TITL.E
RLC CIRCUIT CALCULATIONS-CRITICALLY DAMPED CASE
REPeAT 1
TABL.E
1
SAME INITIAL CONDITIONS AS PREVIOUS RUN
CONTROL.
'2'
1
"S.O'
COEF
1
o
2.0
LIST 1000
WAIT
TITL.E
RL.C CIRCUIT CALCULATIONS-UNDERDAMPED CASE
.. T
"''''-REPEAT
CONTROL.
1
1
TABLt::
SAME INITIAL CONDITIONS AS
COEF
1
LIST
B~FORE
1'5 b
1.0
lOU
END
o
7'0.
cW"W ""
=
""
RLC CIRCUIT CALCULATIONS-OVERDAMPED CASE
DYNAMO
INFORMATION TABLE
15 C
THE INITIAL CONDITIONS FOR THE RUN ARE
A)
1(0)=0
B)
V(O)=O
E) THE CIRCUIT POTENTIAL IS 4 VOLTS
VARIABLE COtFFICIENTS
NUMBf:.R 1
RESISTANCE
NUMBER 2
CAPACITANCt:.
NUMBER 3·
INDUCTANCE
OUTPUT
TIME IN SECONDS - FIRST COLUMN
CURRtNT IN CIRCUIT - THIRD COLUMN
VOLTAGE
ACRO~S
CAPACITOR - FOURTH COLUMN
o
9/,
RLC CIRCUIT CALCULATIONS-OVEROAMPED CASE
2
15 D
o
LIST OF ENTEREO COEFFICIENTS
(OEF. NO.
1
THE VALUE IS
0.30000£ U1
COfF. NO.
2
THE VALUE IS
O.'lOUOOE '01
COt:.F. NU.
3
THE VALUE IS
0.10000t:. 01
o
-
---------------------------------------
--~~
3
RLC CIRCUIT
CALCULATIONS-OVERDA~PED
C)I RST COLUMN- I NDEPENUt::NT VAR 1 ABLE (X) •
0
CASE
O.OOOOOE 00
-0.12472E 00
0.10904E 01
0.65341E 00
0.19999E 01
-0.29337E 00
0.82378E 00
O.1~220E
3000
0.29999E 01
-0.21542E 00
0.56805E 00
0.25112E 01
4000
0.39999E. 01
-0.14813E O()
0.38813E 00
U.29837E 01
5000
0.49999E 01
-0.101181:. 00
0.26493E 00
0.33063E 01
6000
0.59999E 01
-0.690691:.-01
0.180821:. 00
0.35266£ 01
1000
O.69999t:. 01
-0.47141E-01
0.12341E 00
0.36769E 01
8000
0.79999E 01
-0.32174E-01
0.84235[-01
O.37794t. 01
%GO
O.S9999E 01
-0.21960E-01
0.57492£-01
0.38495E 01
O.OUOJUE:. OU
U.40UOUE 01
100-0
0.99999E 00
2000
,
~
DEPENDENT VARIABLES(Y) FOLLOW SIX PER ROW'
au
0
..
15 E
O.OOOOUE
01
o
9~.
_.... _- --'---'--.- ..
__._-_.
- - - - _.. __.__._......................_....__.
~---
5
,-.--...-._-......._,.- .. ----..--... --'_. -- ---Rt:€-€tREtt 1 f (;ALCULA T f ON S"C R 1 TIC AL. L. y DAMPED· CAS·t:· -.- '.'-' .. -.- ....---. _.". --- ..-.- ..
I
15 F
&VNAMO INFORMATI·t}N TABLE
._., .... - -- ... _.....--- ..--... -----SAME--. ffll-f IAt..CONO 1 T IONS AS PRt VI OUS' RUN
o
o
6
RLC CIRCUIT CALCULATIONS-CRITICALLY DAMPED CASE
()
LIST OF
cNT~RtD
COEfFICI~NTS
NU.
1
THE VALUt:. IS
0.2UUOOE U1
COl:.F. NO.
2
THt:: VALUE IS
O.10UOOE 01
COEF. NO.
3
THE
VALUE IS
O.10UOUE U1
CO~F.
15 G
9S.
---------_._ ...•......
RLC CIRCUIT
_....._............. ...............-_.. _......._-_._................. ... .
"
"'
CALCULATIONS-C~ITICALLY
FIRST COLUMN- INDE"'ENDi:.NT VARIABLE(X).
UAMPEU CASE
7
15 H
DEPENDENT VA~IABLt.S(Y) FULLUW SIX PER
0
U.OOuUOE Ou
u.4UUOuE 01
O.OUOOUe: OU
O.OuQUOE 00
1000
O.99'i':1':1E 00
0.2381':1E-06
U.1471!>E 01
0.10569t:. 01
2000
0.1':19'i':1E:.. 01
-0.!>4134E 00
O.10826E:.. 01
0.23759t:. 01
3000
0.29':1991:: 01
-0.3982':1E 00
O.59744E 00
0.32034£ 01
4000
0.39999E 01
-0.21978E 00
0.2930?E 00
O.36336E 01
5000
0.49999£ 01
-U.IU780E 00
0.13475E 00
0.3ts3831::. 01
6000
O.59':1':1':1E 01
-0.4'1575E-01
0.59490E-01
0.3.244191: 01
1600
0.15999E 01
-0.689U5E 00
O.20398E 01
0.~6491E
01
01
99
RLC CIRCUIT
FIRST
COLUM~-
IND~PENOENT
CALCULATIONS-UNDERDAMP~D
VARIABLE(X).
CASE
DEPENDENT VAkIABL£S(Y) FOLLOW SIX PER ROW
10200
0.10199E 02
-0.28095E-01
0.15698E-01
0.40124£ 01
10300
0.10299E 02
-0.26781E-01
0.12952E-01
0.40138E. 01
10400
0.10399£ 02
-0.25338E-Ol
O.10345E-Ol
0.4CJ150t: 01
10500
0.10499£ 02
-U.23791E-01
0.78887E-02
0.4U159E 01
10600
0.10599E. 02
-0.22165E-Ol
0.55902E-02
0.40166E 01
10700
0.10699E 02
-0.20484E-Ol
0.34573E-02
0.40170~
10800
0.10799E 02
-0.18767E-01
0.14946E-02
0.4U173E 01
10900
0.10899E 02
-0.17035E-Ol
-0.29559E-03
0.40173E 01,
(Jo
01
0
o
j(}{).
------~.-.--.-
...
---.--.~---
o
Do
.
o
~
+~ER
Ct-EMICAL cr:R=mA TICN+
cr:R=mATE ENGINEERING
.
5
5
~",..".._ .... - - - t. ..... __ ~ ....
!
4
RLC CIRQJIT ANAL '(515
UNOERDAMPED CASE
//
CURRENT XXXX
/
vo rAGE ••••
,
. . . ... !.._,
•----_
t.. ____
4l
c___"
L ____ '- ____
L __ - - l. __ - - t.. ___ -
/
I
I
/
I
3
. "'x
2
3
I
/
es
/
I
.f-
~
2
.~
JH
f en
!
fo
l~
1<
·z
o>-
......... ~ ...
I
.4
~
l~
...... ~ ...
/
d
. 11
~ ......... ~ ...
I
~
I-
- ...
/
I-
8
l-
1',/",...
1
B
I " "\.,
/ "/
iI,~!', (,'
~
"
Jt'''-,},
f'
\
\
\,
"",~
1
'"
""0
Dr~----------~~-----------"'.
~
..--~"
, , , , , , , , \' , , , , , , , , !ii' , , , , , , , , 5' , , ,
,"-;-1<--....-"--"--
.4
'-....'-.....
I , , , , ,4 , , ,
ELAPSED TIME IN SECONJS
I I I I \;1 , I I I , , I I
6' , ,
I I I I I'
'7 I
iii
.JL...--lL-L-L--lI..--Jl.
I' , ,
,0'
I I
a'
I I I , I , , ,
9'
I I I I I I
SY!iTDS 0GJNEER1Ni rR1.P
tr4
I
Ilhooker
16
I would like to
Wlderstandlng and
paper.
th~nk
encourage~nt
my $upervisor Marliss'O. Bird fot; his
throughout the preparation of thil:J
The suggestions of G. A. R.
considera~le improve~ent
of the
~rol1ope
D~AMO
have resulted in a'
c,pabi1ities.
o
o
;(.'l..
II hooker
17
c
APPENDIX A
DYNAMO INPUT LANGUAGE
#
o
o
II hooker
18
The following is a list of the DYNAMO input commands and the
o
functions they initiate during the program execution:
A)
INTEGRATION and STORAGE controll
CONTROL
M
This command causes the program to accept M data cards.
These cards
may be used to define: a) The number of equations describing the systam,
b) The integration increment (delta),
d)
c) The total integration domain,
The initial conditions files to be used,
e)
storing the integration data on the disk, and f)
The frequency for
The di.sk record where
the first X-Y data are to be stored.
COEF
This command permits the entering of the values for the variable coefficients
of the differential equations.
B)
o
INITIAL CONDITIONS control:
START
This command permits the entering of the initial conditions for the first
step of the calculations.
REPEAT
M
The program will return M-1 steps back and it will begin calculations with
the initial conditions of this step.
RESTART
M
Depending on the value of number
M
an interrupted run can be restarted at
different points.
RESET
The last X-Y record of the previous run is used as the initial conditions
of this run.
Other integration variables can be Changed through the CONTROL
c01lDDB.nd.
-
---~.--
.....- - - -... -.- ... -...
.........................
----~-
...
- - _._----- ..
__._-_..
-
Q
II hooker
19
C)
IDENTIFICATION controll
TITLE
The contents of the card following the TrTLE card will be printed as a
heading to all output pages.
TABLE
A maximum of twenty cards with information pertinent to the input data or the
model can be read and stored with this command.
This information is latar
printed on the 1132 Printer.
D)
OUTPUT control:
LIST
M
The calculated results from every Mth integration step will appear on the
1132 Printer.
c
PLOT
M
This command specifies that M sets of plotting instructions follow.
FILE
OP
= 'Y\
START
=m
END
=
k
This command indicates some operation on the data files, depending on the
value of the OP code. (storing, listing, punching).
M and K indicate the
starting and ending records for listing or punching only.
When OP is
negative no calculations need preceed the file Operation.
The number of
variables to be dumped or listed is controlled by the CONTROL command.
E)
utILITY commands I
SETUP:
This command resets the standards files during program execution.
LOG, NOLOG
These commands control the listing of the input data on the 1131 console.
o
Ic.:j
-~---------"----------,--.-~--~------
II hooker
20
F)
END INDICATORS
WAIT
Stop input and proceed with execution.
"Return for more data.
HALT
Stop input and pause.
~en
the START button is pressed resume execution of
the input.
END
Stop input and proceed with execution.
Return to monitor.
o
II hooker
21
APPENDIX B
PLOTTING INSTRUCTIONS
c
THE CARDS FOLLOWING THE
II
PLOT
M
t.
COMMAND
CONSTITUTE A SIMPLIFIED APPROACH TO DIGITAL PLOTTING.
A SET OF PLOTTING
SPECIFICATIONS.SIMILAR TO THE ONES LISTED BELOW MUST BE PRESENT
FOR EVERY ONE OF THE M PLUTS SPECIFIED.
PLOT
2
*** DYNAMO COMMAND****
NUMBER = l TITLE' PLOT NO.· l '
X AXIS
TYPE = 12 TITL£' INCHeS OF WATER' MIN • .001
SHIFT = 0
VARIABLES
FILE NO. =1
*
MAX =
.2
C
Y LEFT AXIS TYPE=12 TITLl'VOLUMl OF WATER IN LITERS'
VARIABLES
FILE NO. = 2 LINE = 1000 GRAPH = 4122 C.OoE= 4251
TAG • DIAMETER 2 FEET'
FILE NO. =3 LINE = 2511 GRAPH = 4122 CODE = 4281
TAG' DIAMETER 3 FEET'
*
**
** INSTRUCTION TO
GENERAL
PARALLEL X
FRAME
STOP
NUMBER=2 TITLE • SECOND PLOT'
•
•
• •
• •
• •
•
• • • •
• • • •
• •
• •
• • •
• • •
• • •
• • •
• • •
•
•
STOP
• • • •
• • • •
•• OTHER
• • • •
• • • •
PERMIT ENTERING
GENERAL SPECS**
• •
• •
•
•
• •
DYNAMO COMMANDS
**
• •
• •
THE PLOTTING INSTUCTIONS ARE USED TO DEFINE
TH~
PLOT TITLE.THE
TYPt OF THE AXES( LINEAR,LOGARITHMIC,POLAR ).THE SIZE OF THE AXES( S.S OR
/O~
11 INCHES) AND THEIR MAXIMUM OR MINIMUM LIMITS IF IT IS 50 DESIRED.
~
INPUT FOR DIFFcRtNT AXt5 IS SEPARATED BY . (*) AND THE END OF ALL INPUT
FOR THE AXIS CHARACTERISTICS IS INDICATE~ BY C**) •
THE FILE NO. IS THE
S£UUENCt NUMBER OF THE VARIABLE IN THE FILE
OBSERVATIONS. THE FIRST VARIABLE IS· TO BE THE INDtPENOtNT VARIABLt X
AND VARIABLES 2 AND 3 ARE TO BE PLOTTED ON THE LEFT HAND Y AXIS.
ENTRY IDENTIFIES THt TYPE OF ~INE( CONTINUOUS OR DOTTED)
THE** L.INE**
AND THE **GRAPH**
ENTRY IDENTIFIES THE TYPE OF PLOTt POINTS.LEAST S~UARES.FILL IN LINES. E.T.C.).
THE **CODE** ENTRY IDENTIFIES THE CODE FOR THE TAG THAT FOLLOWS IN THE NEXT
CARD.
THt GENERAL SPECIFICATIONS FOLLOW WITH THt **STOP** CARD INDICATING
THE END OF DATA INPUT FOR THE FIRST PLOT.
C:
iii
ItI
I'
·1
Ilhooker
22
o
APPENDIX C
FORTRAN-ALGEBRAIC EQUIVALENCE
OF MA'rnEMATICAL MODEL
o
I/O
II hooker
23
0
,
1
'!oll
The two
systems
of differential equations .that were presented in
PART TWO have to be converted
DYNAMO
Simulator. The FORTRAN
i8 shown
below
CASE NO. 1
into function subprograms for use with the
format
of the differential equation
for both of the cases il1ustated on
model
pp. 14- 16.
The CSTR battery
This is an example of straight forward application. The derivatives of
of the concentrations of the reacting component in each of the tanks with
respect to time are identified through subscript J and a computed· GO TO
s tat e.men t :
o
I I DUP
*DE~£T~
F
I I FOR
*EXTENDED
P~~CISION
*ONE WORD
INTEG[RS
*
A~L
liST
FUNCTION F(J.L)
COMMON JCARD(80) ,LC( 70) ,Arvl,L>ELX,COEF(40) ,INl.)IC(4U)
COMMON DIS T ( 6 ,10 ) , X ( 5 ) ,y ( 20,5 ) , I X ( 2 ) , I Y ( 20,2 )
GO TO' 1 9 2 ) ,J
1 F= 1.
- Y(1.L) *(1.+ .5*Y(1,L»
GO TO 3
2 F= Y(lt~)- Y(2,L)*(1. + .5*Y(2,L)
3 RETURN
END
/ /
OUp·
*STORE
WS
UA
F
III
-~-
-----------
II hooker
23a
,"
C
"c
CASE NO. 2
The RLC ci rcui t
This is a case where the higher order derivative must be reduced to a
first or-der derivative. The reduction method can be outlined as follows:
1. Set Yl equaJ to the fi rst order derivative
"2. The derivative of Yl is the second derivative and it can be expressed
in terms of the differential equation.
3. The second equation, describing variable Y2 is the first order derivative,
that is, d/dt(Y2) ::: Yl
The same type of approach will reduce hi.gher oder equations.
I I DUP
*DELETE
I I FOR
*ONE WORD INTEGERS
*
F
LIST ALL
*EXTENDED PRECISION
FUN CT I 01'.1 F ( J , L )
COMMON JCARD(80) ,LC(70),AM,DELX,COEF(40),lNDIC(40)
COMMON DIST{6,lO),X(5),Y(20,5),IX(2),lY(20,2)
GO TO (1,2,3},J
C ••••••••• FIRST EUUATION IS THE SECOND DERIVATlvt
1 A=(1./COEF(3>)* «1./COEF(2)}*Y(2,L)+COEF(1)*Y(1,L)
F= -A
GO TO 4
2 F= YC1,L)
GO TO 4
3 F=(1./COEF(2»*Y(2,L)
4 RETURN
END
I I DUP
*STORE
WS
UA
F
liZ
II hOOker
24
c
APPENDIX D
LIST OF SYSTEM'S OPERATING STANDARDS
DYNAMO STANDARDS FILES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
GENE~ATOR
NUMBER OF E~UATIONS
0
PROGRAM RETURN CODE
2
FILE INITIALIZATION INDICATO~
a
FILE NUMBER FOR THE FIRST INITIAL COND '0
INTtRRUPTION POINT INDICATOR
0
RUN TITLE INDICATOR SWITCH
0
PLOTTING INDICATO~ SWITCH
a
INITIAL CONDITION INDICATOR
0
LISTING OPTION SWITCH
o
LISTING AND/OR PUNCHING OF THt RESULTS o
LISTING FRE~U~NCY FOR 1132 PRINTER
100
NUMBER OF PLOTS RtUUESTED
o
CURRENT INITIAL CONDIT. FILE RECORD
o
CONTROL COMMAND INDICATOR
o
LISTING OF THE TABLE OF VARIABLES
o
UNEV~N FILl ALLOCATION INDICATOR
o
NEXT INITIAL CONDITIONS FILE RtCO~D
o
INT~RRUPTED RUN INDICATOR
o
NO OF INCREMENTS OF THE NUMERICAL INTE o
LAST CASE REGARDLtSS OF ALL INDICATORS o
DISK USAGE INDICATOR
o
STORE FREutNCY( RESULTS TO FILE)
o
RESERVED FOR SY~TEM USE
o
RESERVED FOR SYSTEM USE
o
RESERVED FOR SYSTEM USE
o
LOW BOUND/INITIAL CONDITIONS FILES
o
MAXIMUM RETURN FOR REPEAT COMMAND
o
START PUNCHING FROM FILES AT REC.
o
STOP PUNCHING FROM FILES AT REC.
o
INDICATOR SWITCH
PLOT NO. 1
o
PLOT NO. 2
o
PLOT NO. 3
o
PLOT NO. 4
o
PLOT NO. 5
o
PLOT NO. 6
o
PLOT NO. 7
o
PLOT NO. 8
o
PLOT NO. 9
o
PLOT NO. 10
o
PLOT NO. 11
o
1132 PRINTER LINeS/LINE OF OUTPUT
PRINTER SKIP CONTROL
FILES STORAGE FREW. CONTROL
INTEGRATION INTERVAL COUNTER
PRINTER LIST CONTROL
NO. OF NEXT ENTRY REC. IN DATA FILE
52
53
54
55
a
o
o
o
o
o
o
o
49
50
51
o
OUTPUT PAGE NUMBER
TYPEWRITER LOG CONTROL
LISTING FROM FILES STARTS AT REC.
LISTING FROM FILES STOPS AT" HEC.
RETURN FROM PLOTTING INDICATOR
--- ..-~-~-------------"--.--.' --
o
o
1
c
o
o
o
/ I-Y-
~--------~-------------------------
------------------------------------------------"""""
II hooker
o
2S
IBM Publications
Manual No.
o
o
H20-0282-0
1130 Continuous Syatem Modeling Program-Program reference manual
C26-37S0-
IBM 1130 Disk Monitor System-Reference Manual
C26-S933-3
IBM 1130 FORTRAN Language
B20-0001-0
General Purpose Systems Simulator III
II hooker
26
o
BIBLIOGRAPHY
(1)
Hamming R.
w.
'Stable Predictor-Corrector Methods for ordinary
Differential Equations t
J. Assoc. Comp. Mach volume 6, 1959, pp. 37-47
(2)
R«lston Anthony 'Numerical integration methods for the solution of
ordinary differential equations'
Mathern. methods for Digital Computers, chapt. 8. pp 95-109
John Wiley & Sons, Inc.
(3)
Milne W. E. and 'Fifth-order methods for the numerical solution of ordinary
Reynolds R.R.
Differential equations
J. Assoc. Compo mach. volume 9, 1962, pp. 64-70
(4)
( 5)
Wal as S. M.
"Reaction Kinetics for Chemical Engineers"
McGraw-Hill New York 1959
o
Miller K.S and
Walsch JaB
" I n t rod u c tor y e 1e ct i cal c i r c i u t s"
John Wiley and Sons, Inc.
(6)
pp. 93-97
New York
1960
Franks R.G.E. "Mathematical modeling in Chemical Engineeeing"
John Wiley and Sons, Inc.
New York
1967
o
/10
~~---------.--.-.-
....... _-
mpDf·'?,W""?
m
.. rWWUTSWRIWUfT' -----YWiSfliSiiHiWi&ir¥bbiHiS"#itit"it¥t"rlridftfijSbi"wWftW··"i"WitbfP"i#&jt'· if" L ·j ... ·.. tr
.
/1,
c
l.-
Robert L. Cushman
ARCON Corporation
Wakefield, Massachusetts
THREE-BAR AND FOUR-BAR LINKAGE SYSTEM WITH
PLOTTED OUTPUT
A mathematical model of a linkage system is controlled by an array of
In addition, the
system simulates turning the entire mechanism about a major axis.
The
output of the system has yielded some very interesting designs.
o
In conjunction with this report, a list-oriented free field floating point
input program has been developed.
This routine enables the control program
to modify only specified iteITls of the system input.
Under break character
control, an autoITlatic ite rative procedure can be set up.
Some sample outputs are shov:. n in the enclosed photo .
.l
~
i·
\:'7
I
J
(.;
.-~---~--
-~---.~-Ar", "
...-~~-
~
,·,r
(/r
r,',
.
r? j'
"I.;.
Abstract for December Common Meeting
inputs to plot out the path of the linkage intersection point.
. ft-
It!!!"e M
' ,rlW''''.I·
L
o
COMMON
San Francisco, California
PROJECT:
Management Installation Division, Operation Project
SUBJECT:
The Systems Reference Library
SPEAKER:
Mr. G. W. Goesch, Manager, Product Publications
IBM Corporation, San Jose, California
Telephone (408) 227-7100
FOR
PRESENTATION:
o
Monday, December 11, 10:30 AM, Session M. 2. 6
8 Pages Text
o
//~
The System Reference Library
o
I'm Gordon Goesch, Product Publications Manager, IBM, San Jose. There
is a Product Publications group at each of the Development Laboratories,
as well as a Programming Publications group at most of them. And, of
course, there is a department in White Plains that prepares Application
descriptions. These groups all provide input to the System Reference
Library, and that, rather than my own department, is the subject of my
remarks.
A manual looks quite simple, and sometime sit's hard for our own field
people and for customer people to understand why it seems so difficult
to get a few manuals written, and then to get them delivered while they
are still fresh. There is a great number of manuals that must be kept
current and kept in stock, and record-keeping alone is a sizeable task.
The System Reference Library is the current method of indexing and classifying system manuals, and is the latest of a series of different methods,
each designed to solve the problems that the previous methods couldn't
handle. Problems created by new and very complex systems. This
method isn't perfect either, as you are well aware, but we are certainly
constantly trying to improve it. (And I have one improvement to announce
today, a little later. )
c
That is the reason I am always happy to talk and listen to groups such as
yours about our publications. It gives us an opportunity to discuss with
you our Publication Library; its organization, its purpose, and its distribution system - because even with a good system you must understand
how to use it if it is to be effective fo r you.
I think that this type of a meeting can be a two-way street for information:
We explain the principles; you provide feedback. We do get feedback from
you via Reader's Comment Forms - we want more of your comments and we
certainly appreciate them.
I don't know how knowledgeable you are about our publications; therefore,
for the benefit of the new members and also for the updating of the veteran
members, I'll run through a few slides which I think will tell you the
publication story.
By the way, you have probably seen the publications display in the main
convention lobby. The display, I am sure, contains publications of interest
to you. Many of the publications on dis play have been published since your
last COMMON meeting. Feel free to examine them in detail but please do
not take them away, as there is only one copy of each and we want everyone
to benefit from the display.
o
1
//;;.a..~::"~"~~:.:1t'~~~~~:'~~~·~.~.........~.acr.~~_~b~:i!ir~".s:"!'L"'~"r:_«~..."'I'J1:?~,£",.f~~·:...~~~~.sr~lf
IIJ}':s;Z/tJ()lJ
x
/1
x'
x
1/
/1
x
~ / fi 1,! ):' i/'.>} IJ J~
i I J ! \/ l-..ll! v' ( /
-/j
x
/1
/)!J
DD
}
IISrp;L
- -
"-'--"-~-'----"'--------"'---------"-"~------'-
x
o
WAD MODUU: . FORMAT
SYS 1.
NUCLEUS
SYS 1.
SVCLIB
. . ,. . I
PDS or
SYS 1.
UIIo1CLm
A('('.!'.
Mf"lhodl.
D6 '380
Trammmt
...."IP
NUCLEUS
Supervisor
006'380
SYS 1.
U~KUB
SYS I.
SORTLIB
SYS I.
COBUB
SYS I.
FORTLIB
SYSI.
PLiLIB
COBOL
FORTRAN
Mudulr5 .
Library
ModulE'S
PL I
Ltbrary
Us.r
LIbrary
MCJdulf'''
Moch......
Module!!
Any Va.er PDS
SYS I.
MACUB
I
IBM-
I U•• r
Sort Mf'rgf'
OUJf'('t
:.u:~~~ : Macros
Pr~ram5
Transient
Us.r
IBM-
Prf'cumpilPd
Sort Mf'rt::t>
COBOL
FORTRAN
RPG
USf'r
IBM-
Rout'"p,
Pro(,fOs5lng
sl.lpphf'd
Pn)(.'etuiifll:
1 0 roullnes
ObJt"('t
SUbr(lUtIM&
SubrwhM'1
SubroJt inP&
Ob)f"ct
5upplird
Macros
Pr<.t:rams
An)' UluPD6
U8er PLl .tatfom.ntl
IIBM-
~~~f'SSUlb : ~Ur:l!::l"
Pr(~rams
Routhlt'6
Any User
PD6 or
SYS 1.
MACUB
ModuJ ••
Modules
Us.,
Ma('rol
Subroutine-.
Uler
COBOL
Ulie,
User
Subroutines
COBOL
Slatf'm.nt.
Uti"r
Statf'mpntl
PrO£rams
A!:tsf-mbler Sublibrary
CORE IMAGE LIBRARY
LOAnABLE FORMAT (PIIa ... )
Figure 8.
COBOL Subllbrary
RELOCATABLE UBRARY
SOURCE LIBRARY
EDITA!1LE FORMAT (Ob),n modul . . )
TRANSLATABL.E FORMAT (Symbolic)
Program library correspondence between OOS/360 and OS/360
ESO
Object module
or load module
TXT
RLD
Input
TXT
Obj ect module
or load module
~:£ule
} Output
RLD's
OS/360 linkage editor
Object module
Input
TXT
Object module
00S/360 linkage editor
o
Figure 9.
12
Physical difference in linkage editor output of the operating systems
Phase
}
Output
BOOK1
CS9 Module
CS10 four
END E
lIBA
BOOK2
CSI2 Module
CS13 six
END F
CS 1
CS2
CS 3
BOOK3
CS6 Module
CS7 three
C58
END G
C512
CS 13
Primary Input Data Set
CS 6
C5.,
CS e
ENTRY START
INCLUDE LIBB(BOOK1)
OVERLAY A1
INCLUDE LI8B(BOOK2)
OVERLAY A2
INCLUE>E lIBA(800K3)
OVERLAY KJ.
INCLUDE lI8A(BOOK1)
OVERLAY A2
CSll
END D Module fivlI
OVERLAY A1
INCLUDe lIBA(BOOK2)
Output Module
librory
CSI
C52
~
C54
ill
C56
C57
~
CS9
Q.l2
lIBB
BOOK1
BOOK2
CSI Module
CS4 Module
C52 one
C!5 two
CS3
END B
END A
Figure 35.
I
eill
CS12
CS13
entry point START
Processing an Overlay Program From Libraries
lI8A
BOOK1
INSERT CSI ,CS2,CS3
OVERLAY Al
INSERT CS4,CS5
OVERLAY A2
INSERT eS6,C57,e58
OVERLAY A2
INSERT eS9,C51O
OVERLAY A2
INSERT eSll
OVERLAY Al
INSERT 012,C513
ENTRY START
BOOK2
CSI
CS2
CS3
CS4
CS5
e56
C57
eS8
CS9
C510
C511
CS12
<:513
END X
CS 1
C5 2
CS 3
Al
C5 4
CS 5
I
A2
CS 9
CS 10
CS 6
CS 7
C5 8
Output Module
Library
CSI
C52
Pdma'Y Inpul Dolo S e l - - - - - - - - - - . .
INCLUDE lIB.A (BOOKl, 800K2)
~
>------I~I
ill
C54
ill
e56
CS7
~
C59
~
au
C512
C513
entry point START
. I
Figure 36.
o
Processing an Overlay Program With Insert statement
CS 12
CS 13
I
~~--~~--~~----~----~~
DO caras required
program execution
IIGO
;;.----
fo~
EXeC PGM-' .LKEO.SYSe-1QO
~/~~~~~--~--~/
and linkage ed1tor control statements
I cb~e~t deck(s)
IIIS't.SLI~j
DO
I 1/
,
Sf' AC [~_.(_1_0_2_11-'.0,,;,,<2_0_0-,,:....2_0...:,.>..;..>_ _ _ _ _ _ _-.
//ISYSU'l!l
DO
UXjl
;'.lS:'~:;::".A )~,F::"(SYSLMOD».
C
/1/
UNIT .. SY:.iL.iA.Dl,:,P'(NEW.PASS)
/1 ISYSLi~~?.e. __-.P.2~A:~;:>'~,::·';;'t~TiG;:·) ,'::PACg· 002 4 • <::;0,20,1) J. C
ll/SYSPRU:T DO
IIEDITGOI
S,{S~~:"_·l_"'.A,-,--_ _ _ _ _ _ _ _ _ _ _ __
EX?<:._~~]..;:;~.[.lj':'~";:'l' . J.."::. F LIST. L&;T. NCAL'
II ILKf.D
I
JOB
1,S~I:~,~~~~EV~~·1
gUre 26.
o
The Linkage Editor step of an Edit and Execute Procedure
Appendix A:
Exarrples of Linkage Editor Processing
43
1
o
las
I.
#?'\~ Vi"f, C... c..t$'-
:>
. (
~
\
~". . ~;
\-~ . ( . \ 0 C
~.~.:~, \ ' . ) )
bv h· Ct~ -:. ~ C, '\ t."
.
\.~ t~ f\\ 0 E.~ {~
--_._-----_._-----
--_....__ ._---_ ------_...
/i
(1 A. \ \
t~\.J
J)
\
\
\
:
,.~l
~
....
-
I
I
~
~.
a
3:
'-.9
<
~
x
j~'"
C~
l-lj
~"
......
A
--;k<..
\'
Ck
'-'
0
.......I
~ l
~
II
-- ,-20
::J
3
C"-
""I
--;-~
tJI
c2.
-~:~
:s
c:
>-
"
l..JJ
1-{
()
\~
Q1.
'-../
tj
ljJ
p..
\!1 if}
P
...,:)
(-
Q
\-r
A
P
\!:.. ,.-..
~
\:1
-
,..-.~
:)
<"'\
0"
-
:.~
'-J
C:L
0
)L
<"'\
~ r;!.
-to- tn
i....tJ
~':. '--"-
'-"
'1
W
0
c:: c::
L: c_
VI V1
..,.~.
LJj
.,.1
-1
'-A
i....U
:ll
QL
f-)
1-
-:><"""\
"
W
'-'
'.J..J
,-j
.c:.:
c.....
V1
r.
--
r<)
~
It
J-
~-
l-l
L.
:J
t-=l
q
p
:~
~
:2
'-../
"
)C"_,(,,)
\::
,..
.J')
0
0....
~
A
/~=J
VJ
?)
-l
CJ.-
X
L1..1
-Z
0
~
~ <
<::;:
~J
~
('
8J
'-::;
~
~
s::
--t
\....,
-..............
~
0,r
<:.;.)
(--
c:
C.
~
~
C
c
~
-~
--"""
\-1
~
~
"'.
\j
~
/37
"I!'_k"
C(-JTf1I.. 0 G·OED
./Jt1rF}
SE;:TS
To - CI)-Tr1/.. 0C- /f - DI1ra SE7: :
/1(;1 PrSTi5'?
i)iJ
H~
To
/..CCIfTF:. 11
~[)
/;Idd nc: Y'/i e.
iJISP::. t;
4
DD
D.D
DD
<..
-------~
-x~
f)fj-jl4
PROG·j?A-fI1
o
//
••
•
//QO ~""'ll\J
I'lt) - irl/lv
f)/)
,1/
//
1/.
II~S{)1?7c}() 7"'
./1
~I
//
" .. \.,.1 ~~ ... I (,9 LJ
fi;~f:nf)/).4).~
t'
/)
/')l;
. f. (6
O/:VI
- __
'l"~ II?;?C/('
J
JS
$$6
7\1-0
8)';Z:SJ.
.
(J Yilil~lllv~/I .
;'
.
CJ/J
/40
.
.~-.
---.- ......
----~.~~-.--
... --
/1
o
r>
~
CHAIN OF SYMBOLIC REFERENCES
COMILED W' TH
CREATED WITH DATA SET
OBJECT MODULE
DeB
J
DCBNAME
GET
COMPILED WITH OBJECT MODULE
",,-,X
~~.
'-..
DATA SET LABEL
DDNAME
DSNAME
DO
CONTROL STATEMENT AS
INTERPRETED AT JOB TIME
SEQUENTIAL ORGANIZATION *
CREATE
QSAM
BSAM
*
,x
~.~
1(3
~
\ti
ADD
RETRIEVE UPDATE
X X X
DA
ONLY
X X X
DA
ONLY
APPLICABLE TO ALL DEVICES
C)
o
o
o
~
,"",_..;;:'
PARTITIONED ORGANIZATION*
CREATE
BPAMBSAM
BPAMQSAM
(or BSAM)
(ONE MEMBER)
""
;I~,X
\.N~
ADD
RETRIEVE '
SEQ.
X X X
REWRITE
MEMBER
X X X
REWRITE
MEMBER
* DIRECT ACCESS DEVICES ONLY NO UPDATE
o
o
o
DIRECT ORGANIZATION*
CREATE
BSAM
ADD
RETRIEVE UPDATE
X
(WRITE-
LOAD MODE)
BDAM
X X X
* DIRECT ACCESS DEVICES ONLY
><
I
W
~
~
o
o
C)
INDEX SEQUENTIAL ORGANIZATION*
,.
CRE·ATE
QISAM
BISAM
--
-t:.
ADD
RETRIEVE UPDATE
SEQ.
SEQ.
X
X X
(LOAD MODE)
(SCAN MODE) (SCAN MODE)
X
* DIRECT ACCESS DEVICES ONLY
,,(jl
X
I
"->
RETRIEVE UPDATE
DIRECT DIRECT
X X
ACCESS METHOD SUMMARY
ORGANIZATION
SEQUENTIAL
INDEXED SEQUENTIAL
QISAM
DIRECT
PARTITIONED
ACCESS METHOD
QSAM
BSAM
LOAD
SCAN
BISAM
BDAM
BPAM
PRIMARY MACROINSTRUCTIONS
GET,PUT
PUTX
READ
WRITE
PUT
SETL,GET,
PUTX
READ
WRITE
READ
WRITE
READ,WRITE
FIND,STON
SYNCHRONIZATION
OF PROGRAM WITH
INPUT/OUTPUT
DEVICE
AUTOMATIC
CHECK
AUTOMATIC
AUTOMATIC
WAIT
WAIT
CHECK
CHECK
RECORD TYPE
TRANSMITTED
LOGICAL,F,V
BLOCK U
BLOCK
F,V,U
LOGICAL
F,V
LOGICAL
F,V
LOGICAL
F,V
BLOCK
F,V,U
BLOCK (PART
OF A MEMBER).]
F,V,U
!
I
,
I
"-
~~
BUFFER CREATION
AND
CONSTRUCTION
BUILD'
GET POOL
AUTOMATIC
BUILD
GET POOL
AUTOMATIC
BUILD
GETPOOL
AUTOMATIC
BUFFER
TECHNIQUE
AUTOMATIC:
SIMPLE
EXCHANGE
GETBUF
FREEBUF
AUTOMATIC: AUTOMATI C: GETBUF
FREEBUF
SIMPLE
SIMPLE
DYNAMIC
FREEDBUF
TRANSMITTAL MODES
(WORK AREA/
BUFFER)
()
MOVE
LOCATE
SUBSTITUTE
-
LOCATE
MOVE
LOCATE
~
BUILD
GET POOL
AUTOMATIC
MOVE
-
BUILD
GET POOL
AUTOMATIC
-
BUILD
GETPOOL
AUTOMATIC
BUILD
GETPOOL
AUTOMATIC
GETBUF
FREEBUF
DYNAMIC
FREEDBUF
GETBUF
FREEBUF
I
-
o
I-
I
I
SYSTEM CONTROL FUNCTIONS
OS/36 0
DOS/360
Device independence
No
,-Yes
DASD extent information
XTENT cards with
absolute addr~ssing
DD cards, DSCB with
only space request
Space allocation-primary -
XTENT cards at
job time
DADSM
Space allocation -secondary
XTENT cards at
step time
DADSM, dynamically
Device allocation
Fixed at assembly
time, but actual
device addresses
may be changed by
job control cards
at execution time
Dynamic allocation at
execution time_
Cataloging data sets
No
Yes
Cataloged procedures
No
Yes
Systems residence
2311" only
2311, 2314, 2301 t 2303
Split systems residence
No
Yes
Libraries
1 source
1 relocatable
1 core image
(aU on SYSRES)
n source
n relocatable
n load
Private libraries
No
Yes
Job control language
Basic, easy to use
Extensive and therefore
complicated, but use of
cataloged procedures
helps
Link editor
Overlay, map
Overlay, map, XRE F t
mixed load, and object
module input
Relocatable programs
Limited to MPS
macro utilities
and LIOCS
Yes
Overlay
Fetch
CALL
SEGLD
SEGWT
(on any DASD)
22
o
/J../')
1
·1
1·1
I
SYSTElvI
CO~.TROL
FUNCTIONS (Cont.)
c·.·,
II
OS/360,
DOS/360
Dynanlic program loading
Fetch
LINK
LOAD
XCTL
LOAD
ATTACH (option 4 only)
E.:nd of progl'am
EOJ'
Return
Timing
Time of day in all
partitions; illtCl'val
timer in anyone
partition at a time
Tim.e of day and interval
time.r in any p::Ttition
Multiple wait
TP only
Yes
DATA MANAGEMENT
DOS/360
.'---
08/360
Sequential files
Yes
Yes
Partitioned files
No
Yes
Index-sequential files
Fixed records only
Fixed and variable
records, automatic
record deletions,
dynamic buffers
Direct access
Fixed records,
undefined records;
absolute track only.
Extended search
(to cyl end only) )
multivolumes
Fixed. variable, and
Wldefined records;
absolute and relative
track.
Automatic insertions,
multivolwnes, extended
search (to cy I end),
dYnamic buffers
BTAM
No burst mode
devices on multiplex
channel with TP
devices. Official
support for 7770/72
and 2780. Support
for 2260 local and
remote
MFT and MVT only.
No support for·
7770/72, 2780 I or
2260 local
QTAM
With multiprogramming
option only
MF'T and MVT only
Graphics
No
Yes, 22GO and 2250
0
.'
DATA MANAGEMENT (Cont.)
o
Data set specification
DTF only
DCB,
DD,
label information
User standard labels
Sequential: header
and trailer
Di reet: header
index -sequential:
none
All organizations
supported mid '67
Nonstandard labels
n program routines
1 system routine
Level of support
Queued (update,
move, locate)
Basic (index-sequential, di rect, and
work files) EXCP
Queued (update, locate)
Basic (all access methods) EXCP
Buffering
1 or 2 static .
(GETBUF, FREEBUF
in BTAM only)
n dynamiC (OPEN,
GETPOOL, BUILD, •
GETBUF, FREEBUF
macros)
LANGUAGES AND SYSTEM COMPONENT SIZES
·0
DOS/360
0
OS/360
DOS/360
Equivalent OS/360
High Performance OS/360
Assembler
10K
18K
44K
COBOL
14K
17K
80K
FORTRAN
10K
16K
P;L/1
10K
44K
SORT
10K
17K
RPG
10K
.15K
Link-edit
10K
15K
18K
44K
88K
Scheduler
10K
18K
44K
lOOK
Supervisor +
other resident
routines for
workable system
6K-10K
PCP 14-20K
MFT: 26-36K
MVT: lOOK and up
Minimum
partition sizes
10K background
2K foregrowid 1
2K foreground 2
PCP: 18K
MFT: 18K
per partition
MVT: 44K
per partition
128K (G)
200K (H)
(
24
J4j
File
Organization
DOS/360
OS/36 0
Contiguous with
cylinder index
Any place on any DASD
device
Called in from
DASD device
l\tay reside in core
~rust be single t
contiguous extent
within volume
Can be multiple extents
within volume
User responsibility
Automatic deletion of
tagged records
Limited to cylinder
boundary
Can search complete
data set
Index Seq.:
,
i
Location master index
Highest-level index with
random processing
Prime data area
Record deletion
!
...
Direct:
1\tultiple track search
.
,
Relative track and
block addressing
No
Track overflow
No
Yes
Partitioned Data Sets
No
Yes
Yes
DOS/360
Oc)'
OS/360-MFT
Ma..-ximum I/O Areas:
.
Seq. file
2 + work
No limit
Index seq. direct
1 + work
No limit
Simple
Yes
Yes'
Exchange
No
Seq. files only with
restrictions .
Buffering Techniques:
;
Buffer Generation
Compile time
Compile t or dynamically
by OPEN
File I/O Characteristics:
DTF
DCB
. Assigned at De',-ice independence
Data Set· Concatenation
Compile time
'No
No
:
Compile or execution time
Yes - within devices using
same file organization
Yes
c
\
j
~.-I
1/
o·
Sequential
Indexed-Sequential
Telecommunications
Direct
Graphics
Read/write
Get/Put
Read/write
(partitioned)
Read/write
Get/Put
Read/write
Read/write
Get/Put
Basic
005/360
SAM
SAM
none
ISAM·
15AM
DAM
BTAM
QTAM
none
VS/360
BSAM
Q5AM
BPA.'vt
BISAM
QISAM
BDAM
BTAM
QTAM
GAM
BSAM*
.......
..,
Format F - unblockt!d
DOS
Format F - blocked
0
DOS
Format V - unblocked
DOS
Format V - blocked
DOS
Format U
005
QSMl
BPAM*
BISA..'-'t
QISAM
BDAM*
OS
DOS
OS
OS
DOS
OS
DOS
OS
DOS
OS
OS
DOS
OS
OS
DOS
OS
DOS
OS
DOS
OS
OS
DOS
OS
OS
OS
OS
OS
OS
DOS
OS
OS
OS
OS
OS
OS
OOS
OS
OS
DOS
OS
• Deblocking and blocking of lo~ical records must be done by problem program.;
Ij
~--------------------------------------------------------Prtme&aar~
~
I
Cylinder
1
I
I
I
I
t
I
I
I
1 50 I
•
I
I
I
I
1125 1
I
I
1175 1
1200
..I
I
I
I
I
I
1100 I
Average seek. 100 cylinders
i:llample 2: 05/360 spUt &ta area
08/360
ODly
o
....- - - - - - - - - P r t m e data area
I
.
I
I
1--------~'"1
J
j
....- - - - - - - - - - - - - - Pnme data area 2
&
I
I
I
I
I
I
,
I
Cylinder
I
I
150
Average seek,. 50 cyl.
Average seek. 50 cyl.
--------..-.f
.,
I
I
175
200
-I
I
l
ISj
ll&&tt:&ILf.JI,A&,;,'""" ,,,',
,.¥,... ,.
P¥',
1
'I
'."1
'~J
o
A Batch Processing FORTRAN System
for a Minimal Configuration IBM 1620
This paper describes a software package designed to
increase throughput on a 20K card system IBM 1620
computer in an environment where a large number of
fairly simple FORTRAN programs must be processed.
The increase in throughput is accomplished in two
ways; (1) programs are batch executed under control
of a loader-monitor routine, and (2) a powerfUl precompiler reduces the number of execution errors and
decreases the requirement for on-line debugging.
The compiler is a version of the PDQ FORTRAN compiler
which has been modified to handle the batch processing
features of the system. Batch execution is made
possible by keeping the entire subroutine library
resident in core rather than reloading it with each
object deck. Object programs, separated by control
cards, are stacked for input. The reading of a control
card by the FORTRAN card read subroutine or the execution of a CALL EXIT statement will cause the next
object deck to be automatically loaded and executed.
The monitor will also terminate a job if the output
line count exceeds a control card specification.
o
The precompiler detects more than seventy distinct
errors based on PDQ FORTRAN specifications. Many
of these errors, such as undefined symbols, are undetected by the compiler and cause execution checkstops. Recognition of multiple errors in a single
statement is possible, thus eliminating multiple debugging passes.
o
~,
/5Z.
- - - - - -..
--.-.-~~----
..
--
'±'HHtWttltt .... i.;,:
'ii±yttii»riril'·'·r ......... t",.. ....,---
.,.-
'Tiri··· rrrr
.. , M .
WIt L I AM G. LAN E
CHICO STATE COLLEGE
. dHICO, CAL IFORN IA
DEAR MR. LANE,
PLEASE EXCUSE THE USE Of THIS TYPING FOR A LETTER, BUT I CAN
PUNCH CARDS FASTER AND MORE ACCURATELY THAN I CAN TYPE. ALSO, I
UO NOT HAVE A SECRETARY AT THE PRESENT TIME.
MY MAIN PURPOSE IN WRITING TO YOU IS TO ASK IF THERE IS A SLOT
I NTH E 'C 0 MM0 N' ME E TIN G fOR T ~J() PAP E R S, 0 N E 0 F WH I CHI j'''l E t\J T ION E j)
TO YOU IN SEPT. THESE ARE THE SAME TWO PAPERS I GAVE II\] SEPT., BUT
I THINK THERE ~JILL BE ENOUGH PEOP'_E \~HO::;()ULD NOT COlvlr: TO THE
MEETING IN SEPT. TO BE WORTHWHILE FOR ANOTHER PRESENTATION.
I HAVE TWO PAPERS, UNE SPECIFICALLY FOR THE 1620, AND ANOTHeR
OF A M0 REG ENE R A L NAT LJ i{ E. THE PAP E R S S H[) UL D f\j (J T 1\1 E ED M 0 RET HA1\1 AB (J U T
10 - 15 MINUTES. THE ABSTRACTS ARE u
1)
'A LARGE 1620 DISK UPERATING SYSTEM,
REASONABLY 7094 COMPATIBLE.'
60~ 1620 HAS BEEN OEVELOPED AT
PRINCETON FOR USE AS A DEBUGGING AID FOR THE 7094 AND SOME 360/0S
PROGRAiVIS. FN II liD ANI) FN IV liD ARe INCLUDED AS WELL AS
PRIVATE STORAGE OF PROGRAI'-1S. ,'v1AI\lY C(}I\ITROL CARl) OPTIOI\lS EXIST.
S PEE DCA N B E 2 TO 4 TIM ES FA S T E R T H A,,' r-H1 1\1- I (I F LJ NE HAS H ARI) \~ ARE
FLT • PT.) I3DTH 11\1 CO[vlPILATIOf\i Af\lD EXEClJTIOj\j. A USERS MANUAL
ODES EXIST. fv1ANY STUDENTS ARE PRESENTLY USII\!G THE SYSTEfvl
ON A COMPLETELY OPEN-SHOP RASIS.
A NEW IvtONITOR FOR A 40K DR.
2) , BAS I C P RO B L E iVl SUr
THE 1620.'
I i\) F (J R f"l ATIn j\1 RET R I E V A'- Ai\1 D A SOL UTI 01\1 Ui\]
SOME OF THE BASIC PROBLEMS OF INFORMATION ~ET~IEVAL AND
IMP L EM E NT AT I ON ARE DIS C US S E f) • T HE T EC H[\1 I (,) U E S fJ F F I L E~ 0 RGAi\I I llx T I 0 I\l
ARE DISCUSSED. THESE TECHi\JIOlJES /~f~E nISCUSSEO fOR THE I'~OST :OIVlfv10N
FILE DEVICES (DISK AND TAPE). A SOLlJTlni\1 IS SHUI.-)f\l FOR THi:
1 62 0 WIT H DIS K TOO EIvttJ 1\) S T RAT E DIS K l JTIL I T Y•
I AM SORRY ABUUT THE LATENESS OF THIS, BUT I HAVE BEEN
OUT OF TOWN FOR SOME TIME AND CAN ONLY GET RAC< FOR WEEK-ENDS Ar
THIS TIME. I HOPE TO SEE YOU IN S.F.
VERY TRULY YUURS,
o
oo~ @~ un
lS;]
NOV 11967
C.S.C. COMPUTER CENTER
~~
C () [IIi PUT I I\l G GRU UP
GUGGEI\JHE If'; LABS
PRIN·::;ETON, i\).J.
08540
,,
o
1800/1896 COMMUNICATIONS ADAPTER
PROGRAMIvlING SUPPORT
Presented at COl\Il\ION, December 11, 1967
San Francisco, California
by
Robert L. Smith
Systems Engineer
IBM Corporation
1602 West Third Avenue
Flint, Michigan 48504
Phone:313-235-6631
o
/51/,
tfh""j"",f iit#5riF· .. ··-PT..T····:--za ""r--W"¥"WW-
o
_n
TABLE OF CONTENTS
Section
I
II
III
IV
V
VI
VII
o
Title
Introduction
1
Terminal Descriptions
3
Hardware Interface
6
General Description of System 7
Line Operation Requests
10
Data Table Layout
20
Skeleton Core Requirements
21
DIAGRAMS
Diag!,am
Title
1.
1070 System
2.
1070 Data Flow
3.
1050 System and Data Flow
4.
2740 System and Data Flow
5.
Data Table Layout
6.
CA Hardware Interface
7.
Programming System Data Flow
o
i
1
as:az:all'iiiilii"ildNtiiibiiUiiiQilii\ihiiffi..J·.dWM"<:r:,i,,,(J·~"l,·WLU-:'I1' '-""'TtL-"",:jo,
\.,F,~,,", '"'
,.";;Z;S:f¥?,, ..
';·1
I.
I NTRODUCT ION
This paper describes the program written to
p~ovide
programming
o
support for the 1896 Communication Adapter used to attach start/stop
terminals such as the IBM 1070,1050, 1030, and System/360 via
2701, 2702, or 2703 Communication Adapters to the 1800 control
system.
The initial 1896 support was written as a joint effort
between the Flint, Michigan branch office and the Chicago North
branch office to support 1800/1070 systems.
The project was
started in February of 1966 and we were able to run 1070's,
though somewhat erratically, by November of that year.
An
1800/1070 Type III package running under TSX II was submitted
a few weeks ago to PID.
The authors are Norm Rawson and
Chuck Reiling of the chicago office, and myself of the Flint
office.
Implementation has taken roughly five man-years of
time, and five years
off
each of our lives.
You are recommended
to the paper being presented by Charles Jonas of The Dow Chemical
Company to hear the other side of the story, the user's.
The original commitment to the customer specified that the 1896
Support would be accessed via FORTRAN and run under the 1800
Time Shared Exceutive (TSX), with little or no loss of the
capabilities of either the 1800 or TSX.
We also required ourselves to consider the following three points:
1.
The 1800 Control System is at least three magnitudes of
speed faster than the terminals, implying that we not hold
the 1800 to wait for a communication function to be performed.
This requirement was reinforced by point 2 below.
1
-------~-----
/5~
o
uawn,
o
2.
An 1800/1896/1070 System may be
contro11in~
more than one
process, including processes interfaced through the 1800
Process I/O features.
3.
The resulting system
mu~t
not use more than 3000 words of
memory (Skeleton resident).
To fulfill these requirements, the system was designed to meet
the following specifications:
A.
All user programming may be done in FORTRAN.
B.
All requests for communication functions will be placed
in a priority queue.
C.
Upon completion of a requested function, a user specified
flag will be set and, if the option is included, a mainline
()
core10ad may be placed on the TSX coreload queue (priority
and core10ad specified for each function requested).
D.
Printer output may be done via FORTRAN WRITE statements.
£.
The support should take less than 3,000 words of memory
(i,e"
F.
run on a 16K 1800).
Multiple terminals per communication line, and multiple
lines per 1800 will be supported.
G.
TSX will be altered the minimum amount possible so that
TSX maintenance by FED will not be sacrificed more than
necessary.
o
2
/5r;
IIGUlIiGilIG.L:;;:"bl'lI1,l/flIui..i£'~","n""';.I:'...."·,,ihJ.L'Ulii1i#ffi" .r..cf\if?!?i¥M/" """,1.,1",,, f+.
"
The above specifications have been met.
(JD
The use of the FORTRAN
WRITE statement for 1053 output required the modification of
either FORTRAN I/O or Skeleton I/O; we chose what we felt was
the lesser of two evils, and modified TYPEN in Skeleton I/O.
II.
TERMINAL DESCRIPTIONS
The following brief descriptions of the various Start/Stop
Terminals available are meant to present the operating differences,
and the capability of each pertinent to the discussion that
follows.
For more detailed information, please see the appro-
priate reference manual.
IBM 1070 Process Communication System (Form No. A26-5989)
o
The IBM 1070 system is a tele-processing system designed to
monitor, supervise, and control widely spread process areas.
Two-way transmission between a remote 1070 System and a central
station is over standard voice or sub-voice grade communication
lines.
A 1070 System consists of a 1071 Terminal Control Unit,
1 or more 1072 Terminal Multiplexors, plus options to allow
analog input, digital input, digital output, pulse counters,
pulse output, 1053 Printers, manual entry stations, ahd display
units.
Transmission speed is selectable from 134.5, 600, or
1200 baud.
See Diagram 1 for a schematic.
Diagram 2 gives
some examples of character exchanges necessary to perform
certain functions.
Analog Input is received as a series of
o
three digits for each point scanned, representing a count
between 0 and 7999 proportional to the signal attached.
3
IS%'
The 1070 Process Communication System will recognize 5
();
function codes:
o-
Conditional take from address 0
5 - Unconditional take from address 0
6 - Conditional take from present address
7 - Unconditional take from present address
9 - Send
In addition, the 1070 can be set to random or sequential operating
modes, and, on send, the printer output channel may be selected.
Terminal Addressing is a sequence of 3 characters transmitted
from the 1800 that specifies the terminal and function wanted.
o
The three characters and their use are:
1.
EO T -
2.' A
©-
Calls to a t ten t ion all
term ina 1 son t hat 1 i ne .
The address (one alpha character) of the terminal
that is wanted.
3.
N
The function (one BCD digit) that is
requested.
The 1070 Character Coding consists of bits BA842l.
The 1071
Terminal Control Unit contains a 3-Digit Multiplexor Address
Register.
Random or sequential operating mode is selected by
the presence or absence of the B-Bit in the
Printe~
o
hundre~s
digit.
Output Channel is selected by a B-Bit present in the
unit's digit of this register.
4
This register may be set by
/59·
- --------,,-,,-••..
"~-~--
--"- --.. --.----------.-------"---,=---.--- .-.------------------
-~-
ash 0 r t SEN D, i. e . ,
© A 9 .. @A
M1M2M3@' . wher e'
a positive response from the 1070.
M2
It • • "
·.;.~.,.c.:;;.;"".;"
•.•." .. ,,~ _ _ _ _ _ _ _ ._._.__.~~_
rep res e nt s
0
Ml is the hundreds digit,
the tens digit, and M3 the units digit that are placed in
the 1071 Multiplexor Address Register.
The B-Bits are placed
in the correct digit by the program before this message is
transmitted.
1050 Data Communication System (Form No. A24-3474)
The 1050 system is a tele-processing system designed to send and
receive data in the form of cards, paper tape or manually.
A
system will contain a 1051 Control unit, plus any or all of
the following units: Card Reader, Card Punch, Paper Tape Reader, Paper
Tape Punch, Programmed Numerical Keyboard, Selectric Printers, and
an Alphanumeric Keyboard.
Diagram 3 shows a typical configuration
o
and data exchanges with a central computer.
The 1050 receives and transmits on standard voice or sub-voice
grade communication lines at a rate of 134.5 baud.
Addressing
the terminal is done by a two character sequence consisting of
a terminal address character and a device selection code.
The
device selection code also specifies whether data is to be sent
or received.
Multiple blocks of data may be sent or received
without relinquishing control of the communication line.
o
5
--~
- - - - - - - ------
/~()
2740 Communications
~erminal
(Fqrm No. A24-3403)
The IBM 2740 Terminal is a tele-processing Selectric typewriter.
As such it has the capability of sending data inserted via-the
standard typewriter keyboard, or receiving data as a printer.
The 2740 communicates with a central processor via standard voice
or sub-voice grade communication lines. Diagram 4 illustrates the
character exchanges necessary to communicate with a 2740.
It is
highly recommended that 2740 model 2 s, with the buffer attachment
1
be utilized for most efficient operation.
2740lS normally operate
at 134.5 baud, but with the buffer attachment, 600 baud rates can
be used.
Note that all three terminals use a slightly different adressing
o
scheme and that all three utilize fill characters (@) to allow
time for mechanical movements on carriage returns, tabulates,
and warm up periods.
On the 1070, filler characters are used
on the 600 baud model to present output characters to the printers
at a rate less than 14.8 characters per scond .
.I I I •
HARDvJARE INTERFACE
The 1896 Communications Adapter interfaces to the 1800 via Process
Interrupt - Voltage, and Electronic Contact Operate.
Each separate
communication line attached requires a full group of each.
Diagram 5 for a schematic of the interface.
The programming
system assumes interrupt assignments to level O.
o
See
The reasoning
here is that with the exception of the 1816 Keyboard, this is the
only device that has the possibility of data over'run without
hardware failure.
/Go/
6
;;WIIULi!I!IaiIiI;;;:;;:;m;;;ru;t",a;.L·.II\\'1D";;;a=,:;"6J:m'ft''''-1naLM1,,:;:;;;;n,,.,~.,"·;:;;";;;;4",;,,,,
,.-,., .""., ..
,4\
"1""
,,"
Each group of process interrupt is wired into one bit of the
level 0 Interrupt Level Status
proceeding through bit 15.
is established for each
Word~
beginning with bit 0 and
In this manner an inherent priority
line.
However, once the servicing of
a line has begun, it will proceed through to the BOSC
instruction before any other lines are recognized.
This takes
somewhere around 750 usec. on a 4 usec. CPU.
Except for the fact that the level 0 exit from MIC has been
modified, there is no other good reason to restrict level zero
to just the communications adapter, although not too many people
will accept responsibility if that restriction is not honored.
IV.
o
GENERAL DESCRIPTION OF SYSTEM
The system may be divided into four logical areas:
1.
Subroutine LINOP
2.
CAISS -- A subroutine containing
a.
Supervisory Routines
b.
ISS - Interrupt Servicing Subroutines
3.
1053 Support
4.
CATST
~
The error coreload.
Each of these areas is described briefly below.
Subroutine LINOP
This subroutine, callable from FORTRAN, is the user's means of
requesting a communication line operation.
The user has a
choice of having an indicator set for him at the completion of
7
o
.. Pif9&
m
n.
T
.-iff"ife) -'jjt¥¥5h¥5!¥6ifW6i'¥ffffiC'¥ 'j""ffPWU53"%Bf5UTlrr
the requested function (LINOP), or he may-have the system
C~,'I
,,'
execute a CALL QUEUE (LINWQ), placing a designated mainline
coreload on the TSX queue in addition to setting the indicator.
LINOP, which is used here to designate both calls, contains
a line-operation-request queue which the Supervisor may
interrogate, plus the necessary queue maintenance routines.
CAISS Subroutine
This subroutine is defined as an ISS Subroutine, allowing the
definition of a separate entry point for each ILSW bit used,
plus a CALL or LIBF entry point.
The ILSW bit entry points are
entries to the ISS portion of CAISS.
The CALL entry point is
defined at Skeleton Build as the servicing subroutine for a Program
4()
Setable Interrupt that is assigned to the Supervisor portion of CAISS.
Brief description of the Supervisor and ISS portions of CAISS
follow:
Supervisor
The Supervisor is not a separate program or subroutine as
in"LINOP, but is a set of routines, contained within the
CAISS package, to provide services to the Interrupt Servicing
,Subroutine (ISS).
The Supervisor is entered by a CALL LEVEL,
said level being below that of the ISS itself.
This allows the
Supervisor to perform its tasks without interferring with
the servicing of 1896 CA interrupts.
The Supervisor builds
o
8
/~3
-".. ...
~"
=._---=._.- _._._ .. -..--.-.... -...
--~.--.=.--'-."-="
'~'-"'-=-'-"=--"-='---'''---''-
-----=~
line operations for CAISS from the user's queue
does data·table
chaining~
entry~
o
halts typing on 1070's to allow
scanning, removes completed line-operution requests from
the queue, saves error information for the error coreload,
and in general acts as an assistant to CAISS and an intermediary between the user and CAISS.
ISS
These routines service the interrupts of the 1896 Communication Adapter~ sending and re~eiving data to and from the
terminals.
To TSX it appears as a special I/O Subroutine.
1 0. 5 3 S.!:U?J?..Q.r,t..
The 1053 Support consists of modification to TYPEN in TSX, so we
may
inte~cept
non~existent
()
typewriter messages just prior to their output to
1053's attached to the 1800.
These messages are
placed on the disc, since the non-existent 1053's are defined as
.'
always being busy_
When a "1 i n e i s f r e e and h a $ n0 que ue e nt r i e s wa i tin g, CAISS i n-i t i ate s
a special TYPEN
fu~ction
to recall these messages from disk.
These messages are then converted to BCD code, fill character (T)
counts are inserted, and a CALL LINOP is performed,_
The 1896
Support System thereafter handles this entry as any other.
o
,
9
o
An additional feature called Type··Stop was put i.n to allow
the 1070 users to halt the typing of a 1053 message, ···perform
control functions and/or input scanning)
then resume the
typing where it was interrupted.
CATST - The Error Core1oad
Errors are handled in a rather elementary fashion to get around
,
core limitations.
When an error is detected, the contents of the
work area for that line are transferred to an error buffer, and
the whole operation is retried.
After a set number of retries,
the operation is aborted and terminal is marked down.
o
queue entries for that terminal are attempted
successful, the terminal is set back up.
an~
Any other
if any are
The abortion S;9n2.1s
for the ·queueing of the error coreload - priority and error
parameter of 1.
This coreload prints out a message on a designated
unit, can dump the error buffer and other information on the list
printer, keeps counters updated, and tests terminals that are
marked down by the simple method of addressing them.
terminril answers, it gets put back up on line.
If the
The error coreload
is also queued up periodically by CAISS for counter update and
te rm ina 1 t est s .
o
"
10
jf£;·S
Design Consideration
The -structure of the support is the resul t of trying to satisfy the considerations mentioned above in Section
I~
0
plus a few
factors which were not known at the initial specification.
These
factors were:
a.
The specific hardware design of the 1896 Communication
Adapter and its interface to the 1800.
b.
The need to service at least six 1200 baud lines.
This speed
means there is only 7.5 milliseconds between characters on
a line and 1/6th of that allows us only 1.25 m/sec. per line
for a maximum service time on the CA interrupt level before
we start to lose data.
On a 4 microsecond 1800, this is
approximately 65 instructions, using San Jose's estimate
of 20 usec/instruction.
For this reason, a high percentage
(90+) of the coding is in one-word, indexed instructions,
o
which run approximately 10 usec/instruction, allowing us
about 120 instructions (which merited a minor sigh of relief).
Because of this time limitation, we have, wherever possible, delegated
jobs to the Supervisor, letting the 1896 - CA idle if necessary.
IV.
LINE OPERATION REQUESTS
,Functions:
...
All line operations are initiated from CALL statements to a
skeleton resident line operation routine.
functions will be
provided~
The following line
o
11
2fi"[fPPi .-
1070 Function Codes:
o-
Conditional TAKE from zero multiplexer address
5 - Unconditional TAKE from zero multiplexer address
6 .. Conditional TAKE from user specified multiplexer
address
7
~
Unconditional TAKE from user specified multiplexer
address
9 - SEND to user specified multiplexer address
11 - SEND to 1053 Printer
A terminal multiplexer may be set to either random or sequential
operating mode, and the 1053 Printer Output Channel may be
selected on a direct send (specify cuntion Code 11).
1050 Functions (Device Selection
~odes)
Polling (Take from 1050)
Sendin~o
0 - Any input component
1 - Printer
5
-
Key bo a rd
6
~
Reader 1
7
the 1050
2 - Pri nter 2·
·3 - Punch
4
Reader 2
Punch 2
9 - Any or all output components
that are in a steady status
2740 Function Codes
Being a relatively simple device, the 2740 functions are relatively
simple.
4C).
A.function of 0 (zero) requests'data from the 2740 keyboard,
its only input device, while any non-zero value will send data to the
2740 printer.
Period.
12
.
. .. p
_..¥4.¥,
/67
Call Statements:
Three line operation subroutines are provided:
a.
o
LINOP (line operation) which maintains an indicator
for the line operation status.
I,
b.
LINWQ (line operation with queue) which queues a
mainline coreload when the line operation is
I·
complete and maintains the operation status
I
I
indicator.
c.
TYPSP (type stop) which interrupts an 1896 message
of priority 255 in favor of a more important operation
on that line, and then somples the message when the
line is again available.
Subroutine LINOP is called either by one of the following call
sequences
1.
FORTRAN:
2.
ASSEMBLER:
CALL LINOP (LF, LP, LL, LT, NDATA, IND)
CALL LINOP
DC
LF
DC
LP
DC
LL
DC
LT
DC
NDATA
DC
IND
Note:
FORTRAN ARGUMENTS
LF, LP, LL, LT, LTSXQ.
may be integer constant
or variables, other
arguments are integer
variables. The assemb}er
constants are the addresses
of the arguments.
o
13
....... -
..
~--
........
~---
/(''P
o
Subroutine LINWQ is called by either of the following call sequences:
1.
FORTRAN:
CALL LINWQ (LF, LP, LL, LT, LDATA, "IND, LTSXQ,
QUE UE, CLON r~ )
NOTE:
Program names QUEUE and 'CLDNM' must appear in
an EXTERNAL statement in the calling progam.
2.
ASS EMBL ER":
o
CALL
LINWQ
DC
LF
DC
LP
DC
LL
DC
LT
DC
NDATA
DC
IND
DC
LTSXQ
CALL
QUEUE
CALL
CLDNM
Subroutine TYPSP is called by either of the following call
sequences:
1.
FORTRAN
CALL TYPSP (lL)
2.'
ASSEMBLER:
CALL TYPSP
DC
LL Address of the line number
Definition of parameters for LINOP, LINWQ, and TYPSP:
a.
o
LF - Function code, 0 - F16 (0 - 15 10 ) stored as a
one-~ord integer, only the low order of 4 bits are
used.
14
b.
LP - Priority to be assigned to this line operation
by the 1896 Line Control Supervisor.
word integer, between 0 and 254.
o
LP is a one-
LINOP uses the 8
low order bits as a positive number:
Zero (0) is
considered the highest priority and 254 the lowest.
FORTRAT 'WRITE' statements to a 1053 are automatically
given priority 255.
c.
LL - Line number, a one-word integer between 0 and 15,
only the low order 4 bits are used.
This number
references the Communication Adapter (CA) attached to
that ILSW bit specified by LL.
The 1800 Interrupt
handling technique implies an order or priority with
ILSW bit being the highest if two or more interrupts
are simultaneous on the same level.
d.
LT - Terminal on the specified line.
1 to 16 with
o
LT can range from
LINOP or LINWQ converting to the alpha
addressing character A-P.
Terminals on a line are
required to be assigned starting with A and proceeding
consecutively.
Neither LL nor LT may reference a
. higher numbered line or terminal than has been assigned
at system generation time or a LINOP error will result.
e.
NDATA - This is the low core (FORTRAN high subscript)
address of user's data table.
Data table format is
explained below.
15
I?O
f.
IND - is an integer indicator which will be set at
various stages of the line operation.
The value IND is
set to depend on the status of the line operation.
IND
1
Line queue was full, unable to enter request
IND
2
Request entered in line queue successfully
IND
3
Line operation complete with no errors
(CLDNM has been queued)
IND
4
CALL paramenter in error, request not entered
IND
5
Repeated line operation failure, request cancelled (CLDNM has been queued)
IND = 6 - Line or terminal down (inoperative), request cane.,
The user may elect to have the real-time clock returned
with the operation complete (3 or 5) indicator.
In this
option, user must dimension the indicator as two adjacent
words (as DATA IND, IND2/0,O/).
o
Indicator value will be
returned in IND and clock in IND2, the next lower core
address.
g.
Format of CALL LINOP is not changed.
LTSXQ - This integer is the priority used if the user
wishes the 1896 Line Control Supervisor to issue a CALL
QUEUE at the completion of the line operation.
may have any value from 1 to 32767.
LTSXQ
It cannot be zero.
h. QUEUE - A dUlmny parameter which indicates the following
argument is a coreload name.
i.
CLDNM - is the name of a core load to be used by CAISS
in a CALL QUEUE (CLDNM, LTSXQ,X).
(X is assigned by
the user at system generation and determines if the
lowest priority entry is replaced or the queuing is
o
ignored when a queue overflow occurs.)
Note that the
coreload is queued on an' unsuccessful line operation
16
171
(IND indicator of 5), as well as on a successful operation (IND indicator at 3).
The type stop call is provided for rapid recognition of interrupts.
It will suspend typing in favor of more important operations on a
line.
CALL TYPSP will stop typing at completion of the current
character and will allow the line queue to be scanned for other
line requests.
When no more entries remain in the line queue,
typing will be continued at the point of its interruption.
This
call should be issued from any interrupt servicing subroutine
that requires line access in less time than the maximum typing
requirement.
For example, a 600 baud line will require in excess
of ten seconds to complete an 130 character message.
o
TYPEWRITER OUTPUT
The 1800/1896 Process Communication and Control System provides
two methods. of typewriter output to the 1053:
1.
Normal FORTRAN 'WRITE' operations with a 'FORMAT'
statement.
2.
Direct line operations CALLED by the user.
FORTRAN 'WRITE' STATEMENTS:
The TASK program of TSX has been modified to handle output to 1053
typewriters on the 1896 system from the normal FORTRAN I/O subroutines.
Typewriters are assigned logical unit numbers for FORTRAN
when the TSX system is built by the user, and the user initiates
a message with the 'WRITE' and 'FORMAT' statements.
In TSX, all
FORTRA~ typewriter messages or message units are buffered to disk,
occupying one sector per message unit.
17
();
Control is returned to
J?Z-
!II!!!tY".
the user program immediately after buffering.
o
Buffered 1053
messages are recalled by the RECAL subprogiam whenever their 1896
line is idle; therefore, typewriter messages have the lowest
priority of all line operations, (priority 225).
The length of
time between 'WRITE' statement and message typing is dependent
upon the number of requests for that line which are in the linequeue.
Messages are recalled from the buffer on a first-in, first-
out basis; alert messages will be recalled before normal messages.
In TSX, alert messages are recognized by tne first character in
the 'FORMAT'
stai(~ment
being a "Ribbon Shift Up".
If a typewriter,
or its line, is down, a backup typewriter specified by the user
can be used.
Error conditions such as disk buffer overflow are
handled by a modified TYPEN program of TSX.
o
A maximum of 16
typewriters, one of which must be an 1800/1053 typewriter, may
be supported by the 1800/1896 modified TSX system.
Because both 'Printer Start' and 'Carriage Return-Line Feed' operations are very time consuming, the user may tal-;e the option to
permanently wire 'ON' his 1053 typewriters and to have the FORTRAN
generated CRLF at the beginning of each typewriter message,
automatically deleted upon recall.
With this option, message
units may not be used, and it is the user's responsibility to
place CRLF where needed.
If the option is not taken at system
generation time, a printer start character is automatically
placed before the message unit or CRLF of the message when the
message is recalled from disk buffer.
o
18
/73
...
' - - '-----.-.•..----..-....
---~
-.-----.~.-~.-
-..... ............. ... ......... ..
..
"
"
"
"
~~-~--
I
:1
0',,:
An emergency 1053 message may be sent with priority over other
I
operations on that 1896 line by using CALL LINOP or LINWQ and a
data table containing the message.
Function code is an 11 in the
user's CALL, conversion code is 4, and random mode bit (10M in
data
tabl~must
l~ft
8 bits and filler count in right 8 bits of each data word.
be on.
Data format is valid 1070 character in
SEQUENCE OF OPERATIONS
The following sequence of operations is initiated by the user with
CALL LINOP or LIN\'{Q:
1.
User CALL from normal or interrupt program.
2.
a.
Request processing by LINOP on same level
set user indicator.
b.
Enter request in line-queue.
c.
Set programmed interrupt for CAISS supervisor.
a.
Recognition of programmed interrupt.
b.
Search line-queue for highest priority request.
c.
Build line~operation tables for terminal and multiplexer addressing.
d.
Trigger first interrupt from Conununications Adapter.
a.
Level 0 interrupt for CAISS
b.
Device addressing.
c.
Data transmit and receive.
d.
Longitudinal redundancy check.
e.
Clear line and set progralruned interrupt for CAISS
supervisor.
a.
Recognition of progranmwd interrupt.
3.
4.
5.
19
~s
user,
17Y
6.
!?~TA
b.
Set line not busy, set user indicator.
c.
Call diagnostic coreload if error has occurred.
User program recognizes operation complete code in
user indicator.
TABLES (Diagram 5)
The data tables required by the 1896 system are modeled after
the normal 1800/1130 data tables.
The address transmitted in
the call sequence points to the low core end of the table, requiring in FORTRAM the high subscripted end.
The first word
in the data table is a word count specifying the number of words
of data or the number of characters to be transmitted or received, depending on the conversion code.
o
The high order bit is
the chain indicator bit if this option is included.
If the
chaining option is not included, this bits presence will be ignored.
The system uses the low order 14 bits as a positive integer.
The second word specifies the conversion code in the high order
- 4 bits, and for 1070 systems, the multiplexor address and mode
in the low order 12 bits.
o -
The conversion codes available are:
Unpacked, one character per word.
1 - Packed Digital, two characters per word.
2 - Unpacked Digital, one character per word.
3 - Analog Input (1070 only)--the converted binary
value of one analog input point is stored in
each word.
C·"·'
4 - 1053 output, the character to be printed is in
the high order 8 bits, the number of filler
characters to follow is in the low order 8 bits.
.,.'
5 - Packed, two characters per word.
20
u·
++WiliIliiUiiii......iiiliMMiGialGII,iI1Gi1&QiQa;;:;;;g,£l1I,:gg;:, ....
_-__ c_. ____ • _ _ _ • _ _ _ _ _ •.•••• _ _ _ . • • _ _ __
i
I
The difference in digital or normal conversion is the handling of
the BCD zero, consisting of the 82 bits, which is converted to a
binary
a
o
in Digital conversion, but left as'a binary A (1010) in
normal mode; plus that Digital conversion concerns itself with
only the four low order bits and ignores any higher order bits.
The Chain Address word must be present and set correctly if chaining
of data table is used.
The contents of this word is used with no
checking as the address of the next data table.
come of an incorrect setting here.
Bad things could
Note that in MPX, a load
address function is provided to allow setting of this chain address that is remarkable
simila~
to the one provided by the 1070
support package.
SKELETON COnE REQUIRE11ENTS
Approximate core requirements for the skeleton resident portions
are given below.
These core requirements are approximate and I
will not be held responsible for them.
-TASK modifications for Typewriter support via FORTRAN I/O:
620 Words.
Subroutine LINOP/LINWQ plus the queue:
The type stop subroutine TYPSP:
330 Words.
30 Words.
CAISS Subroutine:
1200 Words plus 100 each for the chain
on Send and Type-Stop Options plus 40 Words per line attached.
The sum is about 2240 words for a reasonable system.
0,
21
17('
w,···j·· . _-jjf···. :·_jl· ..¥ no-b··
#¥t-_ ¥r---p&""5fW·
Uu
Momentcary
An,aloq
Contc3cts
sign
drS
--:'
'---,~-----------t V Q
--------Q. .-4)
o-(J&.----1
Process Contcact
Alert
sense
~~
--.~_ _~_ _ _ _ _ _ _ _ _ _~
Pulse input
..
-
KY.
A~_ _ _ _ _ _ _ _~
-4 Pul;"e ~utput
JJ14
to
li-IO-73---2--1
DIC
Counter
termindl
I
1073-3
~
r
~~
DigitClI pulse
con vert(;~r
cus'J-omer
1h~ devices
·''-0
.J
~
~
~
1078
Pulse
counter
1072
Termin,;d
--
1073-1
Leltching
C.O.
Multiplexor
~
/077
MC1nucJ!
entryDecImal
~
r,---
' - - -_ _~r--r-'t--
1074
BinCJr y
Displ,ay
.:::l--
1075
Decimdl
DiSple1Y
<:!--
I)
~
i
~
0
i
I
1076
u
Mcanucal
entryBinary
-
I
io-.
more
1072's
-
,.
1053
Printer
1071
Terminc31
Control
Unit
fmc if tcrr.,inol d~tec:ts
AddrCH dIed:, Vi:C or lRC
error dvrin:) first messo:;e.
©
1070 addre:;s reller
Q
7 used f~ begin input
7
(M lM]M3 set aLove)
unconditio:lOlly.
1070 Address
@
If oddre:;; register is H;t on output
©O
cC=:::J
~
01 current
multiplexcr cddre;s if
Pro::t.s Alert indico~or
®
@
}
X
T
S
End of block
End or tro:w.Jclion
©
c:J
[)o:o (rom odJrcs~ tS')
1070 Add:e.s
@
LRC
Or'!
Process elert indicok·r is on
T
E
D ••• lRC.
--
. .,. . . ~
S
0
po@ or non-ovoilable p-:>inl, ther>
rc;ponse is scnt imleod "f
T
E
X
T'
®
End of hamadi,,,
6
0:1
®
R
,
~
~
.~
1070 J..ddrcss
0- Stort seen cf e::i(!'c;s C2)
if Process Ale:t ;ndicc:c~
Q
LRC
scon. of current O-:1dii~SS
OF 1070
M'i
M3
End of block
COEDITIOl{!~L POLLlt~G
End of block
C
Ie:t ...... I
lRC
_.-.. -..----q._-D-IA-G-.R--A-A-4--2-:-/C-)?-~O-D;.&.-~-T-P\-F-L-O---~~----·_m77f.=--·
.
----.--------~-.--
READER I
C~\'
.
- - - - " " ' (...!'~---,
INPUT
105r
Control
Unit
OUTPUT
() POLLING
1800
SEOUENC~
© A7
CA7
\
I
\1
\1
\
SENDING
(0
'\
'\
I
\
\
@ (D<;dd)®L~
1050@
(0
I
\
I
\
I
I
SEQUENCE
New
'Address
Address Printer I
II
@) (Oc3tc1)@
CAl
I
\
\
1050
I
@(D<1,t<1)®\/_---=-@=-'_ __
Terminal
1800
\
I
\
@B 7 •..:...:.
I
\ I
(0
L
L
©
~_@{Dd1;a)~Rc
\
\
\
,I
I
I
C0
L RC Ol(
I
\
\
I
\
0
I
I
I
©
C9
I
\
I
\
\ I
®
Terminal
Not ready
-orParity ch eck
in dddress.
DIAGRAM 3: /050 SYSTElvt
171
.-
PRINTER
~
tdtion
KEYBOARD
contro~~
Checking ,feeltures
CONTROL
required. .
fBuffer 0111
LModel 2
J
POLLING
1800
\
\
I
\
\
\
\
I
I
®
2740
I
L /
I
\
\
I
I
\
/
D.atca ®~_~©~
Not
Recady
EOB
Key
____
EOT
Key
SENDING7
1800@@ab@Text®\ Text ®\_---=©~C_._.._ _ _ __
\
\
2740
I
I
\
\
I
\ I
\ I
Y
Y
I
\
I
\
\
I
I
0
DIAG,RAM 4: 2740- SYSTElvt
/fo.
DATA TABLE LAYOUT
FOR /800//896
SUPPORT
CAISS
o
I
2·
---------------/5
DC/tea caddress in
LINOP CALL
rc fer s to this
.address
10\,.,
core
5
4
---/5
~--------~--~--~-----
Conversion Ran 0
Code
scan
HeClder Word I
1\;1u/-riple)(or
Address (bincar )
CONVERSION CODES
ddjr.~------------~
Hedder Word 2
O. Unpacked} one chc3r/word
Next
Fir 51'" Da"fa Word
*
Packed) two digits/word
2 _ Unpacked) one digi,,/word
3. Anc3log input one MP)(
I.
.Roin t/'Word
4.1053 output) one chClr/word
5. Pdcke~ two chClr/word
~-t-A
character is jl-he 6-bi-t
Be D chdrac-l-er used by -I-he
1070 Termil1.a/s.
high L Clst D~tC1 Word
core Chd1in AddreSs 1 0
Cldclr:
Next Table
~
Chdin c3ddress is]
required if chain
bi-J- is on.
r
Jl.Dp •..,lc/f..Cl.,rl
0-'-9/0-/2-/5
BCD- (x--xBA84·m
Digi1"d/-
J)(
)( x 842/J
....Ea.c ked
0-2-4--78-'0~/2-/5
mAC42 I xxBA842]]
I.»o( x8421 x x )( x8
427J
1053 Output - 0-2--78 --/5
o
[XXi3A842/IFLCNiJ
rNo.
(j)'s J
~f1-e r char.
rls
DIAGRAM 5
IfI.
.~
."
:~O
Q
(d
N··
~
DiT
u.:
~
=>
J-
<:
Ui
u..
0
u
L&J
BIT
BIT
011
BiT
o -'-
~ . ~ ift;\t!SN! i
3
d
RESET C A
:.: :-~
=1- CC:.:~)lrr::~
R~CfHI!:; };nD~?1
READY
::~l
MODE
"~OT ,..P.EADY
.........._),
orr ,~ _L C.:-:-··,..,~:..,..
........ u ":7;;"
t.:;..'I. II.
BIT 6 -~ K~ST~R CONTROL:::::.4
DiT 7 ~- SLi1V~ CC/:TROl.
>!
.~(~) ~
-J
BiT
BIT
6 ,-{'
; 5
BtT
BiT
~~-.r
"IT
un
~ ~':rnIT
t..._
-!<- GOOD
DATA ReADY
-~-::.::::- VRC/LRC ~::SSAGE
'4 - '.~.__
- -"- T I .....
M~
,.,
0 UT
I
--'<-
10 -'<-DATA BIT B
~
IEIT
12 -'<-DATA
I~
I Bi I
14
I. ~
~
I
0
I.
_
U!.IT 15
<- DATA
-i'~
SET
r~-Rr:C~IV£D
C;"?'A
~ ~i"\
2020..,
OR
CO!/"~~"II
• ... V."
IIT Y pcC 2 BI
1:<-RiNG INDICATOR~L I NE
.
CO·;l!'"l!~I'I.''''O~'~'''
.... _.~ Io.Inl I
I'~
I8
9~DA:~~~ L
rDA'A
I
I
I
ADi\P t:1
JeRPQ C083 I)
TERM.RDY-~l .
-----J
L}
BII 2
DATA SIT 1
-------IJ
F- COH.'\mN --------1
~
<:
---.
ltJ
Q::
;;f
:5:>
~
~
-C
~
.
I
[HT 8 ------..../!
LL:
C(
\D
I
I
;«
.
i
f
I
I03J
DATA
[ii'.At;SHIT DATA->
i
I
"1
S~7
i
....
P.'[:A.DY:J--!MODEL 202Cr
CC=·:?LETEJ
IB'Tl1~-:::::;-DATABfTA
B!~ 13 j<-OATA Bi!
:J-'
TO
SG'M!)
_.....
I
... ,~
'
4 -t<-7AANSMIT ReADY
5
v
7 .~ RI"'''' 1"-"'~,\·'l">.U;)T
8
VRC~ E~~~~;'h. ,
9 -~<- DATA BIT C
~
I
f
F~SSAGE COX?LETE --I
IB~T
_
~
'EHT
0
1
2
]
_&\'::::-~CCEPJE
"'-
(::-Dr.'\TA.
~J.
I
~.-----I~
817
!..
I.c--C~
r.AQ '40 5;'7·'D·· ~ -.....
......... '4'.
_.,.
>.-!
:;1
I'"
~R\~QH":'Si
•
---J
_ .....-'
D~~.~
U
I
e
~
llJ
,•
~
,1
DIT ;;,I
BIT 10
D. ~1.~7A B~I 0
a-Dtt':A B!TA:::::~
BIT 11
ro .,. 12 - ' - C.",',,; D!T
:'>1
... l l
·-, _1._
r':' I~ It( - . ; ; >
~-~ ~
I D.I\i... ";:".",
0"
~ a
BIT 1;;
SIT 14 ;_
.. , BI! 2
BIT 15
~.~~:~ ~Il I
vU .... o.!
' - -_ _ _ _ _~.i
Ii.
~
I
I
!} - : -
~
·P. I. (v) BITS 8 THROUGH 15
DO NOT CAUSE AN INTERRUPT
REQUEST IN THE 1800 SYSTEM.
«
a::
~
~
Q
LINOP (USER)
CALL
o
C/~/SS
ZERO
LINOP----_._~
FOR[]\~
QUEUE
ENTRY
[ CA LL LEVEL]
INDICATOR
SET TO 2
-----_-----oj
.~
GETENTf1Y
QUEUE
READY
t'VORl( AREA
1
interrupt
CALL INTEX
o
HANDLE
CHAR.INT.'S
AS THEY
OCCUR
-~.
VvHEN
DONE,
IDLE LINE.
BOSC
CLEAH
INDICATOR
(iET TO
3
VvORK
AREA.; QUEUE.
-.--------
/f3.
*IWl&iIiiIiiiiiWiiiiiiliOiaiiOiiIiiiiiIWiIiiliM&ii!lild&tD!Gl:::&ItU&Q&\&&&IC!ibM:,rrl,,,:u:;;,,,,A....w4r.l,,....... ,"'j." ..;.,
, , 'n,,"
, ,.
,.. "
1"1., " " .. ,
',"
I
I,"
"
$.,
I
,f¥4 , .. ·,4?f?+¥P+
~-~--
..
-...•---...-................
···r
ABSTRACT
FAST FOURIER TRANSFORM
l'}ITP APPLICATIONS FOR THB 113M 1800
T11e Fast Fourier Transform technique as developed
by Cooley ancl Tukey has had a widespread effect in the
field of time series analysis.
HOlvcver, some difficulty
has been encountered by potential users in determining
exactly
!10\\T
the technique works.
An effort to explain
the derivation in detail will be made.
Also, a G.cscription will be given of an analysis
package prog ram ill. 1;}li ell th;;:
r as t Fou r ier
TrllI1S
form
technique is used to yiel<.l amplitude/phase spectra,
po\Vcr spectra, cross pOhcr spectra, anJ auto
correla~
tions.
o
/Pf
;',,\,
0
".,"
.'/
FAST FOURIER TRANSFORMS
The
algorith~for
the computation of complex Fourier
coeffi'cients, as introduced by Cooley and Tukeyl, has had
a widespread
~fect
in the area of time series analysis.
IIowever, some difficulty has been encountered by potential
users in determining exactly ho\" the technique works.
We
will derive a special case for transforming a sequence with
two factors and attempt to generalize the case for r factors.
INTRODUCTIOi\
It is advantageous to review briefly the classical
Fourier transformation in orJer to fully appreciate the fast
Fourier trans fo rm (F FT) .
FUll
r i cr I s theorem says tha t any
time series XCt), which has a finite number of discontinuities
and is periodic (i.e., X(t)
XCt+T)), can he consiuered to
be a linear superposition of sine anu cosine terms \",hose
arguments are integral multiples of the fundamental frequency
W0
=
21T / T .
I\'la t h cma tic all y t his is to say
X(t)
=
!a
o
+
r
n=l
(uncos nw t + b sin nw t)
0
n
0
I
),.,'"
(1)
If we let
cos
t
0
=
sin nw 0 t
=
llw
1 ( jnw
e
0
2"
t
+ e - j nw 0
t)
and
1 ( j nUl t
c
a - e - j nw 0 t)
..,."
L.J
We lIlay rewrite (1) as
00
XCt) = Co
+
L
11=-00
c Jl ;j
w t
n
(Z)
1
l/Z J"=r--r, w =nw
have let c o =-a
c
=(a
2+b 2)
2 0'
n
n
n'
n
o'
Equation (2) is known ~5 the complex fourier series of
Here
fet).
we
Upon examining (2) it becomes evident that we can
describe the sequence XCt) completely in terms of the c n
and wn' TIlis is known as the frequency domain representation
of the time series XCt).
The transformation from the time
domain to the frequency domain is Fourier's transform and
is written, where T is the time length of the series X(T).
(3)
The function Few) is in general complex, and
F(w) = R(w) + jI(w)
=
IF(w)\
e~
.~)
.t.
~
c
,
o
The NIMS monitor system was founded as a TASK offline system
with certain protective features in the core-load builder modified
to allow the use of interrupt service subroutines, the masking
subroutines, and timer subroutines with non-proces s coreloads.
Thus, becaus e TASK is the basic building block, the TASK
support programs such as TRACE, CORE DUMP, DISK DUMP,
etc., are available with this system.
The core-load builder is constructed such that the special
purpose operation codes associated with different type core-load
builds are separated in their table by delimiting constants.
For
example) a process type operation code could not be used in
anything but a process coreload.
It was a very simple thing to
change thos e delimiters such that the operation code tables for
the process and interrupt coreload functions were included in the
non-process operation code table.
The final modification negated the test in the non-proces s
supervisor for the type of coreload being executed.
Figure 3
represents a table of the modifications necessary to provide the
consolidation of process and non-process coreload functions.
4.
TRACE DEVELOPMENT
The TASK utility programs have been included as a subset of
the NIMS monitor system.
All utility programs were considered
adequate for debug tools by the data reduction programmers with
the exception that it was felt that the TRACE output was too
difficult to interpret.
o
020/5
;;;;;;na;c;:;-==a:;:;;;;A&LilIll&=======au:J~m"£{"'.oil!"£.&tI1{,~-1kr.rL';G;;"':r.l&m~£.G,cJt.;,U"",IJn:::Kr; .,,;·A,.,; ."";;,,+fu%;,,4i§.,.,$,, , .,;..,;g; .". ,g" ~';;,." ."
., ..,."
'0 "" """; """ -
r" "n·; ,.Y\$, "" .. 9~-.-4tf.+hr;?f., ·,,-r·
¥
, '"\' )4\fiiflf,,¥4f¥.S¥A4 ..
.. ,,,f4,@?2.,
GAT
47 1548
Table of 1\1odifications to Consolidate
Process and ~~ oru~]roce$s functions
1.
CLB - CORE LOAD BUILDER PROGRAM
• MINOR MODIFICATION TO OPERATIOr\l
CODE TABLE LIMITS
2.
NPS - NONPROCESS SUPERVISOR PROGRAM
• MINOR MODI FICA TION TO REMOVE
TEST FOR PROCESS CORELOAD
\\
\"
fv
~
p
~
@
c
{]
,
This problem lead to the development of ins erting additional
coding in the TRACE routine in TASK to provide a mnemonic format
conversion for generating the output for the 1443 printer.
An
example of output generated while running under the TRACE
mode is shown in Figure 4.
The prime features of this development
are a much easier
format to read and the calculation of the
,
effective addres s that the instruction operation code will be
referencing.
5.
SOURCE PROGRAM FILE STORAGE SYSTEM
Significant effort was expended in providing a source program file
storage capability in conjunction with the NIMS monitor system.
This capability provides prograITlITlers the means for storing
source language programs for subsequent retrieval during checkout
c
activities.
The programs are stored on the file such that alter
cards could be read through the card reader to modify the source
program before the assembler performs its function.
was accomplished in two pnases.
This activity
In the first, the systeITl COITlponents
suCh as TASK, ASM, FTN, and DUP were ITlodified to accept input
from ITlagnetic tape.
In the second, a series of programs were
written to generate a systeITls tape, provide an alter capability, and
perform a periodic update of tile source program systeITls tape.
The key to providing the systeITls routine with the capability of
accepting input froITl magnetic tape involved including the coding
of the MAGT routine in the TASK skeleton and providing a one
word logical switCh in fixed core for interrogation by the input
o
7
c203
47 1549
Trace Output
LOC
OPCO
F
3eBF
LOX
L
38C1
2097
2094
BSC
I
3880
STX
LOX
LOO
2
OF
1
LO
1
80
lE
00
38B1
3883
3884
~
2
80
EFFA
MEMORY
ACCU
MO
0040
38AE
0040
1674
2097
70FC
0000
0000
0000 0040 0008 2083 NO
0000 0040 0040 2083 NO
4400
2097
0000
0000
0000 0040 0040
2083 NO
38AE
2094
38AE
00
L
0000 0040 0040
2083 NO
0000
0000 0040 0040
2083 NO
0000
0000
0000
0000 0040 0040
0000 2096 0040
0000 2096 0040
2083 NO
2083 NO
2083 NO
BOOO
0000 2096 0040
2083 NO
2083
38AE
0000
3802
1170
7001
0000
F1F1
IX *1
IX*2
IX*3 CO OFLW
AOOR
FC
MOX
BST
00
NO
NO
NO
NO
NO
NO
NO
NO
0008
145E
0008
0000 2096 0040
NO
38C3
38C3
0008
NO
NO
0008
80
38C4
38EO
0000
0008
0000 2096 0040 2083 NO
0000 2096 0008 2083 NO
NO
NO
L
20
3A8C
4C80
0000
I
FF
80
02
3A8C
38AE
3A86
2096
3888
BOOO
7402
0000
0000
0000 2096 0008 2083 NO
0000 2096 0008 2083 NO
0000 2096 0008 2083 NO
38AE
38AE
2095
0000
0000 2096 0008 2083 NO
NO
NO
NO
NO
RTE
CC
oc
3887
STO
LOX
2
80
3889
aSI
2
3AS8
BSC
MOX
38B5
3886
~
CJ
TAG OISP
3A8A
3A8C
esc
38B8
MOX
L
38BO
LOX
L
1
00
0040
1180
4480
0000
0000 2096 0008 2083 NO
NO
38BF
LOX
L
2
00
0040
0040
1674
0000
0000 0040 0008 2083 NO
NO
38C1
BSC
I
80
38AE
2097
70FC
0000
0000 0040 0040
NO
2083 NO
@
oo~
C)
0
segments of the systems routines.
c··"
in tne system to read tape drive
,I
supervisor after reading a
This switch forces the I/O routines
a when
set by tne non-process
I / TAPE card from tne card reader.
The
switCh will be reset for the I/O routines to read from tne card
reader if a / / CARD image is read off of magnetic tape.
It was also necessary to modify each of the I/O routines in the
assembler, FORTRAN compiler, disk utility program, and the
non-process supervisor.
Since CARDN is included in the NIMS-
TASK, the coding of CARDN in eaCh of the systems routines was
replaced with the coding necessary to test the card/magnetic
tape switch and perforrn the magnetic tape input operations.
Depending on the routine, it was necessary as part of the magnetic tape
input function to convert the card image from tape either back to
card image format, a special one character right justified EBCDIC,
C\
or leave in standard two word EBCDIC.
,J
Figure 5 is a pictorial r epres entation of the NIMS rnagnetic tape
system showing hOw the support prograrns interface with the systerns
programs to perforrn a complete job operation.
The CDTST-Card to Source Program Systern Tape Prograrn is used
to place the initial version of a source prograrn on tne systern tape.
Tnis
procedure involves searching the system tape for the last record, backspacing tne tape to prepare for the write operation, and writing the card
irnages from the card reader in blocked tape forrnat of 200 cards per
block.
The terrnination card of the prograrn causes a new end-of-tape
record to then be generated.
o
9
r:2,OS
47 1550
r~~~JJS T~~J AS$®mb~y Pr@codur~
SOURCE
ALTERED
SYSTEM
TAPE
_PROGRA~,A
IN
TEMPORARY
STORAGE
TAPE
NPS, DUP
NPS,
ALTER
CDTST
RELOCATABLE
UNBLOCICED
PROGRAM
CLB
ASM
-'
ALTER
CARDS
/ / TA PE ~ -"'M";'
CARD
CONTROL.
m.l8
CARDS
@
~
C)
~-
I-'
0
C
(j
0
The ALTER-Source Program System Tape Alter Program has the
C~. · '·"
'
,
function of allowing the programmers the capability of altering source
statements as it is unblocking the program from the source progra:rn
system tape to a scratch tape in preparation for input into the
assembler program.
Figure 6 shows the three data card formats
that can be used with the ALTER program.
The alter function of the
alter cards is determined by the character in column 72 as follows:
R
means replace the card i:rnage on the source program system tape
having the corresponding sequence number in columns 77 -80 with
this card image.
I
means insert this card image in front of the card image on tape
with the same sequence nu:rnber in columns 77-80.
o
D
means delete the card images on tape inclusively between the two
sequence numbers in columns 73-80.
If only one card image is to
be deleted, the sequence number should be punched in colu:rnns 73-76.
The output magnetic tape on drive 0 is generated in 80 character
EBCDIC for:rnat for direct input into the ASM system assembler.
This
is triggered by a / / TAPE card following the last alter card in the
card input stream.
After the assembler has finiShed with the asse:rnbly process and
stored the relocatable program in temporary storage on the disk, the
next record is read from the tape input stream on drive O.
This
record is a / / CARD image transferring control back to the card
reader.
The program that has just been assembled can then be
executed or permanently stored on the disk.
o
~07
11
47 1551
Sampl@ Alter Card
21
XYD
SW1
Fonn~ts
27
32
35
46
72
77
MDX
L
ABC, -4
DECREMENT ABC TABLE
R
0051
LDD
1
TEMP
R
0193
STD
2
TAPE-1
R
0194
I
0512
I
0719
I
·0720
N0P
LIBF
MAGT
DC
/5000
REWIND TAPE
00999
010041021
@
~
~
~.
o
~
NC
{)
,
The third progralTI associated with the NIMS tape asselTIbly
procedure is UPDAT -Source ProgralTI SystelTI Tape Periodic Update
o
ProgralTI.
This program will utilize sublTIitted alter decks as input
to generate an updated version of the source progralTI systelTI tape.
The progralTIs that were modified by this run will be resequenced
on the new source program systems tape.
6.
SYSGcP - NIMS PROTECTION PROGRAM
Early in the operation of the TSX system, a programmer
inadvertently destroyed the system on the disk to a degree that it
was necessary to do a new system generation from cards.
It
was decided at this point that the system disks would be copied
on magnetic tape with a capability implemented to restore the
disks from this tape.
The SYSGcP program was written to provide the disk to tape
operation, the punChing of a self-loading execution card, and the
subsequent tape to disk regeneration process.
The first step
of the execution of the program involves storing the first 12
locations of core in a table for the regeneration phase of the
program.
A split-binary self loading card is then punched that
is used to load the system skeleton and the SYSG program that will copy the retnaining tape records
on the systems disks.
Figure 7 presents a schematic representation
of the operation of this program.
7.
MINOR SYSTEM DEVELOPMENTS
Additional features have been added to the NIMS monitor systetn
that extend the perfortnance of the system or provide a more simplified
operating procedure.
111"",
C~...)
The standard ASM Assembler Program has been replaced by an
IBM released MACRO Assembler Program called MASM.
This has
been found to be a powerful tool for the programmers providing
extended programming capability.
The NIMS-TASK program was modified to provide the programmers
and operators with the ability to execute stored coreloads by typing
in the program name on the typewriter.
This activity is initiated by
pressing the console interrupt button with no sense switches on.
The
typewriter will request that a 5 character program name be entered
through the keyboard with the following message:
TYPE IN FIVE CHARACTER PROGRAM NAME
o
c2/0
14
o
o
~
47 1552
.svs~f~
Slf~)'i@ril1 St@r~g~ ~~@GQd~r[!
DIST TO TAPE PHASE
-_.-SYSG0
..., T APE LOAD
CARD
[
>!Ia!!ft
.
Dlsrc
T
REG~:t~ERATION
AP~-L~·~~;]·. ~{. ~O/l..D
CARD
PHASE
SYSTEM
~.i ~RO:'~4 TAPE
-----
;._ .. .:-_L:.:..W::::::-1:.:::...
•
WRITE CORE
TO TAPE
REQUEST DISI(
DRIVE TO BE
COPIED ON TAPE
~::::·iI
'7
'3ft-~~.~ _
_~
REQUEST 0151<
DRIVE l\r~D
PACI( LAElEL
FIND LABEL
Of~ TAPE AND
COpy TO 0151<
COpy CONTENTS
OF DISK DRIVE
.. ON TAPE
.
~
'"
-.........
:.
'
@
~
lJI
When the name is typed, the NIMS- TASK program generates a
o
card image to execute a fixed coreload and transfers control to the
non-process supervisor.
The non-process supervisor with the
card image in its input buffer then performs a FLET search and
proceeds to execute the coreload.
The data reduction programmers requested that a current
date be printed on the assembler and compiler output listings.
This
was implemented in NIMS-.TASK and the non-process supervisor
such that any control message printed by NPS would have a date
loaded from the fixed area of TASK included on the end of the line.
The date is stored in TASK by executing a program called DATE
where the date is requested by the typewriter as follows:
TYPE IN DATE
-DD MMM YY-
When the day, month) and year are typed
in~
the program converts
o
them to print codes and stores them in the fixed area of TASK.
8.
SUMMARY
Every attempt was made in the implementation of the NIMS
monitor control system to provide the data reduction programmers
and operators with a convenient system oriented to telemetry data
reduction.
This endeavor has been accomplished and further
activities will be directed to additional facilities for the convenience
of the programmers and operators.
16
cZ/r2,
INVESTIGATION OF
.'
ABNORMAL OPERATING CONDITIONS
...
o
COMMON MEETING
San Francisco
Dece,mber II, 1967
Kenneth R. Ande:--sen ~,,'
Senior Field Enginee:--i:~g S:;:'2c:2.~i.s:
Area 12 Technical A.sS~5~c;-4(;e G~-c-~~:;
San Francisco
o
c2/3
Investigation of Abnormal Operating Conditions
o
A few short years ago many of our data processing system error recovery
procedures were hard'ware dependent.
That is to say the program hU:lg up
until the hardware circuitry recognized that the condition causing an er::-or
had been corrected or reset. 1401 systems stopped with a Process
I
Reader
Punch, Printer, or Tape error-indicator lamp on the operator's console
I
I
directing the operator to inspect a particular device.
The reader may indicate a "reader check" (i. e., card read or checked :n
error), "punch stop" (i. e., card misfed or jammed), or a tape t.:::-:i-;: mc.y
ready.
The system could not continue until the operator had
cause of the error by refeeding the error card
I
:".0: ::'2
correc~ed L~2
correcting the feed proble:n
I
or making the tape unit ready.
The point I'm trying to illustrate is that the problems were fairly obviot.:s a:1d
the improper restarting of the system difficult when viewed from todays' s
environment.
<.
~""
"
The design, architecture, and flexibility of System /360 makes th:s type of
operation impractical. Error information is presented to the program and
further visual indications to the operator are dependent on the system con-:=rol
program.
Since, for example, a system /360 Model 30 can run under BPS,
BOS I TOS, DOS,
~-.--
-~~~----
as,
.. -....---------------.....-..
emulators, or a user written control program, an
-------~---
..-.-- ..
--------.~---
...
-------------~-
.. --...- - - - - - - - - .. ---...
-~-
Page
l
operator must be knowledgeable of the specific system in use. Let me give
you an example. A model 30 operator has just observed the system in the wait
state.
Hitting the start or interrupt button produces no effect on the system.
What has happened?
There are not outward error indications!
No messages
have been printed on the typewriter! If BPS or an Operating System is controlling the system certain bytes of main storage will contain a coded
,message indicating the reason for the halt. Only pr,ior kno'wledge will
direct the operator to display this area.
Furthermore, interpretation of
this data can be in zoned decimal, binary, or packed format.
Basically there are 3 main areas a system can be divided into.
.
o
Processing Unit
ll
.
(CPU), or heart of the system; the
II
Channel,
which the CPU communicates with I/O devices; and the
II
Units." A fourth area could be "Non Operational Units.
\I
The
II
II
Cer:tral
over
Input/Output
These may be
off line units or devices never installed on the system. Any address not
O~
the system will result in Non-Operational Status if used.
I would like to define the. effect of problems in each of these areas.
<,
\
./
"-
",
Central Processing Unit errors are caused by bad data on data busses or in
key registers of the CP'U and result in a logout via system micro,;...p;:-ogramming.
From 12 to 256 bytes of information and status at the time of failure are recorded and the new machine check PSW is loaded. Appropria te indica tors are '
Page 3
0
set in storage by the control program as operator information and the
1 . ·.••. '
I
".,.."
,--,j
~;
system is placed in the wait state w,ith all interrupts disabled. (J~ Channel
failures are errors that occur during
an I/O device.
communi~ation
between the CPU and
This could happen if the CPU signals a device and the
device fails to respond within a specified time. A programmer requestirig
a unit to read or write no data could also result in a cl1.annel error.
Channel
failures may result in a micro-program logout with appropriate indicators
set in storage as operator information. -This area should be investigated
closely as fail"\.lres can result from improper channel programming.
Input/Output errors pertain to specific devices (tapes, disks, reader,
punch, printer, etc.)
to that unit.
an~
This sense operation will transfer up to six bytes of
from the unit to
tl'~e
o
always result in a sense command being issued
informa~ion
CPU for analysis by the control program. Error recove:y
will be performed dependent on the error. Tape read errors result in a backspace and reread with a TIE if possible. Each time this sequence has been
performed eight times, a tape cleaner routine is executed.
backspaced 3 records and forward spaced twice.
The tape is
This runs the record over
the tape cleaner blade re'moving any excess oxide. After 100 rereads a~~e
performed, it is considered an unrecoverable I/O error and the job may be
cancelled.
Tape write errors result in a backspace, erase and rewrite.
Disk' read or write errors cause a retry unless a defective track is found.
A
defective track will cause a seek to an alternate track and a read or write.
Unit record devices r,equire operator action at the device and a proper
o
'Page 4
response to the programming system.
·0.',:1
The operator should examine the
"/
sense information for the cause of error.
If the job had cancelled under the Disk or Tape Operating System all the
pertinent information ·will be logged on the typewriter; however, it must be
decoded with respect to the specific 1/0 unit to obtain all the meaningful
information.
The same error on BPS or BOS would result in the wait state.
Unit record type devices will require operator intervention.
Restart pro-
cedures require the proper reference SRL. Operating guides for all devices
should be available to operators.
.
o
.
Proper maintenance of materials can save countless hours and headaches.
Initialize all disk packs before use.
Notice that the plant has affixed
a label indicating tested bad tracks.
These tracks should have alternate
tracks' assigned during your initialization.
This is necessary because
surface analysis performed prior to shipment tests the complete usable
surface for each track.
Subsequent use of these tracks will be unpred~ctaDle
~"""
,.
and dependent on individual.R/W head tracking on each disk drive.
Tape must also be properly initialized
I
even in an unlabeled shop.
Tapes
are always checked by logical IOeS to ascertain if label information
0"'"
,'.,
.
exists. An. improperly initialized tape my result in a permanent tape read
error and cancellation of your job. This can be avoided by writing a TiVI
:2/7
~--
-
----------._._-.----....-------._-----_._-_.---- .....
--------..- - - - - - - - - - - - - - - - - - - - - - -
"--"--.... ~----"---.------ ..
Page 5
at the beginning of all unlabeled tapes.
Remember the seven or nine
o
track tape marks can only be read error free on their respective drives.
Mark all reels plainly for use on one tape drive (i. e., seven track
Drive 183). Operators should clean tape drives as often as necessary
to minimize tape errors.
Card decks should be given proper care. Plastic edged decks are best
in high usage circumstances. Master decks should never be used except
for reproduction.
This is your best protection against card decks that ran
yesterday and program check today.
Be certain you have an IBM card
gauge and check for punching off registration periodically. A high L'lcidence of reader checks or one card that cannot be read also require
chec~L-:g
with the card gauge.
What methods should be used during isolation of an abnormal operating
condition?
This depends on a great many variables; however, a certa:n
oasic attitude should prevail.
How can I obtain the maximum usefill aJ.".c.
meaningful error dala for analysis by the programmer or IBM engineer?
11(,
.,
\'
.;',
The first point should go without saying, but all too often goes unheeded.
Stop!
Determine what the system will tell; you may have to ask (i. e. ,
display appropriate core locations). Remember
I
a hardware
fai~ure
does
not neces sarily produce a red light indication. On DOS or TOS only by
displaying Main Storage locations 0-3 can it be established that an error
Page 6
has occurred. On BPS and a BaS hex location X '32' contains the error
indicator. -What action the operator should take is described in the operator's
manual .. If a machine or channel failure has occurred
I
a logout will take
place. The system is placed in the wait state to allow the operator to
run SEREP. The
II
System Environment Recording Edited Printout" program is a
stand-alone card deck that will interface with IBM control programs
I
determine
the type error that has occurred, and print out all pertinent information and
. status of the system at the time of failure.
This program will be provided by
your Customer Engineer at installation. SEREP will print out from 12 to 2S 6
bytes of logout area (dependent upon CPU type)
I
old PSW' s
I
new PSWI s
I
CAW, CSW, GP regs, Floating Point Regs., or sense information.
C
';l
It is especially useful when a customer engineer is not on ·site. It is
essential that all operators be versed in rocognizing a SEREP request! A
core dump or re-IPL at this time will destroy any i.nfonnaUon of value.
SEREP will pinpoint invalid use of channel programming
I
use of non-opera-
tional devices, and bad 1/0 devices. All SEREP runs should be saved for
, ."
.-....
analysis.
I ' .. ,
Only after the operator has collected all meaningful data for subsequent analysis
should a run be attempted.
This is particularly important on intermittent
failures.
o
If a SEREP run was not requested a core dump should be taken. Save all.
core dumps I console logs I output, input, program decks.
Documentation
.
\
Page 7
rf",
I
i
'-'
to the extent of recreating the failure is best.
/
Try to isolate the culprit by changing devices, restoring disk packs
I
using other card decks, or if possible run the job on another system.
Of course if the operator believes a hardware system failure exists
I
he
should follow the normal procedures for notifying his customer engineer.
If the problem seems to be in programming all the pertinent data should
be returned to the programmer.
Experience indicates the best approach when system time is available
is to rerun a failing job under the supervision of the person or perso:1s
responsible for analyzing the problem.
Th~s
quickly determines any
procedural or operating problem and gives you first-hand knowledge
of the events leading up to a specific problem.
Many times what one person feels is irrelevant is the key to the problem.
If your programmers after,thorough analysis conclude that IBM's
~
p~ogra;,!,~.s
>"'.
are not meeting specifications or a new release level produces a change in a
program's operation
I
contact the applicable IBM representative.
If your programming system is not at a curr·eni level, check the latest an:lour;.cement letters. The problem may be fixed by updating your system. Many
~d-O
\
I
Page 8
0'·;
problems are now fixed with product temporary fixes (PTF' s).
These are
,"
patches or modules available to correct problems before they can be
included in an official release of the programming system. If no PTF
is available Field Engineering will determine a temporary fix or circumvention and submit an APAR to resolve your problem.
A meeting with both your hardware and software customer engineers at
installation time to establish adequate PM schedules and error recordin;and reporting procedures can insure a maximum system availability.
I
~
The
smooth running of today's system requires quick and accurate diagnosis of
hardware I program, or operational problems.
Our best solution to this is continued co-operation.'"
..........
.~;;.;.;:;..;;;:;=
..
·,_~.;;;; ",,;,:.:;:,,;,:;;;;,,=c;.:;,.~;:"'.;.C;;
...•·.;".. :.;·•. :.:.C.: •.:•..:.;..::..:._;.;;..• :..•::.::.: ••.:;':' •.C.;o;.;.:,;;.:::,:;',.;;,:.;,.:,
INVESTIGATION
OF
ABNORMAL
OPERATING
CONDITIONS
«"'ijtth' f"j'""Y"fft'f" shiH '+ii'Mtr
o
SYSTEM
CPU
I
NOT OPERATIONAL
UNITS
i
I
(OFF LINE-NON-EXISTANT)
:--------...-
o
~
8- ---
. I/o ·/--T---r
11
.
r--~
, II 0
f
I
I... _ _
o
I
1- - - (
t
.J
I
E
L
---t'_ _- J
..........._.. _-_... _._....- - - - .
,
I
I
o
CPU PROBLEMS
CHANNEL PROBLEMS
LOG OUT VIA MICRO PROGRAM
a.
12 BYTES MOD 30
b.
256 BYTES MOD 40 UP
NEW MACHINE CHECK PSW PLACES SYSTEM
IN WAIT STATE
APPROPRIATE INDICATORS WILL BE SET IN
STORAGE
a.
b.
OPERATOR
SEREP
G'
_ _ _ _ _ _ _ _ _.lIIIIiiIInriiilliillCWllliiillillrWiIiiIINmlillliilii)
RIilllifiiiW@"IIIiiiii"""IiIIiIiII-·".'W'--iIIIIiIlJ·llIiIiiiIww"rliiiiliir"l"f_'ftf""IIiiiiiiIHiiIiiiii"aifbiiilij,'eie_"ijShiMbiiiilil-Wiiilliiijijb±ioiiI!ii·'rl±rtiiiiiljjjj········iii/Iiiiij",jJr-fiiiiiiii·
.. ifiliiilllittrifiliiiil··Wiiiliiii···{·.·wM··ttifiiiliii
... fflbii\iiilii"tlteiiiilliitrriiiiiii""O#_Wrlfeu··
.. ·triiliiiii"tH_ritfifilillillbW".""j··IIiIiIi·$-6_ri##tiiiiiiii·
. ·lijlliliilt·ij_h±6i1iil11td'Wlll/ltrl·j_"ttftiiMjij
.. ·.f"tlllilliliWIlll!lili·'fi_'''"i_ttitt_ttllilllllllillillllilil''#UIII"$, _'.- ." :Vb
1
cL..
o
I/o UNITS
SENSE INFORMATION IS READ FROM THE 1/0
DEVICE IN ERROR
APPROI?RIATE ERROR RECOVERY IS ATTEMPTED
TAPE READ ERROR
a. TIE IF POSSIBLE
b. BACKSPACE REREAD
c. TAPE CLEAN ROUTINE
o
TAPE VVRITE ERROR
a.
BACKSPACE -- ERASE -- REWRITE
DISK ERROR
a. RETRY
IF DEFECTIVE TRACK SEEK TO
ALTER NATE - RETRY
: b.
UNIT RECORD
a.
b.
o
OPERATOR ACTION AT UNIT
OPERATOR RESPONCE TO SYSTEM
CONTROL PROG
- -
~
~-
-
---
-
-
~
- -- --
- ...... .:.;,. ';;::;:;::;;" ...;;.,.-.;.;.,;;: ..;;::.;;,-••• ..;;~.;::.:.:-~-;;..:.; .....;~.;~;;;;.; .• ;;-;,-".
.-.
.
....~~= ;:-.:.;.;:~.:;,;;:::-,:;-
:~-:.;;;~.::..-.;;.;::.;.:.~:~:.;:;-.=..;;;;.;..;;.;
. "-. ._ .. ----.. ----.. _....... -
o
TAPES
MUST ALWAYS BE PROPERLY INITIALIZED
TAPE MARK AS FIRST RECORD
VOLUME AND HEADER LABLE
FOLLOWED BY TAPE MARK
OPEN ALWAYS CHECKS FOR VOLUME LABLE
FOR MAXIMUM USER PROTECTION
UNREADABLE RECORD WILL RESULT IN
UNRECOVERABLE I/o ERROR
7 - 9 TRACK TAPES
ONLY READ ERROR FREE ON RESPECTIVE
DRIVES
o
a. INCLUDES TAPE MARKS
MARK ALL TAPES CLEARLY
EXAMPLE: 7 TRACK MODE x '90'
7 TRACK MODE x '68'
.1
WORK TAPES SHOULD BE PERMENENTLY ASSIGNED.
EXAMPLE: 7 TRACK WORK DRIVE 183
o
DISK
INITIA1rZE ALL PACK PRIOR TO USE
FLAG ALTERNATE TRACKS AS INDICATED ON
TEST LABLE
o
FAILURE TO DO THIS WlAY RESULT IN
UNPREDICTABLE ERRORS ON THOSE TRACKS
TESTED AREA
BACK UP TAPE ( CARDS OR PACK)
FOR ALL SYSTEM PACKS
c·
I'
o
CARDS
HIGH USE DECKS PLASTIC EDGED
MASTER DECKS FOR REPRODUCTION ONLY
BEST PROTECTION AGAINST PROGRAM
THAT WORKED YESTERDAY - PROGRAM
CHECK OR MALFU~CTION TODAY
CHECK PUNCH REGISTRATION PERIODICALLY
WITH CARD GAUGE
DONIT OBSERVE PRINTING ON CARD
FOR REGISTRATION
READER CHECKS REQUIRE CHECKING WITH
CARD GAUGE
..
'\:
"
1 '.
",.
o
S ystem
C
E nvironment
R ecording
E dited
P rintout
PROGRAM TO PROVIDE PRINTED ERROR INFORMATION
AT TIME OF FAILURE
LOG OUT AREA INFORMATION
a.
12 BYTES TO 256 BYTE DEPENDING
ON CPU MODEL
.
CORE LOCATIONS 0 - 128
a.
b.
o
OLD PSW'S
NEW PSW'S
c. CAW
d. CSW
e. ~'? REGS
f. FP REGS
STAND ALONE CARD DECK
SUPPLIED BY CUSTOMER ENGINEER
ESPECIALLY USEFUL "WHEN CUSTOMER ENGINEER
NOT ON SITE
PROGRAMMING SYSTEM WILL REQUEST SEREP RUN
CPU FAILURES
CHANNEL FAILURES
BPS - BOS UNRECOVERABLE
I/o
ERROR
BPS - BOS NOT OPERATIONAL UNIT
o
..
..;::.::. .. -"':"''':'':'':''..:,:,:,.:..::... :.;,'::
'-:'::~:....-.:..: ;;.:
o
INVESTIGATION
ON ALL ERRORS GATHER ALL AVAILABLE DATA
PRIOR TO CONTINUING OPERATION
LOOK IN APPROPRIATE CORE LOCATION INDICATOR
BYTES
BPS - BOS BYTE X 1321
DOS - TOS BYTES 0 - 3
ONLY INDICATION MAY BE SYSTEM STOPPED
IN WAIT STATE ALL INTERRUPTS DISABLED
TAKE SEREP RUNS WHEN REQUESTED
IF SEREP NOT REQUESTED TAKE STAND ALONE CORE
DUMP
SAVE: CORE DUMPS
CONSOLE LOGS
PRINTED OUT PUT
INPUT DATA
PROGRAM DECK
TAPE
DOCUMENT THE FAILURE AND IF POSSIBLE SET UP
A RUN THAT WILL REPRODUCE THE FAILURE
ONLY AFTER FULL DOCUMENTATION SHOULD
CIRCUMVENTION BE ATTEMPTED
c'
0230
PROGRAM SUSPECT
C."
!.'
RETRY NEW COpy FROM MASTER DECK
RESTORE SYSRES ( DISK - TAPE)
TAPE ERRORS
CLEAN DRIVE
TRY TAPE ON OTHER UNIT
NEW TAPE
DISK ERRORS
NEW PACK
OTHER DRIVE
RUN JOB ON SIMILAR SYSTEM
HARKWARE FAILURE - NOTIFY CE
BEST APPROACH IS TO RUN JOB UNDER SUPERVISION
OF ALL CONCERNED
a.
b.
c.
d.
e.
o
HARDWARE CE
SOFTWARE CE
SE
PROGRAMMER
OPERATOR
~
,'-
..
01.··.
PROGRAMMING SYSTEMS
ANNOUNCE~IIENT
LETTERS
PTF
APAR
INSTALLATION: MEET WITH CEIS ( HARDWARE -. SOfTV/A?.2 )
PM SCHEDUALS
ERROR RECORDING
REPORTING PROCEEDURES
SMOOTH OPERATION REQUIRES
QUICK AND ACCURATE DIAGNOSIS
HARDWARE
SOFTWA.~E
OPERATION
<:
1'..."
.t. "
BEST SOLUTION
COMMUNICATION
CO-OPERATION
"
UNIVERSITY OF MASSACHUSETTS
COMPUTER SCIENCE PROGRAM
CURRENT RESEARCH TOWARD THE STANDARDIZATION AND FORMAL DEFINITION OF pL/I
by
o
JOHN A. N. LEE
Prepared for the l.finter Meeting It COMMON Users Group
San Francisco. December 11-13, 1967
o
c233
o
ABSTRACT
In April 1966. a subcommittee of USASI Task Group X3 0 4.2 was established to
investigate the standardization of PL/I.
Since that time, two working committees
have established with explicit charters:
Group I:
The Resolution of the Language pL/I
Group II:
Formal Definition of pL/I
and
This paper deals specifically with the work of the committee on Formal Definition
and its relations with the definitional task forces within IBM.
This paper will
discuss the techniques of definition as proposed by the committee. the uses to which
the definition is to be put and the implications of the work of this committee on
the standardization of other computer languages.
A familiarity with PL/I is not presumedo
o
The need to
O
~rogram
digital computers in abstraction as opposed to the
·~'1
,
,
subjective wiring of boards has lead to the development of a variety of programming languages o
These non-natural communication systems are classified, broadly
according to their orientation toward the human user or toward the moronic
detailed instructions of a machine o Whereas the human desires a language akin to
that which is either his mother tongue or one of the other artificial languages
with which he has been associated, many of our current languages have been developed
from machine dependent codes towards these desires and have not yet reached the
half way pointo
Althou~h
there are many proponents of the concept of ultimately
using natural language as the means of describing a computational (or more
specifically, a transformational) process, these forms of input suffer from two
major defects:
a) they are verbose, and b) they suffer from many possible ambiguous
contextural translations o
4()
Further, we currently do not thoroughly understand the
syntax and semantics of natural
languages~
In any
case~
mathematical notation
(language) has developed over the past few hundred years in spite of the existance
of natural languageo
Existing analyses of programming language, such as found, for example, in
the ALGOL 60 report [l]j place emphasis on the syntactic form of the concepts
and definitions u
X304~2C2
In contrast, the approach being proposed by the subcommittee
of the USASI, concentrates on the role of constituents of a language as
commands or instructions when the program is executed in a machine o
This implies
that the programming language is to be considered in relation to a machine 9 be
that machine actual or abstracto
Classical abstract machine models such as
Turing machine or finite automata, which are obvious bases for the definition of
language semantics, lack many of the salient features of an objective situation
o
which are relevent to the description of the execution of a programo
For example,
Turing machines lack the random accessibility of current memories, thus throwing
the burden of (say) table look up to the wiles of the definero
2
Conversely. some models l.rnitate machines more accurately but are unsuited
to the general definition required in the specification of a standard o
o
For
example. if there exists ail executer (compiler, interpreter. or translator) for
PL/1 on a particular machine, then that executer is de facto definition o
In
paraphrasing De vCarte (I thinkp thex'efore I am) a compiler can be said to execute
and therefore to define o
On the other hand i an assembly lising of a compiler
cannot be said to constitute a nationwide standard and in any case contains many
algorithmic processes of
which are irrelevant o
analysls~
recognition and (1n particular) generation
Further as the state of the art compilation advances p
it must be expected that the efficiency of compilation and the optimization of
the generated code will
However. a compiler
improve~
thereby presenting an improved definitiono
transfo~ms
an input in one non-natural language to
another non-nat.ural langto.age which is not a standard"
Further only the execution
0
of this target language can reveal the semantic nature of the input (or source)
language
0
We must, therefore. raise the question as to how a natural language
description of the syntax and semantiCS fills the basic necessities of a specification o
In particular. a definition which may be used as a standard must
possess the following
qualities~
- it must. be
rigorous
complete
- it must
structure
underlying concepts
~isplay
- it must distinguish between
language defined elements
implementation defined
elements
undefined elements
Current experience with ALGOL 60 [1] and FORTRAN [2) shows that the natural
language descriptions of a programming language do not meet all those requirements o
Ambiguities exist in the current documentation standards of these
languages which have been inserted as a result of the use of a natural language
as a .descriptoro
t..Jitness the list of "trouble spots" reported by Knuth [3]
0
o
3
Similarly~
natural language specifications tend to overlook the possibility
of the effect of executing a procedure in which the values of certain entities
fall outside the domain of the specificationo
That is, current specifications
only define the semantics of a program that exactly fits the task for which it
was designed o
For example p the
~ORTRAN
specification (sec 70102 0 l03) states:
"701020103 Computed ~O TO Statement o A Computed
r.O TO statement is of the form~
00000,
k ), i
"""11
where t.he k S are statement labels and is i an
integer variable reference~ See 10 0 2 9 8 and 10 0 3
for a discussion of requirements that apply to the
the use of a variable in a computed GO TO statemento
Ij
Execution ot this statement causes the statement
identified by the statement label ~ to be
executed next~ where j is the value of i at the
time of the exeuctiono"
o
However. no mention is made of the actIon to he taken when the value of the integer
variable
i is outside this range o Yet this situation can occur during execution
without protection or detection at compIle time o
In my own shopv we use four
different versions of FORTRAN which treat this problem of values outside the
range of definition in different manners o
If a description is to demonstrate the structure of a language both at
the syntactic level and the semantic
level~
the elements of the language muse be fully
then the interrelation of both
expanded~
~or
example, consider the
interrelation of COMMON p DIMENSION and EQUIVALENCE in FORTRAN o
planation of this interaction 1s verbose and contains many
!l!,
A vel!bal exelses,
excepts i the subject phrases of which are not readily accessibleo
~
and
On the other
hand, an algorithm for specifying the interrelation of these statements is
definitive though lacking the ease of readabilityo
4C)
Yete why should
a specification be generally readable?
In no way can we
consider that specification or a standard to be a teaching instrument and further,
023?
4
a specification or standard must be aimed at the designers and inspectors
of standards and not the general public.
For example. a building code is
couched in the standard terms of the civil engineer or the building inspector
and is not of general concern to the person usinl (aay) an auditoriumo
o
If
such users have questionl regarding a building code, then they rely on the
technical ability of the experts.
Thul a Itandard specification which i8 based on the technological tools of
the profe.lion is • valid Ipecification.
However, in the computer industry
there are insufficient people trained to have an understanding of the theory of
programming to bs able to act as the experts for the
g.~eral
user.
This is
the fault of our educational .yatem and cannot be uaed as a deterrent to the
development of atandarda baaed on the Itate of the art.
The purpose of a standard definition exceeds the bounds of merely forcing
a minimal conformity upon implementera.
A definition should also act as an
information Iystem to an.war such questionl aa:
o
"Is •••••••• a legal part of the lanauaI8?"
"What happenl if •••••••••••••••••••••••• ?"
"If thil ele.ent ia added to the language, does it create any
ambiguitiel?"
"Does the compiler provided by the manufacturer conform to the
stal\dard.?"
Thus in some respects a standard must be an information retrieval system,
in other respects, it needs to be a model translator and executero
However, ther.e
is a point beyond wh1ch this .tandard cannot pa.I, that 1a, there i. a limit
to that which can be thought of a. closing the credability sap.
For example,
arithmetic mUlt b. considered to be implementation defined and thus outside the
Icope of the definition.
How.ve1:, the que'tionl, "What doe. 0.1 + 0 0 2 mean?"
can b. answered a. "0.1 + 0.2 i. an expre.lion, where 0.1 and 0.2 are real
decimal fixed point con.tantl and + ie the infix operator reprelenting the operation
o
5
o
of addition
Then the r'luestion\} fVWhat is the result of evaluating the
Q "
expression 0.1 + 0,,2?" might be "The result of evaluating the expression
0 0 1 + 0 0 2 is a representation of Oc3 to the same base and scale of the operands
and the precision of the greater of the precision of the operands + external procedure
+ O,I*
O,I~:»
+ DELETE (3,3* [ I (O,l* event
option»])
where the enclosure of metacomponent names in braces eliminates many of the
problems of ambiguity in the metalanguage.
The parenthesis groups are of the
general form of an implied DO list in a FORTRAN input/output statement with the
variation that the limits of the repetition precede the list and the increment is
always unity--thus the construct
+(l,3*A) means that the metaresult may
be formed from 1 to 3 concatenations of the terminal A.
The same metalanguage
may also be used to define the syntax of the canonical intermediate text and it
is further anticipated that a transformational grammar can be developed which uses
the same basic nomenclature.
However, the terminology of the analyzers, synthesizers and the executer
cannot be readily expressed in Backus Naur Form though some variant of AMBIT [7]
would seem to be appropriate.
The "execution" of the input text using the
schema shown in Fig. 2 is merely one approach to the problem of definition.
An
alternative technique would be to define a fundamental language in detail
according to Figo 2 and then to define transformational grammars on
languages to transform them to this fundamental language.
o~her
The choice of this
fundamental language could be one of the biggest political problems to face the
data processing section of USASI.
In fact, if PL/I is all it is purported to be, then it is the obvious funda-,
mental language.
However, the growing band of advocates of APL (or Iverson
o
9
o
Notation) have a candidate for this position which has already been shown
to be adequate for the formal description of a machine [5] and hence should
~H~
adeQuate as the target language of syntax directed translator o
T~e
".'.:t1d
type of linear executer described by the Hursley and Vienna definitions
hy definition executer of X3 0 4 o 2C2 or the translator described in the
previous paragraph p cannot adequately describe a language which is salf
modifying or self generative o
"inter~'reter"
However, a feed back loop from the ttexecuter" or
to the syntax analyzer will solve this problem o
On the face of it. a definition executer does not appear to be as efficient
as a
stnnd~rd
structure
Q
compiler for the Rtrnight forward task of recognizing an input
This conclusion is based on the premise that no matter how efficiently
the specification is designed, the analyzer portion of the executer will spend
some considerable
4[)
~ercenta~e
of its time exploring blind alleyso
Further~
the
keys of string recognition may be camouflaged beneath the wealth of associate
syntax p every element of which must be verified before even a preliminary recognition can be made o
For example, the first three characters of each BASIC
[6] statement are a key to the type
of.;.~:.atement
following.
As another example.
a compiler implementer has at his disposal certain machine dependent information
which he may utilize to improve the efficiency of the recognizers o
In particular.
the sorting order of characters permits the testing of characters in a relative
manner so that a binary tree can be built. in which the length of the maximum
branch (representing maximum number of comparisons to be made in recognizing
a character) is a minimumo
The maximum tree branch needed in a USASI FORTRAN
compiler for the recognition of a keyword is fiveo
A syntax directed translator
does not have this ability and thus must chain through a linear sequence of
alternatives till the list is exhausted o
0
That is. for example. to recognize
that the character I is not an element of the alphabet requires only five comparisons using a binary tree search in a compiler (Figo 3) but 26 comparisons in a
..._._.
-_....._ - - _ .
""---..----..-- .. -..•.....
........ "--.--.-...".- •.•...
~
"~~
10
syntax directed analyzer.
Under these conditions, a definition executer may
o
never replace the specialized compiler and with the given purpose of definition,
the speed and efficiency of execution is not in competition o
As programming
languages become more permissive, the task of defining a language in prose will
become more difficult and there will be a corresponding increase in the complexity
of a compiler.
However, the definition execution and syntax directed translator
can handle this expectation.
o
11
Sf''' ~.._~,4.RY
-...-.._
.....-....
o
.
In this paper. I have tried to explain the concept of a language definition
based on the techniques of syntax
directe~
'~mpiling
and translation.
Whilst
a standard definition of a computer language may be incomprehensible to the
majority of the programming community of today, those who are concerned with
implementing systems conforming to the standards will be members of the
minority (by definition).
A definition executer as described herein will not
only define the syntax and semantics of a language but also outline the translatory techniques of a compiler.
In comparison. a building code contains
parallel partitions whereby the design criteria
~
techniques of construction
are specified.
Programming languages are intended to be languages of precision.
0,
,Ii,
are relying currently on
langu~ges
Yet we
of imprecision to specify the syntax and
semantics of those same programming languages o
If for no other reason than for
consistency and perhaps to maintain the image of the industry, the formal
definition of language will repl&c.
o
~he
current prose definitions.
~--
,
.....
----
t
t
Concrete
Text
Abstract
Syntax
I
•
&.--- ........ _ ... '
IlL
~
~
Translator
Concrete
Syntax
~
, Abstract
Text
'
,
t
~~
,
f
Data
Sets
•-- ....
_
...
~
-
I
,"'" - - -.
.... ..aa _ _
--,,
I
-- - - v- -- •
o
--
,
~
...,
~
Interpreter
..",
Storage
I
......
t
,
I Informatj on •
_-_ ... ,
,varies with
,source prograit
..
o
Processors
~nformation
lis dependent
~n language
IBM Definition Schematic
(Figo 1)
--,
o
I
t
-~
8Ml
declarative
source syntax
,--
I
I
source program
and data tree
....-.. -.....
~-.....
b;J
,
.....
~r---i-m-p-e-r-a-t-iv-e------~
l.::l ~
'\- .. - --'
translator
canonic program
and data tree
-
8.
Ml
- - -'.
'.'
canonic pro gr am
and data text
,
.... -
.-
Key:
-_ _... __..... ..
,
,
~
.
-._._--_..... -
.......
-
.....
I
____
cOo
...........
,
~--- .
~
- -
Information varies
with source program
---
~
"output" t
or
• "results" •
---------------------,
'1
rid
-.----
declarative
canonic system
'-----
~--
imperative
executer program
~~--
,,
SM - Synthesizer Machine
AM - Analyzer Machine
EM - Executer Hachine
(These are ffxed over all
lan2Ua2es)
Information varies
with language being
defin~d.
DIAGRAM fOR THE X304o2C2 DEFINITION EXECUTER
o
(Figo 2)
2'-17
o
p'
Alphabetic Recognition Tree
(Figo 3)
REFERENCES
o
o
o
10
Naur. p~ & Woodger. Me (Eds)~ Revised report on the algorithmic language
ALGOL 60. Comm~ ACM V6 pp l-20~
20
Anon~,
3c
Knuth. Do E~,
pp 611-618
FORTRAN vs Basic FORTRAN. Comm. ACM V7 1)1' 590-625.
The Remaining Trouble S1)ots in ALGOL 60. Commo ACM VlO
0
4~
Abrahams, n~ et ale Report of Ad Hoc Task Group on Formal Definition,
Private Communication to X304.2C2, May 18, 1966.
50
Folkoff. Ae D~ et al.
Systems Journal #3.
A Formal Description of System/360,
IBM
1964~,
6~
Lee, J. A. N. The Anatomy of a Compiler. Reinhold Publishing Company,
New York, 1967 Appendix C.
7~
Christensen. C~ Examples of Symbol Manipulation in the AMBIT
Language. Proc e 20th Natl~ Confe ACM, Cleveland, 1965.
Programm1n~
.. -.-.-...-.-.-..--.-~-~-
.~-~~----
A "SMALL" OS SYSTEM
/
As Described to Wade Ao Norton (1125)
By Max L. Allen (1164)
o
Rust Engineering (1164) of Birmingham, Alabama, and Pittsburgh,
Pennsylvania, a company which specializes in engineering design and
construction, has recently become a division of Litton Industries.
Their work includes (1) Design and construction of process
mills--paper mills being a good example, (2) Renovation of office
buildings, (3) Maintenance and construction work for government, and
(4) Design work for governm~nt, including sow£ launch facilities at Cape
Kennedy.
Rust has offices at Calhoun, Tennessee; Montreal, Quebec,
Canada; Vancouver, B. C., Canada; Brussels, Belgium; and Mexico City,
Mexico--in addition to their principal offices in Birmingham and Pittsburgh.
Rust employs about eight hundred people each in the Birmingham
and Pittsburgh offices; and, even though the mix in the jobstream is
quite different at the two installations, they are able to generate a
single OS system to meet the requirements at both places.
In Birmingham the mix is approximately 65%-35% engineering
vs. business data processing, "{,Thereas in Pittsburgh the ratio is about
15% to 85% (in favor of co~mercial work). Some of the work in Birmingham
--quite properly classified as engineering--is engineering management
information, rather than design or analysis.
()
In Birmingham it is possible to schedule very little of the
work because report periods vary for different contracts. Near the end
of a report period, they likely are making several runs with their
management info tools.
In Pittsburgh one of their big jobs is labor costs, which is
often run several times a day.
However, staffing is not too different in the two localities,
with each employing about ten or twelve analysts and programmers. In
Birmingham they have two machine operators; in Pittsburgh, four for two
shifts, therefore .employing two per shift just as in Birmingham.
On the other hand, there is a distinct difference in the number
of keypunch operators--due to the difference in the amount of input
data. In Birmingham three keypunch operators are employed by the computer'
center, whereas in Pittsburgh (where the mix is 15%-85% commercial) the
accounting department has ten to twelve keypunch operators on its payroll.
o
You may ask--and properly--why I present this paper. I am
the first to agree that the real author is Max Allen of Rust's Birmingham
computer center.
Unfortunately, he has been unable to make either the Cincinnati
meeting or this one. In Cincinnati this paper was presented by title
to the OS Committee. That group agreed that the subject matter is far
more appropriate to the full S/360 Project. Thus it appears on the
agenda for today.
Rust's computer department exhibits a certain rugged individualistic spirit--of just the sort we want displayed by COMMONers--to dig
into the uncharted and build there something of real value.
An~, having mapped the previously unknown, they have demonstrated to the fullest a second trait we wish of all COMMONdom--the
willingness to share with others their knowledge, their system, and,
indeed, test time.
o
I can truthfully say that we would most likely be in some stage
other than the final one of conversion from a 1620 system to S/360-40 OS,
had it not been for Rust's pioneering of OS and their willingness to
share with us the fruits of a then hard-won prize. (My only reluctance
at making the last remark is that it might be construed in some quarters
as implying that this paper is in payment of a moral debt. That is not
the case! At no time has Rust ever implied that somebody owed them
anything; furthermore, this paper was requested by your former OS
Committee chairman, Mrs. Barbara F. Young.)
And certainly, Rust's work deserves reporting--OS is for the
MOD 30 and MOD 40, and both COMMON and IBM need to be told this fact.
The remainder of this paper fills in the details as told to
me by Max Allen of the Rust computer center (and observed by me in the
course of Southern Services' conversion efforts)o
Max's position at Rust is as the Number One man on the systems
and applications side of the activity, and in the absence of the manager
he assumes these functions, also.
This manager, who was sweating the dollars and cents of OS
pioneering while Max perspired over the bytes and bits of the work, is
Bill Mylius. Bill is in charge of computer facilities at both Birmingham
and Pittsburgh. And, while this story is basically one of implementation
and told by Max, the policies reported in the telling of the story are
Bill's.
o
Rust tried DOS and it did not give the flexibility they needed.
This was especially true with respect to structuring of programs larger
than one core load. CALL LINK under DOS FORTRAN helps, but it does not
solve things. In reality it creates greater rigidity.
If
.
-3Another initial consideration in the choice of as is that the
froject Management program and ICES (MIT) were both written to run under
OS. Later developments of ICES push it out o"f 64k and only turther
justify the choice.
o
Core requirements for OS of slightly less than 16k were evaluated against both the 6k DOS requirement and the 10k needed for DOS-2
with fixed scheduler partitions. It was decided, Max said, that the
additional core devoted to OS more than paid for its dedication in a
more powerful system, "particularly since the linkage editor is so
flexible in how one may structure his program."
Still another item which influenced Rust's decision at the
time was that they were enthused over the prospects of using fullblown
PLI. Their experience to date has been that it is clumsy in both compile
time and run time, but very easy for the programmer. At the time
(November 1966), PLI was not supported under DOS. It is known that PLI
has been and is continuing to be improved. That as yet they have not
gone to it as their primary language has not caused them to regret their
choice of OS.
In fact, they would make the same choice today--OS over DOS-which they made more than a year ago. (And, I might add, today they
would have a lot more support from IBM in its implementation.)
Rust considers the following as disadvantages of as or, more
properly, as part of the price paid for the more powerful system:
as
1.
There is lots more to learn about
system for driving a 360.
than about any other
2.
OS can do lots more, but this flexibility also creates problems
of choice.
3.
To applications programmers, JCL presents great problems.
There are no clear-cut logical patterns to followe.
4.
Operator requirements are not easy for any method of driving
S/360. It is a complicated machine. But with as, the training
of operators is considerably more difficult.
5.
Rust paid the price,also, (in November 1966) which IBM's lack of
field knowledge about OS cost. IBM Birmingham was reluctant
to send their p~op1e to school. Rust's CE worked on as when
he had nothing else to do, and APAR anSvlers were such as, HIt
looks like you got a bad tape." Rust started to SYSGEN on
November 8, 1966, with Release 6. They obtained their first
running version on December 23, 1966, with Release 70 In
between there was much effort expended--mostly by Rust peop1e-before IBM finally imported a man from outside the local office.
o
o
-4Actually IBM did not really begin to support OS in Birmingham
until two Mod 40 insta11ations--A1abama Power Company and
Liberty National Life Insurance Company--were being imp1emented o
(Southern Services, Inc o , (1125) whom your speaker represents,
has merged its operations with Alabama Power, an affiliated
company.) It turned out that much of this trouble, which took
more than a month to find, was caused by the Linkage Editor
which was losing a CSECT out of the nucleus.
o
6.
When Max went to OS school in Cincinnati, there was not a single
course offered by the New Orleans region. Because of the aerospace industry, some training ~n OS had been offered in Huntsville,
but this was not under region sponsorship. In Max's opinion,
the course at Cincinnati was a good course. Since that time,
IBM has been schooling their people regularly and offering
courses on a broader geographical basis. The know-how is
available now; the marketing emphasis for as with the Mod 30
and Mod 40 is lacking. In both Max's and my opinion, this is
something IBM should correct.
Among the things they bought were
l.Improved Job Management.
o
2.
Better Data Management.
3.
More powerful Linkage Editor.
4.
Identical systems for Birmingham and Pittsburgh even though
the hardware and addressing differed slightly. In Birmingham
their punch is a 1442 addressed at OOA, while in Pittsburgh it
is a 2540 addressed at OOD.
5.
Under OS, Rust has found that it is almost impossible for an
operator to lose control of as by a misstep at the console.
6.
Combining items 2 and 3 has made it possible to catalog programs
and data and put them on disks for exchange between Birmingham
and Pittsburgh.
7.
Without the power of as to put programs elsewhere than on the
system residence volume, Rust's programs would have required
multiple syste~ residence packs a 1a the 1620.
Rust did not run extensive timing tests, because they were
not felt necessary in weighing OS against DOS. Enough was done (with
programs which constituted a large part of the workload) to bear out
the already known generality that os FORTRAN(E) compile-link and compi1elink-go were both superior to their DOS equivalents. They used only
.
FORTRAN because this was the only language in which they felt sufficiently
o
... 5-
proficient at the time. They have since used COBOL some, too, while
they wait for a better-performing PLI. They ·checked compile-link using
a Plant Material Thermal Balance program, which is their biggest job.
They found that OS did the job in 8 minutes compared to 30 minutes for
DOS. Using a Steel Stack Design program, they found a compile-link-go
advantage of 22 percent while using as.
o
When they started, all they wanted was enough system to do
some testing (and it was the hardest to come by). In the interval
December 1966 to December 1967, Rust generated five releases. SYSGEN
is now easily done in a manner that suits Rust's needs better than the
cookbook method described in the SYSGEN manual. However, this is stated
as a plus for Rust's experience and not as a criticism of followi.ng the
book--which procedure will get your as for youo
You can have a basic system for about 14k of core storage.
This will include PCP (the sequential scheduler with all options transient instead of resident), support for a reader, printer, punch, one
disk (for system), and four tapes. It would include the access methods
BSAM, QSAM, and enough BDAM to access programs and compilers. It would
not include a timer J indexed sequential access methods, nor direct access
methods for applications data setso
Rust's system requires 3F0816 bytes (or 16,136 10 ). To be
exact, it is slightly more than 16M bytes, but slightly less than 16K
bytes where K = 1024 = 2 10
In this amount of storage, Rust provides
for PCP, interval timer, three disks J no tapes, two.punches (as described
above), a 2501 card reader, and a 1403 printer o They have the standard
access methods QSAM, BSAM, BPAM, BISAM, QISAM, and BDAMo They have
resident SVC entries and trace table entries. If IBM ever supports just
a plain timer, Rust will take it over the interval timer and thereby
save a little more core.
0
The full direct access method, the indexed sequential access
methods, and the trace table entries were included, because Rust believes
they buy enough in supporting FE work to justify their inclusion.
At the moment, Rust can do the work required of their system.
But, looking to the future, they see more core needed, greater processor
speed, and more devices to support. And they look upon their experience
with OS as a great boon to future growth. A possible first step might
be the addition of a second selector channel, if spooling under MFT-2
is shown to give reas~nable performance at the 65k level.
Presented to the 360 Project
at the San Francisco Meeting
by Wade A. Norton, COMMON 1125
~
. "
H
bi'l'YNit"d!i' 't" "T' "\
' '!lJte'
'Lf 'f'UItl»WfiWWb
b"!
tfP'It",w"Hr't" wt .
H"tW1'.'"Wu"flee&NHm"Hi
'II t ·
MW; etmLn/"b"'WHd* . '+
~.
/
.
I.
7
(Q/
o
o
1)
Working title of the paper:
The 1620 as a Data Collector or
Software, the Crutch Hardware Boys Lean On
2)
Author:
3)
Address and User code:
4)
Position:
5)
Time requirements:
30 minutes
6)
Special Equinment~
(lantern) slides.
Projector for 3 1/4 by 4 inch glass mounted
7)
Level:
8)
Machine: 1620 II was used, but the presentation will be for
general information.
9)
Abstract: The paper will cover the data collection methods
used at the Sacramento Peak Observatory. The principle reasons
for designing a computerized data collection system were:
1) Improve existing methods which used summary ounches.
2) Gain experience and insight into oroblems associated with
data collection in nrenaration for third gener~tion equipment.
Guy A. Gallaway
Sacramento Peak Observatory
Sunspot, New Mexico
88349
User #5053
Programmer
.......... .
~~. . . .r~
~
Intermediate
The paner will also cov~r the roles ulayed bv the prograMming
staff in this type of oneration. Namelv, software as a dia?nostic
tool for the hardware boys in developing data reduction/collection
instruments, and software as a useful and flexible tool for the
research scientist.
Three methods for data collection and reduction bv comnuter have
been tried.
o
1)
On line, one point at a time,
testing each point for
parity and structural errors and storing it on the disk
before the next point is generated.
2)
On line, a record at a time, locking the source device
out until the data has been tested and stored.
3)
Off line, using 7 track incremental tape recorders, with
testing and reduction heing done during slack times.
/
SESSION T.l.7!
USER #
COMPANY REPRESENTED
& ADDRESS
652-2694
Monsanto Co. ,
Decatur, Ala.
(314)
MAl-4000
Monsanto Co. , 1700
2nd st., st. Louis;
Mo. 63177
285-6655
Hooker Chemical Corp.
Niagara Falls, N.Y.
(215)
KI5-7500
Catalytic Const. Co
1528 Walnut
Philadelphia, Pa
(815 )
895-5194
Anaconda Wire & Cable
N. Cross street
Sycamore, Ill. 60178
S. Barr
(212)
571-3977
Western Electric
222 Broadway
New York
N. Kurek
(213 )
341-3010
Canoga Electronics
Los Angeles
D. Adams
(415)
961-1111
Ames Research Center
Moffett Field, Ca
A. J. Bill)S
(805 )
296-6681
Ryan Aero
2701 Harbor Drive
San Diego
R. E. Buckley
G. F. Schoditsch
3438
George Polyzoides
Theo. E. Bridge
E. J. Uczen
'I
c
PHONE
NAME
j'
),.
f'
1100
0
SESSION T.l.7
v
COMPANY NAME
& ADDRESS
(213 )
WE3-7356
Human Factors
Research
1112 Crenshaw Blvd.
Los Angeles, Calif.
(412)
621-3500
Ext.7138
Crystallography Lab.
Unive of Pittsburg
Pittsburg, Pa 15213
R. L. Ferral
(916)
442-1201
USWB Sacto RFC
Sacramento, Calif.
J. W. Wheeler
(919)
199-2311
Ext.357
Western Electric Co.
801 Merritt Drive
Greensboro,N.C.
Michael Robert
Hale
R. D. Rosenstein
o
USER #
PHONE
NAME
L. V. Punder
Joseph Sloboda
1117
Biological Station
Nanaimo
British Columbia
Canada
7072
(313 )
238-1631
Ext. 450
Flint Community
Junior College
Flint, Michigan 48503
~5?
-_._------_.
__.._ - - - -
o
1620 DATA COLLECTION
or
SOFTWARE, A CRITrCH TUAT HARIMARE BOYS LEAN ON
Guy A. Ga 11away
SaGramento Peak Observatory
Air Force Cambridge Research Laboratories
Sunspot, New Mexico
88349
o
Presentation at the Winter Meeting of
COMMON
in San Francisco, California, 12 December 1967
o
1620
DATA
COLLECTION
(or, Software, a Crutch That Hardware Boys Lean On.)
This paper will describe, in varying degrees of detail, three
different methods of data collection. Each of these methods was
developed and used at the Sacramento Peak Observatory, Sunspot, New
Mexico. The Observatory is part of Air Force Cambridge Research
Laboratories and performs basic research in solar physics. Two of
the methods were on-line to a computer and the third, still in use,
used an incremental tape recorder as intermediate storage.
The overall goal of the project was to experiment and compare
the use of real-time collection with remote collection on an intermediate device, such as mag tapes. We originally intended to continue
this experimental approach with our third generation equipment before
committing ourselves to either method. But now it looks like we will
be doing a combination of both.
To this end we felt we should implement these systems on an
established machine. In this way both the hardware and software
people could, and did, gain experience and insight into the problems
involved.
The immediate goals were to develop a more reliable system and
establish a complete operating system. An earlier method made use
of a summary punch and was very unreliable. We also wanted to try
and develop some systematic procedures for handling this type of
data, and to have available a standard set of reduction/utility
routines for processing the data. But this approach usually meets
with opposition from the individual scientist who wants a "tailormade" system for his project, the definition of which will usually
change every few months. Since we are planning on implementing
the same devices on newer machines, this attitude means that we are
faced with the propsect of never-ending systems development. We
have found that after the initial development an equal amount of
time will be spent in making changes to meet specific desires.
The goals of the system were always being jeopardized by
frequent modifications in "mid- stream" . From the time we started
to implement data collection, up until now (last week) there have
been continuous changes to the system. These cover the full range
from the data format to reduction procedures. None of the systems
were developed as a unit and then examined and evaluated. This
lack of a commitment to full development has been demonstrated to
waste considerable manpower and machine time.
The source of the data was photographic films, typically
filtergrams and spectrograms, of various solar phenomena. We
use a scanning microphotometer to read the films. Film densities
are converted to a 3-digit decimal number and sent to the computer
.----~-~
.~--.--,,-
..
--..- - - - - - - - - - -
2
I,
or the tape recorder. The computer used was a full 1620 II with
binary features. The two on-line systems employed the paper tape
channel for data transfer. We have an on-line plotter, but no
conflict occurs since the plotter uses the write paper tape instruction and collection uses the read paper tape instruction.
Termination of the data was done by sending a record mark on the
End of Line (EOL) line.
SYSTEM
tftl
The first method was the only one requiring a real time
response. The characteristics of this system were:
1.
2.
3.
4.
A 1-2-2'-4 BCD format.
One data point at a time transmission.
Packed data.
Termination of the data string by a counter control.
This strange BCD form, 1-2-2'-4, was a "left-over" from the
previous system that used a pulse counter with this format. This
added the problem of encoding the data to the normal data handling
problems. This was the first instance of a hardware problem for
which the software provided the solution. Figure 1 defines the
formats used and shows how the paper tape lines were used. A
minimum of 20% of data handling time was devoted to the encoding.
The data was stored on the disk in one sector blocks of 31 points.
Only every other point was flagged making the data appear as 16
6-digit words.
(123123123123)
The coice of two points per word of data was made to achieve
minimum disk storage and still insure disk compatability with
Fortran. That is, given a fixed point word size of six, and the
array IDATA dimensioned at 16, the statements, RECORD (J) IDATA
and FETCH (J) I DATA , will reference one sector of 96 digits.
The unpacking of the data was done in a Fortran subroutine.
The maximum data rate, 20 points per second, left ample time
available for the programming required to handle each point. Total
prograrrnning time requirements using the longest encoding routine,
worst case, took only 2640 microseconds. The greatest time component
was disk revolution time, worst case conditions being 44 mils.
Figure 2 gives a detailed description of the timing involved.
Originally there was to be an ID number sent as the first data
point. But in order to speed up the process of implementing the
system we used the ID number as an input argument between the mainline Fortran program and the collection program. This ID number was
then the first point of each disk block, making each block 32 points
long.
Termination was to be done three ways. Two of these were
similar, both using sense switches. The setting of one switch
would indicate that the collection had been completed normally and
reduction should begin. The other switch would indicate an operator
error and for the current collection to cease and re-start. The
third method, and the only one implemented, was by having the user
specify in his Fortran program the number of points to be collected.
This number was another input argument for the collection routine.
What turned out to be the most significant advantage in getting
the system going was the "off-the-shelf" hardware. Our engineers
are" time shared!! wi th the whole Observatory. They are responsible
for building and maintaining a very large number and variety of
complex equipment. And as frequently happens under such circumstances the person that yells the loudest gets his work done the
fastest. So by having working hardware we were able to complete
the system sooner than if the hardware had to be built.
The disadvantage of this system was its slowness - not only
the data rate but the over-all system. Considerable time was
required in the Fortran programs to get the data from the disk and
unpack it. This, however, is a familiar story to programmers,
minimizing storage at the expense of execution time.
One problem in the early stage of development was the 1620.
We had learned from IBM that each character signal should be maintained 10+ microseconds. The first try was a signal of l2~, which
got "lost" in the CPU. A diligent pursuit of the signal with
scope and probes convinced the EE's of the soundness of their design. The data left their black-box and entered the computer. To
them it was unfortunate that the program, or programmers, couldn't
find the data. Even elaborate overlay and dump routines failed to
completely convince them that the data wasn't reaching core.
Finally, the extension of the signal to 2~'s maintained the characters long enough for them to reach core. We were then able to
collect data.
SYSTEM 4f:2
The second on-line system used new hardware and software.
The new hardware was designed and built by our own engineers. We
wrote new but similar software. This system had several entirely
new features ~
1.
2•
3.
4.
5.
o
BCD compatability.
Transmission of records of data.
A source generated 6-digit ID.
Variable data rates.
Starting the source device with the computer.
BCD compatability solved the encoding problem and cut down on
the required programming.
d.tal.
4
By allowing the transmission of a record of data we had to
limit the total number of points collected in each record. We
decided to use 1/2 of the available core, 24,000 digits. This was
equal to 4000 data points three digits long, each point being
converted to a six digit fixed point numher. Conversion to the
six digit size was done in the collection program. Considerable
time was saved by eliminating the slow unpacking routines in
Fortran. We still limited the size of the data blocks on the disk
to one sector of 16 points.
o
The hardware generated ID number was an aid to the file
management problem. It was to be the first six digits of the data
string. The number would be chosen by the user and selected on
thumb wheel switches, and read out of a register.
Also, a control was installed to allow the user to digitize
at a variety of rates from 1 to 100 points per second.
The collection of variable length records required the "locking"
out of the source device. This was to be done by allowing the
device to send only upqn receipt of a computer generated signal.
The Exponent Overflow indicator was chosen as this trigger. This
would require turning the Exponent Check light, EXP CK, on for 5
mils immediately preceeding the read instruction.
The software was quickly written and debugged due to the
similarity with the previous system. The construction of the
hardware took considerably longer, mostly due to the manpower
problems mentioned before. And before the system was operational
there were more than a few changes in definition.
()
The first problem was the EXP CK light. The 5 mil signal was
not long enough to turn on the data source. A testing routine was
quickly written to vary the time and 250 mils was determined to be
the shortest reliable length.
The next problem was the ID number. The ID register read out
too slowly. For any data rate greater than 20 points/second the
first few data points would be lost. Designing a faster read-out
didn't completely solve the problem, so the ID number was moved to
the end of the data string, the last six digits.
Moving the ID solved the problem of losing points but caused
another problem, which is still in the system. For some undetermined
reason the ID number is frequently, in half or more of the cases,
seven digits long. In the majority of these the extra character
appears to be the seventh, last, digit. But there are many times
when the first and second digits are the same J both flagged, making
the first character appear to be the extra one. But with either
case there are six good digits which can be selected, by programming
for the ID.
o
')
/
.,
5
Here were two hardware problems, ID generation, and device
control which required a software solution for implementation.
In the early development of this system there were random
parity problems. Several test programs were written to display
these characters and examine them with the binary instructions.
There were also structural errors in the data, that is, one of the
three digits would be dropped. The design was changed and these
problems reduced. (They still occur, infrequently, and are a good
indication of some temporary problem in the hardware.)
We were still having problems with the EXP CK trigger and
decided to use the computer logic of the read instruction. With
the able help of our IBM CE this was done, and with considerably
less effort than expected.
SYSTEM 1ft3
This was now compatible with tape recorder design and the
recorder was hooked into the system.
o
The person collecting data could now work at any time, tE ing
the computer or the tape recorder. This was, operationally, a
better policy since the computer wasn't tied up for long periods
collecting data. The tapes were reduced in the evenings and during
the day when the computer wasn't in use. The same software handled
the data tapes by using the read paper tape instruction.
With the installation of the tape recorder we were also able
to reduce binary tapes from a different source. By this time we
had learned not to write a full software package before the hardware
had evolved into its final form and was debugged. Two short routines
were enough to reveal various failures in the hardware. These
problems were similar to the other system, lost points and digits,
and a new one, a data point in the middle of the ID information.
USE OF THESE SYSTEMS
The second system was modified for one user to eliminate the
use of the disk. He collected a constant 500 points. The collection
program was changed to format the points and place them directly
in the COMMON area of Fortran. This was a considerable time-saver
for his application and he was able to collect about 7 million
points in one month. This was the only production work done using
either system.
Most users wanted their data plotted as soon as possible to
determine if they were operating the equipment correctly. Since
the Fortran plotting was pretty slow we wrote a short machine language
program to read the data and plot it directly from core. As the data
6
is defined as ranging from O~999 no scaling was required. The
result was plotting done four times faster than in Fortran. We
were now inspired to try and hook an oscilloscope onto the computer
for even faster plotting. We even went so far as to develop programs to convert the plot commands into a continuous string of
digits corresponding to the pen commands. In this way plotting
was done by dumping core, with a pen command coming every lO~ .
Manpower shortages prevented the hardware boys installing the
scope.
The amount of time required in developing utility and debug
routines for the hardware was greatly underestimated. Until the
hardware worked, a programmer had to be available every time the
hardware was tested. One four hour session resulted in the
writing of five short machine language programs. In addition
an immeasurable number of dump routines were composed and executed
at the console typewriter. I do not mean to cast aspersions on
the hardware fellows. Their basic design has proven very reliable.
It was a lot of furuand very satisfying, to build the system into
a usable tool. The point is, both the hardware and software people
must work closely together from the very beginning. It is not
unreasonable to expect the consulting of some computer oriented
person when designing any data system, since the data will eventually
reach a computer.
Our error was in planning on just writing the software package
and not budgeting for time to write the programs to debug the
hardware. They could have done it with scopes and a bread-board,
but a more reliable "unit" is developed when both sides are involved in the operation.
Of the three methods we used, the best, in my opl.nl.on, was
clearly the use of the tapes. The procedures the user goes through
to set up the source device are usually very time consuming. And
he will often be reducing more than one type of data at a time,
requiring additional set-up time. When the computer was on-line
it would often be used less than 50% of this setMup time and
never used fully. Production runs could often be done only in large
blocks of time, four or more hours. This forced the user to use the
computer at night. And more importantly, the on-line system
required someone familiar with the machine and programs to be
available at all times.
(Typical reason for this was the user would
send too much data and clobber the program, but he still wanted to
save the data in core. This and similar problems required the
presence of a programmer during data collection.)
The advantages of the tape system were mostly operational.
The user was only competing with other microphotometer users, 1 or
2 people at most and frequently no one, instead of many computer
users. The scheduling for the microphotometer was done by the
day instead of by the hour as for the computer. The user nee~ed
only to know how to run one machine instead of two. And most
----_ .. _..----_. ---_
... _-------_. __ ..
c
",.muwn
. ?tIWWf!!!
7
c
importantly, more efficient use was made of the computer. During
the set-up period the user could use the computer in 5 to 15 minute
blocks as required to test his data and operating procedures.
During production runs the bad records can be identified in advance
so the programs will ignore them, a considerable time saver. Also
the reduction of the tapes can be done by a competent computer
operator, with the special cases (extra long records) being handled
by special programs, without loss of data.
These reasons would be less important with a machine capable
of handling foreground/backgrourid programs, but cannot be entirely
dismissed. Generally speaking, the data would still be placed on
an intermediate device such as a disk or tape before reduction.
Under such circumstances careful consideration must be given to
each method. Cost is not a small factor. For a few collection
operations the extra cost of a computer hardware to do the job
would be proportionally higher, when compared to individual devices
and controllers at each site. But for a large number of such
applications, which can be mutliplexed, the cost is reduced for
each device, when done with a computer.
o
Another important consideration is site locations. Transmitting
data at high rates and control of collection devices over large
distances is still pretty expensive. This tends to force the collection instruments to be located near the computer. When this is
impossible then the on-site system using an intermediate device is
one solution.
One way of evaluating a system is by asking, "Would you do it
again?". There is no doubt that we would, and gladly. The one
production run mentioned before, with 7 million points collected,
would have been impossible to do under the previous system, in less
than four months, if at all.
We now know, more clearly, what to expect of the basic collection
routines, what to expect them to do and what not to have them do.
Our new computing equipment will be in use a minimum of five years,
probably much longer. Those of us in programming are desperately
hoping that the research scientists will completely develop a
working system before changes are even discussed. The single biggest
factor in extending the development time of these systems was changes
in the basic definition of what the system would be.
The only advice I would offer to anyone considering doing this
type of work with a General Purpose machine is to see it through to
the end, and then evaluate the whole system. Even if you have
unlimited manpower and time to continue re-doing the same thing, it
is much easier to work with a system which has a few, well-defined,
operational peculiarities, than to re-educate every user every time
he uses the system.
o
Sacramento Peak Observatory
Data Collecti9n System #1
Source
Paper Tape
Conversion
EL
4
,
Check .
~'~
(flag)
X
,,,
.
0
,
.J..
"
2 - 2
"
1 -
• 0
, Check
.,
0 270',.l
3 - 3 "
6 - 4 4401-1
8
7
-
5 3801-1
4 - 6
1
,.
.L
26011
5 - 7
"
o - 8
"
"
I"
7- 9
1320;..
Worst case point 3*440
~{orst
case Prograrr.ming
•
1320~
2640~
o
Digits
0
1
2
3
4
5
6
7
8
9
~:o::-~lal
BCD
Re7.)resenta:lon
1 2 2' 4.
Represen-::al:lOn
1 2 4 8
--"'1"--
1
0 0
,.J.. 0
0 .L
,
+ 1
0 0
1 0
0 0 0
1 0 0
0 1 0
1 1 0
0 1..... ....
1 1 1
0 01
1 0 1
,
0 .J.. 1
,
1 1 .J..
0
0
0
0
, 0
..11 0
C
1 0
1 1 1 0
0 0 0 1
1 0 0 1
~-
0
0
0
0
2
l.'i
,
0
0
0
0
0
0
1
1
1
1
Pauer
Ta~eSChannel
.
:1PU,cllcme
1 2
i
0
1
0
1
0
1
0
1
0
1
0 0
0 0
0 0
0
0
1
1
1
1
0
0
1
1
Read As:
F
0 0
,
..10
1 0
1 1
1 1
1 1
1 1
0
1
2
3
6
7
7;
"5
6
7"
o
f#riit±.w '(Iff
T"fT 1i"""!"If"TTf'"'IWWrm "lfHi'#t?tffl
f
!
I! -~"..'-
0
b
f
I
~
I
I
~
/\
fJ
o
\..::::I'S.Y"
0:~/
;--
~j:-J
:::L
l
i-
t
II
I
0
~
1O
N
E
,t:
~
!
r,.. .
~
.
,j
-~~~
-
~~~:~J
~
~
.-:;:
--,-,-- " -~',..-------'------------';'.-.
\
>-._ _ _ _ .\1
~
~
71~-----
1
.....s. :
~l~_--____
Jj ______
1
~
~
I
-<
<
a
t-
-Ii!
Counter Start Pulse
------------
[
=--]=
~---.~-""""-~~.
DIGITIZF -'.
-CIDCK
_ _ '"" ••T ...... ~~ .......... "- _ _ _ _,-~~· ...
V_"
'''=~C;;~~:;~~~~}-count
COUNTER
er
_<
,.a" .. "t...-
S to p Pu Is
e ,'L:
r"-"·-;;;;;fl'~;~~~·-·-]··· . .
------------.•- - - - - - -•.
._;r·-·-·~-'
SYNC
GENERl~.TOR
[
. ·Imrrll~--···-·
,iii.
r-"---L----··_·L
l~II~~~~II~~~~s········ .. L A~:~~~ER
in 1,2,4,8 Code
J---
r-·-·--·--·-------
PAPER TAPE
J .--------<,'.--------------..
"-,,~--~-~~---~'~."~'
--rJ
INPUT TO 1620 II
--_..__._._._ ...._.-._.
1620 II
[-'~~E~J~
-RJ
------_._-------_u. _______
.---------------
MAIN FRANE
~t··H
'''-.
~
co
t)
A
~~)
o
i
,7 1/
.' ( ..... i
Catalytic Construction Company
Walnut St., Phila., Pat 19102
~528
- Programs for Inverting a Large Matrix on a Small .~1achine
by
T. E. Bridge
In figure 1, we have plotted, time to invert a matrix, versus the size of
the matrix. We have plotted results for two programs, one using partitioning, and
the other using sequential files. The same form of test matri~ was used in all
cases. Double precision arithmetic (17 digits) was used,
0
Size of Matrix
25
60
120
Partitioning Method (Table 2)
Time Minutes
RMS Error %
2
0,000
19
.356
115
1.252
Sequential Files (Table 3)
Time Minutes
RMS Error %
3
29
0.000
213
0,000
29
0.000
231
0.001
Direct Access Files (Tahle 6)
Time Minutes
ro1S Error %
4
0.000
In this paper, we are presenting three
Table 2
Table 3
Table 6
pro~rams:
To invert a matrix by partitioning the maximum size is 3520
To invert a matrix using sequential files. The maximum size is 300
To invert a matrix stored on disk. The maximum size is 300.
Two other programs
also given:
~hat
were used in testing the above three programs are
Table 4
A program to build a large matrix for testing.
Table 5
A program to multiply the test matrix by its invert, and then
compare each element with the corresponding element in the
unity matrix. The RMS error is calculated and published.
o
[over please]
0
_._-_.._----------_._--------"._--_._----- - - - - - - - - - - - - - - - - - - - - - - -
- 2 -
o
Description of Partitioning as-Used by 'the 'Program in Table 2
The following formulas, that we use in Table 2 to partition the matrix,
are given in "Linear Algebra", by G. Hadley -- Addison-Weekly Publishing Company,
1961 -- on page 109.
A large matrix may be partitioned into four smaller matrices. Let-a, b, c, d -- represent four small matrices arranged like this to form a large
matrix:
b
d
a
c
Note that a and d are square matrices; while band c may not be.
Let the invert of the above matrix be represented by:
Let:
e
= the
A
B
C
D
invert of do
The following equations may be used to find each partitioned piece of
the invert:
A
(a - bec)
B = -Abe
C = -ecA
D =e
ecB
:I
0
-1
-
The program given in Table 3 uses these equations to find the invert of
a large matrix. We had only 5K bytes of storage -- 629 double words -- left over
with the partitioning program in core. Since 100 words are required for data
handling, we find that the large matrix must be partitioned down to a -- 23 x 23
before it can be inverted. So, four levels of partitioning would be required to
invert a -- 352 x 352 -- matrix.
The levels of partitioning required are:
Matrix size
Levels required
23
46
92
a
I
2
184
3
352
4
We found that the minimum amount of partitioning gave the shortest amount
of run time when using this progrma. In other words, it took 20% longer to invert
a 100 x 100 matrix using 4 levels rather than the required 2 levels of partitioning.
We don't really know what would be the optimum amount of partitioning on a larger
machine.
o
cQ?a.
CLEARPRINT PAPER CO.
7
8 9
TCL:~.7'R/"
10
f-,"TI
No. T323. LOGARITHMIC: 2', BY 2 3',·INCH CYCLES
3
2
---.;.....-"._-
4
·1····'·
........-.----.......---
III
1'"
.I
: I i I
-+=\-, I _.-!
. . 1....
!
+····:·--t-"-t
!
1
I
!
I---I'--r-+
: . .t '- ... -+.. " l
l
I
•
I
i
i' ~
I
I
l
I
l
!
I~~TI~
1A91
f
.
I
,
'I
-~ -... ". -' ti . . . .• .. -1 ."- -.. t .... t.
j
!' I . I
I
I
f
=---t--=;~
4
r·I
I
I
"" I
1
............ - - - - - + -
I
I
II! ,
' I
~
I I
1
I
;,
I
-+-
---1- ---.. --~. --.--.. -~.--~--. +-- I
~=R=-'~ --'~J--~r! I! I--±+t+~~~~3
5
6
7
8
9
/
2
3
4
-L.__
I ..•
9
T-J--f-+-+---- . _. _ ._.____
Icujm; L_1 uLJ-!
t·
1
mwlili
O
5
PRfNTE.D IN U.S.A. ON CLEARPRINT TECHNICAL PAPER NO. IOOOH
I
. _ _ _ _ .• __ . _ _ .
~ _ _ _ _..
Ii . --I-t-;
f
I
~
,---~=r+-++-Lt
- t··
f
i
..•.....
,1
f
; .... 'r
i
1
'
t
I
!
I
-- .. ---~".-------.--4-
!
i
, .f
- -+_.. _._-+ ..._--+..
.1
_____....___.Li
..,..
""
i'
...
, 50'
.h-r:
'WI"
J
TABLE
1
LISTING OF PROGRAM
TliE FOLLOWING PRUGRAM IS INCLUDED ONLY TO TEACH THE METHUD OF OPERATION
IN IH~ PROGRAMS LISTED IN TA~LES 2 AND 3
US~D
C
THIS PRUGkAM WIll SOLVE
C
\11
N
= i''i + 1
= i, N
= X(K,K)
F
DO 1 J= K,Nl
3 X(K,J)= X(K,J) / F
=
2 X(I,J)=
001~
1,N
2,1,2
0016
X«(,J)
-
X(K,J)
1 C0 I\J T I i'-J Ut:
END
*
0011
X([,K)
0018
0019
0020
STATEMENT '- AbCVE IS THE HeART UF THE PROGRAM. IT JS IN .THE INNER Leop
AND [S EXECUTED N ** 3 TIM~S
~~UATIUN
K
ll~
STA1 EMENT
IN
0023
0025
:;
i
, fVERY TERM IN THE KEY EQUATION IS ~ULTIPLIEC BY
THE .~ SUr) TRAe T[0 FRU('1 TH t: COR k ESP UNO I NGTE RMIN EQUA TIC N I
THIS WILL G[ Nt RAlE A l F KU I;'-J COL UM,'~ K. cAe H S~ [ EP 0 F r HEM ATk I X wILL
Gt.: NERA TEA eeL Urv: r,~ 0 F l r ,( 0 S I;~ CGL U Mi ~
(K)
EXC [ P T FeR THE TERfv'i UN THE
UIAGCNAL WHICH WILL BE
1.0
.
BY PERFURMI~G A SE~[ES JF VAllD UPE~AT'ONS ( NOT CHANGl~GT~tE E~LALITY)
WE GENE~ATE THF UNITY MATRIX. THE ANS~ERS WILL THEN APPEAR IN THE LAST
COLUMN •
2)
0021
0022
0024
lS THE KEY EQUATION
THIS IS CLNf
o~
oU~'
0008
0009
0010
0011
0012
0013
0014
SIMULTANEOUS EQUATIONS
00 1 K
00 1 I
IF (I-K)
0003
0004
0005
STATl~ENT
A(\~ C
x ( I , K)
I
0026
0028
002Y
0030
0031
g~
o~
0035
0036
0037
TABLE
LISTING OF PROGRAM T~" INVE~T A LARGE MATRIX STORED ON
DISK USING PARTITIONING
2
IlaTION CATAL
ASE BJPL$04,S
II EXEC FOK1RAN
COMMON N3, N4, N5, N6,JIM,NM,IT,IS,NJ,NP,NC,NX
C
N3
IS A wO~K FILE ON DISC
C
N4 IS A TEMPORARY SEQUENTIAL ~ILE
C
N~
IS A SEQUENTIAL FILl: FOR STURING MEMBER INFORMATION
C
N6 IS A OISC FILE IN WHICH MATRIX IS STORED
C
JIM
IS THE SIZE uF THt SMALL MOSAIC OF WHICH THE BIG MATRIX IS
C
NM
IS THE NUMBER O~ MEMeEK$
C
IT IS fILE TO ~HICt1 PRlJGKAM WILL RETUR.'J
C
IS
WILL CAUS~ DIAGNUSTICS TO BE PRINTED
C
NP
IS PAGE NU
NC IS NO or COLUMNS IN HIG MATKIX
C
NX
IS ;~COF CCLU~jN~ IN CUMPRESSED t-1ATRIX
DEFINE fILE
12(160Q,200,U,I2)
DEFH-Jc FILE 13( 8()O,200,U,13)
OUUBLL PRECISION X (S2Y)
IF(NX '3) 1,1,2
1 CALL ISSl ( NX, 1,X)
J CALL LINK (O~)
2 IF(NX-4b) 4,4,?
4 CALL 13b11 (NX,l,X)
GO TO 3
5 IF(NX-92 ) 6,6,7
CALL T~~12(~X,I,X)
O IF(NX-ld4)
GU TU 3
7
8
CAL L
1 B l» 1 3
GO T (j
S,e,g
(I-.l X , 1 ,
x)
3
9 I F ( NX - ~ ~ 2) 1 0, 10, 1 1
10 CAll TBiI4(NX,1,X)
GU TO 3
(3,100)
CALL EXIT
100 FURMAT (
2GH MATKIX
11 WI{ITE
TOO LARGE
Ei"'40
1*
TEB
II EXEC FORtRAN
SUBROUTINE'
TK$l(~~,NC,
X)
DOUBL~
PRECISIUN
X(23,23),~
COMMUr~
1>.43, N4, N5, N6, ;\JX
13 ::: 1
NP ::: j"",JC +
i'~ 1
I ::: 1,1'1
K ::: NC + I - 1
CALL DATA (1,1,X,13,
13 13 ::: 13 + 23
DO 1 K ::: 1,N
DO 13
IF (X(K,K)2,1,2
F ::: 1.001 X(K,K)
X(K,K)
DO 3 J
::: .101
::: ,1, N
NC,
NP, K)
~ADE
0040
0041
0042
0043
0044
0045
0046
0047
0048
0049
0050
0051
0052
0053
0054
0055
0056
0057
0058
0059
0060
0061
0062
0063
0064
006':>
0066
0067
0068
0069
0070
0071
0072
0073
0074
0075
0076
0077
0078
0079
0080
0081
0082
008.3
0084
0085
0086
0087
0088
0089
0090
0091
0092
0093
0094
0095
&?3
*
3 X(K,J) = X(K,J)
F
PO, 1 I :=lJ_"!~~_",«".. -~","".,~,
IF (I -K) 4, 1 ,4
4 F '= X'-'_',K"
X ( I ,K) = o. DO
0096
.' .,....... , " .. «','. .....
13
.'
*
=
- - -..
... ..
.,ort'J '
0101
----0102--'-.~_v_
. ____.: . . 0103
~·
~.
=
13 + 23
0101
-, -(f1"08----·0109
._-_.- '"'--"~"'--Ol fa--
0111
'--0'11'2 -.---.-,
0113
TEB
--0114------'
II EXEC FORTRAN
SUBROUTINE DATA( JIM, JOE, A, 13, NRl, NR2, NC)
DOUBLE PRECISION A(529) , l(IOO)
COMMON N3, N4, N5, N6
IF ( N~) 3 I , 30 ,31
30 WRITE(3,32) JIM, JOE, 13, NR1, NR2, NC
31 CONTINUE
32 FORMAT (/618/)
NF = N6
NB
0104
0105
. "0 fO-6-----"
RETURN
END
1*
0115
0116
0111
oiItf'------
------- . . -" ..... '
0119
0'1'20
0121
'~--·-o
=0
~J1s
0126
0121
.' 'Crf2 8 -- ....-.
0129
0130'
LE
O1.32-·~·
MOO(NRB, 100) + 1
~~OC(NRE,
100) -+
0133
'----'01 3 4_~_"-N"
0135
1
ME
I~K + NRE/fbo+' 1
MB = NK + NR8/100 + 1
DO 33 M = MB,ME
lS = 1
IF
11
12
13
14
6
10
7
=
, '-'0-[36
0137
100
IF (M-MS') 11,11,12
LS= LB
IF(M-ME) 14,13,13
LF = LE
READ (NF'M) I
GO TO (6, 7) ,J't M
DO 10 L = lS, IF
'A( I ') Lfr'r -"_-_"-'-"'"
.
I = I + 1
GO TO 33
00 15 l
= LS, IF
l ([) =' ATll'
=
15 1·= 1+ 1
~
0131
I = 13
=
=
=
122---'-'
0123
GO TO (1,2,3), JOE
3 NB = 200
2 NF = N3
1 CONT 1 l.. UE
NK = NC*4-4
NRB = NRI -1 + NB
NRE = NR2 - I + Nfl
LB
....Q9 9 1.".,_,,___
-------~
F
00 14 I
It N
K = NC 1 + 1
CALL DATA (, 2 ... A,X, 13, NC, NP, K)
14 13
..
·0098
DO 1 J = 1,N
X.(I,J)= XlI,J) - X(K,J)
1 CONTINUE
NC.l = NC-l
NP
NC & N -1
=
=1
._.._ ....._ ...' ____..._
....
.
---'--01'3 a-----'
0139
---'-~ ·-'---·-·-OT4'(}·-~·
..."
0141
~---
.. ---~----
0143
.-'0-144 ---- i
.-.- . .-.-.-.-.'-. --..-~--.-g~:~
---'-~-----'
..
0147
~1Jl-48---
.---,---.. ~,-g(l---,0151
I
I
EntIre
WRITE(NF'M) Z
33
=
13, J
END
TEB
II EXEC FORTRAN
SUBROuTINE TG~11 (N,NA,X)
DOUBLE PRtCISION X (23,23)
=
CALL T8$ltNL, Ne, X )
CAL L 1 B ~ 7 ( I\J , :~ .A , X)
CALL TR)l(NU, NA, X)
CALL TB'8(N,NA,X)
RETUR~;
EI'-JD
1*
TEB
II EXEC FURl RAN
S U 131<.0 UTI NET B i» 7 ( N , NA., X )
DOUBLb PRECISluN X(176,3)
NE = NA ~ ;~ NL = I'~ I 2NU = i\ - Nl
j\JO = r~A f. NU
NUl = NIJ - 1
.-J2 ;: B
C
1
*
=
1
DO 3 K ;: NO, I'J E
olJ
1 l
DU
2 J .-: 1, I\j L
;: 1, t\J U
1 X(l,3) ;: 0.00
o IN (2)
CAll UATA (l,l,X,177,ND,NE,K)
BIN (1)
CAll GATA (l,l,X,l,i'.. A,j~Dl,~+ j\lq~
DO 2 I = 1, NU
X(J,2)
2 X(I,3) .-: XCI,j) f. X(I,l)
(3) IN w2
C
C
*
C
'CALL DAfA (2,2,X,353,1,NU,Kl)
3 Kl = Kl + 1
A ;: A - W2 * C
DO 6 K = NA,NDI
C
C
A IN (3)
CALL OATA (
C IN (2)
C.
1,1,X,3~3,NA,
NOl, K)
CALL DATA (l,l,X,177,NO,NE, K )
DO 5 J
W2 IN
;: 1,NU
DO 5
;:
(1)
CALL DATA (
C
0
5
I
X(I,3)
=
".""j"
rti":*t"#mt!IHMtt'h"~I::JJ
1,2,X,l,1,~U,J)
1, NU
X(I,J) - X(J,2)
0154
0155
0156
0157
0158
0159
0160
0161
0162
0163
i'J L ;: i'..J 12
NU
f'-J NL
NO ;: ;\JA + NlJ
Kl
r% "f""·" rr-V"·" Mriti .. "in W""" ttl> f" f "t"
0153
1*
o
."
0152
CUNTII~Uf
IF(N5) 42,41,42
4C)1 J = [3 + NR2 - NRl
WRITE (3,100) tAlK) , K
100 FORMAT ( IPI0013.6)
42 R~TURi'~
c
ill
*
¥'J,~~
0164
0165
0166
0167
0168
0169
0170
0171
0172
0173
0174
0175
0176
0177
01-'8
0179
0180
0181
0182
0183
0184
0185
0189
0186
0187
0188
0190
0191
0192
0193
0194
0195
0196
0197
0198
0199
0200
0201
0202
0203
0204
0205
0206
020 I
C
(3)
IN A
0208
0209
0210
0211
6 CALL DATA (2,l,X,353,NA,ND1, K
KETURt-4
END
g~
TEB
1*
II EXEC FURTRAN
SUBROUTINE TB18 (N,NA,X)
DOUBLE PRECISION X(176,3)
NA1 = NA - 1
NE
('~L
NU
=
=
=
i~ A +
i'i/2
i~
;\J
0214
0215
0216
1
-
0217
021H
NL
-
0219
NO = j\JA + NU
NOl = NO - 1
B
-A
w2
K1
1
*
=
C
=
DO 3 K = NO, t'~ E
OU 1 L = l,NU
1 X(L,3) = 0.00
C
W2
C
IN
(2)
CAL L
DO 2
U A T A (1, 2 , X , 1 7 7 , 1 , i'J U , K 1 )
J
1,NU
A IN
(1)
=
CALL OAlA(l,l,X,I,NA,NCI, J+ ;~Al)
DO 2 I = 1, NU
2 X(I,3) = X(I,3) - X( 1,1) * X(J,Z)
C
(3)
IN
B
CALL DATA (2, 1,X, 3'):1, NA, Nul,
=
3 Kl
K1
W2 = u
K1 = 1
C
07
C
Me
Nt\,
= l,NU
= o. DO
DO 5 L
C
+ 1
=
DO 4 K
5
*
X ( L, 3)
K
I~Dl
0239
C IN (2)
CALL OATA (1, 1,X,
DO 6 J = 1,NL
177,~O,
NE,
K
)
C O I N (1)
CALL DATA (l,l,X,l,ND,NE, J + N01)
00 6
= 1,
I
=
6 X(I,3)
C
( 3)
I 1\-.;
~~
NL
X(I,3) + X( I,l}
*
X(J,2)
2
CALL OArA (2,L,X,J53,l,NL, Kl)
4 K1
C
C
=
=-
WZ
UO 71\,
DU 8 l
8 X(L,3)
C
A IN
C
+ 1
Kl
=
=
=
*
A
NA,
;~Cl
1,NL
O.DO
(2)
C
0240
0241
0242
0243
0244
0245
0246
0247
0248
0249
0250
0251
0252
0253
0254
0255
0256
CALL DATA (1,1,X,177,NA,NDl, K)
DO 9 J = 1,NU
0257
0258
W2
0259
IN
(1)
CALL DAfA (1,2,X,1,1,NL,J)
9
0220
0221
0222
0223
0224
0225
0226
0227
0228
0229
0230
0231
0232
0233
0234
0235
0236
00 9 I
X(I,3)
=
=
(3)
C
Ir~
1,NL
X(I,..;)
0260
0261
-
X(I,I)
~(
X(J,2)
D;
!"II
C
CO...
0 = D - W~ * 0
7 CALL OATA (2,1,X,353,ND,
DU 10 K = NO, NE
N~t
K)
DIN (3)
C
CALL UATA (1,I,X,353,NO, NE, K )
B IN (2)
CALL JATA (1,1,X,117,NA,ND1, K)
C
W2
=
00 11 J
IN
1,NU
(I)
CALL DATA (1,2,X,l,l,NU,J
UO 11 I = 1,i\lL
11 X(I,3) = X(I,3) - X(I,l) * X(J,2,
C
(3)
I~
0
10 CALL DATA
RET UR'i
END
1*
TEB
II EXEC FURl RAN
(2,1,X,~53,
NO, NE, K
TBi14 (N,NA,X)
DOUBLE PK[CI~IUN X (25,25)
NL = \j/2
NU = ;~ - NL
SUBRUUTIN~
ND
=
i~A
+ NLJ
CALL fB$13(NL,NC,X)
CALL TB$7(N, NA, X)
CALL IBi13(NU,NA, X)
CALL JB~8(N,NA,X)
KETURr-J
END
I
TEB
II eXEC FURl RAN
SUBRUUTINe T~h13 (N,NA,X)
DOUBLE PKfCISIUN X (25,25)
NL = N/2
O
NU
=
NO =
CALL
CALL
CALL
CALL
i~
-
;'4L
NA + NU
TB'12(NL,NC,X)
IB$7(N, ~A, X)
IB~12(~U,NA, X)
T818(N,NA,X)
RE JURi'.
END
TEB
II t:XEC FGRTRAN
SUBROI)TINE TI31112 (N,NA,X)
DOUBLE PRECISION x (25,25)
NL = N/2
NU = N - ~L
NO = i\JA + NU
CALL TB$ll(NL,NO,X)
CALL TB.~7(N, ,'4A, X)
CALL TOill(NU,NA, X)
CALL TBi8(N,NA,X)
RETURN
1*
,
EI'JD
0
1* fEB
awe n rVTIIu::::rrnj.1
0264
0265
0266
0267
0268
0269
0210
0271
0212
0213
0274
0215
027b
~
0211
0218
0279
0280
0281
0282
0283
0284
0285
0286
0281
0288
0289
0290
0291
0292
0293
0294
0295
0296
0291
0298
0299
0300
0301
0302
0303
0304
0305
0306
0301
0308
0309
0310
0311
0312
0313
0314
0315
0316
0317
0318
0319
c2 ??
IJ
III{
TABLE
LISTING OF PROGRAM
3
TO
IN~ERT
A MATRIX STORED ON TAPE
II J08 BJPL:lt
II OPTION CATAl
PHASE BJPL$13,S
II eXEC FORTRAN
C
BJPl$ PROGRAM TO INVERT UP TO A 300 X 3CO MATRIX STORED
CULUMNwISE IN FILE 12 • FOUR RECORDS ARE USED FOR EACH COLUMN
COMMO~ N3, N4, NS, N6,MIJ,NM,IT,IS,NJ,NP,NZ,NC
OEFINf FILE 12(1600,200,U,IN2)
DOUBLE PRECISIUN F,G
DOUBl~ PRECISION C(300), 0(300,2), 01(300)
EQUIVALENCE (01(1), 0(1,1»
DIMtNSION NF(Z)
NF(l) = 4
NF ( 2) = 5
M
1
N
2
OU 14 J
=
=
J4 =
=
l,NC
J*4
REAO(N6 t J4-3) (C(l), l= 1,100)
REAO(N6'J4-2) (C(l), L=lUl,200)
REAO(f'i6 t J4-1) (C(l), L=201,30U)
14
WRITE
(4)
C
REWIND 4
READ (4)
REW Ir-JU 4
(Cl(l),
l
= I,NC)
Du 8 K = 1,NC
F = O(K,M)
IF (F) 4,15,4
15 DO 2 J = 1,NC
KEAD (NF(M»
(C(l) , l
GU TO 19
=
:::
1,NC
l.CO
READ (NF(M»
GU TO 22
16 CONT Ii\U E
READ (NF(~)
(C(l), l ::: 1,NC)
IF (NS) 23,22,22
£:3 WRITE(3,11) J,(C{I), I = 1,I~C)
22 C ONT I f'JU E
=
C(K)
C(K)
G ::: C(K)
DU 1
I
:::
*
F
1, l~ C
IF(I-K) 3,1,3
3 C(I) = C(I) - O(I,M)
1 CONT II'-lUI:
19 IF (J-K-1) 20,5,20
5 DC 7 L = 1,NC
7 O(L,N)
=
20 CONT I !\U(
ell)
*
0330
0331
0332
0333
0334
0335
0336
0331
0338
0339
0340
0341
0342
0343
0344
0345
0346
0347
0348
0349
g~Q
l,NC)
21 CtL) ::: 0.C0
C(K)
oM
0350
16,17,16
=
g~;
0351
0352
4 F- = 1. cc, I F
00 2 J = 1,1\ C
IF (K-J)
11 00 21 L
0324
0325
0326
035~
0356
0357
0358
0359
0360
0361
0362
0363
0364
0365
0366
0367
0368
036g
0370
0371
0372
0373
G
0314
0375
0376
0317
gD
dtt'f j\iiitEff tfrill" BNt ""tiiFttti?HfC3b"bbbt trl'ltttttW*itiif?BiiWWfflffBRr . [ i'tXnr-fFifiHtitiWriWfWf "3f689fWf*E "" ... "" .. r WWFrffWhwrr' "TIT W'WftNtWi¥:iirlt1ll!t1'iWP2rr
CONT l;'~UE
wKITE (NF(N»
2 CONTINUE
(C(l),
L
= 1,NC)
REWING ')
=M
M = r'J
N = L
(C(L), L = 1,NC)
J 4 - 3 ) ( C ( L ), L =
i\~F(M»
WkiT E (1\16'
WRITE
(~6'
J4-2)(C(L),
J4-1)(C(L),
WRITE (.\J6'
30 CUNT P~UE
CALL LINK (05)
END
1*
TEB
INCLUDE uVERLAY
II EXEC LNKlCT
o
I.". .
"PI
•
I
0389
OJ 30 J = 1, ;~c
J4 = J ", 4
(
•
0388
l
REAl)
".. . . . .
0385
0386
0387
KEWIND 4
8
-g '",."W'TFP-WW2stibiMIiiI
0380
0381
0382
0383
0384
IF(N5) 9,10,10
(3,11) J, (C(I), 1= l,NC)
11 FORMAl (151 (lX,lP10012.5»
9 WRITE
0 .,10
I
L
l
=
=
1,100)
101,200)
201,300)
0390
0391
0392
0"39j
0394
0395
0396
0397
0398
0399
0400
0401
0402
TAB l E
4
l I 5) TIN G 0 F PRO GRAM
USE 0 'T U G ENE RAT E 1 EST MAT R I X
II UPTION CATAL
PHASEBJPl$lO,S
II EXEC FORTRAN
C
C
C
C
C
C
C
C
C
C
=
N3
N6
=
=
O.UO
13
12
READ (1,1000) N5,IS,NX,(A(K,I) , K
1000 FORMAf(ZI2, [4, IGA4)
WRITE (3,1000) NS, IS,
L4 = NX I 100 & 1
DO 1 K
00 2 L
=
LK
K
KL = L
A(L,l)
IF
*
*
=
(L - K)
6 A(L,1)
GO TU 2
3 A(L,l)
2
=
=
00411
410
409
COMMON N3, N4, N5, N6,JIM,NM,IT,lS,NJ,NP,NC,NX
N3 IS A WOKK FILE UN DISC
N4 IS A TE~PORA~Y SEQUENTIAL FILE
NS IS A SEQUENTIAL FILE FOR STURING MEMBER INFORMATIUN
N6 IS A DISC FILE IN WHICH MATRIX IS STORED
NM IS THE NUMB~R OF MEMBEKS
IT
IS FILE ro WHICH PROGRAM wILL RETURN
IS
WILL CAUSE DIAGNOSTICS TO BE PRINTED
NP IS PAGE NU
NC 15 NO OF CCLUMNS IN HIG MAT~IX
NX I 5 l'~ 0 0 f COL UMNSIN COl'" PRE SSE 0 t-1 A TR I X
OEFINt FILE
12(16nO,20Q,U,I2)
DUUBLE PRECISION A(100,4)
DU 10 K = 1,400
10 A(K,l)
0405
0406
0401
0408
i~X,
=
1,16)
(A(K,l), K = 1,16)
l,NX
1,NX
0412
0413
0414
0415
0416
0417
0418
0419
0420
0421
0422
0423
0424
0425
0426
0427
0428
0429
0430
0431
0432
L
0433
NX - NX & K
434
435
0-0436
r~ X
-
lK f.
fJ X t.
1
3, 6 , 2
=
A(L,l)
=
KL + 1
*
0437
0438
0439
0440
0441
10.
CONTIi"UE
=
K4
4*K -
4
lHJ 4 L = 1,L4
4 WRITE (N6' K4f.l) (A(I,L) , I;: 1,100)
IF (1\4'»
5,5,1
5 WRITE (3,l001) K,{ A(I,l) , 1= 1,NX)
1 COi'-fTIhUE
Nl'v\ = tOICT I (viE (:.)
WRITE (3,1001) NM
0442
0443
0444
044'j
0446
0447
0448
0449
0450
0451
0452
0453
0454
CALL LINK (IS)
1001 FURMAfI/Ilb/( 2X,lOFI0.1»
END
1*
II
TEB
EXEC ASSEMHLY
MCTIME
1*
TEB
STAKT
0
US Ii'JG
*, 15
0455
GETIME BINAKY
0456
lR
(1,1
BR
END
14
0457
0458
4'i?459
4 60
U
-If·j·IiE-i····);· 'e"rS''''Y'''" t· .tW:tittt.M"f
TABLE
5
OF PROGRAM USED'TO CALCULATE
INVERTING A TEST MATRIX
lI~TING
RMS
ERROR
CATAL
'-'PHASE BJPl$05,S
II eXEC fORTRAN
COMMON N3, N4, N~, N6,JIM,NM,IT,IS,NJ,NP,NC,NX
C
N3 IS A WOKK FILE ON DISC
N4 IS A TEMPORA~Y SEQUENTIAL FILE
C
C
N5
IS A SEQUENTIAL FILE FOR 5TORING MEMBER INFORMATION
C
Nb IS A DISC FILE IN WHICH MATRIX IS STORED
C
NM IS THE NUMBER OF MEMdERS
C
IT
[S FILE TO WhICh PkUGRAM ~IlL KEf URN
C
IS
WILL CAUSE CIAGNOSTICS TU UE PRINTED
C
NP IS PAGE ~O
C
NC
IS NO CF CULUM~S IN 3lG MATKIX
C
NX IS NO Of COLUMNS IN COMPR~SSED MATRIX
DEFINE FILE
12(lbOQ,2JO,U,I2)
DOUBLE PRECISIUN A(lOJ,4)
1033
=
NM
=(~M
f'-1 C TIM [
( U)
+ 30) I
i"l M
-
60
wRITE (3,1J33) I\lM , NX
FLJRfliAT ( 11, 2?H i"llIJUfES TO Ii\lVEkf SIZE
SUM = O.
=
L4
;~ X I
100
+ 1
= IfNI(
K*4 - 4
L = l,L4
OU 1 K
K4 =
DO ?
READ( NotK4 + L)( A(I,L), 1= 1,NX)
IF
15,15,14
(N':»
15 WI{ITE (3,10.1) K, (AtI,l), 1= 1,NX)
(2X,lPI0012.4))
103 FORMAl (/1101
14 CUNTINUE
00 1 I = 1,Nx
E = o.
DG 2
J
01 J -=
*
2
E
=
IF(I-K) 1,3,1
3 E = E- 1.
1 SuM = SU~ + E
E
=
S CR1
*
050':)
0506
0507
O[J
0508
*
E
I ,'~ x
) *
0509
0510
0511
1 (I 0 •
E
0512
CALL LINK (10)
(
0495
0496
0491
0498
0504
(S U MIN X
(3,100)
100 FURMAT
0493
0494
0503
NX + J +1
E + A(J,l)
WRITE
047~
0476
0477
0478
0479
0480
0481
0482
0483
0484
0485
0486
0487
0488
0489
0490
0491
0492
0500
0501
0502
[ + 1
(I-J) 4,6,2
OIJ .; Q[J * 10.
GO TO 2
4 OIJ .; NX*l -
I)
04t
046
046
047
041
047,
041:
047·
0499
= 1 t r"x
NX J - ;'.. X +
IF
6
,15
04~
04t
I~OPTION
NM
WHEN
18H IU-1S
E hiD
TEB
1*
II EXEC ASSlMl:3LY
eRROR,
PC
.;
, Fa.3)
0513
0514
0515
0516
STAKT
0
I't\
USIi\JG
*,15
~
GETIME OINAKY
0517
0518
0519
LR
0520
~CTI~E
v,l
BR
14
END
TEB
I[\JClUDE UVERLAY
EXEC LNKEDT
!III ~
0521
0522
0523
O~
o~J5
I:
i
TABLt
I'~B
I
,"PT
PHASE
6
LISTING OF THE SHORTEST PROGRAM THAT I COULD WRITE TO
INVERT A MATRIX
STORED ON DISC.
BJPL$
ION CATAL
BJPL:i»14,S
II EXEC FORTRAN
THIS IS SHORTEST PKUGRAM THAT I COULD WRITE TO
INVERT MATRIX STGREO ON DISC
COMMON N3, N4, N5, N6,JIM,NM,lT,IS,NJ,NP,NC,NX
C
C
DEFINE:: FILE 12( 1600, 200, U , 12 )
DOUBLe PRECISION 1(100,3), 0(100,3), Z(1CO.3), F,G
=
L4
(NX-l) , lOG + 1
DU 1 L
1,300
=
1 l(L,l)= D.DO
=
00 2 1 K
1, j\j X
K4
.-\*4 -4
DU 2 I-j :: l , L 4
2 REAOIN6' K4 + ~)
=
F
=
l.CO I
Z(K,l)
Du 3 t-\
=
=
(O(L,M) , L
=
1,100)
O(K,l)
1.00
1, L 4
3 WKITE(Nb' K4 + M) ( Z(L,M) , L= 1,100)
l(K,I)
DC 21 J
=
=
0.00
=
l,NX
J4
J*4 - 4
DLJ 4 !"i = 1,l4
4 READ (~6' J4 + M) ( T ( L , ~I
I(K,I) = T(K,l) * F
=
G
T(K,1)
00 22 I
=
j\t\
=
L
*
G
1, L 4
5 WRITE( N6'J4 + M)
2 1 CON T I ~,; UE
CALL LII'lK
END
= 1,1(0)
1,j~X
IF (I-K) 23,Zt,23
23 T(I,I) = 1(1,1) - C(I,l)
2.2 CONT Ir'JUE
DU 5
),
(T(L,M), l
=
1,100)
0528
0529
0530
0531
0532
0533
0534
0535
0536
0537
0538
0539
0540
0541
0542
0543
0544
0545
0546
0547
0548
0549
0550
0551
0552
0553
0554
0555
0556
0551
0558
0559
0560
0561
0562
0563
0564
(O~)
TEB
INCLUUE UVERLAY
II EXEC LNKLCr
1*
fEB
1*
TYPICAL DATA FUR RUNNING PROGRAM
0565
0566
0567
0568
0569
0570
0571
0572
0573
1/ EXEC
1 4
113
I1J
1*
It
BJPL$lU
60 X 60 ~ATRIX INVERSION BY PARTITIONING
25 X 25 MATRIX INVERSIUN BY BY SEQUENTIAL FILE
60 X bO MATRIX INVERSION BY BY SE~UENrIAL FILE
0574
0575
0576
0577
0578
0579
,I
./
"
/'
,
/ t{ )
I
o
COMPUTER REGISTRATION AT CHICO STATE COLLEGE
by
Neil C. McIntyre, Jr.
,,("I'I
l~Ji
Faculty, Computer Sciences
Chico State College
G
c
COMPUTER REGISTRATION AT CHICO STATE COLLEGE
Since the inception of computerized registration at
Chico State College in 1965, the College has received
numerous inquiries from other colleges as to how our
system works and how it might be modified and implemented
for other schools and other hardware configurations.
Since
no general computer-oriented overview of our system has
been undertaken previously, this paper is an attempt to
describe the salient features of what we at Chico State
College believe to be the only successful large-scale
genuinely computerized registration system.
Like most California colleges, Chico State has recently
()
experienced enormous growth.
From an enrollment of 2,200
full-time equivalent students in 1961, the College has grown
to 6,738 FTE this fall.
Such rapid growth (about 15% for
several years), with no abatement in sight for the near
future, demanded that the College implement some form of
automated registration.
Perhaps the best overview of the impact of computer
registration at Chico State can be gleaned from a description
of the process as executed under essentially a manual system
and a description of the corresponding automated systems.
Prior to 1965, scheduling of students into the many
sections of each course was effected by what we refer to as
o
"arena" registration: on the Thursday, Friday, and Saturday
preceding the first day of classes, the college's gymnasium
1
2
was equipped for mass student registration.
processing equipment.
o
To register in the College, a
student, having previously applied to and been admitted to
the College or continuing from the previous semester in good
standing, would be given a packet of punched cards which
included a clearance card, study list, housing card, and
the like.
At the hour appointed for his class level and
the initial letter of his last name, the student would
proceed to the gymnasium.
There he would wait, seemingly
interminably, to be admitted to the arena.
Inside, two
major functions were performed--first, a student would
select the classes he desired, proceed to the proper table,
and, if the class he desired were not full, would be given
an IBM card representing the class.
If the student were
exceptionally lucky, all his desired class sections would be
open.
If he were extremely unlucky, a class would close
during the class-card gathering process and he would be forced
to return his unwanted class cards, re-work his schedule with
the help of an academic adviser, and start the class-card
gathering process again.
When he had all the desired class
cards, he would proceed to the second phase: paying of fees.
While the lines to pay fees were shorter and the time less
than the waits to be admitted to the gymnasium and to
receive class cards, the student was understandably grouchy
after spending what was ordinarily about five hours in the
o
registration process.
---
,1
-ii:
For some time,
record-keeping functions were assisted by unit-record data
1
- -------------
o
It can readily be seen that the process of registration
by the arena method for 6,700 FTE will require more time and
facilities than for 2,200 FTE: something had to be done, for
our time and facilities were simply running out.
During the arena years, the class schedule was prepared
entirely by hand, typed, photo-reduced, and printed.
The
secretary to the Vice-President for Academic Affairs would
invariably spend the entire Christmas vacation period and
the same amount of time during the summer manually preparing
a list of all rooms and the classes to be taught therein
during the upcoming semester.
Inevitably, conflicts, empty
rooms, and classes without rooms would appear which would
frequently not be resolved until after classes began.
()
As the
College grew, physical plant utilization therefore dropped
and confusion was the order of the day.
students would meet with an academic adviser in their
major subject area and, together, the student and his adviser
would prepare a program planning sheet for the student's use
at arena registration.
By 1965, this process could not be
completed conveniently on the Monday, Tuesday, and Wednesday
preceding registration allotted to the advising procedure.
Moreover, all of the necessary but often irritating
paraphenalia of student records, such as class lists, packet
cards, grade lists, grade reports, and the like could not be
conveniently prepared in the allotted time.
o
Chico State's computerized system revolves around three
distinct but interdependent systems: schedule preparation,
, ",' M"":U,, ,2,. 1"""·,·"k",M""",g;,,,x",;::.x,,,,,.g$l4,Ji';;;;,tc",'~rnm4;Mi;;;,,..niTik1!,;;;W.J;·M:r&:::·"t,1IlJlM,W:;;:;U;:;:A!£!il1iG11':;&.W...iWTW
student scheduling, and student record processing.
One of the major effects of computerized registration
has been the expanded time schedule for the entire process.
o
Using the arena system, the major functions were performed
from just before Christmas until mid-February for the spring
semester, and from early August to late September for the
fall semester.
Computerized registration has so drastically
altered the scheme of things that registration processes
are underway at all times during the year.
Chico State's registration process now begins with the
computer center preparing, based on history and enrollment
projections, a schedule request worksheet which is sent to
the various academic department heads for their use in
planning their course offerings for the upcoming semester.
The schools and divisions of the College receive these
o
worksheets shortly after the beginning of the preceding
semester.
Onto the worksheets tbe department heads write
the requisite variable information, such as instructor
number, class meeting times, building and room codes, and
so forth.
The completed worksheets are returned to the
computer center for keypunching and verifying.
The data are
then submitted to the first ina series of major programs.
This first program checks the schedule request against
master lists of courses, rooms, instructors, valid times,
etc., and produces listings of errors, omissions, and
inconsistencies.
The error listings are returned, together
with a list of classes vying for the same room at the same
--.---.~--
---
-
- - - - ..
-.-----~---
.. - " - .. ---- ...
-------~----
0,
-.-----------------
..
5
hour, to the department heads for thier corrections.
The
entire class schedule is stored on a disk pack once, and,
as corrections, additions, and deletions are made to the
schedule, additional error checks are performed.
During
the first week of November for the upcoming spring
semester and during the third week in April for the fall
semester, a computer program prints a copy of the tenative
class schedule containing class number, department name,
department course number, course title, number of units of
credit, instructor names, and times.
Building and room
are not printed in the tenative schedule.
The schedule is
then sent to a print shop for photo-reduction and insertion
in every copy of the next issue of the Wildcat, the free
o
student newspaper.
The week following publication in the Wildcat is
designated as "pre-registration" week.
Each student
meets with his adviser and prepares the successor to the
program planning sheet.
On the request form the student
lists each class he desires, possbile alternates to it,
the number of minimum and maximum units he will accept,
states whether or not he will take any section of the
course in order to get it, states any time restrictions
he has, and similar information.
The program planning
sheets are then sent to the computer center for keypunching
and verifying.
o
~.jjl;lh!fu_<@!}__
&l.!!:!",
. .¢L.i!!O!!U·,,!l"",4··!E··_··.24Ii4_~.i
From the student request cards, a tabulation of the
number of students requesting each course is prepared and
_k'"'_ _
.f:i!."""'!\3'+"'E!'.~'.~,.$_,_,
... _• .,_h#-4I~P5.¥Pi'#ii"l'i'l!,,,,F.!!!._a
..',i...._.£bHil,J¥&!III!I,III\\lil
..i\.IIiI'!!iIiI!I/\¢..lIIIIId_ .... _4442"""'
.._. _ _ _~+."~_F~,,,"""'.'-Pi:'l'l'i".t'¥.¥~',m"#1l:1i'diliG_Q,"U;;_Lut&_tJ~t&I_"_llLmGl_.m""",""""",,,".,_.iliI:c:wwaaaazza
........._
.....---------~
--~--
....
-~--'-"
_ ..
~~-"'--'''-''''--
6
sent to academic department heads and college administrators.
Based on demands for each course, departments may revise
the schedule as required.
During all phases of the schedule
preparation, updated room lists, conflict lists, and lists
of available space are frequently prepared, thus providing
a great degree of flexibility and control over physical
plant utilization.
As an example, virtually no room
conflicts now exist the day classes begin, whereas many of
the classes had no room or met in a room occupied by another
class before computer registration began.
When it is determined that the schedule is final
(usually about four weeks before the beginning of the
semester), computer registration is actually run.
Most registration programs for relatively large-scale
use are "computer-assisted."
Emphasis is principally on
relieving facility and personnel problems of scheduling
rather than on giving the student more choice and greater
convenience in class selection.
In this respect, our
registration program is, I believe, unique, for it permits
the student great latitude in course selection, time
limitations, in fact, all of the advantages of ideal class
scheduling with few disadvantages.
In the spring of 1965, two Chico state College students
were commissioned to determine if a computer program could
be written for the 1620 to enable registration by computer.
The students determined that a program could be written by
writing one!
written for, and still operating on (in slightly
---_."_
.. " - , , ...
"---,,
-'-"~---'--'''--'----''
..
_-
o
7
c
modified form) a 20K card/disk 1620, the Chico state College
registration program, as we call it, should actually be
called a "scheduling" program.
The registration program
takes into account both student requests and the state of
the schedule at the moment a student is processed.
The
process of scheduling is markedly similar to arena
registration, but the computer does all of the work
frustrating to students: in accordance with the student's
requests, the program tries as many as 15,000 possible
schedules before determining that it cannot find a
workable schedule within the student- and college-applied
restrictions.
Where the student may spend upwards of three
hours to find any workable schedule in the arena, the
computer registers between 200 and 2000 students per hour
on the 1620, depending upon the state of the schedule and
how unrestrictive the student's requests are.
When the
program finds a schedule for a student which satisfies all
requirements, a record of each class in which he is
scheduled is punched, the disk records identifying the
classes are updated to reflect the new enrollment, and the
next student is processed.
Three major characteristics of
this program set it apart from most others seeking to
provide the same sort of service:
(1) the program allows
great latitude and flexibility in any student's schedule-a student may apply sectioning, time, units, and course
o
restrictions to prevent the program from giving him an
unacceptable schedule;
(2) when the program, as happens
very frequently, is called upon to select a section from
those available, it attempts to schedule on the basis of
class size, from least full to most full, thus balancing
the size of the sections on a continuous basis.
Only in
the event that no section of a class will fit will the
program attempt to delete the class and sUbstitute an
alternate if the student has listed one; and (3) the entire
state of the schedule is held static during his processing,
thus attempts at alternate schedules are unaffected by
other students scheduling.
The only condition causing a
student to be unscheduled occurs when the program cannot
satisfy the student's listed minimum units.
Irrespective of whether or not a student is successfully
scheduled by the program, any errors the student and his
adviser commit are logged and the program attempts to take
corrective action.
o
Among the many niceties in the program
is the ability to include courses which have interdependent
parts.
For example, a Biology class may be divided into
two hours of lecture, two hours of laboratory, and one
hour of discussion each week.
The student must register in
laboratory and discussion sections which have the same
instructor as the lecture section.
The program insures that
conditions of multiple-dependency such as the example are
properly scheduled.
When all the students have been processed, the output
cards are sorted alphabetically on student name and run
through a program which prints, in class-schedule style,
C',
a list of the classes in which the student is scheduled on
-_.
- _ .__.•. - - _ . _ - - - .. _. - - - -.• - - - -
-
..
.-
.-. . - - - . - - - -.....
_--_..... _---
.....
~
~;ii!iLi!i#
continuous-form IBM cards.
The separated cards are inserted
in the proper student's registration packet together with
his other registration materials.
For the spring semester,
the packets are then made available at the College Library
for a period of several days.
For the fall semester, the
packets are mailed directly to the students.
Lists are prepared showing how many students were
scheduled into each section, how many seats remain, and
the percentage of students entering the process who were
successfully registered.
The student now has the option of either accepting
his entire schedule or rejecting it and participating in
arena registration.
()I
It has been proposed and will soon be
adopted that each student prepay materials and service fees
and be required to accept the computer-generated schedule.
At present, those who could not be registered by computer
are required to register at a conventional arena.
An arena-
like arrangement is provided for spring semester for students
to pay fees.
Clearance cards from this fee-paying arena or
from students who, during the summer, return their fees by
mail, are matched against output from the computer
registration run, class cards for the students who accepted
their computer schedules are punched, and the remaining
class cards for each course are punched, interpreted, and
arranged for arena registration.
Arena registration was held for one day only this fall,
1967, for the first time in many years.
. _AA¥ __ #44.##¥441...¥§If#..
. ..!._hT.M21t. ..
Due to computer
c193
assistance in room utilization, class card preparation, and
class status reports, more than 4,000 students were
processed in the arena from 8 AM to 8 PM this fall.
Of
those, nearly 2,500 were entering freshmen and transfer
students who cannot currently participate in computer
registration.
At the end of arena registration, class cards from
both computer and arena registration are merged and class
rolls are prepared.
Updated rolls reflecting added and
dropped students are prepared at regular intervals during
the academic semesters.
Chico state also has, in the form
of the student record-keeping system mentioned previously,
the usual computerized grade cards, grade rolls, permanent
record labels, lists of the standing of each student, and
many additional reports and services.
Due to the size limitations of the 1620 and the
probability of a new, larger computer on the campus during
the next few years, we have set several design maximums on
all the systems.
We have provided for a maximum of 100
sections of anyone course; 2,400 different course types;
3,000 sections of classes; 750 instructors; 240 classrooms;
9,000 students entering computer registration; and 47,000
class cards.
While the design maximums can be very much
larger within the capability of our 1620, we do not feel
that the allocation of the additional space on the system
disk pack is desirable.
For example, we presently sort,
store, modify, and otherwise process virtually all of the
ii,"
11
c
data within the system on one disk pack which also stores
all of the programs to use that data.
To expand the size
maximums of the system would require us to revert to card
resident programs and/or additional disk packs.
A new, more capable computer system, such as a
System/360, will permit us to add additional flexibility
to a request set for any given student.
In addition to the
obvious advantages of increased speed for all processing,
a larger system would eliminate over 500 hours of sorting
class and packet cards each semester.
We would be able to
generate an optimized class schedule based on history,
enrollment projections, physical plant utilization,
personnel idiosyncrasies, unique course requirements, and
C')
student desires, thus relieving the administration -and
department heads from much of the routine work now necessary
for schedule preparation.
Among the many features we will implement with a newer
system will be student request by course only, with programs
to select and schedule the corresponding laboratory and
activity courses required.
We have so often been asked about modifications and
implementation for other machines and schools that I feel
a few remarks might be useful.
The system we have designed is restricted to a college
employing a semester-to-semester schedule and student planning
basis.
Modification for the 1620 for a college or other
school desiring to plan college and student schedules more
---_._._.__ __
.....
..
-"~
I~
I']'
12
than one semester in advance seems unlikely, for we have
stretched the capability of the 1620 nearly to its breaking
point.
I am certain that long-range scheduling is possible
and desirable for any school having, say, a large-scale
1401 or a 360 system having a disk.
I have carefully
examined other IBM and some non-IBM systems for simplicity
of adaptation and implementation of our system and I can say
certainly that the system as we use it could very easily
be processed on a single-disk 12K 360 model 20.
It should
be obvious that a system with capabilities as great as
the model 30 lends itself well to expansion of our system.
While the restrictions on size we have imposed upon
ourselves may seem important, it is highly unlikely that
any college much larger than Chico state would want to
o
implement the system if the largest machine available to
the school were a 1620.
We have been asked if the system
can be implemented on an 1130 disk system.
It can indeed
be implemented successfully, albeit on a smaller scale,
due to the size limitations of the 1130 disk.
Chico state's system is based on a six-day week, 360
lS-minute
periods~per-week
system of time.
This arrange-
ment is not fundamental to the system, however, and could
be changed without difficulty: while our system is designed
for a six-day week, Chico State is, at present, largely
a five-day-a-week school.
By far the most frequent question we are asked is if
any part of our system is dependent upon the characteristics
c
I:
13
of the 1620.
Let me assure you that, even at the time the
system was designed, we considered the 1620 at best an
interim machine and, as such, we have always insisted to
all who have worked on the project that it remain free of
machine dependencies.
While re-programming would, of course,
be necessary for implementation on other hardware, conversion
or modification for a new computer will not be at all
difficult.
Significant in the development of the entire system
are two facts:
(1) many ideas were conceived and all
programs were written entirely by students at the College;
and (2) when the idea for the registration program was born,
IBM Systems Engineers were asked if it could be done on
our 1620.
Their reply: not even on a 60K, four-disk 1620
model II.
Chico State College is indeed fortunate to have
students rebellious enough to doubt all that they are told
is and is not possible.
c2- 9 7.
--~.,.-
...............,......
_._----_......._-_...._-._-_._ .._..........--..-.--
...
---------.--~--
'l?-)
o
STUDENT INFORMATION SYSTEM
AT
CHRISTIAN BROTHERS COLLEGE
Presented at the
Winter Meeting of COMMON
December 8, 1967
by
Brother Jerome David Wegener
Registrar and Dean of Admissions
Christian Brothers College
Memphis, Tennessee
c
STUDENT INFORMATION SYSTEM AT CHRISTIAN BROTHERS
COLLEGE USING AN IBM 1130 COMPUTER
In 1963, a system of records and grade reports adapted to the 1620
was initiated.
This was a minimum configuration consisting of a 40K
memory, a 407 for print-out, an 082 sorter, and several key punches used
also for interpreting and reproducing.
This was basically a unit record
operation with the computer used for calculations.
There was a great
deal of hand work involved in every operation.
This summer, an 1130 with disk, 8K word memory and an 1132
printer was installed.
The immediate problem, then, was to reprogram
the system making it adaptable to the 1130.
This new machine had many
advantages with its on-line printer and its disk storage for direct access
to student information.
It should be noted that IBM has prepared a
Student Information System for the 1130 available from the COMMON
library (3. 0.003).
For several reasons, it was decided not to use this.
One important. reason was that there was no provision made for storing the
student's address.
Since there is a great deal of correspondence between
school and student, it was essential that this be included.
There were a
number of other variables in the old system that had not been included in
the IBM Student Information System.
Because of the shortness of time, it
wa s des irable to keep as many of the features of the former method as
c.··'\
possible. Another important reason for not adopting it was the fact that it
!,
_4"""424A_
. ¥4¥#_
.. __;;;:_.
==_._____
. ,_.q~~~~~". ,;_"~,"_. .",=,,,",·=J:;·,;~,d::tI.\_,=_t,"1=.Id.fl.1'1itrubiLi04\:".J.1jJL·.-·~WfuliLc8.(~"'iJIUiiiliiiOJWiiiil4A:&lUJU4U~"""'------
""'!4+-"#!'...¥J1(2!1!11'
J!IIm
I;W«_!:';;;_4"A-':;:;P'!l!i\ll{
,((\\'I!'\'!I
J~,.&ff¥im
•• ,MII'!~~._fi;;"milli.¥.I!I!!II.a
m,
...T2l1i1l1!i$_MXi!!!l'lLt.tfilM_
.. _iiW
___
IIi!ll
..¥l!Iffi!Ol!fl~""",.r-Mi_"_ _.._
-had been written in Assembly language. While it is probable that it could
have been adapted, it was felt that there would be many problems and
difficulties involved in trying to revise a large assembly program.
There-
fore, it was decided to start from scratch and to write a system in Fortran
incorporating all the data forms of the previous system. This probably
runs a bit slower, but the ease in reading, debugging, and revising seems
to make it well worth the time.
The first problem was to layout the student file.
It had been more
or les s arbitraril y decided to attempt to store two students per sector.
However, most attempts at planning it ended with more than 160 words
required.
This difficulty was solved by adapting several of the subroutines
from the Commercial Subrouting Package (1130-SE-25X).
similar to the Forcom routine for the 1620.
These are
One basic advantage that these
new ones have is that, with two or three exceptions, they are written in
FORTRAN so that they can be modified with a minimum of effort. They are
called in the same manner as are FORTRAN subroutines.
interest is the Pack/Unpack s,ubroutine.
Of primary
This takes an array in AI format
(one character per word) and converts it to A2 format (two characters p:>er
word), thus halving the amount of storage required. A GET subroutine
converts from the A-format to decimal form for arithmetic operations, and
a PUT subroutine accomplishes the reverse conversion.
The student file finally decided upon was set to contain 320 characters.
When operating on the file, these characters would be in AI format, and
30()
r"
l~
L.WUDK.'
_.
"ewr-'WfFf#ift¥tRth··· t ."t""h' ,t K',*,t!,», de It
then packed to 160 A2 format characters for the purpose of storage.
!H " .
'j
h'riHNt,etM wIt~tij
Thus
two students occupy one sector on the disk. As can be seen in Figure I,
this was more than ample for our needs, and it had space for additional
information that might need to be added as time went on.
The system works in the following manner. As was mentioned above,
as much as possible of the old system was kept intact. From the application blank, the student's name is punched and he is assigned a student
number.
order.
The student number is a five-digit number assigned in alphabetical
In originally setting up the student file, they were read into the
computer without regard to order.
The computer then sorted them and so
as signed a file number to each student. At the same time, it stores the
o
student numbers in a separate file in the same order. Whenever a
particular file is needed, the student number is read in.
The computer
then does a binary search using function subroutine RSEEK to locate that
particular student's file.
His file is read into core
I
unpacked, and the
necessary information postioned, usually by means of an EQUIVALENCE
or a PUT statement.
The file is then packed and written back onto the
disk.
At the time of registration
I
the students pick up their own class cards
after their schedule has been approved by their adviser.
Positioning these
behind the student's name card, they are ready to be read directly into the
computer.
o
The computer reads in the clas s cards for each student, alpha-
betizes them, and stores them in the student's file in the manner described.
30/
.. A!44A4I¥._ ...
-. This has eliminated the long process of gang-punching all of the class
cards, and then sorting them into order first by course, and then by
student.
o
The next program reads all of the clas ses from all of the files I
and sorts the students into order according to their classes. Class
lists are then ready to be printed and after that, the student information
sheets.
The latter are distributed in toto to the Dean, Guidance Office
and other administrative offices, to the dormitory prefects according to
residence hall, and to the department heads according to the students'
majors.
Adds and drops are handled very easily.
The very tedious job of
pulling drops and inserting adds by hand has been eliminated.
It is now
necessary to simply place the add cards after the student's name card,
followed by the drop cards (coded with a 1 in column sever) and read
them into the computer which then rearra.nges the student's file.
running of report cards has also been simplified.
The
It is no longer necessary
to gang punch the marks into each of the grade cards.
The cards are
simply sorted by grade and then read into the computer, switch settings
indicating the type of mark that should be stored.
Honor rolls, failure
lists and grade statistics- are produced very easily, also.
Permanent
record labels are printed at the semester.
New students can be added very easily once a student number has
been assigned.
The program finds the proper position and moves down one
c
30~
,f
o
notch all those files that will be behind him, much as a Disk Utility
Progranl does.
Likewis e, a student's file can be deleted with thos e follow-
ing moved up one.
Bills for each student are run off for the Business
Office. The Student Directory, Teachers Schedule and various types of
I
statistics have been programmed also.
Since the memory is quite small,
each of these different operations requires a separate program, and in
some cases, two or three linked together.
This is, however, no great
difficulty, since once they are compiled and stored on the disk, a
simple XEQ command calls them into action.
Improvements will be made when the two additional disks that are
on order arrive. At that time, there will be more than enough storage to
keep the student's entire four year course of studies accessible by the
computer.
The 1132 printer is quite slow, but the operations will be great! y
speeded up when it is replaced by the 1403 printer •
.As presently set up, the system will handle 1200 students, 100
instructors and 3 00 courses.
Figure 2 shows the layout of the cards
used as input. While no claims are made about the completeness or
efficiency of the programs, the system does work well. With proper
modification, it could probably do the job at other small colleges as well.
0,': )
0'
." ,-" . . 4$' . Z .. .P.4,·.. 4i¥.
I
~
·.M '." ,¥It49ffi41i.£JYIM',;;Y;;:-·4".UOU..,Zlk.2miilii!%"hi_44.$4. .Qi£¥44! .. "'"
303
~
I
1 A
,----"-'
'-2-D--
33
34
35
36
37
38
39
4041
42
43
44
8
1
H
0
!-----~.~Q
H
2
I
4 L.
.
5 P
A
3
I
6t:1-L.
4
!I
7
L
9
_
8 W
L
1
9 c=-I_ _
D
9
1---'-"10 L
2
1
11 L
5
P
7
2
12 I
2
A
2
0
45 4
71
6
i
13 A
46
78
1
0
1----i~M
47 C
79
5
d
!--16F
48 H
80
1
1
i
17
49 E
81 1
6
1
18
50 Y
82 9
1
19
51 N
83 0
1
1
20
52 E
84 6
1
1
I~_.--?_l__ _
53 y
85 4
4
0
22 M
54
86 3
8
0
55 R
87 6
5
8
23 R
24
56 D
88 4
0
I
25 W
57
89
0
r--'--------"~-'
58
90 M
1
26 i"i
59
91 __
A___
0
I
27
1'''"'---28-F
60
92 U
0
61
93 R
1~1 2
I
29
1----30-- A
62
94 E
1
1,8 0
!.oc_"_._-.ll_~Q._, ______________ ~ 3__ ~_~ ___________~}______b ___ ." _________ _ l2.1 __1______ 15.9__...L___ _c_ _,
32 0
64 P
96 1
128 8
160 8
I
r
65
66
67
68
69
70
71
72
73
74
75
76
L.
P
R
1
N
G
F
I
E
n
_______
!
.~--~-~.-.~ .. ~-.-.~-- . . . - - , - - - . - - , . .
---.-~.------.- •• - - -
97
98
99
100
101
102
103
104
105
106
101
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
¥ ' .••••••
A
N
129
130
131
132
133
134
135
136
137
138
139
14v
141
142
143
144
14 5
146
147
148
149
150
151
152
153
154
155
156
- - -__ . ___ "____ "_ -_._--_0. __ ... _.. ____ ._. _ ______ . . ___._. _._
_.~
161 1
162 8
163 0
164 2
165 0
166 8
167
168
169
110
171
172
173
!74
115
176
J. 7 7
118
119
180
181
182
183
184
1ij5
186
187
18S
189 0
190 3
191 .1L_______
19~ C
193
194
195
196
191
198
199
2UO
201
202
203
204
2
~
1
E
4
2
3
D
2~5 9
226 8
2~7 F
228
260
2~9
~61
230
231
262
263
2j2
257
~
2'~
~59
1
0
1
7
0
1
1
U
1
6
~64
233
234
235
265
266
267
2~6
268
237
269
2j8
21U
2J9
271
240
272
2~tl___t.1 'J
242
274
243
,15
244
27c
245
271
246
27b
L41___
__ _i.l~_ __
246
280
249
2bl
250
282
251
28~
252
284
253
2d5
2 54
266
25 2 Q
2d7
"0 9
2Sa
4
20~ 5
206 6
207 E
208
2 09 6
210 5
,11 1
212 0
~13
214 1
215 0
21~ 2
~11 C
21d
219 7
2~O 6
i~l d
22;' F
2,3
224 1
2d9
290
291
292
~93
294
29~
2~6
291
298
2~9
300
~
30~
303
304
3' U5
30b
307
30~
309
310
~ll
312
313
314
~15
310
317
31 tS
-31 'i
320
-~---------------------------.---'
~---------------------------------------,
.. ~,._.~_ .... ,_"~ •• _ _ _,_~ .... ~...._~._ _ _ _ _ _ ~
~
_ _. _ _ _ _ _ ~ _ _ _ _ _ _ _ .• ,
l~____"__.."___ .,__""__c,c"-:"__
\__
"_~
I\~
r--'~-----~
C
. - " .. ~~ _ _ _ _ _ _ _ , _ _ _ _ _ _, _ _ _ _ _ _ _ . . _ _ '~' _ _ .
r
Figure I-a
-""--.'--,----,,----------
~-
~
,------------~o--~
_--I
•.
,.·"·"...""·7'~
__ ...... _ • ...."."- _ _ ,--.... _ _ _ _
~_
......."'"_ ..... _ _
~
___, ...
~
........._ _ _ _ r _ _ _ _ _ _ •
... _~ . . . . . . . _ - _ .......
" .. _ _ _ _ _ _ _ _ _ _ _ _ _ _ • _ _
- ;Wefm?n
DISK STUDENT RECORD FORMAT
C
Word
0
1-21
22-42
43-62
63-80
81-85
86-105
106-113
114
115-116
117-119
120-121
122-129
130-135
136-145
146
147-148
149
150-151
152
153-154
. ISS
. 156-158
159-160
161
162
163-164
165-166
167-175
176-179
180-182
183-185
186-188
189-248
o
249-251
252-254
255-258
259-261
262-264
265-267
Description
Student Name
Parents' Name
Address
City and State
Zip Code
Local addres s
Local telephone number
Graduation code
Hours being taken
High School rank
High school courses
High school average
Date of birth
ACT scores
Dormitory code
First time status
Probation code
Number of cours es being taken
Special student code
Religious preference code
Bill ing addres s code
High school code
State code
Year in school
Boarding status
Major code
Action code
Blank
Cumulative quality point ratio
Cumulative hours
Cumulative credits
Cumulative quality points
Course numbers and grades
In groups of five:
1-3 course number
4 grade in cours e
5 former grade if repeating
Rank in cIa s s
Number in class
Semester quality point ratio
Semester hours
Semester credits
Semester quality points
Figure 1-b
_ ;;4
-
_roo"!
'-~~-~'---"----""'--~--'-' ~----
.,,-,,--_._------------
o
CARD INPUT FORMATS
Student Name
1 Stu
C$.4
:3
~
+I
II)
....
....
'0
CleO
....
tJt)
'0 .0
e
~
+'
+I
rf rf
:SO':
5U)z
CI)
CI)
+'
t.l::3
....
~
=
"""
0
C
':>
a
u
~ Q..
o~oooooooooooooooooooooooooooooooooooooooooooooooooooo00000000000000000000000000
1
8
J
,
1
4
5 6 7
Stu
a
9 '011
11 13, ,. 15
~s '1 !3 19,"0~' 77. ~r "4 1~ "'6 '7
•
--
_J • • , - .
Parent's Same
."
" 111.LO)1)1.o·'
~
!CJI18·940~1414H4~5464'~8~95051525JS45S5657~5nObl 2636" ~66F.i !69'on'213741S76nl'H!~
Address
Zip
City & State
I
I
9 Stu
Loe al Add resS
Phone
No.
~~~~~~~~~~~~~~~~~~~!~~!~~~~~~!~~~~!:~~~!~!~~!!~~~:~!~~:o::~!~~~!~:~~!:~:~!!:!!:!
COURse; DESCRIPTION
0810'8008000000000000000000000000000000000000000000000OOOOOOOOOOOOOOOO'~OOOOO.tl
l2J45'1,."n~DUeMnGd~nnn~~~V3N~~n~~~.nu~.~~c«~«VqUW~~~~fi~~UU~~QQMU"Van~nnn~~~nna
, 11111
11 1 111111 1111 1111 \ 1 1'1 ttl
" 222222
222222222222222222222222
333333
"'444444
"
Christian Brothers College
Memphis, Tennessee
333l33333333333331J3331!
44444444444444444444~444~
555555555'555555555~S5t~
'.5ItS&S
STUDENT GRADE CARD
&666616'6'66'66("6666~'
,
,-,'\
111711~~~~~~~~~~~~~~~~~~~~~7771111171717711111ii1~1 •
Figure 2
.~--
-----~-------
-----._--- -'-
---
--
••
MID-TERti
GRADE
----~------~-----------
--------
____~
"My
o
0
. Appendix A.
!W
-
--
"1TIliWjUif nh5
'j ......
I("f'tftiib in ..... t 1" iiti 'tiiW';bt, '&#dt#"j 'bNttWh#tirl':ittw
Programs used to set up and maintain student file
RDNAM
Read in name cards to initialize students
RDATA
Read in students data
RDSKD
Read in students schedules I any order
CTSEC
Count setions and prepare class lists
PRLST
Print class lists
INFO
Print student information
ADDCS
Adds and Drops
COREC
To correct student information
ADNAM
To put a new student in the file
BILLS
To prepare bills
BILSM
Billing summary
PRTOT
To print totals in each section
ENROL
Enrollment statistics
DRTRY
Print student directory
I
I
,
o
307
'...ffi.f/f... .ib.... ,t,$. . .l.J$g
.\\!i$l6!!¥$.- ,;·iiG;ik!liM·M.4;;:i4QAi. .4•• :r#>I¥""'F¥-.,;;;;r.,.
.
.se. :"'1.P..3i-iA·,,,,· \
.\ ;,; ...;#: .. 4!LfAi·",iwfuibii'i,.~'bt!Ilil!t'.m1'tfJiiiJi!.U¥L"'i&ii&.ZZ2
..44 .. &$
.....
N
Programs used at report time.
Appendix B.
o
PCHGR
To punch grade cards and print grade report forms
CLGRS
To clear grade area and set Orientation mark
RDGRS
To read in grades
CKGRS
Check to see if all grades are present
CORGR
To correct missing grades
GRSUM
To prepare grade totals
GRSM2
To prepare grade totals (second half)
PRGSM
To print grade totals and statistics
CTACN
To separate and store actions
PRDL
To print Dean's list
FALUR
To prepare failure list
CLSTD
To determine class standing
o
o
30P
--.--.~---.-
...
-
.. -----..
--.-.----
--
-._ ....
__ ._..__ .• _ - - - - - _.._---_.. _ .. _. __.._-_._--_•...- - - -
o
o
Appendix C.
V,tility
~rog~ams
RDCRS
Rea~
RDTCH
Read in teachers' names
CTTCH
Prepare teachers' schedule
PRTSK
To print teachers' schedule
RDABM
Read in abbreviated majors
RDACN
Read in actions
RDEPT
To read in department names
RDPAB
To read in department abbreviations (4 letters)
RDPAA
To read in department abbreviations abbreviated
(2 letters)
RDFIL
To read in entire student file from four BOAl cards
RDMKS
To read in and store the 12 possible marks
PCHCD
To punch class cards
in master course cards
Print list of all courses
o
PRFIL
Print out student files
PCHFL
To punch entire student file
STPRB
To set probation code to zero
3G9
--~
Appendix D.
-~
,
..., , , . _ . , - - -
o
Mark Conversion
The twelve possible marks are stored in a file in a specified order.
In the student file they are given a code letter as follows:
Mark Number
Mark
1
2
3
4
5
Mark Code
A
B
C
D
F
S
U
A
8
I
H
9
WP
WF
Ex
Au
I
6
7
10
11
12
The mark number is read in.
B
C
D
E
F
G
J
K
L
o
The following steps set up the mark
code:
IF(MKNUM 1
10) 1,2,2
MCODE + -16320 + 256*MKNUM
GO TO 3
2
MCODE = -14528 + 256*MKNUM
3
CONTINUE
The reverse process is used to determine the mark number from the
mark code when printing reports.
o
3/0
...-._ ...
-
....
_ - - ..
"--'--~
--------------------------------,
o
Appendix E.
SUBROUTINES
G ET(RLVAR, JFLD , I, J , ADJST)
To get a real variable from an array.
IGET(INTVR, JFLD, I, J)
To get an integer variable from an array.
PUT(RLVAR, JFLD, I,
n
IPUT(TNTVR, JFLD, I,
o
To store a real variable in an array.
n
To store an integer variable in an array.
RSEEK(RLVAR,RARAY,N)
To locate position of a real variable
in an ordered real array.
ISEEK(INTVA, IARAY, N)
To locate the pos ition of an integer variable
in an ordered integer array.
SORT(RARAY, N RMODE)
To order a real array.
IS ORT(IARAY , N MODE)
To order an integer array.
PACK(LIST, I J, LPKD, K)
I
To convert Al format to A2.
UN PAC (LPKD, I, J, LIST, K)
To convert A2 format to Al.
I
I
o
3/1
M\ ..r. ..v«up.?A1iC&&!i!4AQiIWZ
4
.L ...
Q
#41#444 .. .4 . . #4 ....
·-M-
..
o
1.
Name of Prime Committee:
Installation Management Project
2.
Subject:
"Use of Psychological Tests In
Selecting Programmers"
3.
Speaker's Name:
David B. Mayer, (speaker)
co-author: A. W. Stalnaker
4.
Company Speaker Represents:
International Business Machines
and Georgia Institute of Technology
5.
Mailing address and Phone No.:
IBM Thomas J. Watson Research Center
P. O. Box 218
Yorktown Heights, New York 10598
(914) 945-1708
6.
Day and Time of Speech:
Tuesday, December 12, 1967
8:30 - 10:00 a.m.
7.
Number of Pages of Text and
Number of Pages of Graphics
Included:
21 Pages of Text
3 Pages of Graphics
o
c
3/2\
c
The Use of Psychological Tests In Selecting
Computer Programmers
David B. Hayer, IBH - Watson Research Center
Yorktown Heights, New Yotk
Ashford W. Stalnaker - Georgia Institute of Technology
Atlanta, Georgia
o
A presentation for the COMMON Organization, San Francisco,
December 12, 1967. Adapted from the tutorial entitled "Computer
Personnel Reseaich - Issues and Progress in the 60's" given at
the Fifth Annual Conference on Computer Personnel Research,
University of Maryland, June 26, 1967, sponsored by ACM-SIG/CPR.
The first session of the Conference was devoted to a review of
SIG/CPR's progress in research since 1962 and some thoughts on
problems which still face us. This review was in the form of
a dialogue between a computer center manager, David B. Mayer,
and a management scientist, Professor Ashford W. Stalnaker.
:;:r~.L"".f,:,j••.i •..tL,.,,.., .
""'!!.""
41"
4. ,g.\.
1.¥.ii'."'.~'Il'@i¥
., .. _,i'!!''''''tUiI''''".'..-1:p.;""'.,;,..,.,
..¥;'!':"!".,",.....
1t..fOO.M1ilffii.
....
", • .'!l!!!;,,$.!m!W:r!iM,&i\IIIiIlIillll
..
_il!ll!
•.C~."'I!I....
!Itl_,_ _ _.......
¥fb?§_
.• _ _ _. ..,..,""".
T4~"""+~,#,.,..",!'i"\1,·,i?,iF.4iIi"""1'i"",f¥.\"""··fu:":!l",\L~_"~~,,,T'l;"""dA1][t::IrnLlA&"""'ru;:;;""'·ujl;:uliiii"""._"'"',r''''''n"i..
.....,.liii/iIiIi_
......__•_ _ _ _ _ _ _ _ _ __
II''I,,
,'
I
Maye~:
Good morning ladies and gentlemen.
I represent both IBM and the ACM Special Interest Group on Computer
Personnel Research. Originally, this group was known as CPRG (Computer
Personnel Research Group); founded back in 1962 as a result of some discussions
at the RAND Corporation. I guess one might say that psychologist Robert
Reinstedt was a little lonesome for the company of some other researchers.
He had been investigating the problem of programmer selection and discovered
that there was indeed very little research on this subject -- except in the
hands of a few colleagues in his general professional area, namely, Dallis
Perry at the Systems Development Corporation, Professors Raymond Berger and
James Rigney at USC, Jim Tupac at RAND, and Dr. Sherwood Peres who was at the
Sandia Corporation. He called these people together and they laid out a
charter for CPRG in September 1962 at the American Psychological Association
meeting. The group decided upon the final design of a test battery to be
administered across the country to as many installations and experienced
programmers as they possibly could find. The test battery consisted of the
Programmer's Aptitude Test, (better known as the PAT) by Hughes and McNamara
of IBM; the Strong Vocational Interest Blank; and a special trial test named
the Test of Sequential Instructions. There was some personal background
material covering education experience, and so forth. The results of this
test battery will be included in the body of my talk, but at this point it
suffices to note that we completed it successfully, and we calculated
correlations on the data, obtained much information, and initiated the kind
of research that CPRG performs.
The initial researchers went on from that point to develop a Programmer's
Appraisal Instrument, which is an evaluation device. This was the result of
Dr. Sid Fine's work with Robert Dickmann (now SIG/CPR Chairman) at Johns
Hopkins University. It was tested there at the Applied Physics Laboratory,
and at 24 other installations.
o
o
The next step in our research was aimed at advancing the state-of-the-art
of interest tests. The Strong Vocational Interest Blank had been revised
(this had nothing to do with CPRG) and, after additional testing, Dallis Perry
developed a key for programming as a separate occupation.
In 1966 ACM approached us -- they thought that we were doing so well
that we should join the computing fraternity, since originally CPRG was composed
primarily of psychologists with only a sprinkling of computer managers. We
agreed that joining forces with ACM allowed CPRG's research to be enhanced
through broader contacts and outlets. Hence, in this survey paper today,
I would like to detail some of that research history, and some of the
existing issues as we see them at this time.
First, permit me to tell you a bit about the use of some of these tests
in selecting computer personnel.
Figure 1 summarizes the results of the Dickmann survey of 1966 which
was reported at the Fourth Annual Conference (1). This figure indicates
that 483 firms in the United States and another 98 in Canada participated
in the survey. In the US 68 percent used tests in some form or another
for selection. This corresponds very closely to 72 percent in Canada. The
number of programmer-analysts actually employed by these organizations was
o
c
over 23 thousand in the United States, with another thousand in Canada.
The number of people who are needed in the forthcoming year as of the time
of this survey was another 25 percent.
Figure 1 also shows the composition of the sample by industry groups.
The kinds of programming being performed were basically of four major
types - BUSINESS, SCIENTIFIC, SOFTWARE or MILITARY, or combinations of these
(Figure 2). Of them, the largest percentages were in business computations
with scientific applications coming in somewhat lower. Military and software
programming were small by comparison.
o
What kind of programmers are there? We classified them at the
time the survey was designed into four major categories: a programmer who
was essentially a junior or trainee; the second one was called the experienced
programmer; a third level we called the system analyst trainee; and the fourth
one was termed the experienced systems analyst. Figure 3 answers the question
as to what kind of education is demanded by the various institutions or
organizations in their hiring practices. In the United States it tended
toward having some college training or a degree - over 50 percent of the
US sample, especially for the programmer-trainee. This tendency is
a little more emphasized as you move up the experience ladder. Canada's
educational requirements were somewhat lower, possibly because they do not
have as large a college population from which to recruit. In Canada, 65
percent of the programmer trainees had only a high school education and the
remainder had some college or above. If you consider the experienced Canadian
systems analysts, you will find that 28 percent had a high school education
or better. However, 37 percent were not reported, so we can only generalize
about the true proportions in this situation. Canada, therefore, is drawing
on its resources of personnel in accordance with what they have available.
Tests were used in many of these organizations, but they were used
differently depending on whether the firm required much education, or little.
Or to put it in reverse, possibly tests were NOT used in many cases, and
hence the educational requirement was increased to compensate. Taking
one category as an example, the systems analyst trainee, almost 50 percent
were required to be college graduates if tests were not used; if a test were
used, only 39 percent were required to have college degrees. This pattern
repeats itself throughout Figure 4. We will not dwell on this except to
note that tests are used in many cases in conjunction with educational
requirements, but in differing degrees.
o
What types of tests are used in these two countries? The tests
reported used were broken down into four major classifications as shown in
Figures 5 and 6. The first major classification is the general intelligence
test with the most commonly used being the Wonderlic Personnel Test -6,0 organizations in the US and 7 in Canada. 13 of these organizations had
undertaken validation studies of this test. Whether the validation studies
consisted of actual on-the-job performance validation or training validation
was not indicated in the survey. The other two principal tests used are
general intelligence tests.
The second type is the aptitude test, which is being used as if it
isolates programming as a separate and special aptitude. The IBM Programmer
Aptitude Test, PAT, is by far the most commonly used test in both countries:
over 282 organizations in the United States -- approximately 83 percent, and
67 in Canada. Of the remaining aptitude tests, the National Cash Register
Test, ~he Science Research Associates Test Battery and several others were
used, but nowhere nearly as widely as the PAT. The Federal Service Entrance
Examination is not a true aptitude test - it is a test, I believe, that is
given to almost any prospective federal service employee, for many different
positions.
o
The two other types found in the survey were the personality tests
and the interest tests. Very few of them are used. Personality tests were
being used by only 10 or 15 organizations and very sparingly at that. For
example, the Thurstone Temperament Schedule and the Activity Vector Analysis
were each used by only three organizations; several others were used in
varying degrees.
For the interest tests, the Kuder Preference Record was cited by only
two organizations. Interestingly enough to me, neither Strong Vocational
Interest Blank (SVIB, original or revised version) was mentioned by any
organization. Yet it is one of the few for which there exists a key for
programming. It is hoped that SIG/CPR will have some educational effect by
bringing the value of SVIB to the attention of those responsible for personnel
selection.
Among the other tests is the 1401 Autocoder exam, which was developed by
Computer Usage Corporation. This test is a form of the Logical Analysis
Device developed by Langmuir at the Psychological Corporation of America.
This latter test, known as the LAD, was not cited at all, however.
Let us take a look at the use of the PAT for a moment. (See Figure 7). 282
organizations use it in the United States. 128 of them use it in combination
with some other test. 154 use it alone. As for position levels at which it
is used, for the programmer trainee - 278 organizations gave the PAT; for
the experienced programmer ---and this is always a delicate subject for a
computer manager who is interviewing candidates--- there were 138 organizations; for systems analyst trainees - 142; and for experienced systems
analysts - only 87. Of these, 71 organizations had performed validation
studies - that is, how the performance of the programmer compared to his
score on the PAT. 22 organizations, or a little less than 10 percent,
actually discontinued the use of the PAT for various reasons.
This completes our summary of the Dickmann survey, and I think it would
be well worth while to go into some of these tests and describe them and
relate them to a sample test which you will take in a few minutes.
Let us consider some of the several tests that have been used in various
CPRG studies ~-- a sample of some of these were included in the small
test battery which you just completed. In the original CPRG national survey (6),
three cognitive devices were included. These were the PAT, the Strong
Vocational Interest Blank, and the Test of Sequential Instructions. The
first of these, as was indicated by the Dickmann survey, is by far the
most popular selection instrument in both the· US and Canada. The results
of the first CPRG study indicated positive correlations between
o
\
3/b
C':·~
'i
the PAT score and actual performance only in a small number of cases. In
the overall summary of the report, no correlation was found between the PAT
and supervisory rankings of performance. The PAT has also been used in two
studies subsequent to the CPRG national survey. The first was by Biamonte (11)
at NYU working with a group of non-credit programming students, and the second
by Gotterer and Stalnaker (12) at Georgia Tech with several groups of undergraduates enrolled in a computer course which had as a major component
programming and systems analysis. In the latter case, as was true in the
national survey, no correlation was found between PAT scores and performance
in a training situation. We emphasize that the work at Georgia Tech was
strictly in a training situation and in no way relates to subsequent performance
in an actual programming assignment.
Does such a thing as a programming aptitude really exist? Well, I have
been trying to find that out from CPRG for several years. You know, after
over a thousand interviews -- which I'm told is the worst way to find out
whether they are a programmer or have potential -- and approximately 200
people hired over my signature, I would say that I can detect what one might
call an X factor; it's probably called aptitude. I do not know of what it
consists; all I can say is that a particular candidate sitting before me
seems to possess "it" and will succeed in the -programming art.
o
CPRG seriously questions that there is, as far as the training
situation is concerned, any indication of a specific programming aptitude.
Our interest was generated by the repeated occasions which were observed at
Georgia Institute of Technology, wherein a truly marginal student -- a
student who was only barely able to maintain satisfactory status in school
was able to succeed in developing a rather sophisticated programming skill
and also a sophisticated approach to systems analysis.
Let us now look at some of the CPRG results with regard to the PAT (6).
As I mentioned, only in a limited number of cases was there a significant
correlation between PAT results and ranked performance. However, I feel that
one of the points that we should pay special attention to is the sort of crossovers or the reversals we get in terms of the PAT. We will notice among the
business programmers (Figure 13) who are graded in the upper half with regard
to their performance, 42 percent of them scored 44 or below the PAT. On the
other hand, among those who were rated in the lower half in regard to their
performances, 49 percent of these scored 69 or above on the PAT. We might note
too that the relationship in this case could be curvilinear. In the scientific
group, Figure 14, the relationship is approximately linear and the degree
of the reversals not as large as in the case of the business group. Here
we note that 31 percent of those who are rated in the upper half with
regard to the performance scored 44 or below, and 34 percent who scored in
the lower half in terms of their performance scored 69 or above.
In the sample there were 534 programmers; 301 of them were scientific
programmers and the remainder were in business programming. There were
25 different installations and companies.
A second component of the national study (6) was the Test of Sequential
Instructions. The author of the TSI states that it was designed to
3/?
-·5224 .. .#...4.4.
hlP . . . AL44 ..
;,
,-:.,4
show that any test that in some way measures a form of logical reasoning
will in some sense indicate the level of performance in programming jobs and
will also indicate in some sense the intelligenc.e of the individual. This
hypothesis is borne out by the results of the national study in which the
correlations between the TSI and the PAT were in many cases quite significant.
C
:1
In my scientific programming group I obtained a correlation coefficient
of .70 between the PAT test results and my supervisor's rankings. But
interestingly enough, on the TSI, the correlation coefficient was .71. So
I naturally asked myself, what was I doing that was right? Could I interchange the TSI with the PAT as part of selection procedure for programmers?
The DTSI, to me, tests the ability to do multiple tasks simultaneously.
To perform well on that test you are holding one task in the back of your
mind while another task is being performed in the foreground. Then, we build
up to 3, 4 and 5, in fact, in the real TSI, I think we have 7 tasks running
simultaneously. The PAT, to me, has no such attribute. The PAT does test
other components which are required in programming --- numerical capability,
spatial relationships, and such. I suspect the two tests supplementary rather
than interchangeable.
The third component of the CPRG national the Strong Vocational Inventory
Blank (SVIB). The SVIB is a test that has been in use for a great number of
years, primarily though, in the vocational counseling area. The purpose of
the SVIB is to elicit information regarding the interests of the testee.
The interests are then compared to the interests of people who have been
successful in several occupational groups, the hypothesis being that if a
person has interests similar to those successful in certain occupations,
there may be some motivation to enter this occupational area. It should
be noted that the SVIB is not claimed by its author to predict performance
on the job; instead, it only elicits information regarding interests.
("'\
\"'J/
In Figure 15 we compare the interests of computer personnel to the
public in general in regard to certain answers to questions on the SVIB.
The SVIB contains 400 items. Notice specifically the item on progressive
people: 41 percent of computer personnel like this type of person -- among
the general public 85 percent of these people like people with this outlook.
On the ,other hand, there is a great flocking among computer personnel to
conseryative people. 84 percent prefer these, while among men in general,
only 56 percent. Another example was thrifty people: 45 percent of computer
personnel stated they liked this attribute. They are much better liked
by people in general -- 74 percent.
Can we look at individual items in the SVIB to see if these indicate
anything in regard to the general interest pattern that should be the pattern
of a successful programmer. The answer to this is a decided NO. We mentioned
two categories -- progressive and conservative people. The results shown by
the SVIB are reversed in work by Biamonte (7) at NYU in which he specifically
considered attitudes. He shows there were negative correlations between
training success and such attributes as dogmatism, conservatism and
authoritarianism, a finding which would indicate the reverse of your
conjecture. The point is that the SVIB questions when taken out of context
have no meaning. One must analyze the complete 400 questions in order to
elicit anything regarding the total interest pattern of the individual.
-
~-----""'
~~-
~
o
The normal scoring of the SVIB is quite complex because it has
to be scored for all occupations. It is possible, I might mention, to
hand-score for a particular occupational group. For instance, since
the interest of this group is the computer programmer, it could be that
the keys could be obtained and thus we could hand-score for this specific
occupation. However, I would like to warn against this; the SVIB is
meaningful only in terms of its total content. When we take an occupation
or a question out of context, the results are questionable to say the
least.
I have mentioned the existence of the programmer scoring key for
the revised SVIB. The series of questions that you have in your sample
battery are from the old SVIB which was the one used in the original
CPRG national study. Subsequently, Dallis Perry (8) undertook another
national study in which he used the revised SVIB. He also worked with
a larger sample than that included in the original CPRG study. Based
upon this work, Perry has developed the programmer scoring key which
is now available to all users of the SVIB.
I think now we should move onto the topic of programmer evaluation -how can we determine if a programmer is actually effective on the job.
The first research in this area was reported by CPRG. This was the Programmer
appraisal Instrument (4) developed at the Applied Physics Laboratory, under
the direction of Bob Dickmann and Sid Fine. This is a multi-dimensional instrument,
which in many ways appears to be more concerned with what might be regarded
as a professional programmer rather than the operating programmer - at least
I think this is sort of a summary of remarks made in the evaluation of the
PAl. There is a considerable emphasis on professional activities and this,
in some cases, has led to resistance. It is composed of four specific areas:
professional preparation and activity, programming competence, dealing with
people and adapting to the job. This instrument was validated by Bob Dickmann
but is not believed to be widely used at this time.
o
The PAl asks a number of items to be scored numerically, such as: how
many societies does he belong to? and how old is he? does he give some on-thejob training -- a lot, or not at all? A supervisor finds that" the items do not
really cover the subject of programming, or programming capability. Supervisors generally shy away from question-and-answer procedures for appraisal
of programmers. Practically none have actually put this PAl or equivalent
ratings into effect. Progressive as I am - I haven't either. My project
leaders were very resistant to it. They prefer to use subjective techniques,
such as their impression of the programmer's output; sometimes they actually
read his programs. But a formal document of this type they resist. In
evaluating a coder for upgrading to programmer, a better procedure would be
to temporarily take a coder out of his current category and put him into
direct programming tasks and to observe his programming capability rather
than evaluating him through a questionnaire. Additionally, I would like to
have some kind of test which will actually show his level of competence.
A test that is concerned with programming ability to indicate readiness for
the move from coder or programmer to senior programmer should be implemented.
There is one further thing that must be done, and that is, as Dr. Paul Herwitz
(13) of IBM has recently stated, the only way a supervisor can tell what a
3/~
i
L# . __ # _ Ai_
-
~
_.
- - ._.
---- .. _.- ....-- .....-...
~"
.................. _....
~-
programmer is doing is by being knowledgeable about the code that he has written.
In other words, he must read the program, and very few supervisors do. That
would be the first step I would say towards proper evaluation.
~
The second step would be to give a proficiency test, an objective one.
In terms of such an objective test, Berger (10) at USC has developed a test
which is called The Basic Programming Knowledge Test (BPKT). This test was
developed and validated with a group of naval training programmers and also
with some outside agencies such as RAND, SDC, etc. Specifically, the test
is designed to evaluate six different abilities: first, logic estimation and
analysis; second, flow diagramming; third, programming constraints; fourth,
coding operations; fifth, program testing and checking; and sixth, documentation. Not only does the test evaluate the person's performance in these
areas, but it is also designed to elicit information regarding his basic
knowledge of the areas. Similar tests for different types of programmers at
different levels would be very effective evaluators. The problem, of course,
is the resistance the programmers show particularly as it applies to experienced
ones who are very much in demand.
We have cited the PAT, we have cited the Strong Vocational Interest Blank,
we have cited the TSI.
Now then, the PAT is an aptitude test, presumably, and the SVIB is an
interests test; the TSI presumably tests a kind of logical capability. If I
use all these tests and get good scores on all of them, does it mean I
selected a good programmer? The answer at the moment is - no. I don't think
it means that I have selected a good programmer - but it may well increase the
probability that I have selected one. We have not been able to show at this
point, with the exception of the recurring interest pattern of programmer
personnel, any strong indication of substantially increasing the probability
of correctly selecting a programmer by the use of this battery.
Very probably if you would use all of the tests to select an individual,
you can obtain a person who has a high probability of successfully completing
your training program. Whether this individual is going to like programming
or will possess the motivation that will allow him to take the successful
training onto the job site is a question that is not yet answered.
Well, then I think we come to a small denouement, and it is that if we
look at the past five years of computer personnel research, the major effort
has been in the development of two sets of testing predictors: (Figure 16)
testing for training success, and testing for job performance. We have said
that we have good predictors for the training phase and questionable ones
for the working phase. Temperament and motivation tests are frequently considered
as good predictors of training success and job performance, but we have no
research using them. Finally, of the "work sample" tests, the only proficiency
test I know of is Berger's Basic Programmer Knowledge Test. I understand
that he will publish a version of it in the public domain, by permission of
the Navy. For evaluation procedure, at the moment this consists of the Programmer
Appraisal Instrument (PAl) which we developed but have not used. This I think
sums up the situation at the moment. A more concise summary of our current
knowledge is that the more intelligent person you can find, the better programmer
~
you can probably get •.. which makes all of us here today good programmers!
c
We might mention this point, that a test does exist that claims
to be in this general area. This is the DPMA Certification program.
However, I think we should note that it is quite questionable that there
exists a relationship between this test and what we might identify as any
level of programming skill. The DPMA test is primarily a knowledge test, but
not at a very sophisticated level. It is my understanding that it is general
knowledge -- at least that is my impression from the groups that were formed
to study for it.
To summarize:
We have called for several new procedures which should
be investigated in the years to come. Despite the fact that I have been told
time and time again by my psychologist friends that my or other people's
interviewing technique cannot be a really effective device, it is still the
one that I, as a computer manager, use in at least 50 percent of my evaluation
of the candidate. Hopefully, if these new procedures can be developed, both
I and my fellow managers will be better able. to select, train, evaluate, and
reward our computer personnel. Thank you everyone for your kind attention
today.
o
o
BJ..,/
BIBLIOGRAPHY
',
C
I
1.
Dickmann, R. A., "A Survey of Computer Personnel Selection Methodology,"
Proceedings of the Fourth Annual Computer Personnel Research Conference
pp. 15-27.
2.
Lothridge, Charles, "Levels of Classifying Data-Processing Personnel,"
Proceedings of the Second Annual Conference Computer Personnel Research
Group, pp. 13-28.
3.
Berger, R.M. and Wilson, R. C., "The Development of Programmer Evaluation
Measures," Proceedings of the Third Annual Computer Personnel Research
Conference, pp. 6-17.
4.
Dickmann, R.A., "A Programmer Appraisal Instrument," Proceedings of the
Second Annual Conference Computer Personnel Research Gr ou...p.. , pp. 45-64.
5.
Bairdain, E.F., "Research Studies of Programmers and Programming,"
IBM Corporation, 1964 pp. 78, 136, 62.
6.
Reinstedt, R.N. et al., Computer Personnel Research Group Programmer
Performance Prediction Study (RM-4033-PR), The RAND Corporation, Santa
Monica, California, March 1964.
7.
Biamonte, A.J., "A Study of the Effect of Attitudes on the Learning of
Computer Programming," Proceedings of the Third Annual Computer Personnel
Research Conference, pp. 68-74.
8.
Perry, D.K. and Cannon, W.M, "A Vocational Interest Scale for Computer
Programmers - Final Report," Proceedings of the Fourth Annual Computer
Personnel Research Conference, pp. 61-82.
9.
Stalnaker, A.W., "The Watson-Glaser Critical Thinking Appraisal as a
Predictor of Programming Performance," Proceedings of the Third Annual
Computer Personnel Research Conference, pp. 75-78.
10.
Berger, R.M. and Wilson, R.C., "Correlates of Programmer Proficiency,"
Proceedings of the Fourth Annual Computer Personnel Research Conference,
pp. 83-95.
11.
Biamonte, A.J., "Predicting Success in Programmer Training, Proceedings
of the Second Annual Conference, Computer Personnel Research Group, pp. 9-12
12.
Gotterer, M and Stalnaker, A.W., "Predicting Programmer Performance Among
Non-preselected Trainee Groups," Proceedings of the Second Annual Conference
Computer Personnel Research Group, pp. 29-44
13.
Herwitz, P.S., "Programmer Evaluation" GUIDE Organization Talk, Management
and Administration Committee, .1966 (IBM, Armonk, N. Y.), 12 pp.
14.
Rigney, J.W. and Berger, R.M., "Computer Personnel Selection and Criterion
Development: II. Description and Classification of Computer Programmer
and Analyst Jobs," Tech. Rpt. 37, Proj. NR 153-093, Dept. of Psych.,
U. of Southern Calif., Dec. 1963, pp. 26 ff.
"~---.-
.. - "
-----_.
- - _.•.
_ ...
_-- ........_._-_._.....
_._...._ - - - - - - _ . -
~\
o
C'
!,,!lim" "
Oi
United States
Organizations Participating
Programmer/Analysts Involved
Approximate Number Hired Each Year
Canada
483
98
23,636
1,083
5,317 (25%)
177 (20%)
United States
Number
C
Canada
Percent
Number
Percent
AIRCRAFT INDUSTRY (Industrial,
Aerospace)
47
10
3
3
ELECTRONIC INDUSTRY (Industrial,
Electrical-Electronic)
35
7
2
2
OTHER INDUSTRY (Petroleum, Metal,
Automotive, etc.)
120
25
34
35
FINANCE (Banks, Insurance Companies,
etc.)
81
17
21
22
RESEARCH (Non-profits, University
Labs, etc.)
90
19
10
10
GOVERNMENT (Federal, State, and
City Civil Service)
50
10
8
8
UTILITY AND OTHER NON-MANUFACTURING
CON'CERNS
60
12
20
20
483
100
98
100
FJtom
V~c.kmaY/.Y/.,
Re.6· 1
FIGURE 1: TYPE OF ORGANIZATIONS IN 1966 CPRG SURVEY
o
3J3
"M
""
Wrtt"it
. _ - - - _ . ".........._-, .. ... _-------_ ... "---,,
"
. ,,,._.-
..- - .
o
United States
Canada
186
58
Business and Scientific
84
11
Business and Scientific and Software
72
6
Scientific
44
6
Scientific and Software
34
1
Business and Software
33
1
Business and Scientific and Software and Military
12
0
Software
6
0
Military
4
0
Scientific and Software and Military
3
0
Business and Military
2
0
Business and Software and Military
1
0
Other
2
1
Business
FJtom V.-tc.fzmann, Re6·
FIGURE 2:
PROGRAMMING STAFF APPLICATIONS
0
c
Progrannner
Trainee
U.S.
None Specified
Canada
Experienced
ProErammer
U.S.
Canada
System Analyst Experienced Systems
Trainee
Analyst
U.S.
Canada
U.S.
Canada
9
14
9
10
7
4
8
7
High School
27
65
19
43
13
32
11
28
Some College
25
3
23
6
12
16
14
10
College Graduate
34
13
35
8
43
9
40
11
Graduate Degree
1
2
1
2
2
7
5
7
4
3
13
31
23
32
22
37
100%
100%
100%
100%
100%
100%
a o t Reported
100%
100%
F~om
FIGURE 3:
EVUCATIONAL
Viekmann,
RE~UIREMENTS
R~~.
1
o
Experienced
Programmer
Programmer
Trainee
Non-Test
Test
Non-Test
Systems Analyst Experienced Systems
Trainee
Analyst
Test
Non-Test
Test
Non-Tes·t
Test
6
10
6
10
5
8
7
8
High. School
16
32
6
25
4
17
3
15
Some College
27
24
19
25
10
14
7
17
College Graduate
37
32
53
27
50
39
48
36
3
1
4
0
4
1
10
3
11
1
12
13
27
21
25
21
100%
100%
100%
100%
100%
100%
100%
100%
None Specified
Graduate Degree
Not Reported
0
Fltom Viekma.nn, Re.tS • 1
FIGURE 4:
COMPARISON OF EVUCATIONAL REQUIREMENTS FOR ORGANIZATIONS USING
TESTS IN SELECTION VERSUS ORGANIZATIONS NOT USING TESTS
(UrU:te.d Sta.,tu Sample.)
o
,
WWP'"em!BWU"· PP"fP&§"*
fl
FREQUENCY OF USE*
United States Canada
TEST NAME
VALIDATION
Studies (Total)
GENERAL INTELLIGENCE TESTS
Wonderlic Personnel Test
Thurstone Test of Mental Alertness
Otis Tests (Unspecified)
School and College Ability Tests (SCAT)
Wesman Personnel Selection Te$t
Ship Destination Test
Lowry Lucier Reasoning Test Combination
Concept Mastery Test
Henmon Nelson Tests of Hental Ability
Schubert General Ability Battery
60
12
11
5
3
3
3
2
2
2
o
o
o
o
o
o
o
o
282
67
83
4
6
7
1
18
4
5
2
o
o
2
1
2
o
APTITUDE TESTS AND BATTERIES
c
IBM Programmer Aptitude Test
National Cash Register Programming
Aptitude Test (E5l)
Federal Service Entrance Exam
SRA Computer Programmer Aptitude Battery
(Burroughs Corp.)
Employee Aptitude Survey
Differential Aptitude Tests
Watson Glaser Critical Thinking
Appraisal
Short Employment Tests
Test of Sequential Instructions
Minnesota Clerical Test
Guilford Zimmerman Aptitude Survey
Bennett Mechanical Comprehension Test
9
13
o
5
7
6
o
o
5
3
1
3
4
1
3
1
1
1
1
o
o
o
o
o
5
2
2
2
2
*For Tests Used Two or Hore Times
FIGURE 5: TESTS USEV FOR INTERVIEWING
PROGRAMMER CANVIVATES:
I
3
2
!
0,'··
,.
FREQUENCY OF USE*
United States Canada
TEST NAME
VALIDATION
Studies (rqta1)
PERSONALITY TESTS
Thurstone Temperament Schedule
Activity Vector Analysis
Rohrer-Hib1er-Rep1ogle Personality Test
Humm-Wadsworth Temperament Scale
Edwards Personal Preference Schedule
Cleaner Self Description
Adaptability Test
Guilford Martin Inventory of Factors
Guilford Martin Temperament Profile Chart
1
1
1
0
0
0
0
0
0
0
0
0
1
0
0
0
1
1
2
o
2
4
3
o
o
o
o
o
2
3
3
2
2
2
2
2
1
3
INTEREST TESTS
Kuder Preference Record
OTHER TESTS
Manhattan Symbol (MAZE)
1401 Autocoder Exam
GCT
LOMA
Personagraph
2
2
2
o
1
1
o
*For Tests Used Two or More Times
F~om
FIGURE 6:
VicQmann,
TESTS USEV FOR INTERVIEWING
PROGRAMMER CANVIVATES:
II
R~~.
1
,
UNITED STATES
NUMBER OF ORGANIZATIONS USING:
IN COMBINATION WITH OTHER TESTS:
ALONE:
CANADA
282
67
128
154
18
49
68%
58%
PERCENT OF TOTAL SAMPLE:
93%
85%
PERCENT OF THOSE USING TESTS:
LEVELS OF TEST USE:
PROGRAMMER TRAINEES:
EXPERIENCED PROGRAMMERS:
SYSTEMS ANALYST TRAINEES:
EXPERIENCED SYSTEMS ANALYSTS:
VALIDATION STUDIES:
DISCONTINUATION:
278
138
142
87
67
27
32
14
71
22
12
F~om Vi~~mann,
FIGURE
7:
1966
CPRG SURVEY
IBM PROGRAMMER APTITUVE TEST
]F.M!.ld .. .#4# .. 74.244%.
m
o
Re~.
1
11
o
PAT Scores
Supervisors' Rankings
Percentage In Upper Half
Percentage In Lower Half
69 & Above
64%
57 - 68
45 - 56
63%
44 & Below
58%
FJtom RUnI.>t.e.dt, Re.6. 6
FIGURE 13:
RELATIONSHIP BETWEEN PAT SCORES ANV RANKINGS
-- BUSINESS GROUP
o
330
"IPUM922!illI;r#tdHw t!ln
C\
"
Supervisor's Rankings
PAT Scores
Percentage in Upper Half
69 & Above
Percentage in Lower Half
70%
57 - 68
45 - 56
44 & Below
fJtom Rein6:ted:t, Ren. 6
fIGURE 14:
RELATIONSHIP BETWEEN PAT SCORES ANV RANKINGS
--SCIENTIfIC GROUP
o
33J
F"'rrmmrrRr,..
I
Like
PROGRAMMERS
Indifferent Dislike
Like
GENERAL POPULATION
Indifferent Dislike
Aviator
67%
21%
12%
30%
36%
34%
Mathematics
90
8
3
69
20
11
Solving Mechanical
Puzzles
63
29
9
39
34
27
Giving First Aid
22
51
27
40
42
18
Progressive People
41
42
17
85
11
4
Conservative People
84
15
1
56
35
9
Energetic People
14
48
38
89
9
2
Thrifty People
45
44
11
74
22
4
Sell Machines
8
27
65
26
32
42
A
B
Few Details - Many Details
Technical
-
Supervision
,~-,,\
~~/
Indiffer- Prefer
ent
B
Prefer
A
Indifferent
Prefer
B
Prefer
A
18
27
55
36
28
36
65
16
19
34
19
47
FIGURE 15: STRONG VOCATIONAL INTEREST BLANK
INTERESTS OF PROGRAMMERS ANV GENERAL POPULATION BY PERCENTAGES
33~
U
JrW'N" RP- WTvue BittbfB f '
c·
TESTING - - - - - - I... TRAINING ---....,•• WORKING
PREDICTORS
GOOD
QUESTIONABLE
PROFICIENCY
Programmer's Aptitude Test (PAT)
BPKT
Buchannon Logical Test (BLT)
Strong Vocational Interest Blank (SVIB)
- Basic Programmer's
Knowledge Test
EVALUATORS
Temperament Tests (None)*
(PAl) Programmer's Appraisal
Instrument
Watson Glaser Critical Thinking Analysis
*Motivation, primarily
FIGURE 16:
5 YEARS OF COMPUTER PERSONNEL RESEARCH
ISSUES - '67
IN COMPUTER PERSONNEL
A
RES~ARCH
EFFECTIVE EVALUATION PROCEDURES
A STRATIFICATION OF SKILLS
•
NEW OBSERVATIONAL PROCEDURES
•
ROLE OF CREATIVITY IN PROGRAMMING
FIGURE 17:
ISSUES IN PERSONNEL RESEARCH - 1967
333
i'l
!~
r--\
Vi
PROGRAMMERS
Like
Indifferent
AUDIENCE
Dislike
Like
Indifferent
Dislike
Aviator
67%
21%
12%
85%
15%
0
Mathematics
90
8
3
99
0
1
Puzzles
63
29
9
25
65
10
First Aid
22
51
27
5
55
40
Progressives
41
42
17
70
27
3
Conservatives
84
15
1
5
80
15
Energetic People
14
48
38
75
24
1
Thrifty People
45
44
11
10
75
15
8
27
65
5
85
10
Few
Indiff.
Few
Indiff.
18
27
Selling Machines
,~'~
Like Details:
Technical
Technical
(vs. Supervision)
65
Many
15
55
Indiff. SUEervise
16
19
45
Technical
45
\V
Many
40
Indiff. SUEervise
10
45
FIGURE 18: A COMPARISON OF THE RESULTS ON SAMPLE QUESTIONS TAKEN
FROM THE SVIB, AS INVICATEV BY THE AUVIENCE POLL AT
COMMON, VEC. 12, 1967. MOST OF THE AUVIENCE (WHICH
WAS 75 PEOPLE) SAIV THEY WERE PROGRAMMING OR OTHER
.SUPERVISION, BUT A VERY LARGE PROPORTION ALSO
PROGRAMMEV WHILE SUPERVISING. THESE RESULTS ARE
VISPLAYEV ONLY FOR INTEREST: NO CONCLUSIONS FROM THE
SAMPLE OR INVIVIVUAL QUESTIONS MAY BE VRAWN.
VBM/12/67
c
",W"
MIWUg'W-g'WtHWlpUTIt"tlli
DTSI
c-
II
~ame
1
2
3
Total
3 Games
A.bel's
390
420
476
1286
~aker's
419
501
427
1347
~harley's
289
394
325
1006
Team
this
IB
In
lA
each
ID
word
IE
letter
2D
has
2A
and
2E
combination
3B
For
3E
example,
4A
the
5E
code 2A
6A 6B
paragraph,
8A
thus:
lID
beneath
6C
word
lOB
beginning
l2B
letters
13E
each
l6A
time
l6B
"a",
l7E
circle
l8A
the
l8B
15
20B
of
22C
otherwise,
24A
.4
halt
26A
your
26B
is
20C
atS:t.x., ,.0'
a
l6C
should
10E
with
l2C
the
l2D
number
3A
beneath
3C
previous
5B
first
7B
combination
8E
you
10D
use
24B
it.
3E
sentence
5C
sentence
7C
of
7D
the
9B
first
9C
of
9A
have
llA
circled
lIB
sentence
l6D
begins
l6E
of
l8D
than
20E
the
2lA
third
22E
word
23A
in
23B
the
24C
third
24D
for
26D
next
l2E
beginning
14D
greater
20D
search
26C
4f.HRf.#,.'JA 1~"
words
14C
codeletter
18C
the
22D
"a."
28B
¢ ~.J;,tzI,'.":r ...
all
l4B
of
14A
Also,
15D
letters
22B
.
code
8D
the
7A
code
2C
has
5D
this
7E
occurrence
9D
2C,
IIC
Q9
number
20A
;
In
6E
a
2B
code.
code
l3D
as
28A
th'e
8C
the.
5A
in
4E
41>
it.
6D
"code";
10C
Now;
l2A
0
"has"
word
4C
circle
8B
the
lOA
of
9E
the
4B
exercise
lC
word
24E
sentences
26E
its
18E
sentence,
l3A
first
19A
number
2lB
the
23C
in
25A
the
l5A
with
14E
with
17A
beginning
27A
vowel
l7C
word.
19B
25,
2lC
"w. "
l5C
except
l7D
Meanwhile,
19C
circle
2lD
the
2lE
if
19D
code
22A
sentence;
23E
sentence.
25C
with
27B
the
13C
letter
15B
any
l7B
previous
23D
this
25B
circle
l3B
"u",
27C
You
25D
may
25E
as
27D
well
27E
the
19E
DTSI
-2-
1
Game /I
2
3
Total
3 Games
Abel's
390
420
476
1286
Baker's
419
501
427
1347
Charley's
289
394
325
1006
Team
Observe
30A
scores
30D
of
30E
teams
3lD
in
3lE
the
32D
Baker's
33E
If
33D
team's
34A
then
35C
circle
3sD
word
37C
"the"
37D
·You
39B
the
3sE
"w"
4lC
letter
4lB
next
45D
sentence.
4sE
total
47E
"40A"
49D
and
49E
with
55B
second
SID
which
53D
an
55C
"n, "
55D
the
36C
of
42B
his
34E
this
42C
those
44D
last
46D
higher
48B
than
48C
Baker's
48D
Otherwise,
SOB
case
53E
in
S5E
you
43B
may
43C
And
43A
beginning
4sA
tasks,
46E
if
47A
total
48E
with
45B
Abel's
47B
then
49A
code
54C
underneath
S4D
r:". '\.
I;t..J'
the
4sC
team
47C
circle
49B
and "42B"
50E 51A
the
54B
codes
49C
unless
SIB
Abel's
52E
lowest
53A
all
54E
words
55A
sentence.
56B
·SI
~1i]
• 91
~Oi]
·~1
~~~
~l£
Oi]~
~ZZ
~9£
~£~
·lZ
·OZ
·61
Oi]£
O££
VO£
·i]1
·£1
·ZI
·11
·01
01~
the
4lA
circle
54A
·ll
OZ~
with
40E
than
52D
Hl~
a~~
the
37B
. beginning
40D
less
52C
ao~
O~~
of
37A
"36C."
39A
circle "·40B"
SOC
SOD
·9Z
. ~Z
·i]Z
·£Z
H9~
page.
33C
highest,
35B
was
52B
this
56A
at
32C
the
33B
game
52A
lowest
5lE
box
32B
code
38E
vowels,
44E
your
46C
3lC
circle
38D
sentence.
42D
for
46B
3lB
occurrence
36E
otherwise
38C
words
40C
th~ee-man
the
32A
of
33A
()
bowling
30C
the
3lA
second
35A
.second
36D
for
40B
all
44C
for
44B
"42A."
sOA
in
53C
end
42A
And
46A
was
48A
was
34D
under
36B
looking
40A
looking
44A
score
47D
game
34C
paragraph;
38B
the
4lE
at
4lD
halt
43E
game,
53B
this
38A
halt
39E
also
43D
Charley's
5lC
code
36A
in
37E
. now
39D
may
39C
third
34B
top
32E
the
30B
alZ
HlZ
ai]Z
V£Z
·6
·S
~61
.~
V61
aSl
Vll
~Z
·l
·9
.i]
.£
·Z
·1
0
"TIIWI"U??::" "" MW&iiHtiMi#fM"w-""r"iH 1 "
T,2o ''1-
Session No.
o
-~~-=------'--
iI~D
Session Title:
'I
L, P\
Chairman:
(l.
CVJ.
(i..
T~~ ~.~ S/3c,Q
R B f!.,QtJ GH
~. ~. G~
Speaker:
Topic::
~~J e-J.. ~J£. ~ 1130 - S/3bO
Sununary:
S~ ~
~
R
1/30
~.()
ter L )
1<4-
0-
fJ.A.-(.
~~ (~~
r~~ -J~ i ""1f,;:t:j
"-
S/3Go (;1{)
CJY
.-fc-.r-}£=--.Q...
:Z:-/$th
IfdV/JQur-
Speaker: _________________________________________________
Topic: ___________________________________________________
Summaryi ______________________________________________
~
o
Speaker: ______________________________________~--------Topic: ________________________________________________
Sununary: ________________________________________________
NO. OF ATTENDEES: )JeT
I< 67V~I/~O
ro
IJA /)
f)a-.&
:
'.
l
~
('
'-...~J u.. (.' '.-"
.
0.
1•1
.,
33)
·1.1'~
1..1 30
'As 'R rE .. mIAl~~
o
FoR. S 3~O
THE
USIIJ&.
S YNeHI.O~OUs
o
0,
-~
_._----
o
• 5
~
-
YAle\iR.o~ou..~
T1llt ~ s IV\ "
I
..
lte.eE.I" ~
( S,..1t)
o· ,:&'~l\R.Y
s 'I ~c." ..o ~Oc.L~
eorY\.W\, Y.Wl e. ~"lO N
tlSC.)
o
1130 Type I:
PR06-RAmS
STR
"St.ATJ.
BS'-
·Sc.RT.t
·SC:AT3
o
-REMOTE ~o8
ENTRY .. WOItt(
SrRTloA) Pflo'lAM
·'llO/:l2.GO trt.f1l'tt.~
S\l~sYSTEMS
o
me:,
SCHT..1
o
STiil
if 11 "JlILR 8LE: No W
*I
NTE~~CA.f'r
SLt15~OtA.T
Sl: k"\C-E
(xss)
tAlE
-* POINT-IO- -POINT
COWl WlUN I C.~I' 0 ~
~ Sta)\TC WEb 01t.
kE~sEl>
NETU)oltK
-* '1- OF- ~
-,
RllNS",'SS\ON
Col>E
(nrilX. ''I )11-'»
C"1t'R.ll~-( E~S)
"
.:p,,4J1ig&1$& M. ,
SIi!E: 107'
,$2ZZZA ...
. ..... 44
on.
•
_
-
•
,
.'
I I"
",',. """
..
"II "" ",,,"; ..,.tL·.">LI1II.oD'r-I,,:AI=,,."':;;illJaw~....
..
ST~
SCl\T.1
StA.Y"Ol.,.E»
.:z:SM
sYSTEMS:
5/3'0)
MOl>EL ~fOJH,~'5i'~ 7St&)\Tt\ T\\E ~-Sft\ 2. 70 I
DftTil lfbHPI E~ U~IT
~Q\l.\ -P~EJ)
W'I""
0
Sy~C\\ ro~OtlS
])1)11
Db ~¥'·E 1l. - TYP E: I
.. ~"BM
5/3"0,
EQ\l\~?Eb
C o~ W\..
mobEL ;1.0
\O\T" ,t\E
\L" \e..'AT
Rbll"PTElt
10 A\ S
(-%073) 0
3l/~
-------------------------
-
c
SC 'AT
i
SoFT\l)}TRE
STR
SYSTems:
$/3'0
'8Os/s ftS
TYPE I
PRo(,.Rllm
RU H 11. ~8LE: ,\)ow
5/3'0
Mo1);lO
CXOCS
T'IPE l: -PRO.ll Rm
R Vll\L.RBLE.: ~ow
5/3'0
o
])OS/o~ Sl1lllm
,.ype,.II: ?~O&~llM
R'J ft IL lTBLE: NOW
o
o
o
'"fI'M
o
SC~T.i
t.fl LLEJ)
STR
BY FI N
L'RtJ'IlRfE
fJSSEW\&LY
PRot;. tt" W\
0
LIBF
be
})C-.
be
SC,RT.i
(Co~IR.OL PflR.lJ~
(r/o flnll JJ}])~~
(E'RRoR ~Cl~,tlE
Hl>1>R.ESS)
o
_.-
...
-~-----.---.~~-
~--~.-~------~
J:bEFIL
STR com.U ~ '(..BTIO~
,~ IBm FOllTRl1H
• 8eb, EBeb Ie, IINllR Y :r/D!
I
Type .m P~o~R.1tm
AVRIL-ABLE: /tocO
c
FILE $I:- 1130 -oJ. t. OO~
FoR. TME Z8M "~o
FOR.""9~ LtS£R.
o
o
SC~T4
.". R~ R'l.RBLE: Now
* .r~T~lkll'T
SE ~ V leE
SUBloll'" AI E
* Po I ~T" ,,0...
(Zs.s)
ft'" NT
1t S"-'IT~tlE~ DJt LE9SEb
o
~ET"'OR.K
1i' TR.1tHSn\'ss",~ eo:DE
.. A\OR."'9L Ea~J),e...
· F"U.~ n9Als"9~EtJ"
-If 51 ~ E:
1o,,"
s C. ~T a
"# "-IJla.ItB&.E:
B$(.\)
Now
-'I- %"TE~1tu.'T
S&Jt.",e.£
SI4~R.oilTIA\E (Z5S)
-If "'ULTI'" ?OINi
TRl1SIlTJl-tty STi1rIOtJ
LE RSE» ~'ET u>o 1t"-.
-If T1ttI ~s W\\SS It) J-) CD]) E
*
.~1tW\llL. E~»Ie-.
• FUL.L- I 1t1l~S~~1tE~"-
c
o
$
c AT.2.-SCHT3
.s YsT5 ,"S
StlPPoiT~)
.". J: a WI
S/3'O}
ItIOJ>£J-
!OJ ~ Sl)) ,~ 7
• k).rtf IBm
o
,se.,
~ 'f)1
s-
1>JfT~
Rl>"PT£R uA)IT
kJ I ,. If 51) R - Jr
elA)lTH .Ism 1703
Tl" NS Vl \ \, to H c:.oN'BOL.
c
UNIT
fA) ITtf
SYNC,NloA)OU S
~ftSE
J:
SC.HT1-SC.ltT3
SoFTWaRE
,~
SYSTEJI\s
bO ~ / aT" m
R\JRILftl'-E: HO'&)
$ / 36 0
.,. Sf 3'()
os f ST R tY\
A'l"'L""lE: R'R,I~
0
30) '"
o
*
BY fJN
RSSEmBLY LR~'''1tG-E
CBLl£)
P~O&-ltliM
-IIo
TR.lf~SM\r
4 ltEeE,\Je
· 'P~ '""2.it M
S?EC.lFIEb I/o 9't.EllS
1> RT'R
O\)E~ L.JrP-PEl>
O"PER.1rTloAl
UlfrH OTt\E"R. ~ DEVICES
o
F\lLL-"'~DA\S'Pi11tE~T TEXT
1I~1tESI"\~TEb
OF
l>ftiW
eob .N(;-
t&).-rtl,N
MESS'RG-FS
'&-RO\l~b 't.u.L.ES'
u-)E.l ( -1)EFI~E.b
ERltDit C\\EC.\(, ~G
c."R.~-
.~
1\\ llL-Tl-
\Co
-Pol~T
o
I
0
o
~ Eft\OTE :ro~ E"'TRy
WO~~ S"'~I'O~
() ?1l0 ~ 1t.H M
o
.... """"'M\i;:;·,~.· ..L1i&ii!l?ii!'.:;y¥.¥.".!§!i..,!i¥i.i ___ .€ ·,·· ...-··f··.. ··.HM::.".,p.LT,4,('!t,t.,J.M!fuYiMiti.i
,.Eli,.,
, • . !Ii.......ii4#¥4 ... .
. ' "".,. ".
,_,,'
·'''.I~.;;;;gkn;r;g;m!l2.:GIH1lhira'''''l\IuiiIiilIL\iWm:wasaawwaaac
C.E~"~1=IL..
5/3'0
STRTIO ~
=
OS/Jt.~E
R\JRILRBLE:
fto I~T-'-0- -po I ~T
~",-y 31) 11~~
c
mUl..Tt-- rOI~T
Oe.To~ER.
II!O b I S I<..
"Et;StO~
3\, 11' r
W\O~
,TOil.
I'
o
0·"
,
I
Flt~TtO~S
• .x~¥~T
oF~"B
* COW\R\ RA\hS
£~,ll'Es
-rK~\l. 11 AJ
R'" fKt\El> :rA)-PllT l>£\l teE
o· ])IlTB nll~SW\\SS'ON
.0 4.
F~W\
• t&\~fTl ~'- of
c:.EA1-~lfL.
50'& O\A..\? IA..\
4 MESS'9&-ES To OCA..,p'4:r
1)E\l'~E~
o
.~--~~"""""-.",,,,,,,
- .. SoLE
KE YI..OIIIiIiIIIR-.-..
''''''-----'
..
o
o
o
o
~E(!L.
-SAW\~
~1l~
FOR.. I ,~o
S'-HTlOtJ
~~
sea,
o Exee:~T
• • 1)JJIB
s/a~o
RTlOJl
Fo2..
1)11\ S,
c...
•• 1>1l11l bft\S, b,XBl1ll) ~88B
o
SEc:to'L
In>b~ESSl
~Loe.K
eOu.~T---"""""
c..O"l10~~L)
351
"~o It. ~E
U)~SIj
BSEo
.! ~Eeo t.l) S / sec.,." t.
·
o
i
I
35,J
~.---.-----~--
-
I
o
\\10
'k-SE
&E~e.k.ltT-\O ~
/1 XE£ Il-S E 4J ell
l 'N E. ~ ) \A ~X ITa
SE~ToIt.
Fo~
,-, 'V\,,..~
\AS £ It-
JM' bia~~
--...
STl>~I~_
b,-.e~-r~))
&444.
btl""
"-0 -,." e
E" IT
o
355
I
II~O
R.~£
E~ EC.\A.TlO~
/1 X£Q.
••
R~
R.~E
SrART
'3"ECl- ST~TEMEAlT~
t 05/3bO ~O~S
••
o
t~ENb
o
•
E~TER.
~OB
~O~~
XA10
~"'1t..EHW\
• EQ\1.\~n\E"'tC-E ~"'k.fl L I ~ a.l\O~
o
• COlV't~TI ~&-
?Ou)Elt. oA\
R ]) EMit ~b
Rlrrt\ E~
~llSl S
T"ll~
llE~U.Lll~
• St\ "RR\~(i. 01= ~ e.OW\.ft\..o)-\
o
"BobY OF
:LN FO R M It\lO~
30/
'
- - - --- ---
'I~O
- --- ---------::....::......:-.:.---=-=--.-.--=----...;..:.-.--.-.------:..:.:.:-..:..-.:....:-..:..:..~-..--.~:...... -~-=------
~-SE.
ST1to~&-
-.---~--~,.'.,.----
aS6
SYSTEm
1>\sa,~ L'~E
• ~o~S
~E(i.u.,'~I~Q-
MPI Y
e~I4~E
-:S-08
FL.Ou.\
• 'So~s
collE
SET--,,}'
SYS~£t\4.
-ro
~E
"I'S&1
~EQ\l.\1l.,tl~
CR~
1t- tAT of
e.A"~E
Jl
s/Ale£
tttDC.ES51~- IS bELl\YE])
U~TI&'" C-ORE..
IS
SYSTEM
?1tJ)~, E~
R-\Jlt I '- ~8L. £
I
I D 22 '0
G~HP\-\!c...
o
SU~S
YS TEMS
10
I
~.
-
~-j
*RVll'Lll~E: FE\ IS, "~'
if
FI~Ell-~ OF
SCA.l'-p.o1l.'
. • 'PRoeE~~-tnlb--PRoc.ESsol.
l>lin ) R.Ti~s M'S~ \()~ 0
*
!
i
• )liTW c.oM~ll'" \ S \L...,.y
S£l.",e.ES+
• S'fIT £ L.LI'E & ltFt? \-\ tc:..
-:SO~ ~1'to C!-E sso R...
o
.1
36L/
,~
II
c·I~'
I)
1130 /2..2. S D
AJETtuoR-K..
'*
t
5/~~O n\obEL WI 5"6,' 5;7
kl.-rtl 4.701 ~R. a70 l.
'1'\0 - ~ t<. oJt L1tk'-E lit
o l&)\-rtt- se.R J)\I~ ~IS"
*
Po, ~r- 1-0SU), -w-" tt-£b
?o,~,
oR. LEft.5£b
.:CF SCOt ,-c.tl£b.l U. ~ER
REQ. u.\R..U ~ llSE
o
ou) ~
~ 1t.o~Eb
E51i1a L. tS H
u.1t..&~ ,-0
c.o~ Ee..-rIO A\
,
_~
••___
'_~'_"."'
.•"_''''_"'- .. _, ..... ". ,. -" -·,......
_.,,_=,~==."'""~,",,'"'''''~'-=,,==--=_='''"
. c:c.=..::c_,,_.:.:,..... ,.: .... :...:.............__
I
PRoe e ss.t> 1ie.. ..10-~ft.O'-E.S~d R
)ft,..~
OJ k ft~SMl ~Sl~
• Folta ..... tlE.
FoR.I1t.."~
- ~I~ 1 L..l\k.
:C,,'"
AI
1"0
u..~E~
'b i\-rJ\
. nft~~M.l~~IOA)
5u..~-.o&A.Tla)L~
F()ll
II~" R~~ S/3'0
• HO .bETftIL..Eb K~Cl)L.El>'-e:
OF ])ITlI u~K noC,E)UlE,So ,
, S »EEbE,])
Oil'S() /22.$ b
r'(1t~S W\\~~'O ~
~lA,~ItOU.-r1 ~
S S
a.TH IT-- ZfJITIRLI2E. FIJR.
COMMUN I e9TI() A)
o
~T~El>- ..
IlERb )JtTR
&T IAltlT - - (A)k\TE. bHTR
~re'-T
- ... CDAl-rR.D1- 'rES'"
~TEAlb - -
o
EAlb
COMWlUAJI~F1t/dAl
11~()/22. S tJ
l>~'-R
~M.~ftrl 'atLI~Y
SER..\J
• ~I t
:0
\e..E~
CoA)"ER..~'4A1
SCA.~~OlA.\' tJE ~ 1-0 ftWb
F'Ro~ \\-"0 A-Nb S/-sitJ
1>R}-R
b Rift
• ALL.
FO Jt.. ~ 1\-reOAl\JER.S,O#J
'E~F01tft\Eb
IA.\
is
5/3'"
c,
3~
!
I
" 3D/.t2. ~o
o
SFiTELL.,rE ~R.RPM I~
-:S-o_
?1to~e.sso~
• E~TEA\SlO~
Toa
o
PtOC.ES~oJt
• $/~~O
usn
~O_
eOA\1-)tOL
R-TTACHEb
R£ MO-r E
• l> E F I a.\ E
P£o~Ril'"
'-0
..
R
"~"
Ff 5/"'0
-ro
~~~U~t..Tlo ~
R-~
~~ltPtl It..
OF
Ru~
IAJ
"'l-ret
"~O ~b"tll1 M
3t,9
II~O/22~O
• bEF I~E.
lt~l)
o
:tNttl~YE
~o~ ~
TO 1t.\L~
:r~bE ~ S 1\\ bE~tL.'/
:rAi 5/36>0
•
ReQu.E~~
-S-o~
C!.o~.,lt.dw
%A)Folt..~ft"'loA\
Fbm. A
USe ~ -rKt.ou.&H
'R
SERIES OF
b,sPL.RVS
• ar
lO
I '/
II~o /.1.2. ro
1t. & rAO -r-e
Ysrs,m S
v '
RTTAC. Me) ~D" S/3'O '
S
o
Public Service Company of Colorado
Uses the 1800 Data Acquisition
and Control System in
Gas Load Control
o
By Earl E. McLaughlin Jr.
Public Service Company
of Colorado
Presented Before the
1800 "Common" Conference
San Francisco, California
December - 1967
o
ABSTRACT
The Public Service Company of Colorado, a Gas and Electric utility,
has installed the 1800 Data Acquisition and Control System in its Gas Load
Control Center.
The Gas Load Control Center has the responsibility of regulating
and maintaining system pressures, contracted gas purchases, and gas storage
facilities in the Company's service area.
The computer system is monitor-
ing variables from the field, logging hourly flow and pressure data, and
controlling system pressures.
This paper will explain the necessity and benefits of a real time
computer system in the dispatching of natural gas.
o
3?~.
\
rna
o
CONTENTS
Service Area
1.0
Gas Supply
1.1
Purchase Contracts
1.2
Gas Storage
1.3
Industrial Customers
1.4
Pressure Systems
1.5
Weather
1.6
Gas Load Control Center
1. 7
Computer Configuration
2.0
Telemeter Input
2.1
Data Checking
2.2
Data Processing
2.3
Operator Inquiry and Entry
2.4
Gas Load Forecasting
2.5
Data Logging
2.6
Low Pressure Control
2.7
Physical Planning
3.0
Future Planning
4.0
Computer Output
5.0
o
3?3.
1.0
SERVICE AREA
Public Service Company of Colorado and its subsidiaries,
Western Slope Gas Company, Cheyenne Light, Fuel and Power Company, and
Pueblo Gas and Fuel Company serve natural gas to approximately 400,000
customers in Colorado and Southern Wyoming.
Natural gas is distributed
in the Company's service areas through extensive transmission and dis- .
tribution systems.
A general system map is shown in Figure 1.
This graphically
illustrates the great distances and rough terrain that must be encountered in order to serve our customers.
An example of this is the Western
Slope Gas Company, Southern Division, transmission line that is 449 miles
in length and crosses the Continental Divide five times.
1.1
SUPPLY
Gas utilized by Public Service Company and Western Slope Gas
Company, Eastern Division, is purchased directly from Colorado Interstate Gas Company.
Gas for this area is transported by Colorado Inter-
state from the Texas, Kansas, Colorado and Oklahoma gas fields and from
San Juan gas fields in the Four Corners area by connection with another
pipeline company.
This represents an investment of over $202,000,000
by Colorado Interstate to serve its customers.
o
Public Service and Western
Slope represent over two-thirds of Colorado Interstate's business and
have a direct relationship in establishing purchase contracts and rate
structures in this area.
Gas utilized by Western Slope Gas, Western Division, is pro_
duced from gas fields in Western Colorado.
The Western Slope Gas Company, Southern Division, produces
the gas it needs from the San Juan gas field in Southern Colorado.
1.2
PURCHASE CONTRACTS
Gas purchased by Public Service Company, Cheyenne, Pueblo, and
Western Slope from Colorado Interstate is purchased under a two-part
rate, consisting of a Contract Demand charge and a Commodity charge.
The Commodity charge is the unit price on the actual volume of gas
delivered by Colorado Interstate and the Demand Charge is the charge for
the maximum volume of gas Colorado Interstate is obligated to deliver on
~"
the peak day.
1'1
I;"
"I
I'
' \\\ ----.-.-. ~YC!MING
L
I
'\\
_
GRal'r
PIVER
V'
@
\
NOV
\.
-'-'-'-
-"BlEROCK
FI_EL_D_ __
- \--r-- _
LEGEND
\
• -
•
~
..tLARAMI£
o
F
, ,.......RAl6ELY
~
'1'''---',
I
:
N OOUGl AS
~
"-,
CREEK FIELC'
1~Ss/
L f ytJt:N GAS('\,0
STCF.AGf. '\J
~
(.!ENVER
~
/~F'T
\
.~ ../ /
~Y~.~
.
WESTERN DIVISION
FIEL~
L/)
MORGAN
/
ST')RAGE
1---------
E<%::f::;:'>'<;:;
.
.' ~..::;;::3%
\~.
~'::;;,:,~.--:::
~~;<;/%
I
51> N
GAS
JUA",-:;:
FIE L C
/'.
"i:p~0:/:/!~~!r;<;:;
~%/.:?'.:;::.' , '.' , .
W::/ ..
1
./.
NEW MEXICO
~~;:
....-
I
I
~
"-.)
<.>,
1
CJ)
FIGUR- E J..
SYSTEMS
..-o
....:J
MAP
I
en
I.
TAHT
Public Service and Western Slope are required to pay the Demand
Charge on 100% of the Contract Demand each day and the Commodity Charge
for the gas actually purchased.
Penalties are charged the Company in the
event that the Contract Demand limit is exceeded..
o
As' example of a sub-
stantial penalty, in 1963, Public Service and Western Slope paid in excess of $785,000 for gas that was purchased over the contract limits.
The justification of such a rate structure is based on the
wide range of load conditions that occur in the Colorado area.
On July 16,
1967, the mean temperature was 65 degrees F. with a 24-hour gas load of
107,000,000 cubic feet.
January 7, 1967 produced a load of 565,000,000
cubic feet with a mean temperature of 16 degrees F.
in Denver on January 11, 1963, was -17 degrees
The mean temperature
F~
It is apparent that pipeline capacity required to meet peak
winter needs becomes idle during warm weather.
The two-part rate structure is needed to protect the pipeline's
large plant investment that was made to serve areas during the peak load,
while the load factor is at a minimum due to conditions such as warm
weather.
1.3
o
STORAGE
It would be an advantage to Public Service Company of Colorado
and Western Slope Gas Company to offset the varying load conditions by
increasing the purchases from Colorado Interstate during off peak periods
and decreasing the purchases on the peak day..
This can be accomplished
to some extent by storing gas during warm weather and withdrawing the same
gas from storage when needed in cold weather.
Public Service Company has converted an abandoned coal mine near
Denver, Colorado, to an underground natural gas storage reservoir.
This
project, called Leyden Storage, is capable of storing 2,500,000,000 cubic
feet of gas of which 1,250,000,000 is considered usable for peak shaving
at a cavern pressure of 250 psig o
Leyden i.s connected to both Public
Service and Western Slope service areas by a network of intermediate
pressure systems (50-150
psi)~
An agreement also was made with Colorado
Interstate for storage gas from its storage facilities near Fort Morgan,
Colorado, to serve the Denver area.
Both of these storage facilities
are used during the heating season to control the load curve of Public
Service Company of Colorado and Western Slope Gas companies.
o
A typical 24-hour winter load curve is shown in figure 2.
o
For
this particular day, all industrial customers were curtailed and 107,046,000
cubic feet of gas was withdrawn from Leyden storage to maintain the contracted limits.
1.4
An average hour load is shown by the line at 15.6 MMCF.
INDUSTRIAL CUSTOMERS
Industrial customers in our service area are under contract to
purchase interruptible gas.
Interruptibl~
Gas is gas that is available
in excess of the amount required to serve the firm customer while still
under the Contract Demand.
The interruptible gas is sold to the industrial
customer at a rate that is less than the amount paid by the firm customer.
This rate class justifies the installation of a
as, propane or
standb~
fuel system, such
oil~
When the predicted load in a service area is estimated to be
greater than 100/ of the c.ontracted demand and above the amount that can
0
be supplied from storage, the industrial customer is given notice that
he must curtail the use of natural gas and switch to standby fuel.
When
the daily load decreases to less than 100% of contract, the industrial
o
customer is notified that he may resume the use of natural gas e
There are over 450 industrial customers in the Company's
service area and every attempt is made to keep curtailment at a minimum
to optimize purchases and sales.
The industrial load is approximately
20% of the contract on a cold day.
The load curve that is shown in Figure 2 is with all of the
industrial customers curtailed.
1~5
PRESSURE SYSTEMS
In general, there are three types of pressure systems--High,
Intermediate, and Lowe
The high pressure system is used for the trans-
mission of gas over long distances
to 1,000 psig.
G
The operating range is from 200 psig
This type of system is usually a function of the trans-
mission or pipeline company, such as Western Slope Gas Companye
The
high pressure system terminates at a city's town border station.
A town border station is the point where the distribution
company purchases gas from the transmission company for its systems
o
The
Denver area has six major purchase stations e
Intermediate pressure systems operate in the range of about 20
to 150 psig.
Their function is to carry the gas through the market area
and supply all of the city's district regulator stations o
In the Denver
8??
li~-
".
,
i
:
:
"
;
I'
:
"
,.,
".
,.
,
'='
.,
"
IQ
'J
+
18,00
17,00
:::-;-
6,000
:J ..
.-:-
Er1 ;\
r
15,000
4.000
:
3,000
2·rnnr
,.
"
"
"
11.000
,-,-:,
,
"
10 ,000 '
';0'
,.
~,
0:
\01
III
III
9 pOO
bI
.J
III
I-
"-
:z:
0..: ~
... '" III
o u :.:
8 ,000
,.,
pa
li
"
!+
W
ell
,
:r.:
h
H
~
;
I
7,000
_III
,
r
N
"
Xx
'l'l
~l
i
~~
;;
W'"
1.1:
I
I.
H
rr' '.
I;:'
rr
.1
~
-;
.i
T.l.. "
~!
•
i-r-
,'f:
1 2
3
4
5
6 7
8
9
10 II
12 I
PM
2 3
4 5
~;
t\-
H
AM
' rt
6 7
l, 9
;f~
:
"
10 II
'.1,
12
MID
3??
area, the intermediate pressure system connects most of the town border
stations with Leyden Mine Storage.
This enables peak shaving gas to be
injected into the intermediate pressure system which will result in decreasing the purchased gas at the city's town border.
Low pressure systems can operate in the range of 2 psig upward
to 60 psig.
Their function is to distribute gas to the individual customer.
All of the pressure systems, along with compressors, valves, and
regulators, provide a link that is hundreds of miles long which connects a
customer in Colorado with the gas fields in Texas, Kansas, Wyoming, Oklahoma and Colorado.
1.6
WEATHER
Colorado area weather is the largest factor in determining the
gas load.
Weather forecasting and gas load forecasting are necessary to
predict gas and pressure requirements.
During the winter, it is not un-
usual for a morning to be 60 degrees F. with the sun shining and by afternoon of the same day have the temperature drop to 10 degrees F., with wind
and snow.
This is caused by a storm front and, unless the front is pre-
dicted, requirements for pressure and volume would be greater than the
4()
system could produce.
1.7
CONTROL CENTER
Public Service Company of Colorado, Western Slope Gas Company,
Cheyenne Light, Fuel and Power Company, and Pueblo Gas and Fuel Company
operate a Gas Load Control Center located in Denver, Colorado.
This
center is manned 24 hours a day, 7 days a week, and has the following
responsibilities:
1 - To calculate and log all of the hourly flow totals from
metering stations for all four companies.
Pressures, differential pres-
sures, and flowing gas temperatures are telemetered to the Load Control
Center from 32 major metering points located throughout the combined
service area.
This information which totals 114 different inputs is
used together with various factors to calculate the gas purchased each
hour at each meter station.
The formula for the calculation of gas through
an orifice meter is shown in Figure 4.
A drawing of an orifice meter is
shown in Figure 5.
o
2 - To monitor and control (via phone conversation) compression
facilities of the Western Slope Gas Company.
Pressures from points locat-
ed on the transmission lines are telemetered to the Control Center.
This
3?~
Q=FbFpbFg [1+(.68462-.26923
0= RATE OF
FLOW IN
g!)(21~PA)J
CFH
QR=RATE OF FLOW IN MCFH = Q/IOOO=
[JhPA+b
(aPA2+bPA+c-I)(~)4.825+ W
J
1+ 2~7PA
]:,767
Fb= BOFB =BASIC ORIFICE FACTOR J~~~OOO
Fpb=PBCF=PRESSURE
d=DIAMETER OF
D=DlAMETER OF
BASE CORRECTION FACTOR FROM 14.73 TO OTHER = 1473
. =
Pb
ORIFIC~J~~~"
PIPEJ!~
1
5
d4 _
I.0229:14.4Pb
I.Q055=14.65Pb
.9804 -15.025 Pb
J
1
(.68462-.26923~)-D4F]o
h=H=DIFFERENTIAL IN INCHES OF WATER AT T(60°F) 1,00·
P=INPUT PRESSURE FROM METER STATION
PB = METERING PRESSURE BASE
PA=ABSOLUTE STATIC PRESSURE PSI = P+14.4
]'014.4
(ALL PRESSURES ARE SET FOR 14.4
NOT AVERAGE BAROMETER 14.4
t> =BPRIM = CONSTANT FOR REYNOLDS NUMBER
].'01
a
0=
b=
) SUPER
L
] '10-6 =0.000,000,046
COMPRESSIBILITY
J~o-6 =0.000,148,500
FACTORS
c=
] :0-6 = 0.997,300
T= TEMPERATURE OF
TA= ABSOLUTE
= T +460°F
FLOWING GAS
TEMPERATURE
]~O·F
(RO) OF FLOWING GAS
G=SPECIFIC
Fo = SPFG = SPECIFIC GRAVITY
FACTOR =
FLOW· SUBROUTINE
VALUES PLACED IN COMMON BY
Vv
~J
1.../
THE
/IiG
]!::;
J~
CALLING ROUTINE: OR, P, T, H, BOFB, PBCF, SPGF, D4F, BPRIM
VALUES PLACED IN COMMON BY THIS SUBROUTINE: OR
~
C
FIGURE 4
0
o
o
DIFFERENTIA~
PRESSURE
ACROSS ORIFICE PLATE
TO CONTROL CENTER
GAS PIPE LINE
GAS FLOW
ORIFICE
PLATE
TO
FLOWING GAS
TEMPERATURE
CONTRO~
CENTER
SPECIFIC
GRAVITY
~
0()
"
STATIC
GAS PRESSURE
FIGURE 5
METER TUBE
information totals 39 input points.
The operator can take action by
ordering pressure output from a major compressor station or by notifying
personnel in the area that needs attention.
3 - To monitor and control intermediate pressure systems in the
Denver area.
Pressures from points located on the intermediate pressure
systems are telemetered to the Control Center.
42 input points.
This information totals
Control circuits, which set or regulate the pressure
system, are activated from the control room.
This totals 38 output points.
4 - To monitor and control low pressure systems in the Denver
area.
Pressures from the low pressure systems are telemetered to the
control center, totaling 61 input points.
Control circuitry from the
control room to a regulator station enables the operator to adjust the
regulator's output to satisfy the demand on that station.
There are 54
output points from the control room to various regulator stations.
5 - To operate all of the above pressure systems in a manner
that guarantees safety and service to the customer without interruption.
6 - To control purchase rates from suppliers and storage
facilities maintaining optimum system requirements without exceeding 100%
of the contract demand.
This requires the operator to estimate gas loads, weather
conditions, pressure requirements, gas storage needs, and industrial loads
throughout the operating day, and make the necessary decisions based on
these estimates.
7 - To provide management with reliable operating statistics so
that the t"uture growth and operating guidelines can be determined.
The operation of a control center of this size requires that the
operator maintain constant surveillance of all variables in the control
room, calculate the flow rate for all meter stations, and predict the
weather and gas load for storage requirements and industrial curtailment.
With these responsibilities constantly increasing due to the continuing
growth of the Company's gas service areas and more complex storage and
system guidelines, a method was needed to relieve the operator of the
time-consuming routine jobs.
The time saved could be devoted to weather
and system analysis.
The 1800 Data Acquisition and Computer Control system was
installed in August 1967 in the Gas Load Control Center.
The computer
is assisting the operator by providing a constant monitor of all system
-----------_ ...._.. _._-
---------.--~--
o
3?3
pressu-res along with flow calculation and pressure cont1;ol.
This system is
also providing management with valuable statistical information and a higher degree of accuracy.
the systems control.
The computer is enabling the operator to optimize
The chance that the 100% contract demand figure will
be exceeded is much less and a higher load factor is being achieved.
The
computer's speed allows all the pressure inputs to be scanned at a continuous rate and if an erroneous condition occurs, it will alarm the
operator as to what values need his attention.
of safety is obtained in pressure control.
Thus, a higher degree
The computer is able to
regulate a station's pressure output at a more constant rate than the
operator can, which provides a lower average pressure throughout the
system with resultant benefit in reducing the volume of lost and unaccounted for gas.
One of the most beneficial aspects of the computer system's
ability to handle the routine duties of an operator is that the operator
has more time to use his experience and knowledge for planning and supervision of the overall system.
The computer's data flow and configuration are explained starting at 2.0
o
o
0,"
,'"i
~~---~""~---"--
GAS
..
------"-."..........--------------------
PIPE
LINE
-------------
PRESSURE
CONTROLLER
o
CONTROL PRESSURE
SETTINGS
REPORT t::t
ERROR
CONDITIONS
1800
COMPUTER
SYSTEM
ANALYZE ALL
INPUT DATA
RECORD
AVERAGE HIGH
LOW READINGS
CALIBRATION
OF
Transmitters
OPERATOR
INQUJRY
MANAGEMENT
GUIDELINES
CALCULATE
GAS FLO\V
THROUGH
METER STA.
HOUR a DAY
LOGS
PROJECT GAS
LOADS
REPORT
PROJECTION
FORECAST
GAS .LOADS
FIGURE 6
TIME SHARE
a
STATISTICAL
DEVELOPMEN
REPORT
LOAD
t - - _.... FORECASTS
c
2.0
COMPUTER CONFIGURATION
~)
The 1800 computer system consists of the following:
1801 Process Controller - 4 micro-second-16K Core
12 Levels of Interrupt
2 - 2310 Disk Drives
1442 Card Reader
1443 Line Printer
3 - 1826 Input Units
2 - 1053 Typewriters
1816 Typewriter - Keyboard
Timer A - 1 ms
Timer B - 8 ms
Timer C - 8 ms
10 PISW Words on Level 0
10 PISW Words on Level 1
IBM "Time Sharing Executive" Operating System
The computer's General Data flow is shown in Figure 6.
The
explanation of the software and hardware used follows.
2.1
TELEMETER INPUT
The American Standards Association defines the word "telemetering"
as "the indicating, recording or integrating of a quantity at a distance by
electrical translating means."
There are many types of telemetering systems
available, and Public Service Company uses an impulse duration type.
This
system includes transmitters in the field and receivers in the Control
Center.
The transmitter contains a measuring element, such as, a Bourdon
tube.
This element is attached through linkages to a cam follower.
The cam
follower rides on a linear cam that rotates 360 degrees every five seconds.
The pressure in the element causes the cam follower to be positioned on
the cam in relation to the pressure in the element.
The cam follower is
connected to a mercury switch and when the follower is on the cam, an
electrical circuit is made.
is broken.
c
seconds.
When the follower is off the cam, the circuit
The transmitter generates an electrical impulse every five
The length of the pulse is determined by the time that the cam
follower is on the cam, which is determined by the amount of pressure
that is applies to the measuring element.
If the measuring element is
------------~~~-.-------~-~--~
.." .
_" ...... ........."...._-_ ......._--_.._.. .._._--_._----_...."
"
......
a Bourdon tube that is connected to a gas pipeline, an increase in gas
pressure will cause the Bourdon tube to expand.
This expansion will move
the cam follower to a higher position.on the cam which, in turn, will
generate a longer time on the cam or a longer electrical pulse.
A trans-
mi tter is shown in Figure 7.
This pulse is transmi tted over· telephone
lines to the Control Center.
The pulse is connected to an electrical
receiver which positions a recording pen on a chart that is in relation
to the pressure on the transmitter in the field.
The length of a pulse
will vary from one second (0% pressure) up to four seconds (100% pressure).
See Figure 8.
The Control Center receives 250 different pulses from points
located on the gas system every five seconds.
The distance the pulses
travel varies from one mile up to 500 miles.
A method of inputting the values from the transmitter to the
computer was needed.
al input system.
At first we planned on using the conventional digit-
Various programming techniques could be used but all
of them required the computer to scan the input data every 3 milliseconds
(ms).
The 3 ms scan was needed to maintain one tenth of one percent
resolution on a 5 second input (out of the 5 second pulse, 3 seconds of
the pulse are used for full scale, 0% thru 100%, 3 seconds
.001 of 3000 ms
=
3 ms).
= 3000
ms
Using a 4 micro second computer, it would take
longer than 3 ms to scan the inputs.
If the computer's speed was in-
creased to 2 micro seconds, more than 50% of the computer's time would
be spent reading these inputs.
computer was needed.
A more efficient method of input to the
An input system using the power of the 1800 com-
puter's interrupt structure was developed called "Time Duration Telemeter Input".
Time Duration telemeter input is intended for use with time
duration signals that have the characteristics of a starting time, an
elapsed time, and a stopping time.
through Process Interrupt words.
account of the elapsed times.
Input to the computer is achieved
One hardware timer is used to keep
If Time Sharing Executive software is
used, one modification to the Timer Section is needed. This modification
enables one timer to run continuously without requesting service.
A signal that is sent from a transmitter is shown in Figure 9,
Part A.
Here we have a starting time (A), an elapsed time (X) and a
stopping time (B).
If the computer is interrupted at point (A) an in-
terrupt servicing program can record the time of the interrupt using
__ .......... _•.. _._...__._-_...__._-
"I
o
o
FIGURE'.
TYPICAL _ __
TRANSMITTER WITH TIMER PLUG ATTACHMENT RI4A AND SCALE PLATE REMOVED
CODE TO FIGURE
f.
2.
3.
ij.
o
5.
6.
7.
8.
g.
10.
II.
12.
Locking Screw
Bear i ng Sc rew
Lifter Plate
Fixed Arm
Cam
fndicating Pointer
Zero Adjustment for Cam Fo II owe r
Locking Screw
Span Adj ust.ment for Indicating Pointer
Linearity Adjustment for Pointer
Zero Adjustment for Indicating Pointer
Linearity Adjustment for Cam Fol lower
7
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Span Adjustment for Cam Follower
Measuring Element
Cam Fo 11 owe r
Mercury Switch
Permanent Magnet
Vane-Arm Assembly
Balance Cam
Bearing Screw
Lock i ng Sc rew
Trip-Plate Shaft
Convenient outlet} Attachment RlijA
Timer Plug
Optional
c'
50%
SCALE
2.5 SECONDS
0%
SCALE
SET TO 0%
I SECOND
SET TO 0%
5 SECONDS
0/0
PULSE OFF CAM
I
SCALE
0/0
OFF CAM
FIGURE g
SCALE
OFF CAM
o
C"FIlfWU"§f'
A
""
1000 P.S.I. CHART
PRESSURE= 500 P.S.I.
CAM BASE=5000
I
I
A
..
B
I
I
Y. TELEMETERED CAM TIME=5000
X =TELEMETERED ELAPSED TIME=2500
A =CI RCUIT CLOSED (TIME ON I)
B = CI RCUIT OPEN (TIME OFF)
C = CIRCUIT CLOSED (TIME ON 2)
"'-'---.-YI -------....
r
t
A
X,
~
-zlB
c
YI=TELEMETERED CAM TIME =5500
XI =TELEMETERED ELAPSED TIME = 2750
Z = ERROR IN ELAPSED TIME = 250
Fx = CORRECTED ELAPSED TIME
Fx _ XI
Fx _ 2750
5000- YI·
5000- 5500
Fx= 2500
FIGURE
9
one of the hardware timers.
as Time On.
Off.
This time is kept in an array for the input
The next sequence would be the loss of this signal, or Time'
Again the computer is interrupted at point (B) and a program records
the time of the interrupt.
By subtracting the Time On from Time Off an
elapsed time is obtained and placed in an array as elapsed time.
The
final sequence for this particular input would be the second Time On or
point (C).
Again the program records the Time On as before, but now a
new value is available, the total time required to send this signal.
In te1emetering systems, such as the one used by Public Service, this
value can be used as a correction factor.
al every 5 seconds or 5000 ms.
A transmitter sends one sign-
In cold weather, the cam speed can slow
down causing the total time required to send a value to increase.
This
would also cause the elapsed time to be longer, thus creating an error
in the value sent.
The total time required for the cam in the trans-
mitter to make a complete revolution is 5000 ms.
If the total time
that was actually sent is recorded, ,a simple ratio can adjust the elapsed time to negate any error caused by the cam either
too fast.
~~ing
too slow or
See Figure 9, P2rt B.
The advantage of using this type of input system is that the
computer is freed from scanning.
Instead of scanning the inputs to deter-
mine a status change, the input interrupts the computer only when the
change has occurred.
In the Public Service Company application 160
values were input to the computer at an estimated overhead of only 10%
of the computer's time.
This can be compared to 100% of the computer
time using digital input hardware.
All of the, input variables are output on logs to technicians
so that they can calibrate the transmitter to the computer.
Ten P. I. S. W. words were wired to level zero of the computer,
these are used for all Time On interrupts.
Ten P. I. S. W. words were
wired to level one of the computer; these are used for all Time Off
interrupts.
One signal line from a transmitter is fed into a type
"c"
relay creating an On and Off interrupt to the computer from the transmitter.
2.2
DATA CHECKING
Every 10, 15 or 30 seconds,
~ependent
upon the desired operation
set by the operator, a program will interrogate the data placed in the
computer by the input program.
This program will check each input point
~C
fo~
any of nine error conditions.
If an error is found in the data, a
report will be made to the operator.
Error Conditions:
1.
Off Scale - Below 0% or over 100% of scale.
2.
Back on Scale - Was Off Scale - now reading is valid.
3.
Rate of change is over limits - The present value compared
to the last value is increasing or decreasing too rapidly.
4.
Out of Limits - Value is out of limit bounds that are
pre-set.
S.
Back in Limits
Value is now in pre-set bounds.
6.
Cam Speed Out of Limits
Cam in transmitter is either
too slow or too fast on each 360 degree revolution.
7.
Cam Speed back in Limits - Cam in transmitter is now
operating satisfactorily.
8.
Circuit Out of Service - Input line is out.
9.
Circuit Back in Service - Input line is restored.
An array of last good readings is maintained ,by this program
and if an input is found off scale, its last good reading will not be
o
changed.
This gives the computer an estimating capability in the event
that an input circuit is out of service.
2.3
DATA PROCESSING
Two types of inputs are processed by the 1800 computer, pressure
and meter stations.
for Control.
Pressures are values on the system that are used
A file is maintained for each pressure input and includes
the following:
1.
Instant High - highest value telemetered.
2.
Instant Low - lowest value telemetered.
3.
Hour Average - last hour's average reading.
4.
Day Average
S.
Scale Range - Range of input.
6.
Off Scale - Number of times reading was off scale for the
24 hour average reading.
day.
Meter Station values include static pressure, differential
o
pressure and flowing gas temperature.
See Figure S.
A file is maintained
for each meter station and includes the following:
393
1.
Instant high flow - highest flow calculated.
2.
Instant low flow - lowest flow calculated.
3.
Hour flow - Hour flow calculated for station.
4.
Day flow - Flow calculated for the day.
5.
Year High - Largest hour flow for the year (Peak Hour)
6.
Projected Hour - If the flow rate for this station remains
constant, projected rate will be the hour load.
7.
Projected Day - If the flow rate for this station remains
constant, projected rate will be the day load.
Constants needed for information and calculation, such as
orifice meter plate size, are also included in each statistical file.
204
OPERATOR INQUIRY AND ENTRY
Data such as gravity and plate size, needed for flow calculation,
is input to the computer through the 1816 Keyboard typewriter.
A program
called "Key Board Monitor" enables the operator to queue up a function
program from the keyboard by typing in the name for that program.
Once
the program has started execution, instructions such as the type and format
of the data are typed out to the operator.
The operator follows the
instructions and enters the data the computer is expecting$
This minimizes
data errors and leaves typewritten copy on what data was entered, by
whom and when.
A sample of the input of a plate change for a meter
station is shown in Figure 10.
The operator can callout any variable in the same manne·r, by
typing in the program name for the function desired.
This type of output has greatly simplified the training of the
operators.
2.5
GAS LOAD FORECASTING
Each year statistics on the last 26 years of weather and gas
loads are applied against a model.
developed y
=
a - bx, where y
=
From this data, a linear equation is
forcasted load; a
=y
intercept; b
=
slope
or forecast load per degree (F) change in mean temperature, and x is
the expected mean (F) temperature for the day"
An equation is developed
for each hour for the remaining hours of the day for all four companies
along with industrial equations for the industrial loads.
These equations
are solved in the 1800 computer each hour and a report is typed out
to the opera tor shmling
the forecas ted load tha t is equal to the fore-
cast for the remaining hours of the day plus the amount already purchased
391/
-
-----.-.-----..
-.~~-
FIGURE 10
07:39:20
PLATE CHANGE
PLEASE TYPE YOUR 3 INITIALS, THEN DEPRESS -EOF-KEY
EEM
INPUT IS BY EARL E. MC LAUGHLIN METHODS DEPT
PLEAS E I N PUT THE MNEMONIC FOR THE STATION YOU WISH TO CHANGE
CHCK
CHERRY CREEK
RUN = 1 HAS A 3.00
DOYOU WISH TO CHANGE THE PLATE SIZE IN THIS RUN
c
INCH PLATE NOW
SWITCH TO ALPHA AND TYPE YES OR NO
YES
PLEASE INPUT THE NEW PLATE SIZE IN 4 DIGITS IN THIS FORMAT XX.XX
04.25
NEW PLATE SIZE ENTERED
4.25
IF CORRECT TYPE YES,IF NOT TYPE NO
SWITCH TO ALPHA AND TYPE YES OR
NO
YES
RUN = 2 HAS A 3.77
DOYOU WISH TO CHANGE THE PLATE SIZE IN THIS RUN
S\t'J ITCH
INCH PLATE NOW
TO ALPHA AND TYPE YES OR NO
NO
RUN = 3 HAS A 3.77
DOYOU WISH TO CHANGE THE PLATE SIZE IN THIS RUN
SWITCH TO ALPHA AND TYPE YES OR NO
NO
7:42: 4
KEYBOARD INPUT COMPLETED
INCH P LATE NOt'!
OPERATORS TYPING IS
UNDERLINED - ALL OTHER
IS COMPUTER OUTPUT
~-------"
... ,-.-..
,-.----~-
-,-".----,,,,._---------_._-----_._-----------_.... -
-
for all four companies.
The output will also list the storage requirements and industrial
curtailment needed to maintain gas purchase contracts.
At the end of the
day, a full report is made to the operator comparing the actual loads to
the forecasted loads.
2.6
DATA LOGGING
System logs are typed out each hour showing flow totals, pressure readings, load forecasts and load projections for all systems controlled by the center.
The format of the logs shows all of the variables
that the computer used in calculating the flow rates.
The logs are held
pending for one hour to allow the operator time to correct the log if he
disagrees with any of the computer data.
~f
a correction is made, an
amended log is typed out showing the computer's figure and the operator's
correction.
The only logs in use by the control center are the computer's
output.
System logs in a schematic format are also output to the operator.
These logs show the pressures and flow rates on the system during peak
conditions, and are a valuable tool for system analysis.
At the end of each day, logs are output showing a summary of
o
the center's operations for that day.
2.7
CONTROL
The intent of the control programs is to enable the 1800 computer
to control the gas regulator stations delivery pressuies under normal
operating conditions.
No provisions were made in the programs to handle
abnormal conditions such as electric outages, line breaks or equipment
failures.
The operators on duty should have control over the system pressures in the advent of any abnormalities, since their experience would
be needed to make the correct decision as to what pressures need adjustmente
Eight low pressure stations were put on computer control.
station has a tail end pressure.
Each
This tail end pressure is located in
the stations service area at the lowest pressure point.
A minimum pres-
sure limit is held at this tail end point (usually 3 Psig) by raising or
lowering the station's output pressure.
Each individual station has its
own operating characteristics, due to the size of the area served and
C
\
"
C.
r··'
the type of load in that area.
The eight low pressure stations were
chosen to provide statistics for next year's planning.
An expansion of
the computer system is under study that would enable the computer to
control all of the low pressure and intermediate pressure systems.
Both the delivery and tail end pressures are input to the
computer to give the programs the information needed to take control of
the station.
Using this information and a target set pressure at the
tail end, five decisions bands are set up to determine what action is
needed.
See Figure 11.
1.
Target Pressure + .250 Psig = Band one.
If the tail end pressure is above this band, the computer
will issue a lower command that is one half of a second
in length to the delivery station, providing the end
pressure is below the delivery pressure.
One-half
second is equivalent to approximately .25 Psig at the
s ta tiona
2.
Target Pressure -.250 Psig = Band two.
If the tail end pressure is above this band and below
c
band one, the computer will take no action.
3.
Target Pressure -.30 Psig = Band three.
If the tail end pressure is above this band and below
band two, the computer will issue a raise command that
'is one half of a second in length or approximately .25
Psig at the station.
4.
Target Pressure -.80 Psig= Band four.
If the tail end pressure is above this band and below
band three, the computer will issue a raise command
that is one and a half seconds in length or approximately Q75 Psig at the station e
5.
Target Pressure -.80 Psig = Band
five~
If the tail end pressure is below band four, the
computer will issue a raise command that is two and one
half seconds in length.
If the tail end pressure is
in band five, this could indicate a program or' equip-
o
ment failure as the pressure should not be permitted to
enter this area.
In addition to the raise command, an
error message will be printed on the 1816 typewriter.
39?
I:
I
!
BAND TWO
NO
TARGET
SET AT
3 PSI
ACTION
o
AUDIO
ALARM
C
FIGURE II
\
39f
3.0
PHYSICAL PLANNING
0"
Eigh t month's prior to the expec ted de 1 i very of the 1800 computer system, engineering and programming efforts were initiated.
Re-
modeling of the present facilities along with designing and installing
the input system were to be completed one month prior to shipment.
The
programming was done in four distinct phases with a test on each phase
at the IBM Test Center at San Jose, California.
At the Test Center,
experienced IBM personnel assisted Public Service in testing of its
system.
On the fourth and last test, the complete system was tested
as a unit.
Upon the completion of the test , all of the operating
programs were stored on the disk packs.
When the Public Service 1800
computer system was delivered in August of 1967, these same disk packs
·were put on the system and Public Service was on Line e
o
o
399
-------------_.._..........._............._._........
4.0
---~.
FUTURE PLANNING
At the start of the 1800 computer project, Public Service
initiated a two phase program.
Phase one would be to input 160 variables
from the field to the computer, log hourly calculations to the operators,
provide operation logs for the next year's statistical work and test
computer control of a regulator station.
At the completion of this
phase, the computer is doing much more for the control center than was
first plannedo
Eight stations are being controlled in the field instead
of one test station.
Engineering logs are being output
gas system's operation during peak periods.
showing the
Operator entry has ex-
panded to give the operator access to all of the computer's data, using
a method that does not require the operator to be a computer expert.
With the successful completion of Phase one, planning is now
under way to initiate phase two.
input
c~pability
Phase two would expand the computer's
from 160 inputs up to 256 inputs.
This expansion
would allow all of the center's values to be in the computer.
Output
would expand to enable the computer to control all of the low pressure
s~ations
along with the intermediate pressure stations.
require 100 output points.
This would
Both core and disk storage need to be in-
o
creased to allow room for additional programs.
G·
gmpnrlW" g II
5.0
COMPUTER OUTPUT
The following four pages are output logs from the computer.
Page one is the hourly system log showing all of the variables
and contracts for the four companies controlled by the Center.
Page two and three are engineering logs that were logged during
the systems peak period o
These logs show flow rates and instant pres-
sures in a schematic format.
Page four consists of system error reports to the operator,
gas load forecast and gas load projection called out by the operator.
o
n[
-
"lampe
t
m_
-
..
I
TU~SnAY
NOVEMBER
7 1967 311
DENVER HOURLY LOADS, AND rONTRACT DEMAND SYSTEMS OPERATION
MCF AT 14.65 PSIA
WSGE P-1(139000.)
CHF~FNNE P-1(46500.)
PUEBLO G-1(53500.)
CONTRACTS- DENVER G-1(375000.)
WS~E P-1(139000.)
CHEYENNE P-1(46900~i
PUEBLO G-1(55100.)
BEST DECISION- DENVER G-1(372300.)
TOTAL HOUR LOADS
FOOT MAPE MAPW
LEYIN LEYOUT C + P
"ENDIV NCOlD
DFNTOT PWPTS
DEN G1 WSEP1 CHYP1 PBLGI
O.
O.
872.
19576. 5275.
20447. 1743.
O. 1469.
O.
(CORR) 18704. 6147. 1400. 2219.
O.
O.
872.
19576. 5275.
20447. 1743.
18704. 6147. 1400. 2219.
COMP
o.
O.
O.
O.
O.
O.
O.
O.
O.
O.
O.
DIFF
ACCUMULATED LOADS FROM MIDNI~HT
16.
79.11952.
(CORR) 128438. 50531. 8617. 13931. 11633.
O. 14662. 131467. 35869. 148781. 20343.
COMP 128438. 50531. 8617. 13931.
7603.
O. 14662. 135497. 35869. 148781. 20343.
DIFF
O.
O.
O.
O. -4030.
O.
O.
4030.
O.
O.
O.
NORF PR .•
AVERAGE PER HOUR TO PFAK.
MESA HR NORF HR GRLY HR
696.5
15241. 5529. 2392. 2573.
O.
O.
O.
4278.
282.
1587.
DENVER HOURLY STATION DATA
NU~RFR OF rALCULATIONS 13.
(CORRF.CTED)
ACCUM
COMPUTER
STAT. GAS AVG AVG AVr, AVG AVrDIFF
HOUR
H2
H3
H4
H5
41MCF H2MCF H3MCF ~4MrF w5MCF
HOUR
ACCUM
ACCUM
(GRAV) PRESS TEMP HI
O.
HAMP 0.656 147. 39.5 22.5 26.0 23.6 24.1 23.8 1972. 2118. 2022. 2043. 2027. 12207. 84565. 12207. 84565.
23.7
2023.
~AS
0.0
LOAD CONTROL
0.0
0.0
CHCK
0.666
150.
60.0
~lTB
0.656
142.
43.0 25.5 21.9 21.2 19.8
ARSE
SIXT
0.656
0.656
238.
145.
44.0 10.8
34.0 20.4
ARAP
API L
0.656
0.656
102.
102.
51.0 32.2
50.0 0.0
0.0
0.0 17.2
CHER
CPI L
0.656
0.656
140.
45.
47.0 0.4
42.0 56.8
0.0
0.0
ZUNI
ZPI L
0.656
0.656
66.
21.
56.0 42.8
57.0 17.9
0.0 17.3
LEYI
LEYO
0.683.
0.683
233.
233.
29.0
36.0
0.0
0.0
0.0
0.0
o.
o.
O.
1825. 1691. 1665. 1608.
54.
403.
123.
403.
403.
836.
O.
O.
O.
38.
O.
O.
712.
27.
O.
O.
O.
o.
O.
472. 400.
57.0 13.5
o.
O.
46.0 0.0
DEN/NO COLO
MEAN TEMPS. REMAINDER OF DAY)
*MMCF AT 14.65 PSIA*
+ OR -EST
EST
LOG
EST
(SY.X)
CURT
FIRM
IND
ACCUM
12.0
0.0
DENVER nlVISION
163.2
45.4
131.4
DENVER G-1
128.4
8.0
0.0
NORTHERN COLO
41.3
17.5
35.8
WEST SLOPE P-1
50.5
-85.0
0.0
CHEYENNE
22.5
3.9
8.6
-45.0
0.0
PUEBLO
18.2
6.6
13.9
LIPAN AIR TEMP 35.0 CHEYENNE AIR TEMP 36.6 PUEBLO AIR TEMP
( )-MANUAL INPUT
CARR
PLAT
0.688 486.
0.688 158.
(ESTIMATED
LOAD FORECAST
O.
O.
41.
O.
O.
O.
6791.
53836.
6791.
53836.
O.
177.
1272.
1489.
8891.
177.
1272.
1489.
8891.
O.
O.
6353.
22.
6375.
7567.
718.
8285.
5461.
216.
5683.
7603.
O.
872.
14655.
O.
7.
836.
O.
836.
38.
130.
168.
712.
27.
739.
o.
O.
872.
O.
o.
6353.
o.
22.
O.
6375.
O.
7567.
718.
O.
8285.
O.
5467.
O.
216.
O.
5683..
O.
11633. -4030.
O.
O.
NOVEMBER
o.
o.
14655.
7.
CHFY 34. PUF.BLO 41.
r.ARR+
TOTAL LOAD
EST
MINUS
CARR+ PLAT
ACCUM
TOTAL
BEST DECISION PLAT
352.1
-20.1
349.0
21.4
14.6
102.8
-21.4
117.5
-96.7
-49.8
-6.2
-61. 3
24.9 FT. COLLINS AIR TEMP 30.2
TUESDAY
'~
O.
836.
O.
836.
38.
130.
168.
712.
27.
739.
O.
O.
O.
130.
o.
o.
CIG
EST
LOAD
398.8
7 1967 3.11
8:00
~.
()
o
CIG
ACCUM
LOAD
189.8
MST
______ ( 2 _ .
o
o
~
-.:
OPERATIONS ENGINEERING REPORT
TOTAL DENVER LOAD AT 14.65 PSI~ RAsr
G-l 19354. + PLANTS
1888. + LEYOUT
u. - LEYIN
O. CARR-PLATTE
802.
TOTAL 22044.MCF
PS-1 GRANTED
O.
IS-2 GRANTED
O.
LIPAN AIR TEMP = 36.6 DEG.F.
COINCIDENCE =14.0%
TOTAL DENVER CURTAI LABLE LOAD
(41.0) DFG. r.1FAN TEMP.
O.MCF
TOTAL DENVER CURTAILMENT LOAD
(41.0) DEG. MEAN TEMP.
10182.MCF
F-S =
O. F.-S =
O. D-S
. O. 24G-S
O. r-s =
O.
TOTAL = 10182.MCF
*
LEYDEN MINE
P 88TH rARR
F-161
88TH PLATTE
-P-----------------R-R---------------+-+---------------+--+-----------PR-------------------------PR*O.MCF
107.*
I INJ.
O.MCF IR 802.MCF
I
158.1
I WITHD.
O.MCF I
515.1
I
+--+
I
I F-340A
I
P F-358
I
87.1
I
I
I
I
1 85.*
1 F-340B
+---p F-l71
I
I
1
89.1
+-+
80.*
I
CHEROKEE
I
I
F-402 P
R*P 131.MCF
I
I
1578.MCF RP---+
I 142.1
F-80
I
1
39.MCF W-E I
I
F-392
ARSENAL
1
I
122.# W 126.*F +---------+---------------R*P---------------R*P
I
1 74.INORTH
NORTH TB RR
179.MCF I
+-------------------P F-210
143.1 I CITY
242.1 I
R 58.ISOUTH
I
6479.MCF I
97.'
I
1
P
+--+----------+
I
I
I
I
I
I
I
I
I
I
I
1
52.1
+-------+-+
721.~CF
+---------+
I
I
L. P. STATIONS
DEL.
4.8
46.4
0.0
23
6.3
76
0.0
F-4
6.7
F-7
F-191 0.0
F-230 7.8
F-269 0.0
END
3.0
23.2
0.0
3.1
0.0
4.5
0.0
2.8
0.0
R*P
F-308
6TH AVE
l083.MCF
*P-----+
I
I
68.*
I
I
I
I
STA 11
P
145.1
+-------+-+
68.1
I F-362,
*RP-------+
O.MCF 1
83.IW 1
2
7
I
F-3 P---I
96.11
I
ZUNI
1
1
1
STA.
I
I
+-----+--+
+---------P F-2
+---+
ARAPAHOE
*P 1036.MCF
+------------+
+-P YALEJCT.
P----------------+-----------------------------------~----------RRP F-248
JEWELL-WADSWORTH
1 100.'EXP.
I
87.1
I 134.IYALE
I
F-12
R
+----------------------------PR-----------------------------------P* HAMPDEN
I
112.1
I 11613.MCF
I
************
I
148.1
I
1
103.1
*
I
I
+---+
1
P F-407
89.1
*
*
*
*
*
*
I
*=MTR STA*
R=REG.
*
P=PRESS. *
#=PSIG. *
1
I
I
+-----+
*
I
************
CHERRY CREEK R*P
6.MCF
0.1 .
'&
~
TU~SOAY
TIME
..
NOVEMBER
8:12: 6
7 1967 311
()PEr~AT
I JoiS Ei'lG IIJEE;{ IljG R~P0RT
AT 14.65 P31~.
lO~OS
AESTERN SLOPE GAS
EASTFn~
CHEYENNE 1I GUT FUE l Pu.iEi{ CO.
0IVI~ION
FRONTt Ei{
+--------R*P
I
76.1J\CF
I
40. t
***lEGE~m***
iWhFOlK
R*P
253.:H~F
640.#
I
I
I
I
*
*=;lTR :>Tfl*
P= PRESS *
H=REG.
*
H=P,:jIG. *
*
*
*
*
*
*
*
+----+
CHEY I.P.
,I
+---P
NO.
30~.#
PIP.
iJ2.11
Fl*P
TEilRY
451. tI
245.i·tr,F
I
I
I A I R Tl-1 P • 2 ~ • 5
15~7.~CF
I
I
I
I
R*P
I
I
I +-------+
+-------+ I
I
I
NO COL
RATE CURTAllABlE
S-f,!C
O. ~·'CF
o .t·1CF
I-NC
O.Mr,F
E-NC
O. i'" CF
TOTAL
&
~
321.#
PS-1 GRANTED
O.
'~-2
O.
GRA~TEO
I
5073. LEYliJ
'-
"
CURTAllEu
u .i·1CF
(j
.1·4CF
EST. MEAN TEMP
31.2
AIR TEMP
35.7
I
I
o.
117.1
G-1
~ATE
F-S
SO COLO POJER
R*P
2244.MCF
CURTAllABLE
O.MCF
SCP580.MCF
CURTAILED
O.MCF
E-S
O.MCF
O.'-1CF
D-S
O.MCF
o.I\",CF
O.MCF
C-S
33.1
+---------+
O.MCF
I TOTAL
EST HEAN TEMP
R*P
AIR TEMP
so TO~IN BORDER
580.MCF I
o .i4CF
U.MCF
41.0
24.1
139.1
1401.MCF
CURT,'\llEO
o .MCF
o .MCF
O. ~1CF
o • r.-1r,F
) MANUAL HOURLY ENTRY
TUESDAY
NOVEMBER
TIME
7:32:37
r-~
O.ACF
uEVI NE
PR------------R*P------------------------RP
843.r·ICF I
BOONE
I
I +------+
I
590.'(16 INCH)+------+ t
I
I I
CARR P 471.#
PLATTE
I +--+
*R
R*P
+--+ I
R. 1062.t4CF
O.r.it::F
I +--P
321. i
LEYJEN *---+
158.*
+----P 428S.MCF
O.MCF IN
P*R MESA
o .!-iCF OUT
EST l1E/\'N TEr',1P 41. 0
1062.
15-2 GRANTED
PUEBLO GAS AND FUEL CO.
I
I
C-P
O.i'1CF
9IlO.liCF
420.*
+---+
+--------+ I
I
I + - - - + 8 3 G• # ( 10 I rH~ q )
I
+---+ I
6135.
PS-1 GRANTED
R*P
STATElI tJE
407.#
o .:,1CF
·O.lqr.F
PS-l GRANTEJ
I
1:>-2 GRAnTED
P--------+-+-+
BOULDER +-+ I
345.*
+----+
P-1
7ti.MCF
E-S
G~EELEY
I
I
FRONTfER
RATE CURTAllABlE
+---P SO. 254.11
I
1205 .,liCF
I
FT COLLI ~!S
!~86.i
P-l
I
I
P
lAPOi1TE
I
+--------P--------+
************
I
I
I
119.*
()
7 1967 311
(,)
o
TIME IS
TIME IS
TIME IS
7:43: 2
7: 43: 7
7:43:16
o
~
ALAMOSA HIGH PRESSURE
CHEYENNE INTERMEDIATE PRESSURE
F -210 INTERMEDIATE PRESSURE
NORTH
OUT OF LIMITS 160.8
OUT OF LIMITS 124.1
NOW BACK IN LIMITS
69.7
(ESTIMATED MEAN TEMPS. REMAINDER OF DAY)
DEN/NO COLO 41. CHEY 34. PUE~LO 41.
CARR+
LOAD FORECAST
*MMCF AT 14.65 PSIA*
TOTAL LOAD
EST
EST
EST
LOG
-EST
+ OR MINUS
CARR+ PLAT
FIRM
INO
ACCUM
CURT
(SY.X)
TOTAL
BEST DECISION PLAT
ACCUM
176.3
48.2
111.8
0.0
10.0
346.4
DENVER DIVISION
109.7
344.2
-25.8
DENVER G-l
44.0
13.7
NORTHERN COLO
18.6
30.5
0.0
10.0
103.2
21.9
WEST SLOPE P-l
44.3
117.0
-21.9
24.2
4.1
7.2
0.0
-12.0
23.6
-23.2
CHEYENNE
19.7
7.0
11.7
0.0
-4.0
34.4
-20.6
PUEBLO
LI PAN AIR TEMP 28.9 CHEYENNE AIR TEMP 31.2 PUERLO AIR TEMP 13.3 FT. COLLINS AIR TEMP 23.6
LOAD PROJECTION AT 7:44:42 MST
DENVER G-1 CARR PLATTE WSGE LEYDEN IN LEYDEN OUT CHEYENNE FRONTIER PUEBLO SCP
HOUR PROJECTION
2250.
580.
1316.
75.
o.
O.
o. 6141.
1118.
18997.
DAY PROJECTION
49802. 13942.
1783.
28149.
7603.
o.
7. 147725.
34421.
425291.
TIME IS
T I ME IS
TIME IS
7:46: 3
7:46: 4
7:46: 9
7:46:21
ALAMOSA HIGH PRESSURE
ANTONITA HIGH PRESSURE
CHEYENNE I NTERMED I AT( PRESSURE
CHEYENNE STATIC PRESSURE
OUT OF LIMITS 158.4
OUT OF LIMITS 169.5
NOW BACK IN LIMITS 108.6
OUT OF LIMITS 420.0
TIME
TIME
TIME
T t ME
TIME
7:46:32
7:46:34
7:46:41
7:46:41
7:46:44
NORFOLK GAS TEMPERATURE
TERRY ROAD STATIC PRESSURE
NORFOLK GAS TEMPERATURE
TERRY ROAD STATIC PRESSURE
CHEYENNE NUMBER ONE DIFFERENTIAL
RATE OF CHG IS OVER LIMITS
OFF SCALE
nO"J ON SCALE
N0\11 0 N SCA LE
CIRCUIT OUT OF SERVICE
T I ME IS
§
\
IS
IS
IS
IS
IS
CIG
CIG
FST
ACCU~
LOAD LOAD
507.7 161..4
~. 1..-;
THE CATALYTIC CONSTRUCTION COo
7¥
(0/
1528 Walnut Sto Phi lao Pao 19102
A Program for Making a Non~Linear Regression
Analysis with up to Three Independent.Variables
By ToEo Bridge
This program will be very useful for curve fittingo" It will develop
coefficients for calculating a dependent variable (or function) when up to three
independent variables (arguments) are specifiedo There are nine options available
involving various combinations of polynomial D logarithmic D or reciprocal relationships between the variables
It was designed for system 360 Model 30 = 32K with
2 disk drives o
'
o
This program will also be useful if you need to correlate a large mass of
test or operating datao
Four numbers and a title are read by the machine on the first card
NI, N2, N3 g N4 g title~
Nl specifies the degree of the fit for the first
and N3 for the third~
argument~
(4I2~18A4)
N2 for the second,
N
Type of Fit
o
Is not perrnitteo
I
Calls for a constant or average value for the
function not influenced by the argument o
2
Calls for a linear fito The machine will find a straight
line that gives a minimum root mean square erroro
3
Calls for a parabolic fito The machine will find a
quadratic equation to give a minimum RMS error
o
0
4
The machine will fina a cubic equation to
give a minimum RMS erroro
Any positive whole number or combination of numbers may be entered for
N2, or N3 p so long as their product does not exceed 100
Nl~
0
You must enter a total number of data cards greater than the product of Nl x
N2 x N30 The more the bettero
N4 may be any digit 1 through 9
It specifies the option that is desired -or whether you want to try for a smooth fit on a graph with linear or logrithmic
scale's, or' a semilogrithmic graphS) or even on a graph with a reciprocal scaleo
0
Let Y represent the function 9 and liet Xlv X2 D and X3 represent three arguments
fitted to a degree specified by N1!.j) N29 and N3 respectivelyo ThenS) the following
table wil19 for each option specified by N4D show the type of graph that 1s assumed
by the programo
C,";
~
MW"W"U
V
!
;
reWf
Regression Analysis Program
c-
N4
Polynomi~l
I
2
3
4
YDXI"X2"X3
XI!)X2j)X3
X2!)X3
X3
2.
Logrithmic
Y
Y"XI
Yg XI"X2
Yj)Xlj)X2,X3
Xl
Xl,X2
Xl"X2"X3
5
6
7
8
Y"X2j)X3
Y"X3
9
XI"X2"X3
Reciprocal
y
Y
Enter any desired symbol in columns 76~80 of the title card. The correlation
coefficients (the answers) will be punched out by the machine ready for insertion
in a Fortran program decko Each coefficient will be indentified by the symbol
entered in columns 76-80 of the title card. It is then a simple matter to write
a Fortran program to quickly calculate the function from any given set of the
arguments. A subroutine (TB$PLY) should be put in your relocatable library for
doing this
o
C in the above call statement must be replaced by whatever symbol you put
in columns 76-80 of the title card. Y is the answer or function, and Xl" X2"
X3 are arguments corresponding to the fit parameters NI, N2" and N3 respectivelyo
Data follows immediately behind the title cardo Each card (or row on the
data sheet) will include four numbers in order =- Y"XI"X2"X30 (4FIO 3"IS)
o
The problem is terminated if a digit is read in columns 41~4S. If a zero
(or blank) is entered for Xl" X2" or X3j) the program will enter the preceeding
value for that argument. Therefore" if you want to enter a real zero" enter
some very small number close to itD asp for example -- 00000001. Of course" if
you are making a logrithmic fit" you must not enter a negative value for the
variable 0
Explanation of the Program
I think that my explanation of this program will be a lot simpler if I
talk about a specific formula rather than about a generalized oneo Consider
the following formula~
Y = Ca
+
Cb Xa
+
Cc Xa 2
+
Cd Xb
-+
Ce Xb "a
+
Cf Xb Xa2
In the above equation!) we are mak1ng ~ 3 point fit with respect to Xa and a
2 point fit with respect to Xb. That is; if we hold Xb constant" Y will plot as
a parabola against Xa; and if we hold Xa constant!) Y will plot as a straight line
against Xb. We will read from a family of curves a value of Y for each of many
different values of Xa and Xb
Our program is to find values for Call Cb" CC II Cd,
Ce" and Cf that will best fit the data that we give o Let Yg be the given value
of Y for any point corresponding to Xa and Xb
Let E be the sum of the square of
all of the errors for N given pointso
o
o
Regression Analysis Program
E
=
r
3.
~O
(Y_Yg)2
:
Differentiate with respect to Ca. Since we are looking for a minimum value
of E, this derivative may be made equal to zero.
l.
dE/dCa
.
= 2.,
but dY/dCa
I
Ca
+
1: Cb
Xa.· •..[
(Y-Yg)dY/dCa
= O.
therefore
'" 1
Cc Xa 2
+
I
Cd Xb
+
L Ce
Xa Xb +
r Cf Xa 2 Xb
=
I
Yg
Similarly if we differentiate with respect to Cb, we can write another
equation:
dEl dCb = 2tCY - Y~;)
but dY /dCt = Xa
I
Ca Xa
+
4
-'
Z
oi
o
0:
W
(/)
(/)
w
r
!-:--+...:...;.-:..t-..;....,--;--l--}--+--l
i
!
i
!
{.
t
j
~
I
1
t
-+::-"'F--+--+--+---+---<---i--~--T-~--t·--r-l,-·-r---1
;--t
1;
,
1
I
T--+-----'-i
f
iI
t--i--t---i--'-+--+\------!--!~;t'
I
,
·~-+--+--+--4--i--j--+--~-~-+-~-+--+--+--~---1
i
I
!
j
-}--i---i
~/O
I
-r!"
-
r
OO,-T--T
---~~I~:$~T Al
_
PRUGRAM BY TEB
MULTIPLE REGRESSION
C
SAMPLE TITLE CARD ASSUMING
X2 ,AND 2 FUR Xl.
.I_HE
N4
C.
1
C
2
C.
3
.04
TEST PROBLEM
IS CALLED Nl,
TERMS IN EQUATION FOR
4
y
1-23-61
'jj
OOOA
.. .0.0.0.7
OOOR
0009
OOIn.
Xl , 3 FOR
aOll
Xl X2 X3
VS
0012
t..AND
..
Ol
.IS
_0013
02 IS. CALLEU.J13
[S CALLED NLt
0014
IS A VARIABLE GIVI~G ONE OF N[~E OPTIONS. IF N4 IS
WI HA\tE A STRAIGHT POlYNOI-1IAl OPTION
W~ HAV~ A SEMI LOG UPfIUN. LOG OF
Y UNlY
WE HA VEL 0 G 0 r Y
AND
Xl
ONL Y
WE HAVE LINEAl{ WITH RESPECT TO X3 ONLY, LOG OF OIHER VARIABLES
C
4
... _£ ...... _ .~ ....2._. w..C . HA..V.E LUG
~
6 WE HAVE LGG
C
7 . WE HAVE LOG
WE HAVE LeG
C
8
C
9 wE HAVE THt:
Xl
OF
Q01.~..
0020
ONLY
X I , .Ai\iD X2
OOl~
.
OF ALL V1\1{ I ABl ES
o~
0015
0016
.0011
ONLY
OF X l , X2 , ANU (3 ONLY
RE:CIPRUCAL LF Y ,AND LINEAR WITd ALL OTHERS.
.0021
002?
0023
002'+
c
.Stu'1.PLE D.AIA CAHD. USE FIELDS Uf TEN ..cOLUMNS£AC1:L ...... _.
C..
c
c
C
ARGU~ENT
IF AN
.~____.... _..THEREFURE~
C
.42
C
C
C
C ..
002Q
003~
[S bLANK THE PRECEEOING VALUf WILL BE USED.
DO !~OT USE A TRUE ZERO. ANY VERY S.MALL NU~.B.E& wILL .. DQ...
USE A lAIL CAKD wITH A
C
0027
0023
X3
4.
X2
Xl
3.9
~ ·~~.1
1
IN COLUMN 41.
0031
_.. 0032
0031
0035
0036.
0037
WE READ DATA AND BuILD jlE
O.4500JE
03
03
03
03
03
03
03
0,3
03
03
03
03
03
O.45000F. 03
o • 45.00 DE 0 '3
O.10000E
O.lOOOOE
O.looaOE
O.10000E
O.lOOOOE
O.lOOOOE
O.lOOOOE
O.lOOOOE
O.10000E
O.loaoo£:
ELeD VS OS
YG
01
01
01
01
01
01
01
01
01
01
O.lOOOOE Oi
O.lOOOOE
O.lOOOOE
O.lonOCE
Q.10000E
2-
01
01
01
ot
0.18000E
O.18000E
U.l8500E
O.19000E
0.195001:
y
~
DSGN
Eft PC
01
01
01
01
01
O.lHGI}2t: 01
-·o.oeq
0.18417E 01
0.19032E 01
O.19488E 01
O.lHOOOE 01
O.1!994E ;]1
0,.20500£ 01
O.2~OOOE Ol.
0.24000E 0t
O.270QOt: 01
O.18000E 01
O.23~()OE 01
0.25100£ 01
u,.2G~2C)F 01
O.219b7E 01.
o• 2', 0 f) 4 E. 0 1
O.27004E 01.
0.124
-0.167
0.062
0.031
·,0.144·
(J ..
2'-}~OOE
0 ..
O.34500t': vi
O.l8002E at
-0.011
\~. 149
-O.Ol.J
-0~Ol5
oa
1 8 00 'I' E () 1
o • 2 2 q 6 2 t t) 1
-0.023
O.t64
J.25169E 01
-O~268
J.29454E 01
O.34511E 01
-0. O~i 1
0»1':51
CL
USER EXPERIENCE WITH 1130 LP-MOSS
An Abstract
U. S. Reduction Co. has used the 1130 LP-MOSS application package
since its first release. Various problem.s have been encountered during
its use; however, useful answers have been obtained. Due to the size of
the aluminum alloy blending problem being studied, a second 1130 had
to be leas ed for full dedication to this problem. Early results indicate
a reasonable payback may be obtained from this hardware and software
combination.
An Outline
1.
Company background
2.
Early uses of linear programming
3.
Problems encountered in implementing 1130 LP-MOSS
4.
Results to date and future planning
5.
Specific application modifications desired
o
··TT'y
nil
o
APPLICATION OF SIMULATION IN CONTROL, DESIGN, AND
OPTIMIZATION OF CHEMICAL PROCESSES
M. J. Shah
ABSTRACT
The problem of superviso~'confTol ana optimization of a chemical process
requires thut the control as well as the optimization programs be provided with a
mathematical reJationsh ip between the dependent and independ~nt variables in the
process.
I~ this presentation we will discuss methods of arriving at this relationship
based On i·he fundamental laws of physics and chemistry. We will make the proper
approximations valid for control purposes to obtain the solution of the differential
'0
equations, describing this relationship. Use of simulation l0.ngL:agesherps in
reducing the pr()gramming time as well as the number of.trials required to attain
successful solutions.
An example of methanol plant will be discussed in SOme detail to illustrcte
the mathematical and programming techniques to achieve .the desired relationships
for control and cptimization. It will be shown how these differential equations with
modification of the obiective function when used in the optimization program can
. lead to optimum design of the plant.
Paper to be presented at the Common User's Meeting, San Francisco,
December 11-12, 1967.
o
4/7
o
INTRODUCTION
,In the early applications of digital computers for controlling chemical processes, the work performed by the cO,mputers was limited not only by the restricted
capability of the computers, but also by the fact that control and optimization
, techniques had not
been achieved by personnel of the chemical plants. Much time
,
and effort was spent in data gathering, alarm scanning and limit checking of
important plant variables. Since the interrelation between the plant variables was
not too well known, or in many cases not defined mathematically., virtually all of
4
the control work was done by using so-called regression models' derived from plant
data gathered by the control computer. These regression relationships were
generally simple enough that computer control could be achieved without overtaxing
the capability of theavailable digita.1 computers.
More recent digital computers in the process-control field are much more
powerful in their computing capabilities. They also provide some very useful readymade programs for performing the various control tasks in a plant over and above
the normal logging, alarming, and scanning functions.
These programs will be
discussed in subsequent sections.
It is important to realize that a control computer must operate in a continuous
mode, iust like the chemical process it is c.0ntrolling. The computer thus must
perform all the routine data logging, alarm scanning, and limit checking functions,
plus fhe functions of supervisory control, direct digital control, optimization,
engineering simulation and other desirable tasks.
In th is paper we will discuss the various process control tasks of a control
computer and, in particular, theuse of sjmulation -- off-line as well as on-line -in control, optimization and engineering design calculations'of a chemical plant
with a process-control computer. The discussion will, needless to say, relate to the
available software for the IBM 1800 Control System, although the technique is not
limited to any particular control computer.
o
lIlt!
C
t.",
i
III
n-n
s·
!tIn
Figure 1 shovv's the variou~ functions which the presently available IBM
control computer performs. The central executive program is the time-shared
executive system (TSX).3 This program
a~locates
the processing time
~vailable
for
the various functions which a control computer is requ.ired to perform. The time
allocation is done in an optimum fashion, in that the calculations which have the
utmost importance have the highest priorities. There are twenty-four or more levels
of priorities available on a computer, allowing any higher level priority calculation
or computer function to interrupt a lower priority level function or calculation which
,is being carried out at a given time. The interrupted calculation's are not destroyed,
but stored in the core or on an auxiliary memory storage such as a disk. Thus, after
the higher priority calculations are c,ompleted, the lower priority calculations
which were interrupted are resumed at the point where they were interrupted. When
any priority interrupt calculation is completed, the computer seeks for lower and
lower priority calculations, until the interrupt calculations are completed. At
that time, it is available'for offline nonprocess calculations or it is in idle" mode.
A core clock is provided to perform routine functions such as data logging and
o
report generation at specified time intervals. The limit <:heckup of plant variables
(violat:on of upper or lower limits) is done by a comparator, and the normal computer
calcula!ions are not interrupted unless a limit is violated, and specific computer
processing action is required as a result of the variable limit violation.
COMPUTER FUNCTIONS
- Let us look at the various functions and steps in process control that are
performed by the computer.
Data logging and l'ilanagement Information
As soon as a computer is instarred r it is used for data logging and management
information such as production of a particular chemical in 24 hours, the amount of
fuel used, etc. The logged data is also important in model verification which will
b~ discussed in subsequent sections.
The importance of obtaining reliable plant
data for model verification cannot be overemphasized. The co:nputer is much more
reliable for logging than a human operator. Furthermore, a programmed fi Iter (such
o
2
r. '10, .#,,,;;,,"'' ' =*:.::::1£,; 11,,41, .... n' itLa',l,,;;;::mu,lJ ..iJZil\litJL4&ittGJ,wPUM4&:1&iLlUll&==GIl"'dZhldl,MWJL&aidiiiiiiJlWiWAii,.............
as time averaging, exponential smoothing, etc.) can be used to sort out the
instrumen t signa I from noise. Other rounne functions such as a larm messages and
o
emergency operator action sequence can be handled as a first step in computer
control.
Regulat~~
A second function which a computer performs is regulation. The easiest
regulation is by means of an operator guide which is defined as a~sequence of
control action messages to be taken by the plant operator when dish.:rbances or
upsets in certain plant variables are detected by the computer. To cite an example,
a cement plant had a sef of four kilns of identical design, all fed by a single raw
material slurry feeder and burners using a single fuel. Each kiln was controlled by
a single operator for every shift.
From a total of twelve operators, two were able
to control their kilns in the best manner. It is easy in a case like this to program
within the computer the sequence of actions which the best operator 1-akes as a
r~sult of the disturbances.
This sequence of instructions is printed for ~ operafors
to follow when simi lor disturbances occur.
In this nlonner, the operation of a II
kilns reaches the level of best operator performanC":'e.
o
Superv isory Con tro I .
In a normal plant operation, the operator makes changes in set-point positions
for one or more control loops when a disturbance occurs.
For a chemical plant with
a large number of control loops, there are unavoidable interactions between control loops. An experienced operator can handle two to three loop interactions and
make simultaneous changes in set points of the loops to bring the plant to a desired
·Ie\,el of control. It is difficult, if not impossible, for the operator to handle more
thon three loop interactions. On the ot~er hand, a digital computer can be programmed to adjust one or more of the set points simultaneously by solving a set of
equations which relate the variables of the plant to the multiple set points (target
variables). A typical equation may look like equation 1 •
o
3
(1)
where F1 , F2 ", F3" etc. are constants, T is the target variable to be "controlled,
M is the manipulated variable, U is the uncontrolled (disturbance) variable and
V1 , V
2, V3
are other measured variables which enter from interactions with other"
loops. Thus, for interacting loops a set ofequations such as equation 1 is solved to
control a set of target variables.
In this manner the computer can not only handle multiple-loop interactions,
. but also make the changes in a consistent manner. In addition, since each loop has
different dynamic characteristics (slow or fast response, short ~r long dead time,
first, second or higher order system) I each set-point adjustmen't can be carried out
by the computer in such a mann er that di Heren t dynarni c characteristi cs of each
individual loop are token into account. Although this technique is in no sense an
optimum dynomic control scheme, it is, because of its practical value labeled a
poor man's dynami c control scheme.
c
Figure 2 sho'NS how th is scheme works.
Let t
be the time at which the computer colculations require t~at C~ adjustment of set points
from say 60% of scale to 80% of scale. The adjustment may be made in severed
steps of time intervals.
Let us say that the interval between two adjustments is
and the full adjustment is to be carried out in five steps.
~t
If the loop response is
fast, the adjustment can be performed in five equal increments, or any combination
of unequal increments which make the total shift of 20% scale. For a sluggish loop,
one can provide incremental steps which includes overshoot as exemplified by the
dotted line in Figure 2.
If there is a dead time involved ,one may incorporate a
delayed action adjustment for the set point-.
In certain areas of the plant, it is possible that one or more control valves
ore changed manua lIy, and no ana log feedback controll ers are provided. The
computer can perform the adiustment of these valves via a steppin8 motor if adjustment equations for changing the valve positions ore programmed in the computer.
Thus, the supervisory control can olso provide
,a
direct computer control (DeC) of
the plant valves. As opposed to direct digital contror (DDC), the DeC function
o
4
only provides proportional control and at more prolonged intervals (1 minute or
o
more). DDC takes on Iy seconds.
SUPERVISORY PROGRAM
The various functions described above, routine data logging, alarm scanning,
limit checking and operator action as a result of limit violation, programmed
operC4tor guide control, supervisory control and direct computer control can be
performed with an IBM program known as PROSPRO/1800 (Process Supervisory Program for 1800). The detai Is of this program are in reference 4.
In brief, PROSPRO program is written in such a manner that virtually no
programming knowledge is required on the part of plant personnel. Howeyer, the
program requires that personnel know the behavior of the plant thorough Iy, and,
more important for our discussion, that relationsh ips for interacting variables in th'e
plant be written in mathematical form. A general type of relationsh ip is available
within the program, and one only needs to fill in the value of constants in the
o
general adjustment equation.
PROSPRO requires that each variable in the plant be uniquely identified by
a number. Then, for each variable, six standard information sheets are fi lied out.
Two of the six sheets, the variable information sheet and the adjustment information
sheet, are shown in Figure 3. Most of the required information is given by entering
a 0 or 1 in certain columns, giving values of limits, actions as a result of violating
these limits, and so on. The adiustment equation relating the variables of the plant
is linear; however, the values of the coeffi~ients F , F2 ••• may be calculated
1
by using a general type of non linear equati on such as
Fj
= A + B [C + (F k /
(2)
D) ] E
o
5
---------,---.---~-~-----,,---------------~
'.
-------- .._------------""
,r'·
o
w' '''In] 'Tf I[ "["Mwr
"J"
fW'ij# Hft. 7"( firtffi • " WittHf'tMtrlt beth Ott t" , "tt "tt t 'tt
hy"" "rl b, 'c I',.*,' i 1M .,i,n')' t!!lwt'rtettieit+.W;;; 'nW'tN'i"'t'!?
i lWWrYrit"hSrl'UHtUMhlt:!l ....W;tWw'iW¥tiMw'WN1wUI'%t'
where A, B, C, D and E are values of constants to be supplied by plant engineers. Once the six sheets are completed for each variable, the information from
the sheets is transferred to punched cards and read in PROSPRO as data. The sheets
provide permanent documentation and ani alteration in information on one or more
variables is carried out by ~ltering the corresponding ~heet and the data card,and
reading in the new card in PROSPRO. Some of the features and capabilities of
PROSPRO are outline-d in Table I. 1
DIRECT DIGITAL CONTROL
In some chemical plants, it may be desirable to install a digital computer
instead of analog controllers, especially when a new plant is being built. This is
cO'llmonly referred to as direct digital control. It is importan~ to note that a digital
computer can not only perform, the function of analog controllers such'asproportional,
de~ivaHve, and integral control, but also the more sophisticated control functions
such as nonlinear control, control of loops with fixed dead time and variable dead
time, adaptive gain tuning, and so forth, at no additional cost. Ana log control
devi~es which perfcrm these functions can become prohibitively expensive as the
number of loops requiring sophisticated control increases. In addition, a digital
computer performs the rO,utine logging, alarm scanning, and limit checking functions.
Direct digital control on a chemical plant with an IBM 1800 is attained by
the standard program, DDC/1800. Again, for details of the program and control
techniques, refer to refernce 5.
~~~
I
~
o
ff4liifuftiti:\; iJ
'.% mu,~
6
' ';" ",4:#,; ..I, ,n "",4 ""; ...:
g ;":;,t;r.m,W'",l';.u IOto.C
-r .... Q.GE""I'"
'8o
~
lOll!
C .... lCVI. ....TC':)
nv
I
MA.INT ..... , N
(;10138,
r,...
R -
1
C 4) N V £ It S!of
IdentHicat.ion UHd lor Me.u,e Output.
_ Engineel'1ni Unit. uled [or Con.ole Diaplay.
Proce .. Unit ot VariAble. (Ol:c..=;:). ll:e::::::;:),
,
G-rid Column/Row - Con.Ole Grid Po.ilion.
2-
Adjuume-nt Equation:
Next Unit)
[QIQ] III
~.
1
~
I
I
I III
I III
I
,
.'
I
-,.
.
Equation Output Re s t rictions.·
9 ajAction taken when Variable M Out oI Service.
I '0 I 7 ,!Action taken when Variable M Of{ Computer (On Operator).
I 0 .1 Maximum allowable adjustment to M per pas ••
2:2 06 41Action taken when Target o{ M already at MAX-MIN Limit.
,
jAdju&tment Re(erence List {or Partial Loop Test. (ZftI\I\I\ or Blank)
2 09 2. , If Entry 14 specifies a Reference List, Entries lS-:a specuy s"bUst.
(-1. -1., -3. _.1) referenced (or all subsequent M f • in control1ttop.
\ II Entry 14 is Dlank, Entries 15-i8 specuy ~ll of the subsequent M .
Variables in control loop.
Target o( M ael to Average in"tf'ad o{ Current for Partial Loop? (1a Ye.)
[TI] I
7 8
, OJTIME
l
4
in minutes (bXXXX) or as developed (-nnnn or 1.nnnn).
10 ,.T 14 '.AM
o
75
- ;z. 5
m~ ~
.'.,3
20
. 5 0
l.IS
I
(1)
(2)
(3) \
(4)
=Cumulative elasped time
'1.T
,. AM
=
Crom pre.ent time
expressedaeapercentoCTlME.
Percent oI AM changt: made alter ~T time •
(Note: Negative ,. AM ia permiaaible ,)
Special Action lnitiated Upon Significant Change in Manif'ulated Variable (M) •
..
7 8
3 I
10 Initiating M Change l.l
1
"'1
H,~: :::::::::::
3 5
3 b
_ .
_ _
1.1 Acti,,"
"0, 2
2
Action may'be apedlied aa follow.:
(.) P,oecial Action.
:.unit Che~
ction.:
----~ .. L;ned Proce . . MAXIMUM Limit.
III . .3.00:'12 (No ViOI.-ViOI.)} Sp~cial Action tuen 'Whenpallinl throu,h
~;
(Viol.- No Viol.)
the Aaaigned Proc'!" Maximum Limit.
~
6 o· •• igned Proce •• MINIMUM Limit.
@ I I: 0 2
(No Viol.-Viol.)} Special Action ta.l. the Alligned Proce .. Minimum Limit.
IO
pper Specid Action Reference Point.'
I~ 7
(Below -Above).} Special Action taken 'When pa .. iD, throllih
l:8
Above - Belowl
the Upper Reference Point.
we r. Special Action Re!e rence Point.
~
I (Above -aelO W )} Special Action tAke'n ...,hen pu.tnt. throulh
31 tj
I (Below_Above) the Lo ... er Reference Point.
)til .
J .jAuigned Delta Limit for OpC'!utor NotUication.
3
(No Viol._Viol.) Sf-ecial Action ta.ken ""'hen DeltA Limit exceeded.
3.~
elta required for A Predictive Adju.tlnent Action.
\ ]. L
Aclion taken {or A Predictive (Delta) AdJultmect.
Taq;et Value ADd Deviation Proee .. ing:
~ I Doe. Vari~ble have a Target? (1- Yet, iollowlcg 10 item. appUcAble)
~~.u;
I 5 Mlllimwn Time bel4c .. n lueete .ive T.1Ti~et CaLc. And/or Dev. Adj.
. 4.ll
3 ~J.eediicd ArlioD to evaluate New Tugel Valup.
413
_ .......___(.:}.Bil:ned Deviation Limit for Oper~tor NotUicatioll.
'-\ ,i
No Viol. _Viol.) !)?e cia I. Action take.n ",hen DeY. Limit exceeded.
,5 .
• ~p \(i:.tiOD re'iuired for :-.lormd Adjustment Action.
.
~~~-1
?""~5.!.!.£..~en tor !\::.rrnal DeviJ.tion AdJuument.
.
'7
.
.
~.l.a"lmum Set Point Adju~tment po:r Pl.,,·
•.;8'
S~t Point Output. (I- CODtroll~r, 1. = DOC, 3" Operator M~UA&e)
~·9
Computet Set ·Contrt.>lIo I DOC Addre ...
~..Ej
~et Point Movement Rate. (1: ~! ••• ute, 2= lIZ Max •• 3;'1/3 M~ ••• tc.)
~
FI.nAI Special Action.
..t:
LEvEL
.--.1
12;
.
m
SI~E:O
Targcot. Current(AYerage)JCalculated Adjustment
,J' 2' I
~ [j ~U1Tlber
~
A.T DE.
~I PUI:..... -rEP~~_Q.I~1'\.l\:_.-_~~~""~CE_ o~'=£')"_"!"_r;~~~Ty_~~-=--_ _ __
f"n,:0 BAoc, Q(QV(5T - CONv("R ~ION : rerOf:O!tW....llD R£QV£S,.· f"EEPAATE
0.3
P (T
2
CONV~R.510N
S
liliiij)
ADJUSTMENT IN.·OHMATIOI'l SHEET
:-l~O-========::::::::===;
0.11
. rg:pJ
V .... lve
(P~oovCT_cO~q:NT_~"""T'ION)
c
o
.
'f
~
• ,. I
o
,
..
o
o
,
)
,'~
I
;" •
.. .
,"
~
,
~i
• , to
•
100 I
NATURA'L GAS
...o(oe:f:f-r~
STEA"~
.-c
- 1
1
(
:j?n~
II!:
I
Y
FUel
J
~
,.
't;
i . CO' "
to
BOIlE~
FEE D
WATER
1·
W~STEHEAT
BOILER. .
,
~
.~.
:
:;
.t
-':.'
~
'!-'.,
';'<
'; •. :
~
Kp
. . .t
I'
",
I
;
:',:~.
r
t
v~
~
. '
~';",~"
•
•
,
;, 1, .. "
L,.
REGENERATIVE
T
[>-J
CO:2:R~MOVAL
·'i·:,
..; ..... .;
.,
.
~ ':'~:
'
':~
"
PUR G E
.
INTERSTAGE WATER COOLERS
0
:~
TURBINES:' :'.
~
I
~
00';' .. ;
ClI1~PRESSOR'
R1::FO ~!v\ '1: R '
.
;
'
"
!
lJlJM.,i,
"A
y
2cil::~~~sSOR
FILTER
...·P.URGE
.' .
'~'r'
'::-;:
,~'
I':':
?
, .
.:'.. :.
o~
.....
~.
:,
. . ,.,
.': 0.
:,~ I.,'~""
,:. :;"" ,"C
', H OH
•
0
t ·/.'.'\T~nA·'
G.....:
'"
.,
.
~
~
,,~
o~l
'TO'
~
.,'
."
'.
t~TDC'VIN
SEP'ARATOR
T/'\i'l I(
...
..
~
--t
'BottER FEED
1l,ETHt\NOt
,WATER HTR
CONVE~TER-
x
-#
Simplified Diagram
of a
Methanol Converter
TI{l)
'1
• CATALYST BED
No.1
TO(l)
. T I (2)
I
,~"-_ _ _'"
\~
CA TAL YST BED
No.2
. TO (2)
TI(3)
.:;.
j-'
~
~r...,---_~
g-
CATALYST BED
No.3
TO(3)
__
II (4)
V-
.,
,~..---.
·.
CATAt YST BED
No.4· .
TO (4}
I
'
~:,
.
....
( ~-i~: 4-\ l~;(Ct~
J
I' ~.-:
Figure
~
~
'"'~
.i
.
.0
rS .
, 1
o
it
FEED'
. ~.
P (0)
,~~~ -:
-)
·v PRODUCT TO HEAT E)(C~ANGf:RS
0
-~x-:
-·~·~~':"=':·--;"·7r·:;:'~,t
o
.;....
o
.... .....
",
,
o
"
....t;":,
.~.
~
,.
..
~. ~
"
.~
~
I
,
.• :. "I: .~
.........
••
I
..
, Tima
'Shared
...
:.'.; ..~.
'.
I,
I
..•..
. ·'f '
....
,
I -I
Data logging
1\ Ic r nl S t.: a n
~. i ri' i t Ch c c kin 9 I
I
~.,
. I .... ;-: 11 C g cr:n en t
1n i 0 r n'l a t lon,
Operat6i.
G u ide: ,',.
Is
Dig
1._
: . Con t r 0 I
I
I
u pc rYi"~O fY
Con t r 0 I ., '.:' '" '. .
,t,
I'
;
y
I
•
I
T
f""'t-.\-av-a-L.~-c--c-d-"""
t
f
Process
0 p fi m i z a t jon
'1
Con t r 0 I
or c c '., ;t .! c; U C S
t IV
. '1 I1
IIDirc~
ita
I
::.: ~
",.
0.rr
,I..In c
Off Line
~ n gin c c r in 9
C a IcuI a t ion 5
w
ret!
~
-
Simulation
,
J
V
,'.'
• ''0. :,
'
•
I
\
w
-
Set Poi n t :.',.Adiustmenf"
~
t'
,
'1 ~.
',",:' ,t·
.. .,
.
"
.
'{
.
f,
Fortran
:-,
,
Dcsi n
OptI mization
,
.,..,avcnccd
Si n1 uta t ion
Lan!Juc~cs
,.;.I
PI fc,c t .
.Computcr
Co'n t ro f(DCC)
Various functions;of:.p:ro~ess control compute.f.
~.
~'
~
I},
f
I
!
..... ; I . '
Vcrir-icotion
fig. 1
I I Ia
. .....
V
' Y ::: . '
•
, . 'I
l.lOWe
"
::',.:
: J'll:I ... ;'
"V"
Y
'I
I
••
'
..
,
..
~ ,
o
....
.
~:.
. ,- .
.. t
I.
•
.e-
I
I
r:
e-
I
~----1
O
n.
.
~=>';l
*Q)
.CI)
0
tn
'.r:
~~'
~
.."
I
.' .
~..
•
~""'.:. '1' ~~
::>
,..,...
.
.\.'
. .
.-
I
I
..
,...
I
••# .
'.:\ ~.~ ;: : ~ : :.- ;:,:.~ ":~~.'.:".;':'
-0
0
.
._
_
-"
r.. .. . _.
'.
."
~. -.
.... ...
......
~
C:>
CO
JLlJ
V')
. ."'. .- . ..,.,
-
'
....
#
..... UJ
Z -"'"
-«
()~
o
'-0
~e
.
Ou
C\..
V)
."
T
~
'~p
0
•
'0
r of - ~ •••:" • • ; .
."
• . ,r
8p
_....
<:. . ~ ~.~ ..:::., '- ,X .~ .....~~. .:.~~~. :-:-:. - .:--~~:~\." .'.<.. ':' .~: <. ~ ~.",::,: '.
/...
:
,
~~
' .
..
.. -,"
•
0..
".
o·
,
;"
"
.
CO!,1}-ION
:'."
.f
User~ 5
Meeting
'
San
Fr~ncisco Meeting 'Decembe~
Laboratory
11, 12, 13
Automat~on
..
by Ray Edwards J. f!, t1
The session on Laboratory Automation -will include a general description
of programming and equipment aspects .of the
applic~tion
oft.he 1809 and
1130 to this newly emerging and fast grovling field. Follovling this
discussion, an application will be discussed in detail. This will include
a description of the type of research to be accomplished, the incentives
of the on-lin-e computer, the programming system to run the instrument
in a closed loop fashion, and other experiences to date \vith the system.
o
• '0.
'}',
'.
... - ',-.--:
/
....... -,'
f
;:. J.:' .~..'"
", ,. ..... ' !.....
..
..
'
.. . ,..
~
j' •
,
.)."/
, I·
)'
,I 1"
......
I'
4'
"','
I,'
.......
r .. \, .
,
.
. i":
•
,I
o
DESIGN PHILOSOPHY
- SYSTEM OVERVIEW -.
APPLICATION MONITOR
~xecutes
Procedures·in Specified Seq.
PROCEDURES
Handles a Specific Task (e. g. Sort)
...
PROGRAMS and SUBROUTINES
(
\
All Written .in Fortran
o
Lowest ~evel Routin.es do General Purpose Functi.ons
(
o
.. -
..•.. - ...
~-.----
..........-..•
--~----...
.~~-
DESIGN PHILOSOPHY
- SYSTEM OVERVIEW -
o
CORE ORGANIZATION
Standard Compiler Allocation
Two Sections: COMMON and WORK REGIONS
DISK ORGANIZATION
Single Exte rnal File
Each Record 0 ne Sector Long
Files Use Simple'-String
o
\
~J\ethod
Via Regions
.."'""'-"""'==.'
.... =..._
, =~~':
,=~=,.=-"-',=-=-=="'."~~'"~~=""'
. '.;:-==-c.·....;.;:...:.:...c..:...."-'.;;;;.;;:..;=_...;...;.:;..;.:..;.......,;;;.:c.
.
.~~~~
-
-~-
---
-----
..
-------
._._--'.,.
.., _. "... ,..
,_._~
_,,·_··".··n....
_'~'.~.~...................
"
1130 '
o
MONITOR
XEQ MOSS
MOSS
Initial ize
Core
Common
MOSSM .
r-----~~ Application
Monitor
t<---------l
Proced re
Called
Called
Procedure
NO
xecute New
Procedure
Return to Monitor UL........
.~
..... ~ . . . .?"'UO.....-ot..~o!l
••
.... I
YES
o
o
DESIGN PHILOSOPHY
- PROGRAMiV1ING STANDARDS -
COMMON •.•
o
K = FIXED POINT
Q = REAL
VARIABLES •••
I = INDEXES
L = LENGTHS
. J = REGIONS
o
\
PROCEDURE FLOW
CALL DOWN LIST
(THRU INPUT)
MOSS
INPur
-
OVER-WRITE CALL DOWN LIST
CNVI
CNV2
-
OVER-WRITE CALL DOWN LIST
CLEAN
SORT
SORT3
- . OVER-WRITE CALL DOWN LIST
CLN2
SORT
SORT3
OVER-WRITE. CALL DOWN LIST
Etc •.
o
\
\
o
CALL EXEC U (LINKl, LIN KL, LEVEL,
GO TO
c
KREPT)~::
kkkkk~:~
LIN K1 - Po sit io n inC aII Lin k Ar roy of Fir st Prog ra m
to be Moved [(3-4n) where
LINKL -
Positi~n
n is co-ntrol Call Link]
in Call Link Array of Last -Program to
be Moved [-(1 +4M) where M is Last Linl< in Array]
LEVEL -
Recall Indicator [if Level = 0 must return to main]
KREPT - Provides Re-entry control [Used in Computed GO TO]
jjjjj -
Statement Number of First Call Linl< in ARRAY
kkkkk -
Statement Number of Statement Following Control.
Call Link
c
\
o
I
.
I
I
- READ-WRITE DISK -
DREAD (NREC, J)
Read a Disk Record (NREC) into Region (J) where
NREC
c
2048
~:;
J + Record Number
.DWRIT (NREC, J)
Write a Disk Record (NREC) into Region (J) where
NREC = 2048 J + Record Number
f.;
DREAD (KNDCT, J)
Read the First D.ictionary Record
\
o
OTHER PROCESSING FILES
KMTX - - MATRIX FILE
o
1.
Derived from SETUP
2.
Scaled for Processing Accuracy
3.
Contains:
a)
Explicit Equation Coefficients
b)
Implicit Row Coefficients
-
-
KVLST - - VARIABLE STATUS LIST FROM SETUP
o
1.
Derived from SET UP
2.
Co nta ins:
0)
Name·
b)
Type
c)
Matrix File Element Seq. Number
d)
Scale Facto r
e)
UB and LB For Each Variable
\
IfSS
Oi
,
I
~
j
j "
OTHER COMPUTATION FILES
KETA - - ETA FILE
1.
Formed by INVERT
2~
Represents Inverse of Basis Variables
3.
An ETA Matrix is added by Each Minor Iteration'
4.
On Each Major Inversion KETA is- tedtJced bqck
to Single Entry
a)
Minimize ETA File
b)
Increase Accuracy
KBETA - - BETA FILE
1.
2.
3.
'Formed by INVERT on' Last Step
Co nta ins:
a)
Post I nversion Activities of Basis
b)
Post Inversion Activities of Bounds
Values are Scaled -'
\
"
OTHER O·UTPUT FILES·
KSLTN - - SOLUTION FILE
1.
Formed by LPSOLLITION
2.
Contains:
a)
b)
c)
d)
3.
Status List
Matrix File Objective Entry.
Basis Activity
Red uced Cost
Values are Scaled
KIANL - - INTERMEDIATE ANALYSIS (LPANAL)
o
1.
Used In Conjunction with KMTX and KVLST for LPANAL
2.
Contains:
0)
b)
c)
3.
Increase - Decrease Cost
Basis
Activity Inc rease - Dec rease
Values are Scaled
KWRKl-9 - - ,ORK FILES
o
1.
Used for Transient Data
2.
Some Used by System, Some
N~t
.".,
.•.
,-,----~
,
....
-~-.~,-
- - - - - - - - _........_-_._-_.,-_•. .. ..,----_.,
, ,--,
'
o
TO RUN LPMOSS
WITH FORTRAN COMPILER AND
ROOM TO WORK IN
{ MONITOR
/1 OUP
':: Define Void Assembler
0100
(: Oef i ne Fixed Area
~:;
Storedota
CD
\
FX
LPMSS
1778
'(&H'*MW*fr'W'Mt'tW'·'I@I.ijjiM"!.jriWt/rM·'M'M~!Wi#:rI#,*HWtf6tM."'a
i'.' ". to' .. jO'
. j I!! Y"'2'M,'rWffW'N'WewWt "tH6 int"!;: 'j'imWtt;r:xi:rw"UW'1ttt yj-ji' "j ..-
$'
d't ij :'1#· .. ·'% .. "it ·"#ttdB....- ft7······
tri
S
•.
O)·ittt#triMffflttfflifb"iiwm··WlliW··· I'm.- j1'V ·..iT-li·Twz .. ·..!)· Vel! .'n...
c
. PROGRAMS STORED IN FIXED AREA
ON LPMOSS DISK
BZERO
NEWBE
CHECK
OPTIM
CLEAN
PIVOT
CLN2
PRIMA
CNVI
PRLCY
0
CNV2 -
PR L44
!
CRASH
REPO R
IN P UT
REVIS
INVE R
RNGIP
INV22
RNG 2P
INV44
SETBO
INV55
SET UP
INV66
SET2
LPSOL
SET3
MERGE
SO RT
MERG2
SORT}
/
0
MOSSM
MOVE
\
SO UL2
ST ATI
- .
1f3W' [ Ttl m,.
.. -
"cR'"
_T
I
- t
, l
Proc.e~J UI'-e
(use
FLol~J
-'.~...".
r~$A r)(.)~v~·~ -List. ~,..
r
STAT]~ I
-- -______ 1
-0
-
,
FLIP.
2
o
klfGIJ (
'R IE G-[o IJ I
----(k IZ·G.i.t\! > J
----- N t
-------
RE?(;··.[DI) L EIJG·l";';
)
f< f'e () l( D IJ U M B E R
T HIS
·1
(KTtiPY.>J)
-pf)E. . ·I
(k.THNX~3 )
N G- Yo 7-
,
.,
\
",.
'f" ("""
.....
-
"I '" . . .
u-- "k";)E.- ...... r·." ,.~
\.
.
\
_ (
\-,.oJ
.,..,-
'~'I-', "".
..I I 1 ~A ;0., ,..
,..,
!. ......-
{-, 1·, 1 '
J. ~
I...
F-=->""
" ~! I'"'- U t·~ l.r~ ,..-.,?
.. E.. I" '" 1"
.- "
'\
""'-~. IJ
r->
"'", ,.f
,
;. ~'"
....• ..
_.,' "..-
• ,
F- 1 LE. I • .1.•)
/\!~. "'-~..J ...... I'~':' E LE I·~, £IJ is. IN REC61<.1) . I
VAf? rtlt.::t.. r 6" .J(R";;'.i)li
L.....I.J p. r·····
r-o (.) r"~ 1s_t FIX c:ItJr)C;{oF pi R£/\L VAl:U,fi-::'- (h QRGN)I
(1< Fl1:2)) :J )
(kNDEL).J )
6':.,10.
P
'{~)rl';"}!· I",
(kFRPx)J)
c., ..•
D·
!. •
(l< FJ< Fl.).:r )
('r< PXLlJ) :f)
(kFLL/,J) .~)
f
l
r
c
l).seJ
6
(kTHRC)J-)
o
NCI\E 1.1£I}1-
S
}
t·
\.
LHJG-r/J
of
FIxeD
L_ El'·J G°rn
or:
'R eli 1", Vfll~r /i ;::L CS
Pel/,IT
I
(W01ZDS)
(
kJ D/:~ D:;,)
J
E L£ 1-:~~!.!._7~.~.J._. __
. -_~_lX_[;_L)_ _.
£ /;..E/~1 E NT 1- - ReA L
..... -.
..
..
.
.,
-.--,.---..,,..~-- ~----..,.-.-.---.............., -~~-.~ ._"...",....':4
_f. LE!'.:.!~',J T.~~,~"_=..EI: .~~ '!:~'::~.~_"'---.. '~______ s-~"<." .~. .,_..-~I
.. ,_"'_ _
E .LE ~~ E~JT
~'r
_ _" ....-..·_ . . , . . . . _ _ "'~_ .... -•. _ ..... _
71..
.. ~.~
-
f:\?iA L
..
·... 4 • • ~l'.....• ·...... ·u _ _ ...... .-.~"....·I".· ..:- ... ~........... •_ _ .~"'J.
""~.-"I_:~·I!·.\·
"",•.•"".......
~~
...
'r-\It<'C.···""";",,,"'.~~lO_·",:
__
...,':.;.·,,~..: ' ....: .. e~:...c·~_:.i. .• '..~-¥ol."'o;1J.~,.....• .
..... =
..".-.===
' .. -.. ,-.-~.~
..,... ,="-===._."'_.:;..... _."'._._ ...~" ...,.. _.. _.."_..
~-
~C_
; \
o
I
i
;
FLIP 3
-;
.
i
'D1CTItJl·J/,f(Y
*"PL p'
A- DDKESS
(N)
P Ro13LEI-'\
N AI/:E (0\.)
(USINq.
SU13PJ)UTINES
RC:61.0tJ):
tfl\\
V
DSAVE ./'-.,"
SAVE£
A
f,i(OF:LChi
r: N T·r."}i2
DDLrE .~-' J)£u:rc
fJ Ii 1t1 L~
NAME
PC.F at):} ,
iN
])l:c.r-'l oAJ Ai? y-
2,1'/' PC.F~
r'-/?(Ji'i,
DIe 7·1.bJ..I/j/:" y.
- DUPJ) ""('-
... ~_~".._ . . ,1'-..
C. ALL
D PL. TE
~f..\L.L
DSF\\lE"
'R E T-U r<. JJ
END
J
. J-Lrp. if
DA'rli'
o
·MAJ.-N·Tl:-JJlfl.JCE
---~----.-
---.-----..--
•..
(1) NAI,\E)
, PRDBLEM
NAME (0<.)
=..,_"'...............J.'-.'-"'.... ~ ____,-__.__~..............,_-"'-'___,.._ ... _. . "'."....~ __. .'". . . . ._.....--.__........._
. , _-=~....:........~..."'" . . _ . ._.~.~~~.......~ . .
~-
(
"":"'I'.::I;o~!t.~~C'.....",..
(
FILE-{;.ONTR.DL-t.X£LDVAR..S
,JFI)
~_~"'L'!'"
s,""'_.o:'...
....
~
.. ..,._.
--""~"'--""l'~-""'''~--~I ___ W''''''''__ ''-'_V''_~'-'-'''''''''~_'''''''''''-' _ _ '_''~~-A_~__''~'''.''''''''':''~JIiII'M.I'''''le"~'''~~''_~.'~_''_''''''';'w_.
~ ILE
, J"F L)
-
Err:LD
C.ONlI<'O L -
VAT!,
.----.-.... .- ..----.-.~.- - - . -..........-~.~ ......- .- - · .... ·...-.......... ~ ........·I·····-=-........·~. ..:."..-QI,"(JO>~',......... l,...___ .."" ..........,
o
FLIP ...S
--,...,......----
~~
kpCFf(
-...........~.~-- ~<>*>o...._r
. - - - - - - - - - - - - - - - - - -....- - . " - - -...
..
.... _
............_
j.
(~F/(/<. ~".JFI)
IfEC:OI-~ D Lo .c/;
. . . 7-rOI.,J
...... _ .
(KLSR.c.~:!.!"r21,--~.Sr__IfE~~P__~~~E~r.!> f,)
~
~~ ......,.........-
FJ.:/?s -,-
"_0 _ _ _ ..
:1.
.l-_~~~~
•
~
..;,.-., . . . __
"'"".._~~~
~ .~
____
~
_ _. _______...
~X/~C.,JF2t*NU~B_:~ OF R'!~I?O"~,_.__
{/(EXFI~.JF1)J KTER.1JA L-..!5: L6 No.
0
* N.(Jv;
Su~PctZl·.s
)1,1 u
L.t, J~
1
,) SINCE
1
-p""
(" k'"
. i ....
•
1130/PtS
(ro·
be
• Ldt~ r-.)
Le O.JJS. F:.s
ttl
'2r.J e.
aria-h Ie
PR ESENTLY OIJLY
V
.J~)
i.. (
SVpfbrt
----
-,
-----_.-
..
I
..
J
kDAT}\
SST
VA f<.S
If./l)£X X).JD Ex
(I)
DATA
,-yPE
(:r)
(£)
z
3
-
SET
INDE?<
f(£Fl:"R.S
VARs
.I
RcF£J'(S
/IJ tl (.; ):-'
to
Ttvo $£, E'J 71":.1'£ S
-----, ________
.........
to
~;:,;,""~~"""'''.:zr.~~"-::-~,.:-;':U_"",, •. ~!~ ~~••
IFF.
. C.OLfJl/l
e
It.
E 'J-r~/t~ Y ..
N
-
.
,
EN TJ~Y.
R6W
DU pL.re/\ T'£; RrrCo~D'
A.~D
LAsr USC.r..).
'~~~""1.~~--------_."...~~t_t' f"aa
~....
LI
."
~
j
AT· l~/JPur
VA«S
A
rl-JDEX
No._, Rf~GAR.DL_CSS Or:: DU fL-lcA·rc
R. Ow
AFTE)~
..
LEA'VI/JG-
ONE
._ _ (x)
"'_I!,~~~_T
1 - (R):"
ASSIG#.'}En
EIJT/{.!£S.
A
011.S
L l-£'A /-1-
U{J)
DtJ P1..1 (! ATE:S . AR.,;
EIJTt?'Y
FoR.
Eli£'.. H
-.---,...-.------.-.....
~-'-""----.
.Is
---~.--.
.
FO~':M"i
--. .
X-II
S£C~u[JJC£~
D61-!£;'
By
Rf.t~1!JV E~D
~-ol'?Dl:J.jAr£.
-0
... ~.---.-----~-. -~.~•.-.....,- ......,-.-,""-.-'-
-"~-.-'--"-'.,......-~--,--
fOr Rn'lL No, :. (tJA Z) =-
A
ForUIl fir
J/~¥
,MosS',
_ _ _ _- 1
At-,oJf;i8ET'
GET
OF1=
D ,r'C. TI:c. JJ F'f P.y
RCC_c.R.D
R..{; cD. . Ric..cRDS
1;(A':~ of' U3t) S~i) ._. . .
*"*
k VIlKS
PRoBe
1
'.'_~~
_ _ /' .....-...~~-....,_.~,, __ ....... _Q_~
K PliTt.. K(){\SS
,), '"
; "'. C>f:;' ~
PROei.
Pt.F
_... _ , - - 1
". )'i
"'\/,,(:.(';"
, ". k r ~ T hA
,,)
. . ".
A"i'-,Ot)~ , p' VI",t,
',\;!
2.
1
1.
:..:.,..;)"..:r.~-'_,
J
I
PI<.cat'·: r~'
.
C,i
@fc.r:
®
Nci-r.:r ..r Fil,·~ (',,~ i~ ;'" fl.' hld.:·/'
)
...
.
IFI =
L'l~j' 1·~
® -r(e.$10~ }'J ;)"
® F-'f.(?--' 1. . :)~·-·
LFL
Go -To 25'
l
...;
'1. D Pc F~
/
100
-0·
N DRIJJ ~""'.''''
t I.J 1):1 c.. Ii ·rt:~~S! R£ i\ 1)
r~ f... t~J)) I; I>
:~_·r:'·~i (+ 1.)
REA D eA::..kL.Jlq~.r) L.lJlleYi (.··--t)
tl}
..------"
..
"
ESSO R!SEARCH LABORATORIES
HUMBLE OIL &REFININ~ COMPANY. BATON. ROUGe RE:FiNERY
BATON ROUGE, LOUISIANA
O·IWXNOUM'ON
-M:>dification of IBM 1800 TSX to Support Six Disk Drives
December
i
4, 1967
av
E. H. Spencer
"LK NO.
1407~1
INTRODUCTION
This is a report on work at the Esso Research Laboratories of Baton Rouge
which added three more disks to the TSX system. The additional drives are treated
in an identical fashion as the existing three drives. The new system has been tested
for some of the options of TSX and all known bugs have been corrected. It is easy
to convert an existing TSX Version 3 system into a six drive system. All that is
required is a TASK assembly, the STOREMD of·four subroutines, a partial system load"
a define operation, and a skeleton rebuild. Although the method of disttibution
has not, bee'n decided, these changes are to be made available to other 1800 users.
DISCUSSION
,~
The Esso Research Laboratories are located only two blocks from the main
office building of the Humble Oil Refinery at Baton Rouge. Work at the laboratories
is mostly process development. A small computing center is maintained in the laboratory.
Through the years this has changed from a C.P.C.,to a 650, to a 1620, and will be an
1800 when program conversion is completed. Large computing jobs are sent to the
refioe;ry 360 system. A study of work done by the labs computer showedi t to be data
acquisition and batch processing of small jobs. This finding led to our 1800 order.
Logic of handling the acquired data indicated that some form of indexing
and a random access to the files are desirable. Hence the choice of disks for mass
storage. HOwever, data volume is such that three 2310 disks are insufficient. Our
system was, ordered with five drives. A similar situation exists in the refinery
where an 1800 computer is being used for direct digital control of a blending operation. Here the system was ordered with six disk drives. Under the press of two
urgent applications, I began the job of modifying TSX to support the three additional
disk drives. I was assisted by 'our I.B.M. representative, Mr. J. A. Albritton (S.E.).
I.B.M. was very cooperative and through Mr. Rob Martin of the Houston -.
DACS Center we were able to get copies of 1401 microfiche tapes of the TSX system.
The work was started in May and completed in September. The modified system
has been in use since October. There are no known bugs remaining in the six drive
TSX. Although all of the options of TSX have not been tested such things as skeleton
puild, core load bUild, compiling of Fortran and assembler langu~ge programs, storing
and deleting of core loads and relocatable programs, and the execution of process and
,non-process programs have been done.
••
Humble has consented to make these results available to·other 1800 TSX usersonce the me-thod of distribution has been selected. The Esso Labs would prefer not
to act as agent for handling the distribution •
,.....~~~.....
---
..
_--------,-----
-2· We started from TSX 3 Mod 1 with corrections through PTF 27. Appendix B
shows . the source level changes that we made to TSX. It should be noted that card
numbers correspond to the microfiche tape for Mod O. This is bec,ause only M:>d 0
was available when we started. ,As MOd 1 and PTF's through 27 became available we
added. them to our MOdO source decks.
o
Our work was aided greatly by the fact that in some areas of the TSt programs
provisions were made for future expansion to six drives. In particular, the M:>nitor
and Disk utility Programs (DUP) record and test the availability of drives by the
use of three bits of a word. in DeOM. Since a word hasl6 bits, addi tiona! bits were
available for three more drives. Even ~re important was the embedding of unused
words in DeOM whenever information referring to a particular drive was stored.
Without these provisions it would have been necess€k~ to modify and enlarge DeOM.
DeOM is the heart ofTSX and is referenced absolutely and repeatedly throughout the
routines. Furthermore several system programs end only a few words from the start
of DeOM and enlargement would require IIDving them to lower core. Such IIDves would
be difficult to organize in sections of EDP and DUP where an extensive overlay
structure has been built. In fact, without the unused wOrds of DeOM, it would not
have been practical to change TSX for six drives and because of this it is not
feasible to consider support of more than six drives.
It is difficult to convey the size of a job of this kind. It took over five
months and involved twenty-seven TSX routines. Fourteen pages of source changes are
listed in Appendix B. TSX logic in handling the· disks was suffiCiently general that
only in a.few cases was the logic changed and often, even then, just for a gain in
~
efficiency. Mbst of the changes or additions were minor. But the problem of locating
them was not necessarily small. It sometimes' invol ved tracing back through several
routines and to do this required a great deal of detail knowledge of the TSX programs.
Because no courses on TSX internals are offered by I. B.M., we were forced to learn
the system by reading programs.
Instructionscoutliriing the' procedure to be used in converting an existing
three drive TSX system into a six drive system are given in Appendix A. It is briefly
to (1) insert the change cards in source TASK and reassemble, (2) STOREMD the four,
Fortran I/¢subroutines, (3) partial system load the changed routines redoing the
aSSignments and the DEDIT, (4) define NDISK and CONFG, and (5) rebuild skeleton. The
define and rebuild skeleton must be done with six drive TASK in core. The de~ine
CONFG and skeleton build are required because a new cold start and an error decision
. program, have been stored. Word count and sector address of skeleton are stored to· cold
start by define and to EDP by skeleton builder.
CONCLUSION
Other potential users of our six drive modifications face two problems.
They are the problems of distribution and maintenance of an IBM unsupported system •
. Corrections to 'ISX have been numerous and many are in the twenty-seven routines .which
we changed. If updated source programs are available , it is usually easy to make the
.corrections and recompile. However, most 1800 users do not have source cards, which
are available only on magnetic tape.
EHS/gth
APPENDIX A
c
PROCEDURE FOR UPDATING
TO
TO SUPPORT SIX DRIVES
o
..
"' · 4
q.,
·47,
7. ..,
',-,. ,\
___ p _ ..
_
_ _¥
..+41\,
._
--~
. . . ."' "-'. .
~--..-.-,.----,
..... ...
~
.. .
,~
~ :",:,.:o..~------
"."11
. . . . . --.. . . . .---"~,-,, -,~---
ESSO RESEARCH LABORATORIES
HUMBLE OIL & REFINING.COMPANY • BATON ROUGE REFiNERY
BATON ROUGE, LOUISIANA
Six Drive Changes to IBM.1800 TSX-Phase 2, 1800-0s-001,
Version 3, ~dification Levell, through PTF 27 .
DATil
December
4,
BY
1967
C
( j
FILE NQ.
E. H. Spencer, J. A. Albritton (IBM)
I. General Comments
A.
These decks represent user modifications to TSX to support six disk
drives. Three drives ar.e standard and three are RPQ devices.
B. Area Codes of 20, 24, 25 and lAC codes of 20, 24, 25 are requiredo
C.
The cards consist of source change cards for TASK and obj ect decks for
subroutines and for system routines. Control cards are included.
D.
After the partial system load, user TASK is required.
now incompatible with Sys G~n Tasko
E.
ASSignments must be made for the 3 additional disk drives. A DEDIT
is required to clear FLET. A DEFINE CONFG and a skeleton rebuild are
required.
F.
Thre~
The system is
additional fields are used on the cold start card (24, 26, 28)
for dive identific ation and on the JOB (21-25, 26- 30, 31- 35) for label
identification.
II. Procedure
A.
Complete the additional TASK EQU's" Insert the change cards replacing
cards with equal numbers and including others in the source TASK cards.
Reassemble TASK.
B.
Copy your system on another pack.
C.
Do the DUP operation required to STOREMD for the four subroutines.
D.
Do the partial system load. Redo the ASSIGNMENTS and DEDIT to initialize
FLET. If the three additional disks have not been assigned, be sure to
assign them. (From here on user I s TASK from Step A is required).
E.
Load user's TASK from cards o D() a DEFINE CONFGo (Ignore the error message
which says TASK in core :is not the same as TASK on disk).
F.
Rebuild skeleton using the new SKA program and the Levell SKB program.
Henc~forth
use the copy.
EHS,JAA/gth
4?{)
.~~--~-.----
._------------
MATERIAL LIST FOR
SIX DRIVE CHANGES TO I.B.M.
1800 TSX-PHASE 2,1800-0S-001
VERSION 3, lvDDIFICATION LEVEL 1, THROUGH PI'F 27
USER MDDIFIED lO SUPPORT SIX DISK DRIVES
(Esso Research Laboratories, Humble Oil Company, Baton Rouge, La.
and I .B.M., Baton Rouge ,La. )
I.
II~
WRITE UP OF PROCEDURE
CARDS
Item No.
Description
Quantity
1
Task (Source change cards to be
inserted in source task)
339
2
Supervisor
80
83
Core Load Builder
Cold Start
51
Disk Utilities (Total - 142 Cards)
(1) Control
18
(2) LetjFlet Dump
17
(3) Store Control Card
43
25
(4) Delete
(5) Dump-Disk to NPWS
19
(6) Write address
9
11
(7) Search Program Name Table
Error Programs (Total - 85 Cards)
(1) Control Program
38
(2) Disk Error Program
11
(3) Error Table
4
(4) C.E.Routine A
14
(5) C.E.Routine B
18
Skeleton Builder
60
Task Utilities
{1) Task Card to Disk
10
(2) Task Disk to Card
12
(3) Task Disk to Patch
7
(4) Task Disk Duplication
6
(5) Task Disk Load for Off-Line
System
13
(6) Task Write Address
33
Subroutines (Total - 31 Cards)
(1) F-I/O Busy Test
5
(2) Disk F-IjO Busy Test
4
(3) Disk F-IjO
17
(4) Fortran Find Routine
5
3
4
5
6
7
8
9
0
1
'/
Deck Ident.
(Col. 73-75)
TSK
NPS
CLB
CLD
DCT
LFD
SCT
DEL
DPl
DWR
SRP
EDP
DSK
EUD
CEA
CEB
SKA
TRL
TDD
TDP
TUP
TDL
TWA
BTl
BT2
MDI
MDN
~7/
o
APPENDIX B
SIX DRIVE SOURCE
LANGUAGE CHANGES
TO TSX
()
.1
APPENDIX B
Page 1 0 f 14
TSX . IDDIFICATIONS. TO SUPFORT _______ .-..:.~._ _ _._____.: ________.
SIX DISK DRIVES·
SOJJRCE:_LANGUAGE ... _CHANGE-..GARDS:--_ _ _ _ _ _ _ _ _ _ _ __
~~.._ _-OORG3.-EQU-__. '-.------0:-.. 1F--THREE-~D.1SK-DR..-~..ELSE-l--___TSK00081__·
DORG4 EQU
_ _---......0....--0 ORG5-.E.Q u.--
O·IF FOUR
DI~K
DR. ELSE 1
TSK00082
--__O._lE._F_IVE...:.......DlSK-_DR .•. _El.SE. __ .L______J-SKOO'083
*
*
*
*
PRIL3 EQU
DISK DR 3 INTER. LEVEL
0/23
TSK00084
_ _ _ ...eRl.L4_.. EQU.__~ __.....:.__ D.lSK._DR._4_1NT.ER. •. -LEVEL.-.-*_0 L2.J---.iL..rSKO 0.0..85..-,-________
PRIL5 EQU
DISK DR 5 INTER. LEVEL
0/23
TSK00086
DC___. __ 2 O_.:... __ .._________ AQQJ_T..tQNS __ TQ __ JA.C__T.ABLE_.____TSKO .to 7 t
DC
DC
IN3
24
.----.---·-·--------6c-----IN4----··-·--··-----·-----.---.-----____
TSK01672
TSK01673
.T 51<:-016-74--
_-.:.DJ:;. ___ ...._. _. ~.5. ___ ..________.______.____._
________t.S.KCU . 9.J...~
DC
IN5
TSK01676
_.___.____.....I.~.D.D____O_C___ .__ ._TAD.D.s&:.6 ___ ... _.. .D1<___DHV._T.AS __ AD.D_Ll\BLE.__ L8.1._IS_I$.O.3.79Q_ _ _ _.____ ..._
TADDS&l
ADD OF LOGICAL DRV TAB 188 TSK03800
LOGDR DC
~_..____._. _____ ..Pj:iYQR_.DC.__.____T..ADDS&.7...__'__._ .. ___~DO._.OF ... _P):tYSJ.<:AL __ DR_I.&6_.1_8_9___TSK0381.0 ._ _ __
DC
TADDS&12
END OF PHYSICAL DR TAB 190 TSK03820
_ _ _ _ _ _ _ ORSUP_DC__.___._5_ .. __
CQD.E. __OE.__ LAS.LDI ..sK_DR.V._1...2.Lt..SKO.383CL______
DC
-6
MINUS MAX NO OF DRIVES 192 TSK03840
'O.<:.. _____T ~ DD S_._ .... __ , . F WA__OE_.DEV.I...C.E __TA B~ 1.____ ._J_23_.I..SKO 3.850 ________ _
TADDS DC I
*&5
OK DRV TAB ADD TABLE
TSK03901
DC_--.:.___TA ..~_._.. ______ ._LOG.tC.AL_.__ DRJVE_O___ ..___·__._. __TSKO.39.Q2..________
---------DC
TAl*DORGl
TSK03903
_D.C ________.IA2*D.ORG1*:DORG2..-___
ISKO_39_0.4-________ _
DC
TA3*DORG1*DORG2*DORG3·
TSK03905
____..DC _____._____T..AA"*:DORGI ~DOR.G2_~DOBG3}~:DORG.4_ _ .
T_SK03.9..0_6_____ .___.____.
DC
TA5*DORGl*DORG2*DORG3*DORG4*DORG5
TSK03907
____D_C _ _ _ _TA ___ . ____ .. _E..t:tY.5_LCA L~.RL VE __O__ _
--.ISK 0.3.90.8_. ________.__
DC
TAl*DORGI
TSK03909
_. . . ,___.____._.. __.. _. ____ ... DC ___ ...... _..... _. T.A2*DORGl *IJORG2. _.' ..... __ ..... _____... _.. _.... _._._ ..._ .... _....... TSK0390A.
DC.
TA3*DORGI*DORG2*OORG3
TSK0390B
___
. D.c__ .___TlL4_~D.QR.GJ_~DOR.G?_~_09R.~.~.~D..9RG~___
TS.K.o.~2_0_C _ __
DC
TA5*DORGI*OORG2*DORG3*DORG4*DORG5
TSK0390D
_ _ _---'C DJ~Se..__ EQU______.DB .sVe.~..C ON____ .D. ISP.~ AC~1.1ENJ.. __ .E.RQr1. __COJ~L_ _ ._LS_I<.O.4 2_4_.1 ____.. __ ~___ .____
TA33 BSS E 0
TSK07390
FF~_._A_ .E.Q.u._. _____TA~.3~ J.A 2 _
.. ________ ._____~ __.____________t.s.K Q.7J!_QQ_____.._______
ORG
FFA*DORG2*DORGl&TA2
TSK07410
_..___.._____~_L~S.EJ~.I___.IHE_ ..E.QL-L.O'WJ N.C;.__ 222..__C_~8D.S.._ ~F TE;B_T.SKO..I~LLQ____-_ ....
ERLOOOOl
___________-LIABLE.._EOR.._DR_IVE__ ..3____________
_~ _______ .__ ._EBLO.O,QO.2_________.
*
ERL00003
_ _ _ _ _ _T_A:1. __ D.C
0___.__
.L ~..s.J__ I .O.C_C._.....%AREAo_______O.__ ERLO_O 0.0.4-____· ___ _
DC
/AOOO
%CONTROLC
1 ERL00005
D.C_ _.._O____.___.______NE;.>-CJ__.l.Q~_C__ .._% ~..REAo __.. ______._LEBL.O_QQQp
DC
/AOOO
~CONTROLC
3 ERL00007
--..- -____ .;.....__.__...DC_. __._..._. ___.. O_. ___.__ ._.___.- ._..... _L AS T.._WD.C T. _________________4.__ .ERLOO 008 . -____ .---.-._--."
DC
0
LAST DK ADD
5 ERL00009
_ _ _._ _ _ _ _ DC_.____ O_._. _____ NEXT_'tIJD __.C..T_ ...____....___---6-.ERLO.OQ.l.O
DC.O
NEXT OK ADD
7 ERL00011
DC__ .._.__ .___ 0 ___ .____·__..... ___._SEEK. __I.O CC_._% ARE.A C. __.______ 8_ERLOO'O l~_ .. _._.___ _
DC
/A400
%CONTROLC
9 ERL00013
.-.D.C_:.....__ ._~_O_... _...___.__ ._______SENS.L.I-O_C.C__ _".. _&.-_SEEK.._TO__. __1.Q._._ERLOO_Q_IA__.___ . ________ ._
DC
/A700
%CONTROLO
II ERLOOOl5
,______ ..~J)c._
J..A3. &:3.6.__ ._ _CtiE C:iS_-L~B.L E ___.LQ.C C__/~A REAOJ LERLO Q QJp________-.
DC
/A600
%CONTROLC
13 ERL00017
DC___.___ - 0. ___._.____._. _______T.P.T..AL. __WQ.,.C_T .LEF..I._____..._L4 __.ERL.O 00.1 S_ ..._____ ._.____ .
DC
0
.
FIRST READ/WRITE'
15ERL00019
-C'I'--
*
o
.
..
-.--lIt::-t4--=o::-::::7--=
...1::-----
----------
-------j.fFPENDlx-B--··--
--.-.-----.--- --.. -.--..----..
..
..
TSX MODIFICATIONS TO SUProRT
SIX DISK DRIVES
• So.URCE ___LANGUAGK_.CHANGK._CARDS._
SAVE l'
-~---.-.--.----~ -------.-~
DC
0
DC
0
Page 2 of
_.-
14
..
-----~---- ----.-------------.--------~-
16 ERL00020
.. _~_~__ O..c._____ . _____ ._______Q..____________ .__________________.. ___.____ ._.____.________ .__ ._____...:_.l7__I;HbQ..Q9_gJ__~ _____._
'-'
ERL00022
_________-----...D~ _____O_____ ._____._____.___ . .____.____.__.. _..._..______.._________1._~___ ERLOO'02__3
DC
-2
VARIABLE &-2
20 ERL00024
._ _ _. _.___D.C __________T A3.&1..6____._._SAVE__ ADD.... _____
__.2.1 __ ERLO Q 025 _ .__________ .._
DC
TA3&17
SAVE ADD&l
22 ERL00026
.
SAVE 2
: 0 \
18
__________ OC ___ .__O_______ ..________ .__ ._._._NEXT __ ABE A. __ A00__& __ l _____2.,3_.E~LO 0.0 27_________· _. ____ _
DC
~_.,-
-8
ARM POSITION
24 ERL00028
__ . ___________ .Q_C ______________ O.._______.____._--__ERROR ___ CI _____. _____________2.5____EHLO.OO 29
DC
0
CALL tNT LEVEL
26 ERL00030
______________ .D_C__ ._____.____7..5.Q _____.___ .___ NQNf.?80CE_~s.. __ W_QRJ< ___STO'RAGE __ 22___ ~.RLQQ'O_3.L__________________ .
*
INDICATOR TABLE
NEG-ON
EVEN-OFF
ERL00032
_._ _.___________ .D_C_______._.___ ._O__ .__.____.______ ...______ I .. l.. ______ ..___________ 1.2_________.__ Za_ERL_O_O_Q_3_~ _______________ _
DC
0
SEEK
WRITE IMM 29 ERL00034
_________________D~. ________________Q_______.____.________ C.tL ..TA~..__________ ._~tL_TAS.__8.J;I.____ ~_Q __~BL_QO.Q.~~_____________ _
DC
0
SEEK FUN
READ
31 ERL00036
______.______ -DC__ . ______ O_______.______W/..RS_C __________._ ...._.w_t..8B_C~_.IN~_.9J?. _~.?__ .~BL_O_QQ.;r!.. _ _ _ _~. __..
DC
0
GENERAL INDICATOR
33 ERL00038
_____________ D_C ___ ~--. _.___ .__0._. _____._. ___________ FJ_ LE__ PRO_!. ___ E.I Lf;: __ P_RQ ___SvL ___ .;3.4 ... f::BL 0 0_03_9__________________________ _
DC
0
IND OP
FIN OLD OP 35 ERL00040
.Q~_. __. ___ J _____ ~____ ._______ .____ l/.O_ ....A8.~./\._c9B __ ~lj~_~_t<; __ I.8§t..._~_29_____ r;:8l.,_O_Q_Q~_t ________.__
DC
0
37 ERL00042
_____.___. __.__._.__ D~____________ ._0__ _ _.'._. __._____._. .E.I.8.~.J.._.I __Q.C.~. ___ .... ___._. ________.__.____ ._.~.t3... _.f_.8LQO.Q~~___ ._._ _.___.__
DC
0
39 ERL00044
tf
DC
0
.
FIRST
WD
COUNT
40
ERL00045
I --";
---.--.---.------.- -----------...-------- -------------.-.. --- - . ---- . - ....- ...--.... --. ---------.--.-.--.-------..----------------......
-- .-----....-..-.-.-.---------'V'
DC
0
-
FIRST DK AD
41 ERL00046
.____ ._______._________.___._QC ____ .. ___ . ____ O. __ .. ___.. ___________fJB__~J._I_QIA_L ___ ~{R__C_t_.L_~EJ. ___~~____E;.8.~00_Q4_1. __ .___ .__________
DC
/A701
SENSE/RESET
43 ERL00048
__.__.__. ____._._____ ... _. __.D.C._ _ _._. ___ ...__ .. __ 'O. ._. __._ ....__ .... _. __..__8EL.Al)_\JE ___ARr·.1....PQ~J.I..I..Q.t'~L .... _._4_4. __ .r::8LO_QO..~_9_._. __.__.~ __..___ .__ .
DC'
0
FILE SWITCH
45 ERL00050
_.____________DC_________._0.. _.... _ .______ .____ .S.K I.P... ARf'.1 _ pO $_ I..T_ L9t-J ___C_ttEJ-::_ts:_____ 4.9_~.8_LQQQ_~_1____________ .________ _
DC
1600
NONPROC WK ST END ADD&l
47 ERL00052
...~ _______ ._________ ._______ .... D.C. ____________O' ____ . _____________ NQT.. LJSEQ____Y.ET_. _________________.. 4e ___ E_RLO_OO_5;3_________________ .
DC
PRIL3
DISK INT LEVEL
49 ERL00054
.__"._.. __ ~___. _____ ._ ....__ O~_._._.. ____ .__ .O ___ .__.___.___ ._._____ .~.. t. O. .S E.I__J.N QJc;..A.. TOB... _______________ ... __~ 0___ ~.8~ 0 .QQ_??_ .________________ "_
DC
0
CALL ADDRESS
51 ERL00056
_____________ ._______ ._P_~. ____________ ~ ~.1. _. ____._______._T AB~E_.R E ~p_____ ! 9 ~~_. ___________... ___ ~.~ __ E8L 0.0 0_:=>.7 _________________________ _
DC
/A600
53 ERL0005S*OISK DATA TABLE--SECTOR 0
ERL00059
~
..---.,,-------------- -------·---OC·-----···---·--·i i .-.---- -.---.-- -- ---'r ABLE 'JJ-ORO-Co-UNT--·---------------S4---E-RLO-O-060--------------------.,..-"
.___ .P.~_____._.O_. __________ ._._~s:_~_l9B__ ~QQ8.~_~~ ________.._.55._ggl:-_QQ.Q.~J_
DC
0
PACK LABLE
56 ERL00062
.____ .___.._________D.C..________ .______.Q____________________ Et8ST .. SEC __ P80 __.WK: __ STO' ______51 ____ ERLO_OO.Q::3_ .__.________ . __._
DC
0
LAST&l SEC PRO WK STO
5S'ERL00064
____.____.:.____________________ DC ___. _____ 0 ____ . ______________ .MOD I F I CAT,! ON _LEVEL ___ ._________ ?9___ERLOO'Ot)_~___________________.. ___ _
DC
0
Q.~. ______________ Q_"_.___ ... ____... ___
FIRST BAD CYL AOD
DC
0
THIRD BAD CYL ADDR
-------------g~
~
.~~~. ~~~~ ~~~
_._._______________________ .DC______ . .______ 0 .. ______ _
TA44 BSS E .0
60 ERL00066
.__ ._SE_~O'~~P. ___~_A,D____<:y.l.. __ AOPR_____§ __L_E..8LO.O'O_~7_.___ .____._-
62 ERL00068
.
-.-. . . -. - -----~~ -~~~-6~6j6-----------C
NOT USED __ YET____
_ _ .. _____..... 6_5.___ .ERLOO'0,7J .. ______________________ .
ERL00072
_ _ _.__F A~_.___E..9U.________ :..T.A44 ::or:.A..~. ___.__ ._... _____._____._ .._..___ .____ .__._ .__.______.____E.8~_O' 9_Q.1_~__ . ________"_. ._ \
ORG
FAC*DORG3*DORG2*DORGl&TA3
ERL00074
-'-'-.----~--:--:-; ----TAf3"LE'-FOR DR I V·f~·····4·-·---~-----· . -----·-------·--------------.. -- .. ----··-·--~----------------------·- . -----~:~~66~~----·----·-~-#--7-¥-
-w'pr"ug"I"",uWr"·"
#lJ~O"'«l
"f
.APPEND}~'. P,
Page 3 of 14
-_·--.--fSX. M) DIITCATle::l!..~ ~~:';._ JUFroRT____ ._.__ .c _____ .._._.___• _ _ _. __. __. ___-,-_ _. _ _ .. --~__.
TSX DISK DR J"'.; fS
.*
.
__._.~O URCE .. LANGUAGE~£r~XiGE-.C.ARDS
'.'
ERL00077. .
.
~-A.4--DC-.-.---...O----~..-------LAS T.--10CC-%AREAIl--------O·--ERLO 0078.-------__
DC/COOa
.
~CONTROLO
1 ERL00079
_ _ _ _ _ _ _ _ _ DC. ______ ._0._.___________NEXT_l0C.C __..%AREAo.. --.---.-2--ERL00080-----_._
DC
/COOO
DC'
0
%CONTROLO
3 ERL00081
--D.C ________.Q. _________LAS T_~WD_..CT ______._._____
. __A __.. ERL 0 0 0.8 2......_ _ _ __
LAST OK ADD
5 ERL00083
c.r ..... _. _. __._._.___.___ .6 .__ ERL 0 O_08(+ __...________
-.D.C.___._.___ O. _____.. _______.__ ..NE XT.. _WD. __
DC
0
NExT OK ADD
.~._
7ERL00085
___.____
. ____ .__ J2_~ _______9_. _____.__._ .._. ____.__~_~.~.~_._J.O.f_~ __ ~_~ 8~~. ~'. ___ ..... ____.____~__ ~gb_Q_Q.o f?_e?_ . _---' _____._.____ _
DC
/C400
%CONTROLO
9 ERL00087
_ _ _._ _._______ D.C_. ____ .... Q . . _ .. _____ ._. _____ .. ___ .SENS_~ __ lQ(. ~__ ._C>:._._SEEK .._tO_ .. _ .._.lO __E.RLOO_O§.8..__._._. ______._._
DC
/C700
%CONTROLO
11 ERL00089
_ _ _ _ _ _._____.. DC. ____ ..._.. _ TA 4 &36 ___ ._.___...CH E CK__ TABL E __ I OC C_._%ARE A 0 ____12 __ E RL 0.0.0.9 0. ....____.______ .__ ._
DC
/C600.
%CONTROLO
13 ERLOOQ91
..~_. ______ .... __
. -DC-.--.----.-O.- -.--.--.. -------TOTAL--WD-C T. LEF-T_._.___ .~_. ___.1.4 .. ERL000.92-----.------.-_. __
DC
0.
FIRST REAO/i.IJRITE .
15 ERLOOo.93
..-·.-.--DC......---- -·O.·--··. -.-.-·-~- ._-s A·V·E--l-.--.-~..-----------.-.---..l.6. .. -ERL 0. Co. 94.________
DC
0.
17 ERL00095 .
- - - - -Oc..--.--.--..-O ........ -...-"--'--"-' - SAVE .. 2.-.--.. --... --..---- . ----.. - ..-- ____ 18 .. _ERLOO0.96. __._____ ._. __
DC!
0.
19 ERL00097
_________.. _._. ___. _DC .. _....._... __ ~ 2~ .. _.. _.' _._._ -___ VA R I ABLE. &.-:-2 _._.______ ._. __________ ._2.0_ ...ERL 0. 0.0.98 _____ ._. __ .___ _
DC
TA4&16
SAVE ADD
21 ERLOO0.99
- - _ . ____________.___ DC ..___ ._ ...___ ._... TA4fs17... __... ___ -SAVE...ADD&l_._. ____ ._._____.___ .22.._ERLOo.IOO_ _ _ _.____ .
.
-0-.---------.
•••••_.
_0_'
_
•••••
NEXT AREA ADD & 1
ERLOOI0l
-~~-.---.--- ~8.-.--.-- .. _. ____. __ ~::~:O~~TI ON_. -- --.. -----.--------~:-~:~~ ~.~ ~~----------DC
0.
.-
23
--.-.----OC. _____ .___ ._.0._ .. _____ .__ .... ____.CALL . J NT. .. LEVEL .. _______.. ____..__ 26... ERL.OQ_1.04__________._
DC
750.
NONPRo.CESS WORK STORAGE
27 ERLOOI05
1 NDI CATOR TABLE
NEG-ON
EVEN-OFF
ERLOO 1 06 .
"*
DC
0.
II
28 ERLOo.I07
12
_DC_____._. ___ O._._. ___ ._________ .SEEK_ .. __ ...____ WRJ.TE_.J~/1M_._.29_ .._ER.LQ'O.108_ _
' ______
DC
0.
CH TAB
CH TAB RET 30. ERLOOI09
_ _ _.________.___ -DC_. ___ .__ _____0 ___ ._. __.____ ___ .__SEEK ... EUN _____ ._. __ ._READ.. __ ._. __ ._._.3.1 ___ ERLo..o. 1..1. 0. ____________ .._
DC
0.
'W/RBC
W/RBC
IN OP 32 ERLo.Oll1
__________ D.C. ___.______..O_._. __.. ___ . _ .. ___.. _. ___ GENERJ..\L._.J_~.DICA_TO.R._._.__3_3.._E:RL..OO'.l.1.2 ...._____________ -.
DC
0
FILE PROT
FILE PRO. SW
34 ERL00113
DC. ___.. ___ .. _._O.. ____ .___ .__ .___ __.1 ND_. _o.E _____.. ___._.J:. I N_.OLD_.QP__ .35__ .E8L.OQ_.tl~. __ .____._____.__ .
DC
1
I/O AREA FOR CHECK TABLE 36 ERLOOl15
__.QC ______._O ... _________--____ .______ ._._._.___ .__ .____ ~ __ ._. ______________.3 L ... .ERL.O OJ .1.6_ _ _._____ ....
. _____ • _ _ _ _ _ _ _ _ .'!.... _ _ _
DC
0
FIRST IOeC
38 ERLOOl17
__.•__________ D_C_. _____..__O. ___ . ___ .._._. __ ._. __... _.......__ .._._-____________39_E.RLO'O . Ll_8
DC
0
FIRST WD COUNT
40 ERL00119
_. ______________...DC____... __._.____O........._________ ._F..IRS_T__D.K_.. AQ ____ .....___.._ .. _A.t_ERL.OQ12_Q _ _.__- . - -.. . -.
DC
0
FIRST TOTAL WD CT LEFT
42 ERL60121
_______.___ .... __.____ . _...... _._.... oC .. __ .... __ ._._/C.70 1._ ...__ .....SENSE/RE SET
.. ___ .____ ... __ .4_;3. ERLO_O 122. ________ ._._____..._-.
DC
0
RELATIVE ARM POSITION
44 ERLo.o.123
_ _ _ _ _ _ _..DC__._.__..___ ._O_. . __.__.____.___ E.I LE . . S W.LICH. __.._. ___._._._________ ~__ ER.Lb O_1.2~____
. ___-.-...
DC
0
SKIP ARM POSITIo.N CHECK 46 ERL00125
_______._______ DC .. _... _. __ .___ ~___ 1.60Q___ ~__ ._ .. _._.NONPROC. ... '.AJK. __ ..ST .. ENP __ ADD&L .... _.~]_._.ERL'OQ12t5 ___--.-.-.-.----.--..
__0._______._._._ ._._.g.~__.____.__.__~_RJ L ~ __ ..__._." __ D7.~~ . . .~~ ~ ~L ~ ~~ ~_
~ -.~~~g~.~. ~ ;---.- --.-.----.----~. . .
... --. ____ .__ -----.-----: ..
DC
0.
$10 SET INDICA~OR
50 ERLo.0129
_OC._..__..._ ... _......O..___._.._. __ ._~__CA.LL_AD.DRESS..____._______ .5.l.__ ERLO.QJ3'O___
DC
*&1
TABLE READ lOCC
52 ERL00131
___ ..___ .. _Q.C ......___. ___ /._c:60.Q__..~ ... _...__._. __... __._ ... _. _______ . __. ___ ._. _____ ..._____.____..___ ;;~.._.E.RL..OQJ.32:. __._____
*OISK DATA TABLE--SECTOR 0
.
ERL00133
#-y.
¥, , ,
",un :';,4"
;;"; i", "," ;.,. ·4,D,Z;X..""',""Qeu::::wt,mW,m,===il&&lC;;:MIIGi..UM&lGAlnii=hIim&(fWSillilu.,IIWtDffiliilQWliiiiiiiiWiiUJliiiiiii!iliiii
APPENDIX---13---·----~-----------_;ag;4-~~-;4
L _ '.---.-.--.-----.---------... -.- -...- .. --.-,_.~ ___ ."'_..___,... __..________ ..:__ ...... _.
TSXMODIFICATIONS . 10 .SUProRT--------...-:------...-:-----.-·-------. SIX DISK DRIVES
---'-.-"-SOURCE.LANGUAGE ...-CHANGE--CARDS.------,-o---.,-------_
.DC
~
ir
TABLE WORD
___·_____-:-____. __._____D.c: ___________. . __9. __.__ . _______ .__.____$f:CI9.8___
DC
0
cOuNt
o
54 ERL00134
___.____________~? ___~8l__ 9_QJ_~?
~_QP.RJ=;§$
'PACK LABLE
56 ERL00136
_:D_C--.---Q.-.---------EJH-S..l.--S~C.-ERQ... _WK._._STQ.______ 5_LEE8G______.__.. *_-4&4*DORG1.~P9L~G2~~QQB.§_;3.~'OQ.8.G~___________ T~~53JJ_.A __________________ _
DC
/4000
DISK 5
TSK53118
,___,_____.______________.Q.<:;._____._____.__ /'~f_ 0 Q__._._.. _.________ s.~_~ ?~._ J..Q 9.C __~___g 8J.Y_t;._~ _________1. ~I$ 5 3 Jl.~._. ____ ._________.
DC
Z1 Z&8
TSK5311 D
...•....._"._... _____________ .___ ...DC._____ ...,_____ .__ .2_5 __ .._...... _._ ... ___ . __ . ____________ .__ ._. __ .. _____ ... _ . ___ .. ____ ,._. ___ .. _. . __ ... _... ___ .. ___ ,_. ____ T.?t~~.3.1.1_:::_. __._______,______ ._
ORG
*-4&4*OORG1*DORG2*DORG3*DORG4*DORG5
TSK5311F
____________________ .. _0_<; .. ___ ..... __ ._._. _I N ~____ ._________ ._ . . __.. _. ____ .__________ ... __ ..___ ._____ ~-.----.---____________________ I;;,~~_~_,+!t_. _____________
DC
I N4
TSK54472
...._______.____ ._______DC ____._____ .__ J N5 ______._________ '" __ . ___ .. ____ .._.... __ .__ "_'_"_'__ '______ ._. ____.T.SK54.47.:,1 _______________ _
13 TADOS&l
RESET DISK DEVICE JABLES
TSK55330
ASTAR LOX
__ ._,_________ .___ L_q?<_.____ 12 T ADD S&.2,.... _._. __ ..... ___ . ______ ._.__ .____ ... __ ... _. _____ .______.__ .._~ __I?K5~.?gO __. _______ ._. ___,_. __.___
LOX
12 TADOS&3
TSK55560
.__ ......____.__________________1:-..9_X ___ J 2 __1" APDS!,:,±__ ... _. _____________.______. ______ .____ ... __ .__________._____________._T S_K5559) _______________.
~'OX
2 0
TSK55592
________ 5. TO _._.__~2 _ I NO 1.. __ .__ _____
.. ________________.____________ IS~55S._~3_. ____ .___ .________
ORG
*-4&4*OORG1*oORG2*DORG3
TSK55594
_______ ~Q~ __ ._J?.. _.T.~QQ~.&_? ____.__._._ .. _ . ._._______ .__ . . _____ ._________.__ .____________ J~~lS..?_?~2_~ __________ _
MDX
2 0
TSK55596
--:--____.___._.sTQ___ . X4 __ tN.D_l ________ . ___ -. ______ .. __ _____ _________... _________..ISK55r.?.9..7______.__________.__
ORG
*-4&4*DORGl*DORG2*DORG3*DORG4
TSK55598
_~--,- ___ .._..______________.J- P_X _____ I ~__ r ADD S & 6 _______ .. __ .... ____ .... _ _ - . __.___.___ .. __. __ T S K 5_5~9.9_______ .. __ ... ________ _
MDX
2 0
.
TSK5559A
_S.I.O_____X2___ I_ND_l_... _._ ..._____..._. __ ..___._______.__._._.__-_______ ._ _ _
1.$5.5_5_598
ORG
*-4&4*DORGl*DORG2*DORG3*D08G4*DORG5
TSK5559C
..__.______ ._________I.PISP ___ LOX .. t_2____ oRSU~. __.._______ :.X82I1DRt VE COUNT. __ .____ ._________ .____ ISK_5_62$'O'__
IDKOA LD
L2 TADDS&l
GET oEV TABLE ADDRESS
TSK56290
~-----,-._-----------~I9 .. 1-.2._1 ADDS&_l._........ _rHAT __ 0 I_~!S_... ()FF ~L I NE_ ... ______ .___ .T SK5646_0 ___________________
A
2 LOGDR-CON
TSK60590
* __ R_f::_~_~_IS::_. I $_~ ? 9_.7. :?, Q__ ...._____.__..__..___ ...._..__________. _......-...-_...____._. _____ ._,_.____... ,___ ._ --- .....--._, ..- . . .--. -----.- -.------SKLLV LOX
12 TADOS(;l
TEST .TO SEE
TSK61080
._O_ _. _ _ _
---_._--_. _ .-. . . _-----.. . _._.__._ . - _._-_._. _-- - - _. _--_._- - _. __.- . . --.--.. - _. . - - --- - _._------.._ . _. - _. - .__ ._-------------- - .- - _. _._-----_. _. -"'-¢/g---------_.. - .. _---_.--------
_._---
--'Page--7oT-T4
--~-APPENDIX-D------'---- - .
TSX MODIFICATIONS TO SUPPORT
.. -"SIX DISKDRlVES------- ----------.. ---
. SOURCE LANGUAGE CHANGE CARro
.~,--~~===-~==.==========~------------------
LDRCD LD
L ~BFLG
IF JOB FLAG IS ZERO SKIP
NPS05230 .A----L-·-----188----..:...------ADDRESS--OF--.OEV-I-CE-TABl.E----NPS221-1 O·
-*
*
DELETE NPS22~70
._ _~-*DEAC__.MDX __ -L_3.__._~_L
~eS2.6820
_ _ __
NOP
NPS26830
_ _ _ _ _ CDEAC_LD___2._CDCON&,2 ___._,__.____________ .. ________~NeS2684Q _______.
MDX
CDOUT
NPS26900
._-_._---------------_._._-----------------_._----------------_._---_._-_.__._._---
NUMBER OF DISK DRIVES
2 6
LOX
CLB07770
__ ..•.• _ _ _ ._ _ _ _ _ _ _ _ _ _ _ ••___ .. ____ . ____ • __ ._______.__... _ •. _ .._ _ .___._____ ._._••.. "_" ___ .' __ ._. __ •._. _ _ _ _ _ _ .e.____. __._ _ _._____.•_ _ _ _ _ . ___ ._. ___ •. __________
. _ _ _ _ _ _ .•
DC
/BOOO
. DR.3
._..._,-...-. _____~ _____._._O$;____.___.___.I.C 0.O.Q._______ ._.___DR,e_4_
DC
CL808291
__s..L-_~Q_8.2_2~.. ___.________
OR.5
/0000
CLB08293
.._.._______ ._____. _ D_C.___ .___.. __ . /80 O.D _____.__ -D.R. 3 _______.__________________ CLB 0837..1_ . __________ _
DC
/COOO
DR.4
CLB08372
__~_.________:__:__----.-DC.-------/DOOO ..-.---..-.-.- ..DR.5__ ._. __._. _________. _________----.CLB083.7.3_____.___________.__
DC
0
DR.3
CLB08441
_.__..____. _________ DC ______O____
DR • .4
CLBOH442
______ _
DC
0
DR.5
CLB08443
.__ 855. __ ..__ 3.45 _____.___.____*I.Hl S_ BSS_MUS.T .. BE __ AI __ LEASI_!__ CLB21 530 __ .. ___ ._._____.
BS~
L
INOIO.-Z
BR. IF NON-PROCESS
CLB23110
__ .___.________ .______ BSC ....__ L. __ .lNO.1 O.E _____ -.BR.• __ l.E .. NO.T_ AN_INTERRUEI ____ CLB23_150_. ________ ._._._.
INOlO LOX 11 192
NO. OF DRIVES
CLB23280
,______________ LDX. __ .. .I.2 _189_________
__O _____
-INO'16 __
_______________CLB23290. _____ ,__._
~~.~_._ . L~ ....:~.!_.________ G.E.T._DEV_LeE _.I.J;.BLE._ r?O.I..NIE£L._~.~_~~_~_~.~~______~_____
OR
Ll WSOVF FLAG BIT%BOHlc.DRV.CO.%Bl-G30
CLB23350
_______________.__ ._S~LO_. ___ .L.1. __ WSOV.F_MAK._._RELA.I_LV_E_._D.1S.K_ADDRE.S.s __._.._._CLB..2_~.:3.qO __________ _
_ ._____ ._ _-EQUO'O._E au ______.__ 0 _. ________.___ ._._.____________ .__..____.____________c LDOO2..1J _______.__
LD
L
CLD03605
189
_ _ _ _ STO---.L_._PRTCL ______ . __________._ . __.________________.___._CLD.03606 _______.____ ._
SLA
STO
CLD03607
16
_s_._______ L. _____192 ..________ ..____._ _ _.______,____
L
__CLD.Q.3.6.0.8_.____..____._
PRTCL&l
CLD03609
.______ LDDBL .. L OD. ___. L .. _.S J GP K ___.____ ._.._. ___ .__ .____.______ .. __..___ .. ___ .___ .. _______.:.. .___c LD 03_6.1 0 .. ___ .. ___.__ _
LOX
11 188
INITIALIZE START OF TABLE CLD05750
_ _ _ _~ _ _ _ _ _.MOx.. __.__L_~.L. __________.---Of-ADDRESS.E5.-Qf=_DE.V.I.C~___CLD_05.1_5_2
STX
1 HMORV61
TABLES
CLD05754
L O__ ._.L._._t9_L____________ I.N..LTIA L.I Z.LO_RL\{E._CO_Q.E-_ _ _C_LDO';L1..;?_6~_ _
SLA
12
NUMBER
CLD05758
S.J.O___ L ____S_A ___.___ ..___ .________.____ ~ __.___._. ______________cJ,-'OO.~_'I:5..A _________
LOX
11 191
SET UP LOO~ COUNTER
CLD0575C
._. ___.____._____._____. _. ..... __. _MD.X___.__ ._.1.__ ...l ._........ _____. ___ .__ .... ________... __ .. _._ . __ . _._. __ ._____ ._____ .________._.__ . ___ .___ ._ ...______ ..__ .______C._L.D05.75E_' . . ____ .__ .___________ _
HMDRV LD
Ll
*-*
PICK UP DEVICE T6L. ADD.
.o.C___ .___ L8 8___._____AD DB E..s s_._W_OEO.S_t9~-=-1..2-2
DC
5
~-o----'_P RT.CL-gg________ --:~!----.-,
WORD COUNT
CLD05760
C_L.DJ~.9_0_·___.___
CLD14470
--.--- ~ ~ ~6E-~;·u~·~8.Q·s~EHY __.DJ~Y._T},_~___~ ~-b-i·:-!~~------------·-
..______._!__.______ .__________________.____.__.____ ._._._.__..._____.____.. ____ .____________ ~___. _._._.___~k- ()_t~~.?~. ___._._._______ ._SA
DC
*-*
DRIVE CODE
CLD14760
.______...;.*;.......-I~~_r~_QY. E_.<: LDJ~ 7. g.Q__~N.O_._ . . ~ LO J ;5_'7}.9__._____ _
REMOVE CLD17140 THRU CLD18330
*
.
N><.T S9--i ~~_J.! -~:~DR·&I:---·--~R 111~DD_.9f._l..QG . . . D.RY-..If\J3 ..·-_..
Q.A41 " ..!¢Z$4-f0.244%¥4A.44WPMM¥JMmwm,:· :;'": "",'0 , ;:",:::ra:.;." .•• , ,t,t.'::',n "·N,,"'·',"" .,W;;".QI",iI";;;;::'::=='=M&'''&:MWffi::tilGiltD1llLlliiAUU;;'=''iAGllGHWGWWLIL\D£i1;IWi&zaatMliniWi/Liiiliiiilli\!:\iIiiIiiIiii...liilii..,iiliiiiiiWiiiiia&i&iiiiiiiliiiai....
-~~6!~}~6----
-'7?-7
i
>',\,;'
-~-,........._C.~#l.tro7
...I·
APPENDIX-.B----'-TSXMODIF'ICATIONS TO SUPPORT'
-----,
'... ~.":","":.~>~~.- •..,---.---.",,::,.---.-•.- - - - - - - . - - -.• -.-,-,, ... -.:..,-.:.,----~' ;'-.c:---..----~---.• - ..:- ··-··---8 IX . DIS K DRIVES . . '---..... -'-" -.-...-.---;-.-....-.. . .- . .-------......--'---.. -'-'SOURCE LANGUAGE CHANGE CARDS
_ ..
. _ - - STX
CLD17160
1 S:OG.oWlil
0
S 1>.5_. __ t_. NQ~LYI"..:..t_. _____._._.______._.______
CJ""Dl.11J_Q ________.
MOX
1 -1
CLD17180
."--_ _--"'$ T>'L __ J_.~.O_LY.I_Ct.5
~LO_L7 L~Q
LOX
11 190
CLD17200
_._-_._______.__MD~C. __l_,,_l __ ._____..._.
, _ _ _~___. __._______kLQJ.1.ZJ. Q_____.
STX
1 SETDR&5
CLD17220
_.___ .__ .__________.__LOK_. __Ll .._ROBUF&1..7_____XR1_P_O.l_~T_S._.. T.Q__CQL_l.8_______C.LD.l.723Q__.._______
LDX
2 0
XR2 LOGICAL UNIT COUNTER
CLD17240
__.____._______NEXCLLDK_......__ ..3 .... ~6. .___._____ .___.xB_3.__ LS_C.QMP.AB.E-.IAB_~Q_I.blI.EJ~_C.J...D_1._7250.. ________
*
CLD17260
____.___-----*--IN..LT-LAL __THE._,.LO_GJ C.IA.LORl vE,.s ....._EBJ-3_0R_~Q_.__..1.O' __._ _.__.___CLD.1.72_7.0_ ..__ ~_,;.. ___ .
CLD17280
-_____ ._ _. ___ .REDOA._LD. __..._____ ._l..._EQUaO _____.__ ._LQ.AD.__ H'OLLERLTtL.DRLV~~QQE._--.-CLD_l.7220. _._ ... __.__._
EOR L3 NUMBR&6
COMPARE TO TABLE ENTRY
CLD11300
._____BSC ___ L __..SETDR •. &_~_____ BBANCtL.O~_L_EQU~Lj_O __S.E.I_QB_.___ C.LD_L'I3_LO______ _
-------MDX
3 1
NOT EQUAL, STEP DOWN TABLE CLDi7320
- ___.__·_.,._____ .__MDX ___.. __ HEDOA_, _____ NQ~T_AT_END_._o.E...J.l\BLE. ,REP.E.A..L~CLD,L7...33.0_____ _
LD
1 EQUOO
END OF TABLE, COr. .1PARE BLANK CLD 17340
----:---.----------.---E OR __.____ . ___ NUMBR&6. __.____ .. ___._.__________._________________ .. CLDL735.0.._________~_
esc L BLNAK.&EQUAL BRANCH TO TAKE OFF L CLD17360
._ _ _ _ _ _ ERCRO._MDX.__ ...L . __ .ERRLO.t.l ____ NO_L._VAL_lD_.CD.DE~No.T __BLAbt~_CLDJJ3:LO ______ __
NOP
GO TO ERROR 10
CLD17380
________.___________ .....6 SC ___.,L_. ____BA.QC.D_______
__.___CL D_t1.3.9.Q.____- -
*
:_S.IQB.E____SE::LECIE;Q__ Of;_V.J.C.E__
*
T~~tL.~_A._QQB~_$_S_I_O_LO_G.lc..t\L......__OB_~.~.~_!-=~_:~.~______.O
CLD17420
_______ ...___ SETOR_.LO__ ._L 2 __ .~..._..~________ .. ____.______.. _._________ . _______
.. _._CLOJ.. I!±_~_Q____.__._. __
6SC
L
ERCRD.&ERROR IF NOT ON LINE
CLD17440
__.____ .___...__. _____L.Q__ ..L.3_..~."7.~_..... ____ .__ ,_GE._T___ Q.~V_i_~f:_ .. I~t3_l,.f;___ ~.QQ.RE:.~~~_~___C.L.QJ__I~t;?_9____~_
FROM PHYSICAL UNIT TABLE
CLD17460
_~-'- ______ B_O'GD_W.. __S.TO.:___ .L2_ . .~ __ ! __.. _____ ._. _________ SIOBc: __Ig__ L_()~J C..A.L __ VN.tI_. _________CkO.. LI4 70. _______ . _.
MDX
1 2
BUMP TO NEXT CARD COLUMN
CLD17480
_ _ . __________ .___ .Mo.x_____ .a.....·:':!+_ ..___.___ ..:_.____Tt::S.I_.EOB_.LA;ST_.. _LQ~_l.cAL._Ul~.LT__ c:LD_L7420_ _ _ _ .__
MDX
*&3
LAST UNIT.GO TO CHECK DUPL CLD17500
______~1.RX____ ~.. ? __.___._...__.__.______.____._____.___ .__ C!-.Pl.1;:UO __._____
NOP
CLD17520
_. ________._______~JtQX_.. ______ NEX cL._.._._.____ ... ~RO.CE .SS__ .~ E2<-T_CO_L.LJ.t1.~ ______.. ___C_L. 0 1,.];5.;:tQ ___________
*
CLD17540
*
___._______.__ ..!__ L~§.L_£Q~_P~_~_~_I CA.ILOti_. I N
L()~ I ~A~___Q.~_.LY_~_CQ.~f§_~~A ~::-____ ~_l:-QJ!.?_~Q_____._____ _
*
ION PUNCHED IN THE CARD.
* ._.__.____ ._._._______._____.__._____
CLD17560
C !:-.Q.!.L~:zQ________
1 5
SET COUNTER FOR OUTER ~OOP CLD17580
CJ·:'-~.RL._STX ______L._C.QNEG_... ___.... ___ . ___ .. ____.__.___ .__._______ ._______·____CLDL7.59_0________ _
LDX
12 CONFG
SET COUNTER FOR INNER LOOP CLD17600
._MDX __..... 2 .. ~ L ___ .._._._.:... _____._.._______ ..... _._ .... _________.___ ._____._ .. _ .._____ CLD 1...1.6 1.0 .. _ - -.----NOP
.
CLD17611
_ _ _~_ _ _ NO.L'{I __ LD__.__ .L.L_.~=.* .._.___ . ____.__.LOAD.....TEST._.W_OBD_·
___CLD.L162_0. . _ _ _ _ _
.BSC· L
LILJO.&IF ZERO. SKIP COMPARE
CLD17630
----.____ ......EQR___ L2_. *:_-..~ __ .... ____ ... ____.COr.~PARE --"- .. _..__.__ ... ___ . ____.. _.. _.. _..___ ...... CL017q40 ... _. ____ .___
.
SSC L ERCRO.&IF EQUAL. BRANCH TO ERROR
CLD17650
.__ ..MO)<_._._ .._2 __-.: 1 __ .... ___ ._ . ___.... B.L.J~P .. C:Of·'PARE . COUNTER. __ ._________c:LD 1.]66Q ____ .. _. __
MDX
NOTYT
BRANCH FOR NEXT COMPARE
CLD17670
LIL~O MDx
1 -1.
NO EQUAL. MOVE TO NEXT .
CLD17680
LOX
0
-----·~1"6x·
, 1 - -.
·
- - - - - - - - - .
1---··---.-'.-..
1
. ---·---cH·ARl--------..----·
--.-~--
-.---*-.~-----~~P.~- .. -____ .. ______E3~~. 8I. _______ .____.... ·-..
. . --.. ---.. --..
CL 01-7 690-------
--~-.--.-.-.----.
--·---··-·--------------i-t~i~-~i-6~-ffo
k'
----:--""'-O"-........,....,.----.;..,,;,~--,--.:~14G7 . . ±l.........
-: _.-----..........,....o;.;..~~
..
...,..,...;.....-~.
page--c-9-·-iD.fL-14~--
-APPENDIX-'-'-B---
-. -
.:··.TSX MODIFICATIONS TO SUProRT.
:..
.
. ~~'"7'-~'~---'--'----'-SIX-'-DISK -DRIVES -.--.-~.------.----'----.--.---:-:----.-SOURCE . LANGUAGE CHANGE CARDS
'. .
-:---_"---..
.
---'------~-------:...~--:.... -~.. ..
_···-,,0_.'
",,",.",,-,-.....---,-_.___:
PU_T_·
..
----:'":'-:"'--.,---
_~_~_~_=~~~_~~!_~~~ V ~_~F~_L_.~_~_E_.____"_________
BLNAK SRA
16
ZERO THE ACCUMULATOR
MDX _______ 8_0GOW_. _____. ___'_ _. ___
~.tg~;;~g-----CL01774Q
C.LO.L725 0 _____ _
*
*
CLD17760
CLDj_.7..1.7_0_ _ _ _ _
CLD17780
_ _ _ _ _ ~.~~.8T __ t__Q.~___t l __S.Y.X.RJ __ ._____RgS._T9B.E.;__ JNQ.s...>5 ___8g_GJ. SI~B_S___~_l,._Q_l12??.P__._____ _
LOX
12 SVXR2
CLD17800
.,..>t.-.. - . - - - - - - - - - - - - - - - - - . - - - . - . - - . -.LDX
13
SVXR3
CLD17810
.. ---.-----.--. -- -- -.- .. _ •.....•.. -..
--. ---- .. - - - - - - -. --.-.-.. - .. - - - - - . - - - - - - - .
---.-------.. -. - - . - - - - f
SSC
I
INTPT
RETURN
CLD17820
~
*
.
CLD17830
.
~·-··--·-·---··----·C C-N F "G---O·C··--···-----;=;-·-·c-·-..-·-----·------·
C LD 17840-----
...:.._.*_._____._____ _'__~___
.---,~----_
:
I
l-'______. __~_U!~p..R . __O_<=-_______~Z_Q_O.O'___ .__________ O_.___________._______..~_~_O_1..7.§~Q..______.__
DC
/1000
1
.
CLD17860
'. ___ .__
. . _____ --___________ .D.c. ________ ./..0800__ ·._.. _~_._ . _. 2.______ .___ .___.__ . .____._.___________C_L.O_L18_1.Q _____.-' _____
.;
DC
/0400
3
CLD17880
-, _..... _____.~ _______ ~ ____ DC... ______._./..O.2 00 __________ .4 .. __ -, _______.
C.L.Q.1..7..82 0____ _
DC
/0100
5
CLD17900
.... ___ ._ ... _ _ _ _ ..__.____ ....________.DC ____ . __ . ____ ~Q_O_OO___ .________ 8I,..AN~____ ._. _____.._____..._._. ____._____c.L_Ql...7.2.1..0_______..
STX L3 SVZR3~1
CLD23401
..._...... __.___.._______ ._. _______ ._LDX ____ .Ll __ .1.ea. ___ ~__ . ______ X8_CJ:~OJ~.r..S__ TO__ LQG___ Of.'-v__ ...IA~ ____C_~D_~~_~~..Q_....:__________
REMOVE CLD23640 THRU CL023840
_ _._________ ._ ... L OX ..... _ ..3_._ 0 __._________.___~_J.N.1..I__ DRY._C_O.D.E ._.:LO __ .ZERQ... _ _.._ ....cL.D.2~36..40__._._. ___
STX
3 DRIVE
CLD23650
_____ .____ L Ox. ___.l 3 __ 1..9...l ____ .____._________.________ .___.__________ C_LD2_3.66 0_.________ _
MDX
3 1
XR3 IS LOOP MONITOR
CLD23670
__..__._________.___. _D.E_~_Q__.______ . . ~_5_.._.~_._._. _______.__._. __ .__.____ bJ:J?_Q_-'-~~_~_. ___ . __
SCT08170
___§.!.Q_._. __ ~~.l _ . ______._..___ .___._____ . . ___.________ .
___~.~ TQ.8 t?O~'_" ____'
LOX 12 *-*
AT XEQ TIME,ADDR OF INST
SCT08190
_ _ _ _ _....:..:..*BEMQ_\lE. __.S_C.I08.?OQ __._.._. _.___ ._.___... __________ ._______.___ .___.___________________ .___
Mf::)X
J 1 DLOGO
SCT09660
.
__. _____.______.__._.S_I.X. . _. ____ .1 .......*.&.1 .. _.. ____._ .. _._..__ ._._..._...._... _. __ .._. __ ... _......_._._.. __ . ._...... ___ .___.___ ._____ SCT09.670 .. _. ___.... _____ . . .
LDX 12
SCT09680
L.D____ L.1.__t.). A.OB.Z(S_C___...~.-NO-J_--s.A.r'1J;.-D-t~K.-£-o.-~PJ,tS
SCI09Jp_2_0 _ _ _ _ .
MDX
11 DLOGO
SCT09830
_.______________..$.I'-'_.. _ ... 1_ .....~& 1 .. _......_. _.... _...... _._. __..... _.____ ..........._. ______ ..... __...__._. _.. _._ ... _. _____.____ .___ ~~I.O_~_C3~.9_..____ .___ . ____ ..
LDX 12 *-*
UPDATED
SCT09850
____f:.:..... ._. __..L.. _. . __.[)_LO.G.O_. . ___. __ ..____._____....__ ._____.___ .__ .______ ._.___._._ .._______.. _.____$~_I.1_~.:3§Q ______ . _
STO
*&1
SCT15390
.l.,.QX..... __J .4 ... _~":".!. ____..._____ ..._... _. __ .____..... ____.__________.___ .~I.l?!!.Q.9_ _ _.____:*REMOVE SCT15410 AND SCT15420
A
L
DLOGO
*-*
_ _ _ _
"
_
... _ . _ _ ._~ _ _ . _ . _ . _ _ _ _ _ _ ~ _ _ • _ _ _ _ • ____ H _ _ _ _ • _ _
·~_.
_ _ . _ •• _ _
!/J;b
•••• _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ " _ _ _ _ _ _ • _ _ _ _ _ _ _ _ _ • _ _ _ _ _ _ • _ _ _ - - - - - - - - - - - - - - - - - - - - - - - - -
.-----,-
o
-----..,,-.-.-------
. ----·. . --------····--··--·-----·APPENDIX--·B---···---------------.
TSX MODIFICATIONS TO SUPPORT
Page 11 of 14
~.-------._._. ---.----.. --.------.-------~.---------.---.--.--.- ---- --- SIX DISK DRIVES·-·-·--------------·-··-----··-----SOURCE LANGUAGE CHANGE CARDS
#1407-1
r'tJ
DC
ONECN
DEL06251
~-~:...------------------·-·----··.. DC·-·-·-·-··-·-·-----ONECN . ---.---..-.- --..-- -·----------·----..:...--·---··---------------·----·------DEL06252---------DC
ONECN .
DEL06253
. ____ .. __________-.G \JJ SAD...._LDX ___. 1.3 DRN UM __ .__-->- _____.__ EOP_025.2_0_______
RPLDK LOX
1 *-*
EDPI0140
eP.LND_._S:IO__.._.__ L_l __ ~__:_~ .. _ ... __. ___ RES.T.QRE ._THE. __ ENTRY______._____ . _____ED.2J'O __Lo,o___.________ .
LDX
1 *-*
XRl N NO. OF DRIVES
EDPI0610
_SIQD~_L p______ L_L _~_... ~___________ .______.______.________
f:_OEJ _0_02_0__________ _
SA
DC
*-*
DRIVE NUMBER
EDP12080
H
_ _ . ____________ •
___________
• _ _ _ _ _ _ _ _•
- _ _ _ _ _ _ _ _ _ _ _ _--"6S.C_~ ____.~______ .___.. __ ._____ ._.1_S ___ A.g_E~._._CJ)QE___ L~S.S __TJ:L~~_tp_O'.$.Ko_tQ~O_______ _
MDX
PSTVE
YES
DSKOI042
..._____
- ___________________ LDX _____ 1. ___5. ____ ..__ . ______ ._ ... NO. XR 1115 __ 1~ __ AREA._CODE~2_5_____ D.sKO.l.043____. _______ .___ _
MDX L GETAB&1.11 SET UP FOR ERR CNTR IND
DSKOI044
--0---------------~J~_Q_X___._._____ ._!_ &..l_. ____________ ._____________.---.--.. ---- .--.-------. -- - ---.----- --.------- _________Q~t5.0_1Q~-§---.----PSTVE LOX
1 2
XR1h2 IF AREA CODE IS 9
'DSKOI048
..~-.------.
----------
----..
.. ___________. s-8A._____.___ ....__l._l_. __.__.____ . ._._. __ J?QSJ.IJQ~_._8.}~;:_~___ ~OOt;:_.AI __8.I.__ Q_~.~Q_LQ_:LO______.
STO
AC
SAVE HEX AREA CODE
DSKOI055
----- ~;~--- --~ETAi::- ------~~s!lHE~ _C:ODE._~_ Q8_?5_ ---- --6~~~+~~6---?Lf3
-~,..,..,--~-.
--:--__'.
'~---:---
APPENDIX B
1~a~e_1.2~r" 14
TSX IDDIFICATIONSc-ro SUPFDRT
---....'"'-'---------
#14071-1· .
___.___. .__._:----------:_-.--...:...------ _. ____
~~_._._.
__. ___ .--..-------.-.. . --. -.--__ SIX DISK DRIVES .
----.--~--
. -_._-,------;0_._-'--_._--------_._--
SOURCE LANGUAGE CHANGE CARDS
'>
- - ' - _ - : - ._.--.. _ _ _ _ _ _ - - a -_ _ _ _ _ _ _ _ _ _ • _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
MDX
1 -1
NO.SET XRl FOR DR CO§lOR4 DSKOI080
.:._____S.BA.____ -' __... _~______. __.___.____.________ . __...._. ___. _____._.____ .___.__.____.______O'_S_I<;.OJ_.O~O. ____._______ .
·sse
E·
IS AREA CODE 8 OR 24
DSKOIIOO
_~_ _ _--,___ J!1DX-:..._l_=~ _________.___ NO__t_SE.I_.xRL_.E..QB_DB. __CDll.o__QR3__ DS_KO_LL1..Q_._. _____
· __
NOP
YES
DSKO 1115
d
_______---'WG.EI.AC--STX--LCVEDP~____SA...YE . __ DRtYE_C_O.D.E _____._ _ _ ~.sKQ.. Ll.2O' ______
MDX
1 12
SET UP FOR ERR CNTR IND
DSKOl125
-LQ_.-----.c_V...EOe
_ LQAD_.DRJ V_E_~.C.oDE
_______ 0.5.KOL1..30 ______.
SLA
12
SET UP DRIVE CODE FOR 2ND DSKOl135
__..STO ___.. L __ .__ NUMER&J ._._____.___ .WORD___OF__ DJ.SK___LO'CC ________DSKO.l.lA'O.__ . . _ ..__.___ .
LD
AC
GET HEX AREA CODE
DSK01145
.SRT___ -_.4.. __.____.___ .__ CONVERI__"LO_... ES.CD.1.C ________..._.DSKO.. 1.150__.______.
SLA
5
AREA
DSKOl155
_ _ _ _ --'-_______ SLT____ .__.._4_____ ._______CQDE ________ ._ ____.____ .0 SK01.l.60__.__ ._____ _
,
OR
HFOFO
DSKO 1160
_'._.______________STO_. __ L .... _.PAREA&6_____ .____.__ STO~E __ J.N___eHJNT_.AREA ____.___. __ OSKO.L120__ ._____ _
GETAB MDX
L1
*-*
MODIFY XRl FOR ERR CNTR
/FOFO
EBCDIC MASK
DSKOl175
.STX ___._l ___CNIR.. _. ____ ._~I.ND __ A.ND__ S.T_ORE. __ LI_..___----DSKO lJ8_0.___ ._~ ___._
HFOFO DC
DSK01565
._ _ _ _ *.REMOVE_DSKO 17.30 __ THRU__ DSKO_L8_0Q_ ... _____._. ____._~_____
._________. _. _________ _
*DELETE DSK 02290 AND DSK02300
___CMIH__ DC_______ ._Q_____..___.. _. ______.. _C91~LTAJ..N.$_._C~_T8_._NQ..£_QR__________ .D~KQ2_5.~;:;__~ _______
*
HARDWARE ERROR
DSK02526
___.•________.ST.O _______ ~A.REA&9.___.__._____CQ.8.ELQAQ __ .N~.M.~ ______...._J~.~J'50;3.5_9.0_. ___. ______
STO
CNTR
ZERO ERROR CNTR INDICATOR DSK03780
____LD________.. _.....Ct-lTR_.__ .. _. ____ .. _._.______ .____.______________ . _______________ .O_S..lSO_=i~_~O___________ _
*REMOVE OSK02830 THRU OSK0285Q AND DSK02890
.__.__._____ ~ __*R~MOVE. __ .D.sIj:______.___._?~.~J~.__ .?~~ __ •____ . __.
.2310 25
•
EBC
CEAOIOOO
____ ._____ _
~.~~~E~ ___________________~~~9_L9l9
CEA02011
_____
~~~.9~.Q.lE
CEA02013
.--_._______EHCI___._..EPU..__. ______ .. /FEF O .. _.. ------.- OUTPUT AREA ______ .. ___ ... __ ..__________CE800 540 _____ .__ ~ ___.___ .__.
LOX Ll 114
TO BE PRINTED OUT
CEB00670
___________LQX-..__t_9_.___._. ____.........._. _MA~ __ NO.__QE_lJ.N.tJ_S__T_ttLS~Ey___C_~6_0_L2...t~_ __
LOX
II 193
CEBO·1250
___
. _.__------.-------- _t,.. Q)< ..___J 2_...18_.7_. ____ .__ ..____ ._.._ .___ ___ .. ________ ._ __._._. _____._______________ .__.______._c:~ BOJ_2 6 0.__. ____. ____ _
.
_~~~_____ L~._.~~FF&8 __________... __ . ____~________ . _..____________ . _____ .__ . ___ ._~~~~_~.;~~ ___________.____ C)
LOX
11 187
CEB01780
____~Q_~_._. ___1_J____J. 2.~ __._. _____.__ . ._._._. ____. _. .__ . _. ____.______.. :_____ .__.... _. . . . ______<;_~~~l_E?.!.~ __._~
LOX
2 7
NO. OF LINES OF OUTPUT
CE601840
-
-.-------------~-~.-----.- ----.~~----.---.-------.---------.----.----.---
. . ------.---.. -.---------.-----.. .-·----~-~~6·~~~-6--·- ¥-M
---.ffl~Q'H·-------TSX M)DI~i:~~;;:3Bro- suproRT _____I'age . ~3-O:L~lt__:_~\
_____,____.____ , ___ , _._____ ....___ ..._______._, ___ .__. __ '... '_' .... . SIX .DISK DRIVES ....
SOURCE LANGUAGE CHANGE CARDS
,________.. ___ .______ .. _.. _____ .. __ .___ ...
__ I
--
--"'--' ,-- -~-." .. ---_.._------_ .... -_.'-,------------K2
DC
2
TEST IF 1053
CEB02920
- - - - - - - K 6 , - - - - - - . D C .. ,-- -.. 6--. --.-.,--.--- .... TEST ...IF.231D-...... -.. ---.------------.---CEB02960 ..---...... - ...,----____ .
S
K6
CEB03840
-------,-.-------.------. L DX.-.I. 1 .. -1 93 ...._. ... _______,__....______ .__ ....__.. _________ -_' ______________ ._ ... ---CEBO 395 1 ,.. ---.--.... ,- ..
STX
1 *&2
CEB03952
_________,_._. __.__ . LDX._ ..... "l. 6 ___ .,_____ ..... ". __ .,_SET, _TO' __ MAX. __ NUMBER____ ,__ ._.... __~ ______.__C.EBO,3.960 __.____.. ___ ~ __. __
HSEEK LD
II *-*
TEST IF DRIVE ON LINE
__.._________________________ .5._.. ____ L_. K6
DIV
DC
USED TO CALC NEXT
/1000
/500 D
_._. __________ ...__.______ .§..~. ___ ._ ..... __ pC
DC
-._. ____ .. __________._.... _ . ___ . __ DC......
DC
'_____._..____ .. _______ ,_... _____ .. __.___ A
._ .. _____ . __ ,._. __ .___
__ L...
CEB03970
, ...... _____ .. ___ C.E8.0.4J 00.__________ _
DRIVE
CEB04191
...... ___ ''' ___ ''''. ______ ,_".___ .__ ._~~_~Q~_?_2.2_.
/3000
/4000.
______ . ___.._.
SKA04000
. ..______. ___ ,, __________.'" ._.______ ._._SKA040 1 0 ___.. __.______._
/5000
188
SKA04020
.. _. ADDRESS OF___ DEVI CE"T ABLE .. _ ._. SKAJ 6140,...... ___ . __
*REMOVE SKA16400
---------__ .. __ .. ___ , .. __ .... _._.... .LDX
. I 1
LOAD ADDRESS OrDF< I VE ZERO
DEVICE TABLE
188
STX
*&1
..--...-.----------.________ ..... LDX . ,1 1 *-*
n--·-.. . .·-..- - - .- ,- -,
-~-
LDX
I 1
STX __ .
.1 ,*&1
I1
LDX
TRL01581
__ .. _.._ ._TRLO 1582 ...._". ___ ..__ .
GET FILe PROTECT STATUS
188
. ___ ...
TDD01290
. __ . ____ ._ .... __ ..._. _______ .TDD0129L. ________ ..
.*-*
TOD01292
r--1DX _.-.. 2 -5 . . .
MDX
STI
LDX.
11 187
STX
1 *-1
-------___ ... _....______ ._.__ .__ . _...... ______ . LD - _ L2 *-*
DCON&l
LD
L
-.~.
_______ .___... _. _____,_, __.____ .____ .... _____ LDX ..... 12
STX
____________________ ._.____.___ . ____.__ . __ LDX.
. ___IRL015S.o . _.__ .. __ ....
188.
2 *&1
J 2._ *-~
-
-.. . __ .. BRANCH,oACK TO_ S11
IE ... , __ .........ID.2.00810. ___ ..__ .. _... ___..
CODE ILLEGAL
TDP00820
GET LOG I CAL __ DR 1 VE ..._. ______..___ .. ___ TD~.O.082L .... _ .. _ ..... _'. ___ .
DEVICE TABLE
T9P00822
ADDRESS
_. __ __ ___
___._TD~00830.
_.... __,._ .
TDP01630
ORIV~
_LOAD .. ADDRESS. OF. _______ .. _. .:.. ____ ._ ,_TVPOO.4.8Q ______... _________ DEVICE
,.. TABLE_ _
__, ______....... _______ .,
SAVE BAD CYLINDER
TUP00481
TUP00482 _____ ... ___________ ...
LOX
13 188
TUP00840
____._____ ... ___._____... _ ..MDX ........ ,3 _..1 _.. __ ...... _, .... _. ___ ADDRESS _._.... __...._____ ._, ____ .. ________ ....T,Ue.0084.l _______ ,__ ...._
STX
3 *& 1
ON
TUP00842
._. __._______ .___... _____ ~ ______.__ ._.LDX _ 13 __
TH I S_
__ ......__ ... ___ .. ___ ._.___ ._... __ T_UPO.Oa~t~____ .__________,_ ...__ _
*-*
L
191
S
._...,_____._______.__..... _____... _ ......L D X ,_." .11 1 88 ...
STX
1 *& 1
__..... _ _ _._______ . ____ ....... _.____ .. LD ... L2
*-*
AND
L
~9~===--=-~-M~S-~~-~B~---·
5
...._____._. ____.-:::...... _____._.______ .___A.
LD
- - - - - - - - _ _-------------._.
..
L
HOOFF
...•... DR .1 V
L
191
187
Xl
0
. _ . __
AND BRANCH TO ERROR
TDL01890
......... _. ____ ".......... __ .... __ ...... _._.... ~, _____._____,_TDLO L98.5___... _ .. _ ... _..... _.. '
TDL01986
BRANCH TO ERROR RO.UT I NE _______ .. TDLO 1990. __ .. __ .,.
TDL02340
E.CODE:~.~~E·~._ .~. ~_:.~.
·.5 •.•·• •~.·-.•.·:~ •.~ T.~~~:~.~~:~~~~.~~.~~_
3 -4·.•
MUST BE 0 1 2 3 4 5
_.. G~l __ DI SKN ... OEVICE . TABLE, A.DD
TWA00850
TWA00865... __ .. ________ ."_
TWA00890
-
.:--.----.------ -------. ,._!t£!;--.
'. ;-.:
~____.__.
.
~'::'"~' ".'.:'."~:: .:-'.~'~.'.'. ·~c~,
-
.
";
.
:.-::'.7 ....;::.. '
'.!
:-~
·-f"';:·''r
.APPENDIX B..~_______._______._·_~'._._'.'... PEige14()iI4,
, TSX- IDDIFICATIONS TOS UProRT. .
. "
..
. ·.;i.~( ;:;~));::<~:.~'{};&1;::.rt:~;t~f.~~
. '. -
. '
..;
.
SIX DISK DRIVES
'.
-"'~~""":~~~'-'-~-..._
.. -:.-.. •.---.----.--.. . -.. . . ---~- .... ~-.-:-.--~-~·--SOURCE LANGUAGE CHANGE~CAROO------~_:_-----·.•-.... "-:.'_.... -~A.J
....-.-.-..,.---.;.--~ ----~-~--~.-.---~-----~-~--:---.------.-.-.:..--:---..
DC
/1000
2310-3
BT1.0073i·
..."
~,.........._ _ ~.C___ ~..A.7.JLO
- ___- -________..IoJB.I.l.cLO...:u~__
·. ___
.. _.__
.•- :
..
.
---
DC
/1000
BTI00733
231q-4
/J:..7-0-01oL·---------_ _ _ _ _ _ _...--_ _ _--'B'-l.UO.QUl,~i:_____.___
DC
/1000
2310-5
~TI00735
n.C-·_ _ _ I-.CE.OO
BliDD-23£)_'"_. _ _ _ _
--..-DC
'.
';
\.1
t'
---------------------,------------.----
DC
..._-----.-________D.c..
.
DC
OJ:
._----- - - - - - OC
/1000
&:1-0_0
/1000
BT2002?1
B TZO_QZa..g_"... _ __
2310-4
BT200283
LC.7_0.0'---__
6I..aCtO..z64____ _
/1000
2310-5
BTZ00285
___."-_ _ _--:-_ _ _ _
. Q~___ ..LCE.Q_O_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _"""'B~T~2CLO_=2_=8_=_6_ __:..._ _
~
-- -------A
.....
~
L
2310-3
MDI02590
188
*
*
___~*...R~O~V..Ej1D.IO-Z.aQQ-----_-----_--------------"---____---;-_
*
*
A
L . 1'88
.
MDN00440
~__~___*REM~YE ~Q~~997Q __________________~.____________________________~.~
.---r-"~-----'-------------~----------
- - - . - - - -..,...--.---..
--------:--
.oC _ _ _
· _ . _.•. _ . _ . _ _
• _ _ .• _ _ _ ....,... _ _. _ _ _
.--------.-----.
.
~._.
_ _ _ _ _ _ _ _ •_ __ _ ' _
--~-
1
-!
g
mq
05/360 FORTRAN'REREAD
o
A.
Known Restrictions
1.
Itl its present form, REREAD did not work for a program which
the E-leve1 compiler suggested subdividing. We got around this
by
.
MAINLINE:
CALL RR
Etc.
SUBPROGRAM:
SUBROUTINE RR
CALL REREAD
RETURN
END
This apparently let FORTRAN handle linkage conventions for a
module small enough that use and nonrestore of register 8 by,
RE~ did not hurt us.
2.
We have not 'tried it (REREAD) with data sets organized as direct
data sets. Author said it will not work.
3.
REREAD will not work with a variable for DSRN:
o
READ (J, 3) list
3 FORMAT (etc.)
is not possible.
4.
As distributed, REREAD is for an E-1eve1 system. If your system
includes G- or H-1eve1, or is restricted to G- or H-1eve1, statement 3 of the Ass~mb1y Lang~ge source code should be changed
to
EXTRN IHCFCOMH
·S.
As distributed, REREAD uses DSRN 19 for REREADing. To clloose
ano~her DSRN (~99), change statement 44 of the Assembly Language
listing (as distributed) to
119 DC F'xx'
where xx is the DSRN you desire to use. Author says this number
is independent of number of DSRN's chosen at SYSGEN'time, and
that a DD DUMMY card does not have to be included.
,
B.
o
Use
1.
It is necessary to issue the statement
CALL REREAD
I
once and only once for each FORTRAN program in which RERE~ing
is to occur. This call must be executed before any REREADt~~
ts,kes place. .
2.
After B.l has occurred, REREADing is accomplished by
READ (19, f) list
where f is the FORMAT under which REREADing is to occur.
3.
Any number of
READ (19, f 1) listl
READ (19, f 2) list2
READ (19, fn) list n
can occur for the same buffer load.
4.
An application:
User reads cards whi~h occur randomly with .
respect to type, but each card contains its type i~entification
in the same field (say, Col. 56). Let DSRN = 1 stand for the
system reader. Then
READ (1, 11) JTYPE
11 FORMAT (55X, II)
GO TO (aI, a2, •.• , an) JTYPE
a1 READ (19, fl) listl
, GO TO b1
a2 READ (19, f2) list2
GO" TO b2
~
READ (19, fn) list n
GO TO
hu
will accomplish this.
c.
·Function
CALL
REREAD
~ubstitutes the address of bbb for the addres$ of
FORTRAN I/O area.
F~OCS
tn the "
Each subsequent
READ (n, f) list
passes through bbb, where the pointers to Q~~fers are saved a~d it
is ex~ned for DSRN ~ 19. If DSRN ~ 19, then the control retturns
immediately to FORTRAN I/O for a physical read. If DSRN = 19, then
the buffers are restored to their condition for the previous READ,
physica~ reading is bypassed, and control is returned to FORTRAN
I/O for FORMATting only.
C"
.
"
'
.
'
,
'
"
.
~
.
_
'
"
.
'
"
~
_ _ _ _ _ _ _ _ _ _ _ _ _ • _ _ _ _ • ___ ,_•••••'''_......w''''..'..''''.•••
;' ?·WMfLj·tWtttfltjf¢ijlw· fr"" ····.,iffii:ttttft¥Cj,fflri·.. ""\"'JtHitRtWt'!lIrrr
i!iI!fWU' W
~3-
D.
Author of Program:
Tom Vaden
IBM
Baton Rouge) Louisiana
E.
Presented by a Successful User:
C
~:·
.'
Wade A. Norton
Chief Analyst) Computer Group
Southern Services) Inc.
Birmingham, Alabama 35202
etJMMO'"
T.
pili'."
.~,
TAi$
J
Pllr.
W&6
p,tlS... .J..J ~ N. ~$
s.
.
~e
S,.,,,,.,..,.~'''lly J."~,,,
it .L,s"'~, of .. /). o. s.
Sa"',.",,"""'''
e.,,",",·N"~
o
•
••,/, ... c.
...1
~
..
tAt- "ac I..
w-s J,·.+rif....,+-.I +. """
J).
:*t .. I ... tied
AS'
.1 ..
4" ~,. ~.f!~.
rt
W&I
B,·',....: .. ,
l4&e w;~ D.D.S.
N.t,.,..J tn..
o. S ..
.••• r/o,,-, •
(',,,.~,.,,,,,,&.,.,
h.~';', #s
~.ftI\
.atl",,'''e~
',.+.41
-,
wit. . . . .
'.+e, p/"'.
P..,e ,.,
J ..... M.II.~
it por
HA.~. II _, $.",."1..,.,,.,
(II..,)
J4D "$& ll'I,,.D
WI'" '/)IJ.'. '6~rItAN.
o
..
H ......
".r.';'~_, ~A. ~III.'AD·
A e"" _, .,.1.;.
.r~M
(t.,....,
",lte C,,,,,·c,;",.., ~,.... J 5." ~r... e l l "
'-.,.,,,
"" ••+... ,s, ........
.,
,,~
-'"
"*.~
..,. .. ", . ..".,. ••
:....,~.>jo;,,<.
.-------~-----.-" -----~- ...
,0 . ·
.... ..
,
1..-....
~
..
:Pusl~
J/er $
/vQp/'. It~:/E/-l j:;;
~l.5
I
-.-;:-
S
v//?h..v'~g<0(/I/4n/ ~. sP;.
:0'
.~
IIREREAO Jba ~SGLEVEL=1
"IISTEPl EXEC A$MFCL,PARM.lKEO;(MAP,lET,NCAl~OC)
IICQ,y'P
EXEC PG~-1= I EUASM
IISY$LtH
DO OSNAME=SYS1.MACLIO,OISP=OLO
IISYSUTl 00 UNIT=DISK,$PACE=(11bO,(400,50»,OISP=(NEW,PASS)
~-IISYSUT1
DO
UNIT=OlSK,SPAC~=(1100,(~00.50»
.. -. __ .•._-----_._- ----.:.--~------- _.. OOOOC010
- ... --.--------....
00000020
---~-.-
... --..
~
,-"".",
... _- ....
-
.. _... __ ..• _....
~
- ... - ,-
00000030
00000040
IISVSUT2
DO UN[T=(l)(SK,SEP=(SYSUT1.SYSUT3»,SPACE:(L700,(400,50)
-IISYSPRINr DO SYSOUT=A
IISYSPUNCH 00 oSNAME=&lOAoSET,UNIT=OISK,SPACE=(400,(SO,10»,
II
OISP=(MOO,PASS),OCB=(,OlKSIlE=400)
IICCMP.SVSIN 00 DATA
lEF23bI ALLec. FOR RfkFAO
COMP
STEP1
If:F237I 'SYSLIO
ON 1<)0
'''IFF237f SYSUT1
ON 190
tEF2371 SYSUT3
ON 190
IfF237I $VSUT2
ON 191
IFF2311 SVSPUNCH ON 191
tE~2311 SYSIN
ON OOC
00U00050
00000060
X00000070
00000080
_____.______ .. M_:. ___.MI
--....!
..---.- ----. --.--.--.-.. -.----..-.-----..-.-.-.---..--..---.-........_. ----.-......--.. . -- . . .J
.. -. -.-.__.-_._-.-_-_. -_.-~===~=~_-=-----.-=~==~===
______
.
_
_
~=.==--~- _ _ _ _ _ _J
.- - - - - .------.---------.-------.--------.
.------._-_.... ---.-- ..._- .--...... _._--._.
I
-------~~~~----~~-~==--J
__ _-------_._-----_ _------------_.
...
.
•.
I
".. _._a .. _..._______ ._. __ .______________..
._---- ---------_._._-------_._.._- ..- ._-_.------_._--------_ .. _-_._._ ..-
'----------_._--_ _.__ ......_....._-..._._---_ _------------ --------_._--------_.
...
I
-""'-- ... - ......
..
. _..
I
__
••
_'M
_
-- "
. . . __ II
__.._._-----:
.
_----.---_. -..----.-------.-.---------------
.- -------.--... -...... -.... -
------·---------~·c----------
.- -.... -----_. '- ._-* ... __... _. .
.- ..
-.~
...
-.-
.----~---
------_._._._-------_.--------------.-----_._--_._-_ ... -_.------•. __ ... -._-------_._._._--_._-_. ---- - - -_._-- ...... --.- ... - ._--- ....
--.- .. _---_ .. -._-
------_...__._-_._...._._._--_ .... __..-.._---_._.. _-_._-- _-._-_.._.
......
. .-.!._ .•
1_ ..... _.......___ ..
4
_
_-----o- .... ---..-.. ___ ._.__. ______________ . ___ ...._________. . . ____.__
-.... _---....
.... -.
~------.-.- .-.. -~ .. ~-------- --_. ".;-!
t __ •....•
---_....:-_-----'
-~-----'"
..._--_ ...... _ .._ - _
...._ - _.. _- ... --- .__ .. ---------
__----_ - __
....
. .... .... --- ...- .. _--.-- ...-
...
.... -..- ........ _.-- ---
-.. -:
. . -~-. --.-~-.--. .-.--.-----.--.---~. -.. ----'-'--'-'. !.:..~-~--:-~-.'-,
.
-_._- .......
--- .... - ..-_ .. _ _.... _--... _-...
.
~
--- .. ..... ..
~
-~
"
...
_.-
p
-S VM B' ot--_·-t y p t'-' J 0-'- -AC0 ~ ... L F NGTH . L'0-'1 O' ._-. - - - - -.. E ~ T_ ~.~ ~ A~__ .~Y_~~9 !:.-~}.~!-I. g-~~~~
.
PC
REREAD
LO
(HCFCOME
f.R
'HtCFl aSH -.- FI( ---
01 000000 000002
000000
02
03'" -
~
..
00'-AGE
1
----- ...('
.30 10/29/61'-"--
_ _ _ _ - : - - , _ .. F. _ _ _ _ _
__ ._----- -.-~
01
"-'-'-'"
.__
._---_._------_._-_.---------------_..._--_._-_ .•..._--_._----_.. _...
•••••. _
•••••
_.
L
__ .-_
..... --- --
-----------
•• _. __ •••.•• _ _ _ _ _ _ _ _ _ _ _ _ ••••••• _ _ _ _ . • _ _ _ • _ _ _ •• _ _ _ _ _ _ _ _ _ _ •
-::--:
.----'-.---------.-..
--~.-
•...
-..--- -----------
---. _._-. -----.-....-.... ---------i
!
.. -
.... ~- ..
----.-....
,,-~
..
-- .. _..._-_........ -_. -----.-. --_.._ _.__ ..... _...-._--_._._-
.---.---.------.-----.-.. -----.-.---.... -.. ·.· .. ·······_·-····· __··_-_·---.-1
...
.....•
".-
._.----.-
1
- .. __ .__
._-_.. _-- ...... _-_.....
__ ._-_.
..
-.- ......-
....
'--1
- - . - - - - - - - - - - - - - - - - - - - . - - - . - - -.•.. - - - - - . - - - ... ------ •.._.,....--1
"'-1
--1
-_._--_._--- __........-._.- .... __..._-_
..
... _.
-.• -_. -., ...••.•. _. --_._......-. --------_._-_._--_._._--- -_.... -----_ ...._-------_.
)
__._--------_ -...
'---'."------' --- -...
I
------~
"1
.. _..,...-
-----0--·-----_. . - . . . -.
_._-.,
. '1
o
o
.,.. .,..:.-.-- "'---1
'.1
~".--
_._.... -.-.
,
---- - ..- .•... -
o
.... .. ·t··· .. _,
··-~"------'--t·--..4"'
..
~
ADDRl AOOR2
LOC" OBJECT CODe
,
....•
,".
STHT
- -.. ·.. ·-~··f-·- ..-···-·-·-~ .. · .............. -"'~' "'-
000000
000000
000002
000006
00000/\
OOOOOE
000012
000016
OOOOlA
OOOOlE
000020
000020
4
'.) REREAD
l08F
f
I
' .
0····
t
I
,
• '.
"
.. . .-.. -.. . . -.. '
•
.. ..
00018
4000"'8000--'00000
BOlC
BOAO
0058
16
ZlZ
""
:.
00000--'0003C
OOOCO ' - .
00078
061967WN
061967WN
.....------... --.......
oooca
?,g~g BO/\8
STM
4,15,SAVE
1
~g~g
~gg~:~ ~!:.; ~g~~
.- ...... -.. .
............
'-";~
g~ggg
g~g;~
.. --.-----.----
. . ... . . .
000018 0000000000000000
oonoco
OOuOQOOO
;00.QOC400QOOO 13 -.._ .... - - ....- ... _-.
i
1---_...
,....
~-
..
~.~.-
"
._
. - ........."
.
---------------.•..---..
11,1
L.
061967WN
061967WN
061967WN
4,15.S~VE
LM
OCR ..... 15,1
....- - - - - - - -........- ... -.. CLC
0(4,14), f 19
BC .-.- 8, BYPASS .-.. - - - -
8AlR
--.-.----.- ..
061967WN
061967WN
061967WN
• .. _., 06196-(WN ...........-_ .. _ . _ - -
ZIZ, 11" . - - - - -..---.-.-.-........-.---.------.---... -.-..--....-.-.-.---.... - ----. --..
4,0"
0(2,4) ,HOOFO
-------.
0,CHK19
. 1, AUF I DC S-------------·····
CLC
BC
061961WN
061967WN' "
---------.--.-.-.-........" ..-.. - ..---...... - ... - ... - .. ---....- '-''''''''''''
061967WN
-,..-----.-.-.-.-.--.-.~ .... -..-.-.---..... 06 t 96 7WN
0,1
...
~~
.-.. 06196 7WN .. _ ... _ .. _.. _-_.-.--
.
... '
-"
..
;~ ·OVPASS···· - ~
g:
~: ~~~O'
--';:~~~
g~~~:~~~'
- - - - - - - - - - - - . - - - - - ....--.------..-----.. - . - - - - - -...... -.-.-
'-'--" .... ~:~::~~~
-----. 39 .. .... -..._.- BCR" -- 15,1
. 42 SAVE
43
"-'-'" 4 /t
45
46
41
.. 48
~~ .
DC
-:
o·
. - - - - -•..- - - - ----' - - - . -- •. _.-
OS· F
SV2
... OS
F
HOOFO
DC
X'OOFO'
. . END'
g~ ~ ~: ~~~
..... -.---.------.. . ---.. -....-.... -...-...... -.- - ......... _..._ . --.- --.-_ ........... , 061967WN
I
............ -
I
·······.. ········_·-·:---·1
··--.. '-.. -······-······-~-I
.- ~ ...: . ~
... - _._-..•;
.
ADFloes
DC
A( IHCFIOSH)
119
. DC -_ ... F' 19'
SVl
I
d"
- 06196 7WN - - - - - - - - - - '
g~~ ~C~IIE )
1 OF'
.
06196 7WN--'--:- .
·061967WN
.
;~ LEAVE -.--- ~~ -----·!:~~~~A·~~-------·----------·-----·- .
z~ ~ g~ ~~o~·
.... , ..• - ... - .... '''--"'1
.
06196'/WN
..- - - -..- - . - - . - - - - - - . - - - -......- ... "" ...
i~: ~~~
00005'•. 50)0 OOAC--··--·----····OOOCC-- 31 - - - - . - - -.. S r - - 3 , SV2 -.000058 47FO 8044
OU064
34
OC
15,LEAVE
OOOOCO
·ooooee
000000 OOfO
."
061961WN
. 06 196 7WN --.--_._-'----
..- - - - - - . - - - - .- - - ' - ' - - - .- - - - -
20
lR
2,14
OOOCO _ ... 29 ·······--·--·-·L ..--. 1, AOF I OC S30
~mH mgg~ ~~
~. .
. . . . . ,~h..
" ... _ ... _._ ............:..._.
..... --. F01JAN67· 10/29/67
~XTRN [HCFIDS~
LR
8,15
................. -
... 17'
. DROP .
lU
LR
1 q .. -- .-.... - .. - US I NC
20
LR'
21 .-----.--.-.
22
2)
.... _--.-.
24
2S' ...
EOOO OOA4 00000 000C4
26 CHK19 .
B03C· .. -....
OOOSC ---.. 27 .......-----.
'00 n06e 01 F 1
~
061967WN
061961WN
061961WN
*,
904F 1050
~~;g
I (
START 0
ENTRY REREAD
EXTRN IHCFCOt·H:·-·---·····-----·--···-·· .-- .... -.-.. - - -..- - - - - - .. ---- .. -.
14
~CR
15,14
... - .. --- l'i .--.--- - _ .. -. US INC'
1
00004C OSU1
g~~g~g
0 .
...
...--.-.--.---..•.,_._-- _.-•• ,.. .. , "P ~~~ ;.. ,
'I
'USING REREAD,8
061967WN
000 C8
ST
1 , SV 1
.,. ...... . ..
06 1 96 7 WN
00070
B·
L
1,ADIII
061967WN
····00074· ...... 9"'--'- ···-··L
·15,ADlfiCOW--------·-.--·---------·
_.- 061961WN 0004A
10
MVI
74(lS),X'50'
061967WN
11
EX
O,74( 15)
. -.....----................. ' ..... - . ._.-.-.................. - ....
"'--'-"'-'''' 061967WN ...- .---.- ..•.... 0004A
12
MVI
14(lS),X'58'
,061967WN
0004A
13
L
1, SV1
....- ...---.--.-----....... - .----.. --.----...... - .. - .... -"""'-'...... --.- - .-._.. 061967WN .".-.------....•......:.. ',-oooca
..
000046 10L(
000040 5810' 8(lAO
gg~g.~~
I III
6
7
SOLO 80CO
'l810 8010
S8~0 0074
C)2~0 F04A
440F 004A
92sa F.04A
5810 80CO
07f-[-
0'0002410Ul
000020
00002b 18'tO
'000020 OS01
00002E '.7UO
0000 32 ~U 10
000016 904F
000031\ 07fl
00003C bsu'~
000042 4180
,
... SOURCE STATEMENT ....... .
1. .
2
. 3 ........ .
If
I
\,
I
00.0000
,
"1
..-.~..-~
.. •... ... .
-----------.
--.------..
_._---_ .._---- ..... -....... - ..
---~--.-
........- _...
-
..1
I
061967WN'
061967WN
'061967W'N
061961WN
. 061967WN
..------.--!
.--'
..... _·_ . .·----1
.-._. .-.__.·__·_·----:--1
..... _...-
.
i
!
_ _ _ ....... ___ . _ .. _ _ _ _
•••••• •••••• 4
_
...
t}C lOt I I:'
M,W'ittW'ij" ;I'¢'i t tn' ·r"" Mi"M"wtf'tiHH"i!irn:'WitHl 11 '$'1'
r·
i
·f.
I
I!
......
o
VI
.
o
i
I
I
-_ .
"Q'
I
I
.!
:i
i
i
I
000
I
I
I
Off' f'";- t· pj -"HrltrttthMtitiff@"rltfil f Pt.ftf .... §tU#"ifitfiii#"· tffff" w r)5¥f-"])TI"j' V" fMwiftfJ"j*["'Ei" I I'll"" jl""lfWf" !IIY..., urn ['rUlIT""wr ,';
t
I
.i
t"
fi
I
.,:
! ...
i
i
I
I
.,i
;0
I
000
m
r
\>.
o
o
:;0
n-..l-.l
m
Ool'-O
VI
VI.
I
I
I
·1i
I
!
it
I
j
I
I
I
I
I
,
i
!
i
!
I
i
i
!
t
i
!
I
!
:~
I
;
I
,m
.,-
I
I
I
I
.0
,>
.n
II
I
i
I
i
IiI
I
;
i
I
I
I
i
I
I
i I
I
I
,I
:.
I
!
I
i
I
I
L
I
I
000
I
:-4
:0
;z
C
n
I
-4
I
I'I
i
I
I
t
I
!
i
i
I
i
i -
i
I
I
O'
i
I
I
I·
I
I
II
!
II
·1
I
I
i
I
I
I
!
I
I
!
!
i·
.I
I
I
I I
I I
II
I
I
i
1
!
!
i
I
I
I
I
I
i
I
I
j
i
I
I
II
I
i
,i
!
i
i
I
Ii
I,
I
!~
I
i
I
i
II
I
I
I
Ii
I
I
-i
i
o
.....
N
!
i
I
I
I'
I
.()
.....
'"
-.l
'i
r
I
!
I
I
i
I
I
i·
I
I
I
I
:.
!
.
I
I
i
I
i
I·
<
I
!
!
I
I
I
I
I
o
I
II
I
I
i
II
I
I
1-1i. !I
I
i
." ·1
>1
C')!
ml
-
I
,
Ii
. I
j.
I :k14~a
-··-·-··~-··-···o····
.
CROSS-REFERENCE
SYMBOL
LEN
VALUE
,.. J\OF IOCS
00004 OOOOCO
ADtHCOM 00004'000074
~AOllI "':"00004 000070
BYPASS . OOOO~ 00005C
CHK19
00006 D0003C
HOOFO
00002 000000
IHCFCOME 00001 000000
IHCFI0SH 00001 000000
i" I 1 CJ
0000 I. 000 a C4
LEAVE
00004 000064
'Rf~EAO
00002 000000
SAVE
00004 000070
SVI
00004 ooooce
SV2
00004 OOOOCC
·zz Z.----.· '0000'. 000020
· · . · · . ·0
.---.. . ··----0---·· - - - .OEFN
PAGE
So
,
-.-:.
-~-.
.... -- . 1
- --.-..... ---.... - .. -.....
....
.'
--.. ---- ..
__._-----_.__
...
0006
0024 0038
0013 0032 .- 0035'
0036
0040 .. _- ..... "'--"'-' .... _-
.'
-- -- .-........ ---_._--_.-
,_ __
..
._"
----- ... -----
'-'
. --
...-
_
~]
I
====1
..
.'- ..... - ...................-.- ..•--._--- ...............- '''-1
--.
•
1
--.~.- ..--..--- .............._.[
---_._---_.
..I.-'-~
. ---... -.---.. --- ....-.-.-...--------....._.....-- . . -. -..................... ........-........., . -·_·.:....._·_ . ·..·..··_·1
. ,I
----_..--_.... -.- .. - ....-. __ ... --- .... -.-... _.- ......... ---_ ...-.- .... -- ................._....... -- ---- ...-----_._-_.. _----,---
!
NO STATEMENTS FLAGGED IN THIS ASSEMBLY
82 PRINlED LINES
I
....... :~.
1 ..... -." -.. - ..- ...... -_ .. .... --.. -_.
0029
....
~
..... -: ......... ~
- 10/29/67
REFERENCES'
0023
00".3
0041
0009
0040 .... 0008·
003')
0027
0026'
0022
0021
00'.7
0041
0003
0004
0043
00/t4 ....
0026
0037
0034
0005
0002
0016
00't2
0045'
0007 .
004~
0033
0016--- 0019 .
..
.• ___ .•..4._. ___._ _.___
j--.- .. _--_ ...
~_J
·.. ·1
I
---.---- ... ---- --.---.•
r"-
~.-.--
......
-- ... ----- .__ ......... -... -
I
!I
f-----.~.-.--.~-
..
.. -" ....-.. ---- ... --.-.. - .......-._ .. -.. . -.-.-------
------.---------.------.
--
~.............
..- .. - --_. ---------------_.-.
-'-'.
__ .-
------_ .. _---------------'-..:- ..- .
-_
··· .. ·· .. ·-1
_________.._____. _____ .____.....___._ ..____.___.__.J
-.. ------.---.. ---.-.---------~-------------------------.
I
·_·_· _· . ._·····__·····---1
~._--
...
I
-----------------_... _.... _-_.----------_._---_. __ .. __ ... -_ .. _._--_ ......... -- .......... ..
_._--_._--_._-----_._--.......... _.......... -.. -._ .....
...-.- .
""
II
-----"--
... -,,--- '-'''-'-''--''--'.
__
...
"-'1
..-
--_.--_._-_ _-....
..
------....__ .. ----- - ... ----_ ....... ... __.
.. .. _...... ""'.- ... - ...........
. ..---..... _-_ ....---_._-_... ------- ---_ ..... _--_ -------_..---•... _... •.._-_.. _- -_._--_.._.. --_._._-----------_._---__ .._---- _ __ __ __ ----- --..... _.....- _.. -.. -.- ... - .... _
....-_ .._....
I
...... - ..... -.
---_.------.
;
~----.
...
_
..
.-.~-
...... , ....
....
.
.
.....
...
~.--.--
'--.-.--~-"'-'-'
~.,---
.......
..
--.--.----....
- - _ . _ .... _ - - - - - - - - - - . - - - - - - - _ .. - . - - - _ . . . .
-.-------.-- ...
~.----.
-
........
--------.-~
_ . _ -
...... -
.. -
...
----....--~
_ _ _ . _ . . . . .I
... -
__ ._-_.
..---.. - - - -..... -p.---.-.. --- ....__ ._. __.... __._. . ._-_ ....__ .__..... _ .... _...................-._---•._._---------------_.- .. _._-_._-_ . ._.-.__.----_._----_._--_. _--_. __ .-............ -.
. .. - '-f
•• _ _ _ _ _ _ _ • _ _ _ . _ _ • _ _
'--"---'-- -----_ ..
~
_ _ _ _ •• _ . _ _ _ _ _ _ . _ •• _ . _ ..
~_
•.• _ . . . . _
........ _
••. a . _ _ _ _ _ . • • _ _ . 4 .. _ _ _ _ _ _ _
~
• •_
• • • , •• _ _
., • . _
... _ . _ _ •• _ _ . . . . .
_
~
.'
.-...- ..-~-~
------_. _---_.._------------------_._._..- -----.-.... ---
_--_.
.
--.-~~
SV$l.MACLfB
.
KEPT
......-... -.--. ...
SER NOS.:·-·OOOO 17.-- ------... ----------------.----.. - -.-.. -lEF2tl5I
AAAAAAAA.AAAAAAAA.AAAAAAAA.AAAAAAAA.00000042 PASSED
.. _-_... .......-..
--- I(F28S1 - VOL 5ER NOS= 000017. ' - .-.-------'--------.
I(F28~r
AAAAAAAA.AAAAAAAA.AAAAAAAA.AAAAAAAA.00000043 DELETED
It:F28~I
VOL SrR t~OS:: 000011.
..-.
- ---------IFF28~I
AAAAAAAA.AAAAAAAA.AAAAA~AA.AAAAAAAA.00000044 DELETED
--'r EF2 US r-·-·· VOL Sf R NO S: 000011.
. - - - - - - . - - - - - - - - - - - . - - - - . -...
IEF2851
SVSOUT
SYSOUT
IfF285I
VOL· SER NUS:
IEF2B5I
LOA05ET.REREAO
PASSED
IEF285I
VOL SER NOS~· 000011.
IILKED EXEC PGM=IFWL,PARM=(MAP,LFT,NCAL,OC),CONO=(5,LT,COMP)
00000090
.•
00000.100·
-'I'SVSLIN" DO DSNM1F.=&LOAOSET,DIS.,=(OLO,OELETE),UNIT=0ISK
00000110
II
00 OONAME=SYSIN
0000012 a.-------------------- --.- ---- - .IISVSUTl 00 ·OSNAMr.::*.COMP.SYSUT1,DISP;(OLD,OELETE)
IILKED.SY5LMUD 00 DSNAM[=SYS1.FORTLI8(REREAO),OISP=OLO
... I IS YSLMOO DO CSNAME=T E5 T. LOM)' 0 r SP=OLO
.. -.. - .---.--00000130 --.---..--..-------.-..---.----.------ ... -.. -..-.- ...-....-.. -_ .... -......- ...- -......--... -,.-115YSVRINT
DO
SYSQUT=A
00000149 - - _ . - -II
... -.....
.... - ..-- ... - -....-.-----.tEF285t
-ft:~28S
I ---;-··VOL
~---
---_._---_
----------------- _._-_._-_._-_._-------
IEF2361
tfF2371
'lEF237t
I(F237t
ALLOC. FOR REREAD
ON 191
SY $ U T 1
ON 1() 0
SYSLMOO ON 192
STEP1
LKED
SYSLIN
-_ ..
_-------------------------_._-..... _-_...............
--._ ..
~-.-
._....___.___ .. _. ___ .___ . _____ . ________________________ ._ .... _---_._0.._-_._-----_._-----_. . _-_._--_.-_._---- ....
_-------------------------_.. _--_._--.-._--_._-_.- ---
------_. _---
.
.----------..----. ---- -- - .._-- ... __.. _----_._-----_. ------ -
-- - .
..
._----_.... _ - _ . _ - - - 1
._._--_.
------------.----:--.........--..
..
.-...,.-.....;.-~.-----.-.--
o
...
~.-.-
...
_._ ....•.
--.-.... --.
. ...
'-.-""-'
.........
_
.. _._ .....
------_ .....__ ......._-._-_ -._._----------------_._-----_._--..
__._---
o
.. -..... _... _-
.....•.
_ -. ....
._---_._----------_.-.._---_._---_.._._-
..._-- _.- ....... --- ...-....... _..- ... '- ... __.__ ..... _..-_._........._-- ........ --.- ..-.... ........ _-------- ------_.-...
-~
".,
_._-- _.....
_~
- --...,.....
. - '.
o
~
..-
---_ .._- ...._-_
..
. .-.. '. :
o
E-LfVEL LINKAGE EDITOR OPTIONS SPECIFIED MAP.LET.NCAL.DC
tr:W0461' IHCFCOMF ._.-_ ........ .
IEW0461 IHCFIOSH
)***REREAO
ODES NOT EXIST BUT HAS BEEN ADDEO TO DATA SET
·0···
-- _----.--- _ - --_
.. .... -... ....
...
~
.
......
--.---.---:--~.-.-
- ._.... _-_._--.......
..... , .... ........ '-'._-'.""
"
""
DIAGNOSTIC MESSAGE DIRECTORY
-
----
"'-"
__ __
.
,-
- ......- --.-......•. __..
....
., ...
-...
~.--.-
....,
"-'~'--'~i
~
~'~r
..
'-~-
....... - ...
.-.-----.--'
----:. -.•.~ .. '--·---1
_
..
_
...... _.... ..... -.
.
...... ..... -...................
,.
_._--_•._---- _____.____ . _.-... _. __.___ . . _. . ._. . . . . ____ . ----1
IEW0461 WARNING - SYMBOL PRINTED IS AN UNRESOLVED EXTEKNAl REFERENCE. NCAl WAS SPECIFIED.
'--"-'''--'--'--'--' ... -' .......... -..... - ..... -.. . ....
--.- .. -
...
.- .. ---.,.
... ····· ..·.. ·..-1
_._._----_.- _............ _.. _ .... _.. _... _........... .
MODULE MAP
"CONfROl SECTION
NAME
-.-~
ORIGIN
.-
ENTRY
--..-.----....--- .-
"-'~"'-
-- .-......._-._..__ .. _...... _.. -.-- ..
.,
....... ---. - - ...
_
_---_.. __ - - .---. -......
.. - .. _... __ .._..
.
...
NAME' "" ·lOCATION ,,---.. NAME '"'-'LOCATION-''''' NAME'''' LOCATION .. ·..--·NAME .. --LOCATION-.-· ..·--·--·--·----
lENGTH .
•
PR I VAT E ..... - ·· ..··.. ·00-·· ....... :. 02 ---.-.-- ·-·--·---~·~R·E·~·~---· .-.
J:NfRY ADDRESS
TOTAL LENCTH
......
--J
.--.-----. -
...l
I
00
....
,
I
._-_._._----_.- _._...-----_.._.
--~-~.
.
02
•
,
--..-_._------------_ .. -
.,,------.----- . -.. -..-... -.... .. . . --.. -----.---..--.-.. . '-' ··--·-1
-_._-_._--_.
---.-... ... _.--..._ .. -- ....-._ .. .:._---_.. _------ .. :_._------------_. ---_._----_._ ... _...... _-_._--.-_ . _- ... ---------_... _--'
_..... _----_. __ ._- _...•. _-_._---------------_.
__
.- ..
_-_ ..__ __.....-._-------.
i
I
-.- .-
"--
--------------------_. -_. _--_._------ -"---------!
. -..-.--..- . .-.- -.. - ... ......... . . ·l
I
-.------------~--.~--~
..~
~··
_------_._---_ ---- _.-_.... _. -- _-_.._-- _ _--_._-_. _.. -_ ... _._- _.._... - -.-'- ---·----1
!
---.--.--...---.- ....-.---....
--_.. _---_ ..._---------_.
I" ....
..
.
....
..
"--------~
."
.
--.----.-----~---
______. ___.______! i
- .. -------_._-
_....... __ ... _._._----- _ ... __._-----_ ... ----.-
-----
....
__ .. -'-
...
.. ... -....
-- .
.
... -.... . .. ........
'
"""
..... -...
-.~
- .. - - - -... -- .. --.. -- ....... - ._ ..... -_ .. _ ....... _ _._ .. - '-"·---"'1
----------_ ... _-----_ ... _-_._;-.---------------_._---_ _-_ _----------_.------_.......... - -_.- - ....... __
--_. .....
...... -_._.-.- _... _. -'- ...
-_._------- .-- ---------_._-_._-_.
-------_ .... - .._.
-~-
..
... _-_ ..._---- ...
.-.--------.-~----.
._----_.. - ._.-.-._--_._ .. _----- ---. --------_._-••.._---------_....
- - .... ---.--.---..
..----..-------..
"
...... ..
... -:-
_._ .. _------
__
...
--'----~_.---
-- ...-.-.... -- .. - .. -- ..
...
.-...-.....
-~
I
~- .. - --~
---------------------
-----~
.._. __ ..
-----_
..
_.. _..
-
.'
..
~
.
i
.-.- . -.. -.' .-.- ·······--·--··i
:
...
;..-.
···· . ·. i
-- ..... _.-" ...... -.- --". --- -_ ... - .... - -
.---~--.-----.-
----- .... -----.--.-------...
_.. _-_._----_._--_ .. _-_._-------_.
lOACSET.REREAO
IEF285I
. I f'F 2
.. _____...... '
as r ....- vnL
IEF2A51
IEF285t
I EF 2 8 S I
If7F28')1
-IEF28sr--'
VGl.. SER NOS: 000017.
SYS1.FOIHLIO
V() L S Ef{ NOS = 00000 'j •
SYSC(Jf
'VOL SER NOS:'
__
-
C'-
----.- --.--~.~---.--.~:-. - ~.-'._--_._---....
...
DELE TEO
-~-- -.. -.- .
S ERN 0 S=' 0000 1 1.
AAAAhAAA.AAAAAAAA.AAAAAAAA.AAAAAAAA.00000042 DELETED
I~F2851
-~
-------_. __ _-_._ .. _.._-----_. __ .- -..
-.-.------ .. -----.
•
-~-
KEPT
--
sy sou r
. _. ___ .._____..____ . _____
-'-'-'- .--_.-.
---.-.-------------..
---- --.---_. -
--------- ..... --.--
--.. ---.-. -
..
~------
---- .... - ----.-. --_.- .-- .-...._- ... -----_ .....--
_---
------ --.--.---------------------------------
.
_.- . __._._---_.---------_.
.
----_ .•_----------_.-----_._.. - _.
-----------------..
------_._-_. ----_...
-- ...
-.-
-------------------------------------------~-------
_-_ -:----------.
•.
.--.- -..__ ........_-- ..--------_._---_._---------------_.
-
...
__._-----------------_._--_.__ _._----_ ..._- ._._
.•
.. _-._-
------------------_...
-_ .........
__._--- _-------..
--.. --- ---- ... _._-----------_._.. _'-'-_.'----- ...._---
._-- _..._- - .. --.............. :.-.. -...
....... .
.
-------~-
- --
-........ _._---,..-
........... _..... __ .
- -...... -.. ---.-.
_.-.._-_ . . . . . ------.-1
- --. _.. -...---. _. __ ... __.. _--_._ ... --- .. -----
....... -... -
_.. -...._._- .-...
--_.... ----- ._--------_.-
o
. - . __ ... ______ .. _ . _ _ .. _ . _ _ .-.-. . . . . - - . - _ .. _ . _ . __ , --'0 •• _ ............... _. __
··'0
o·
-·c·
----~.-~-.~~~ ~~~.
-.'~::
------.. . -.-----.. --'. ----.----... . "1-. .. ---~-~
o·
'/~Rf.REAO
Os MSGLEVEL=l
IISTEPt EXEC FORTECLG'-'-
IICOMP EXEC PGM=IEJFAAAO
IISYSPRtNT"f)D SYSOUT=A
IISYSUT1 00 PNIT=01SK,SPACE=(1024,(200,lO»,DISP=(NEW,PASS)
IISYSUT2 00 UNIT=OrSK,SPACF=(fRK,(30,10»
IISYSLIN 00 OSNAME=&LUAO,DISP=(MOO,PASS),UNIT=OISK,SPACE=(TRK.(30.10»
-1 I COM P• S VS t N DOD AT A
... -_.... ...
tEF2361 ALLoe. FOR lREREAD COMP
STEPl
IEF237t SYSUTl
ON 190
tEF231I SVSUT2
ON 192
IEF231I SY5LIN
UN l~l
lEF2371 SYSIN
ON OOC
00000010
00000020
00000030
00000040
00000050
.
.. ...
.•.....•.••...• _co...••. __ ..
.
..
.... .......
.
..
•.•... _ ...... '. __ ........1
I
.
I
. ""1
f
_ _ _ •••• _ _ _ _ _ _ . . . _~ _ _ . . . . P _ _ .. w _ _ _ • _ _ _ _ ~ .. _ .
.
,
-.. ~.~-:~~__--. _ .~_.~~~ . ~ : . . . . .
.J
. . _. ._~. ___.____.___.___.____._.~_. _______-_J
j
..
- -- ..~
I
-"--'--1
.
. tEJOOII COMPILER OPTIONS IN EfFECT: SOURCE,MAP,LOAO.AOJUST,PRfRM,SlZE=45056,LINELNG=132
"4'
••••• _
• . _ _ _ . _• •
_
... ~_ _ _
._~.
_.
____
...
_
P
. . . _ _ ........._ _ . _ .
_
...._
I
--1
•• _ _ _ . _ _ _ _ _ _ _ _ .... _ _ _ ._ _ _ , _ _ _ ,
i
!
- ..- - - - - - - - - - - - -... --.-.-... --.- ....--.---.--.--.--.---.----.-.... -- - ..------...... ------.-. ---·----..- - - - · -.. ·-·-·1
--- _._---. - ----..... - ...---_..
__._
.. -
.'
------_ ...---. -- -- .---..
. '.
I
_... _.........
.... _ ..- ... --"
)
. - - -..-------..-.--~
'.
-
""
.
- ..
-
....1
'
1
...........
~---.------.-.-
-----.- _.-...._._ _...._.. _--------.- .... ---_._----------_._----------_.
..
._-_._._-------_._._--._.
--'-'---"--'--"
.....
"'-'
.... _.. __ .. - ........
. ... .. - .. -.-.- ...
,
i
.. _---: -~----------!
....
'-"'-
__
-----.------~
..
. ..- ......
-.--
_._--------------------------------------.
-_..._ _._---.- .......__.__._----_.._-----------
....
-.-- .. -
..
.
,."
------------------
.
-~.-
.... , ....
....
__ __ ._-
---- ..
~
,
._------------_._ ...-
- --------_._.. _---._---_. -----_..... _-_._-
......
-- . . . _.. _.'-"- .. -....
1
-."'-~--
..- _... _. __ . -----_ .......... ---_ ... --- _........ _
...
---------_._--_. __._---_.
1
.- -
_...... __.-... --_._-------_... -.----- -.. --
." .- .... _....
..... .
----- .....
-------------~
i
_______.______. '" ___________________. ______ ..-_0_--... ·_- __ ...... _. . . _....
.. -_......
~
.-----..---..
,------------~---- ---"7.-~-
-
.....-.. --:----.....-. .•... -'-"--' ... _.. ... ---_._--_._.
- -.--~ .........
-
--~
....
-~--
.. --
.•.. _.
__ ._--- -_ . --- -. -_. - - ..- ....... - ...... .
_ • • • o_ . . . . .
.!_
i
,:".~
.•. ____
J
'i
.
I
__ __ ...._--------------_._-----..- _---_..-----------------------_._----_._-----------{
.
..
..
o
o
o
--........... ,...
~-
. -""'' ' .-. . --.. . .-.. _. . . . _.-_. -.. . -.-- .. __ _
-----.---..
··O·.~
.
,.~
-.- ........
~'
.-.------~J---
... -
................_.......
•.
..
.. --
. ----..
----.-::-~--
. .--..
S.0002
CALL
~. OOO'~
..
~EAO
$.0004
.-~.
REREAD
(5,1) A
....... -.......... -.----.... -.---.---
S.ooon
061961WN
...... -------..- - -
061967~N
FOR~AT ·(3F6.2,I4)
R F. .\ 0 (19 t 1 0) X , y , Z ,J . .. --.-----. - - _ .._ 1)0 5 1=1,10
5 WI{ 1 T E (b t 7) X , Y , l ,'J .-..
1 FORt-1AT (' ',3F12.2,II0)
S T UP
..... ----------.-.
10
S.0009.
5.0010
S. 0·0 11
5.0012
5.0013
5.0014
061961WN
----_._------_. - ....... -- 061967WN
REL
A
ADR
0001BO
.......
NAME
REL AOR
REREAD
000168
c=:~_Q~_':t_~!!.,_~:..~.~U} VAL~NCE
_ _ _ -._-_.
REL ADR
NAME ..
TAG
...
...
_.. - .. _.... _-----_._--. NAM E" •.--.... _.
REL
.
- - - - - - - - ----............-.-..-.-... --'---.. -----•.•.- .
----------------.
'
REl ADR
NAME
TAG
......-
---I
REL ADR
y
, ........ :..-- .-- .. _:o-__..-_ .. ~ __ J
OOO~Ae
NAME----.---.
~_. :E-l-·~::------~--~~~.~.--"-=
...... -.- .., ....-.------.----............. --..............- - - - - -.... -----.. --
ADR'
_------,------
i
EX:::N: ~:ReF ::::te-s~~~~A~·R-
......... - ....-......---------------- ... CONSTA·N'rS""··
~'MME
. ..
)
0001A4
X
OOOlA8
.... _... -.- ... --_.-.---. -.. 000164 ...- - . --.-------.------.--- ...
I
J
000154
'l
TAG
NAME
"'I.
061961WN
.------. 06l961WN ---.------.. ------.. - ...
END
TAG
Loac'
061961WN
S TORAC.~___~.A~ _ _ _V!,~. I A8Le.s_~.!_~~.~.~.
NM~E
..."
-------.-.----.--.-.--'.~-
'-~06l961WN-'---'
00 ] 1=1,10
3 WIUTt: (6,~) A
"S.0007
.. _ . - - - . .
061967WN
t FORNAT (20A4)
S.0006
.-------
06l967WN
·----:---·061967WN·-
2 FORMAT (. ',20A4)
0005 ...-------.
-"
"-'-"---'--.--DATE': 61.302
061967WN
MENS ION A (20 ) ...--...---.----.--......-.-----.-...; . - - - - .- - - - - - - - - - 06196 7WN-
or
'''·s.ooor·
. .----..-.-_. --' ........ 0.' ,..... ---.---.-... _
-----
------- ......-.-
_l:.~_V..~.~.!"'}.J.U~.66.~ .. ___ ._.. __ ...._...,..._.. __ .. ___.____!.~.~ ...Q~!~.6.~_.~.~S.. ~.~~q~!.~~~~V~~~~~.~!_I..~~!}.~.~ ,_ __
C 061961
-~-~-~
.................. -....,_ ......... _......-. :--·-1
.--..-----: .... --.-r--:--·------~r
..- - - -
- NAME·---·--···------ RECADR-NAMC-----;--· "-REl ADR··.
NAME'-- .-.- ................. -.. REL AOR
···~ ..
-:--:I
. ..... -...... --.
"..--.-.-_._-.-."' ........... ---.-- ..... -:.
... --....... -......... _. . ., .............-......": ..·. :-1
... _-------_ ..IMPLIED
- . . ... ..EXTERNAL
.. ......... REFERENCES
-- ._ .. -.. _.. _. __._--,
.----------.-- .....---•.-..
-'------~
-.
..-
----_ .•-- ...........
-
,--.-~..
.....
.
..
NAME
REL
IOCOMN
OOOlFC
STATEMENT NUMBER
,~..
~
00'002
00005
REL
,
----...,....-~-
AOR
ADR
NAME
REL AOR
NAME
REL ADR
.-
STATEMENT
oootce
00001
0002B8
00007
SIZE OF COMMON
NU~BER
PROGRAM
OF CaMP I
LA T I ON
..
_.,
......
00003
000248
._ ....... _._._ .. ~ .• _._.
_._ ...._..._._-_ .. _. __ ...
STATEMENT
. __ ..
.. - ... _.. NUMBER·
.•.. - ....
.-- _.-._
__
--STATEMENT
_.---_NUM6ER
.. -- ..... _...... _- REL.. __ .. ADR
..
AOR ..
000104
000lE8
000000
END
REL
NAME
REL ADR ._
.................... _ •.. ·.M ..... :_ ..
...
,
,
_-_
00010
_-_..REL
ADR
...... .
...._... -.......
OOOloe
..-
---.,
..
000182
MA I N·----..·---·----..--·--·-·-----··--·-..·-------·-·
• ...... . ..... ...... ........
.
- ---.-...-.
-- .... ...
--.~.,.-
.........
- _..... ----...•
. .. ~
........ ·7 - .......~~.:-.
>
. . '. .
.
'" ,,_w'.-"'; ...
,
'::1'
.._._-_.-.._._.-.--_. __. - .... _............. _.-:-,.--:---:- -.--.-.
-~!
"1
.
:mm· m:n~ ~~!~~ ;~~
A
A. AAAM M A. A 4 A A AAA A. 00000049
::::~:
-"- •...
-~-"
..
~
....
..•. _..
~
_.. _-----. _....•...
- .. _--"-.--- _.-.-._._.,_..- ... -, ........__ .. '.0._ .. - ---..
~.--
__......_-_._----_..--.-- -- .......-....
......
-~
·_--tl
-R)
-"---'••- - " - - - ' - - - - ........ - ... -.... " ., ...-
I(F285I' VOL SF.R NOS; 000017.
"J
- . . . - .... - " - - - '
. . . . . . --... - ......... _.- ..... - .. -._.-.... .....
.
. . . . . .IEF28S1
hAAAAAAA.AAAAAAAA.AAAAAAAA.AAAAAAAA.00000050 DELETED
IEF28SI
VOL SER NOS= OOOOO~.
lCF2USI
LOAO~r~EREAO
PASSED
IfF285I
VOL S~R NOS= 000011.
IILKED ExrC PGM=IEWLtPARM~(MAP~lET,ltST,DC),COND=(5.lT,COMP)
00000060
- _..... --. ----_ .._----- ... - ... _IISYSLtA 00 OSNAM(=$YS1.FORTLIB.DISP=OLD
.. 00000070 .. - ..... _.
IISYSUT1 00 OSNAME=*.COMP.SYSUTl,OISP=(OLO,DELETE)
00000080
IISYSPRINT 00 SYSOUT=A
00000090
IISVSLIN 00 DSNAME=&LOAC,OISP=(OLO,OELETE)
00000100
II
L> 0 0 r. N AM f = S V SIN
- ...._-. --. - - - - - 00000110 ...----------.-----.
IISYSLMOO 00 OSNAM~=&GO(MAtN),OISP=(NEW,PASS),UNIT=OISK.
XOOOOOl20
- .. -..... _.. _..- -.-- --.....
_... - ... ----. _. __.._-" .. _- ......-. -.. .. ---....----- ....... - ..
II
SPACE=(}072,(20,10,1»
00000130
IEF2361 ALLOC. FOR TR[REAO LKED
STEPl
_._._---_.- ..__._--_._--- ..._.. _-_.-. .... - ..._. ._------- -..
I[F2311.SYSLIH
ON 192
I~F231f SYSUT1
UN 190
.. - ..... -- - --.---..
IFF2371 SYSLIN
ON 191
IfF2311 SYSLMOD ON 190
---_.... _..... __ __ .__ ._ .....
-"'-'
~
----.-.------~-----
.... -
......... ---. '-T'
.
~--.-----.---.-.-----
----
-.--- --... -_. -_._ ...... --_......... --..-." -. _...... __.----.. ---..
__---_ _-
"-"--"- -...... .-._,
_-_.
.~.-.- ~.~~
...
__ .. _-_._------- -. -.--------.------
.
.-------.-------------------------------------~--- .. ~-
.. -
_
..
------_._--_ .._.• _ _-,._------
..
..
~-.--
__
__
. "-
.._-------
_.. ---------._---_._--_.
_-----------',-----------------------------------------------._-_._..-_.--_.
._._ .. __ .;:-- _-_ _--.-_.
...._-- -.............
.
_..
-. --.. ---.. ---------------..--
..
......
~-...
-.-.~
--..---- -.------r--------._'-'.-.--" _._.- .
---.-._-_.:._._-_._-------.
- ..
.
"-'-
..
-.--.--- ........--... --- .. _..
_.. _.__._- _._ _--_.-._
..
....
..
_---_.
._-------_..__ .._----_._--_._--------_._._.. _._ ..... _---._... _---..............
_._._-_ .._-_._. - .....--------------.-..-----.
---.-------~-~--
........
-.------.~--
..
- ..-........
------.-- - ...--........ -..
.. -.. -................ _.......
. -- -
~
-
.. -.- ...... .... - _._ .....
_-
----------------.--~-----------.-----
_ _. _ . _ .... _ . _ _ • _ _ _ _ _ •
_ _ _ _ _ _ _ _ _ • _ _ _ ..... _
•• _ . _ _ _ _ • •• _ _ . . . . . . . . . _
••• R .. _ _ _ _ _ . . . . . . . . . . . . . . . . . . . .
-_._-_. __....._........... ---_._----------_ ... _._-_._._---_ ........_--_ ... _._--_... __ .... _.. _._._------_ .. __ .. _-
-'-.·.-·~O-··-·--· . --.-:-. .-.---.. . . . -.. . ----- . ~-. .----.-..-.. . - - . .-..- . . . .----.---"-.- - . -.-----~
------:--.-._--.. ---_. . --.. . . --------_.-..---..--O-··-.-·-..· -.
~
..:~·~·-··-·
i ,...
~..., 0
.
-.--.--~:--....
. - --.. . . . . -.--..--- --.----
. .......
'"
-.. ------::------------------------. -0 . . . --' . ..:~~
0
_ _-_ _---------
.-~ ......... -.. " '.......__ .__ .- ..•._._ ..... __ ._ .. ...
....•..
..------:--~--.-.-. -----.~
..•.
..
,
E-LEVEL LINKAGE ·EDITOR OPTIONS SPECIFIED MAP,LET,LIST,DC
-.-*.·*MA(N .. _....... - DOES NOT EXIST BUT'HAS BEEN 'AODED 'TO DATA" SET-....
_._ ...._...... __ .- ...........
,
~
.
-------_._-_ __
....
.
.....
--.. - .... _........ _--.. - ...... -. - ....-..
ORIGIN
tlfd N .
..
_--- ... - .. --- -.
--- ...
..
..
...
.
_ _._-_.- ._._.,-
...
----.-
•..•
.....
ENTRY ADDRESS
TOTAL LENGTH
'.
'.
_--_.._-_ __.-._. __ - _-...- _----..
'
..... ...
.
....
MODULE-MAP
-- .... _.. _--_ .... _...
_........
._------_ _-----_.,----
-
..
LENGTH
00
NAME
'LOCATION
. NAME
NAME
LOCATION
LOCATION
NAME
LOCATION
lOE
02 - .-.-.--.---....,...----. -.. --.... - - -.. ----.--....
lE8 ..... -..-14.64...... -.. -...-.-----.-R~REAO
._-----_310
..
-
.
--'-~
IBCOMI#
lEa
FDIOCSfI
CF2" - .. -----... --.-.--.-..
-....- ...--.. ._ _ _ _ _ .FIOCS#
.1a~O____
-IHCUATlH':*--'-'~-2548 ----... 138
IHCFIOSH*
_.. __
' _ _. _
ENTRY
....-._-. --_._... _._ ....
-$·PRIVATe·-.···- '310
IHCFCOME*
"
_._ _-----_ _ __ --
CONTROL SECTION
NAME
--.- ..
.
,
.
9F4
--------_. -------_. --.. -
1850
-_ .........., ........ - .... - - -- ..... - -_ .•.__.... -
._._--'
---_..._--_._-.--.--...---
____. ______
00
2660
. -.
_ ....
'"''
_._---- ._- ......... -.... _--------------------_.
~-----.-'-- .... -... ---.- .. -.- ..-.--.-.-~=~~-.---------------_._--_._---------_._----------------_. __ ._------ .~=-~~-~:~--.~-=--~~
.-- .. - ..... - .. _---._--- ... - ...-._------_... _----_._-----_ ... _-_._-. __...._--_.
'-'''-'''.'
---.~-.--------.-
•..--...-... --- --
...
-- ..
,.
-_.- '--' ------------_._------_._--_._...
----~--
..
-.------ ... -
...
_....
-~
.....- -.
------_ ...
_-_ _
.. .. _ ..- .... "---',--
-_ .. _-- _._ .._._---_ .... __ ._-_.. _.... -
_._-_. __ _-_.-
.
..
~.-
-
........ -............. ...........
-----------------------.-~.----
_... ----_
..... --.
--- ---- --. '--- "-'--. _ .. ---- .. _0_ ... -_.'.- ...... '-
,........... -.....- --
.............. "-" ............ -_ ... __ '_'1
'I
l...:-_ _ _._._._._ ...... _......__ .... __ .. ___ .... -.' .... _.... _ .. - .......,----__ .• ______ .. _. ___ .. _.. ____ .___ . ______________ .____ .. _ .. _______ ._ _ _ _ _ .____________.
.. -- ...--- -- ------. -.-. -
..
~---
--------.. ----..--·------iI
- -. --.-. · - 1
. .- .... - . .. .
. . . . . . . . - -. . . . . . - . - - • • - - . . - - . . . -
.. - . - . - - ..- - . - . - • • - - - - - - - - - . . . . . . .
---~
. . . - -. . - . - - . - • • -
.. - - - - - . . -.-.---.-.------.-~ . . - - - . - . - _ _ • • 4 . . . . _ _ _
. "'1,
---1
~--:-- _ _- - _ - .
_ ......
•
-
• • __
-
_______________ • _ _ _ _ _ _ _ _ .0 _ _ _ . _ ......... _ _ _"
.... - ........ -~...
•• _
i--~.-
-..
IEF28ST
SYS1.FORTLfB
SF.R NOS=·000005.
KEPT
AAAAAAAA.AAAA~AAA.AAA~AAAA.AAAAAAAA.00000049
DELETED
-----------------_.- --
TEF28~r-""VOL
I~F2851
IF.F205I··
IfF285I
IEF285I
IfF285I
t EF2 0 r -. .
IEF2051
'VOL SFR NOS= 000011.
[tF2851
VOL SER NOS= 000011.
~
~,
SYSUUT
VOL SFR NOS:
SYSOUT
LOAO.TRf~EAD
DELETED
GO.TRE~EAo
PASSED
VOL S f R NOS = 000 a1 1 .' . . . -......-.. -- .. -... ".-- -. ... ..---
I/GO EXEC PGM=*.LKEO.SYSLMOD,COND='(5,LT,COMP),(5,LT,LKED»
IIFT01F001 00 OONAME=SYSIN
/IFT02FOOl DO lJNIT=PllNCH
//FT03F001 DO SYSOUT=A
IIGO~~T06FOOl DO SYSOUT=A
IIGC.FTP)FOOl DO OllHMY
/IGO.FT05f001 DO DArA
STEP 1 I(F236I ALLOC. FOR TREREAD GO
IEr-2311 ~G~=*.OC ON 190
- .. ---..-- - ._......... -.-. --_._.
'TFF2371 FT02F001 ON QUA
tEF2371 F10~FCOl ON OUC
••• _•• __ ._ •• ____ -"_. _ . __ .. ______ .• _ . _ ... _..... _
•. _
••.
_~
_... ----_._-_._--- .......
00000140
00000150
00000160
-'·00000170 ...... ----.---.----.--
__._------_._-
.....
-
..
-
- ---
.
-
\~}
.:...
-- ---~--~~--
.... _ • ___ "_JO
_._ ..- ...... - .. __ .._.. __ ._.- ... -_._.
_.-_._--_._------_ ..... - ... --- .-- .-.--'.
-~
.. ..... --..
~
---_._-- --------------_.
123~5670912345678q123~5618912345670912345678912345678912345618912345610912345678
.__#_-._--- _.- ---.---- .......- _. _...... -- . -.
123456709123456789123456789123456789123456 789123456 789123456 7 89123456 '/89123456 78 -. --.-.---- -...-.-ll34567891234S67891l34567U91234561891234567891234567891234567891234567091234567H
1 2 J 4 5 b 7 fI 9 1 2 34 '> & ., 0 9 1 2 34 5 67 Po <} 1 2 34 56 109 1 2 "34 56789 1 2 "345 67 8 9 1 2 34 5 6 7 891 2 "34 5 6789 1 2 34 5618 .-.-----
.~
1214561891234567091234567091234567891234567891234~678912345678912345670912345678
123~5&78912345678912345670912J45b789123456709123456789l2345678912345618912345678
1l14567091234S6781123456789123456189l23456789L234567891234567~912345678912345618
12 .~ 4 5 &.18 9 1 2 j 4 56789 12. 3 It 'j 618 9 1 2 j 4 5 6 7 U<} 123456 709 12 3456 7 8 q 1 234') 6 7 09 1 2 34') 6 ., 0 912 3 4 5 6 7 U --.- . _ - - ----.---.-12145670<}1?34567U912345610q1234567891234567891~345678912345678912345678912345678
12345670912345670912345670912j45676912345678912345618912345618912345618912345618
1234.~6
7091.l3
4561.89
1234
1234.56
7891.23
4567.89
1234
1234.56
7SQ1.23
4567.09
1234
1234.56
7891.23
4567.89
1234
1234.56
7091.23
4567.89
1234
----·1234.56· .,. 7091.23
4561.09
1234 -.------------------------.-1234.56
78~1.23
4567.89
1234
1234.56
7H9l~2'3
4561.89
1234 .- ...... --- . ..------.------.----.-.--.........- ... -- ...--1234.56
7891.23
4567.89
1234
1234.56
1891.23
4561.09-' 1234 ...... -.-- .. -.----.--~--
---._---_ .. _._-_._--_...._--_._._ ....
-,
....- .. ....
_
-".' --'. .0.-- .- . . . -----.-.-.-. .
-. _... _..-.- _ .......
...... - ...... -. _......... _....... _ ...... --- -... -
_. __ .._-----_._-_._---------_. ._- ---' -------_ _--- -----------._- _-_._---- ...--.......... _. . ...
_---- .._-------_ _-_._- -_ _._--_._ _---_. --. -_._------------ _...... __ .._--_.----------------.----------_.- -.... _--_ ... _. __ ._-_ _._---_. -'--' - _. __.. _. -. -.---- ..---- _. __ __ .. _-_. ___ __. _-_.-_._............ ...-..
.
........ .
.....
-".- . -....
..
_.....
..
.....
-~.---
....
...
-
.............
..
...
...... ,--
•... .
_- --
_._---_._._--_._----_.-
.---.
...
......
._-------_.............. - ... -- --- .- .. -- .. ---.----- ---.
----------
....
-
.
....
..
.
~
......
--._-. -··0-----·---·-··· .--:----:----.-. ---._----------.---.----...-0------ .'-.:-. -~--.-'--
-•.. ~TOC'OJEC'f~!oOe-~~lj:_ri~R~io()ti1>~s rtlK1~'f}js ~~
-~·OOOOOO
. 1 REREAD
START 0
·2
ENTRY REREAD
..... _.-, ..... 3
'--"~~''-'-
EXTRN
I JTAeOM
.. ----·-··54 -_.-. __ .............. .: EXTRN
US I NG
IJTflOS
*.
1 ')
000000 5800 F060
00060
000004·',5810". F06't . - - - - - -...--..- 00064---··
000008 5000 1816
.
. 00818·
-oooooe O'lFE:' ......... - ...... :............ ' .......
. ." ... _..
l8Ul
6
7 ......... _......
8
9'
12
13
onOOOE
.
... -
000016 0503 4060 8082 OOOOO'OOOCO
O,AOlZZ
ST
O,~%o.la
14
15
L
, ... _- .. _ .... LM
~--
24
.. 5030'-'S'OAE
00004U 47FO U04()
00004C, 5820 UOAA . . -. -_.......... _...
00008
OOOBe
.
00054
OOOBS
'
000064 00000000
"DC'
37 AOIBCOM
DC
.. DC
41 SV 1
.....~.:- ................... ~- ....- .............. " . ,,-, 42' SV2 .
. OOOOCO,OOFOOOOO
.
43 HOOFO
--._.
---.". ...
".... ...... .............
"... ... .. .......... ' ....... 44
DC
DC'
OS
,OS
DC
. END'
. . ....
............_.. -_ .... _ ..... - ._- ....... -
~@QQQQ@
3,SV2;' ..- - - -........--.--..... ".....\..... - .. -.-..-.-- .. ---~':.----.-.~.--- ...
,
3,SV2
36 ADZlZ
I
~
X@OOFO@
L
LM
.
. .. ;........_......
-.......I
ItADFIOCS
2, S VI
.............-..... BeR'
I
0,1
Be
L ..
... LA ... -.
000060 o~pOOOOO
39 AOFiocs
--OOOOB4'~ OOPOOOOA .. ~--·-.··-· ..~-· . ~ ..· _.. _ ... _
......... _..... -:._.. 40 199'
,
2, 14
15, LEAVE
-"000068 0000000000000000·· ...... _· --" ................... 38 SAVE
000068·
--'00008t"":'
LR
30·
00005S'984F B05A
00068
34,
00005C' . 0 1 f 1 "
-._-.. __....... ,--------.............·....---.... 35
OOOObO OOOOOOOE .. ...... ...........
. 15 t 1
. 0%4, 14D, (99 ....-... - ,
8,BVPASS
2,SVl
'.
.
BCR
28
ST
29 ..·-----·-·----·, ST
000050 5830 BOAE'
OOOBC
32
-'00005'4 4114. 0004 .. ..... ' .. ~..... -...--:--. 00004' .... - 3 3 LEAve
00005E 0000
1,AOFloes
eLC
BC'
DC
. DC
31" BVP.ASS
.
~.
... 4,15, SAVE' ....,...........
l
26
21 ...... -..... ..
.
...... _............... -:
.
.............. --_.-..
H ,Ct-IK 99
'Z5'-"-'--':':BALR'
... ......- _n ....._
000040 5020 'BOAA
I
O%4,4D,HOOFO
.
... .... ------..-.-.. _-_ ..... _- .... -...... - .. ---....---.---.... - -----.----
·eLC
000028 a-1Ft
.
"
20
00002A D~O 3 EOOO" SOA6 ~OOQOO . 00084 ... _._ .. 21 CHK99'-'"
000030'4780 U03E
0004C
·22
-0000.34"': i'82E ...... ' ......... -.. ,_._-,_._.-.-_.. ,.. .
2'3
.... ---.-.... -... -
~·OOOQ·44
I
ADDRESS OF Zll
ADOReSS'OF~IJTACO~'
.......-..............
USING lll,11
4, a
16
OOOBe
11,1
. LR
00000
18
00068 ....... 19'
•
. ..
.~.
....
1
LR
000020 56l0,BOA2:
000024 984F 60~""
000D3e OOFO
00003t: '0000
15, 14
DROP
.. ... ....
0002 A''''-'': 1 '1 ... -...._.... -.....__.-.. BC
000036 5810 DOA2
00003A 0501'
......... --._-,-, .... _......
red. (..~er"~ ),' .
LOAD
1 ,AD 1BCOM
.......... HCR
i"
0 WfTH
(OAD 1 WITH
L
L
00001 C 4 180 801 C ........... _-._....
.......
~~...d
•
*,
........ '''-'''- ....
000014 1840'
0- . •.
10
US I NG
1
. ~
.
.
I
11 ll1..---"· S fM.
4 t 15, SAVE .......------..... -._............__ ._.•..... ··~--··. -··· .. ·-··-·· .. - -.. -··-··- .... ·I
OOOOOE
.
OOOOOE' ·'q·04F"·1·05A--.----·-----..00068
000012
F-"~4- S -*ep-{.7- .
rl2iJ5 f:f-:3'_'?fI~i
.... _. c._I,.~_o:\."les. f yo.:>m .. g,.$ .... e ~.s ;.,.." ___ ~ ' ..
;<1l"
.
............. .
-'06 a a00 .-.. --,..-..
TEeg ~
-
1 t 4%4D'-
4,15,SAVE
15, 1
... _._.-................." .......' .........- ................._...,,--'. . ...-..... _-"..-... - . _ ...... : ....... ".:. t- .. ·-_·,-- _... ..,.--- .,. - ......... - '-'
A%llio
A%IJTACOMD
18FGlO@
A~a JTF I
"'--"
OSc
FG110ial
F
F
XQ)OOfOruuLQ.Gl
;
...
....
SESSION REPORT
o
Session No~
Session
/,
J
__~/~~~/~~~o~'~~~;Tt~'~~c~-LL::
___
- _________________________________
Title~~/~/_>~?~~)
Chairman:
Speaker:
i,(,'..
p /1 Ii
~2
Ip
IJ
~
at'de" l kl..z'~,
o
Summary;
-::-
14/1./
/d (
Topic:
!
5';::/0
II~, ':'
)I;,i~>
.s:c \-.
f
f
,../'
J~,' hd-jr.
c·
". - : [ /
--t/... cft,"b\";: 'Ok! (
C~~"./
.,
. 7/.:-r-z/-~~t:/
Speaker: _________________________________________________________________
Topic: ___________________________________________________________________
Summary: _________________________________________________________________
NO. OJ? ATTENDEES: ~~~~: X
k0
o
50S
"
·IAJ·.~ ' / ~
,'r. .
·--~.,'f
, ',j"
,
o
i
I
I
t
r
I
, OVERLAPPED, PRINTING ~OE IBM 1130 C~MM~CIAJ.I
APPLICAT:!=ONS USInG FORI'RAN WRITE STA'I'FJ.1ENT
by
B. J. SHAIN
THE SHAHINIGAN -ENGIN~gERING COl'1PANY I.IMli"£ED
Box 3010 Stat,ion B,
Montreal,
Cana,da.'
S06
TABLE 'OF
COlrr~~TS
ABSTRACT
i
DISCLAIMER
1i
TEXT
·1.
Use of Fortran for Commercial Applications on the
IBf.1 11-30
II.
1
Programming Philosophy Using COInmercial
. Subroutine Packages
2
III • Difficulties Presented by the Usc of Commercial
Subroutine Packages
IV.
3
Use of Fortran \'lRITE Statement for Output in
Conwercial Reports
-
Definition of
5
Alphabet~c
Literals Using Fortran
8-
,v.
Performance of the Subroutine
9
APPENDIX I .-
Fortran Output
Users-Guide
\-TRITE
~ou.t~n~,
with Overlap .
12
12
REl'..D
o
,
'
SO?
cont'd
Page·
APPENDIX II
Loading the Overlapped Fortran Output. Routine
16
APPENDIX-III
-Source Listings
(
17
.,
~.
,
,
'V
.1
-1- -
ABS,]~RACT
A method is -descr:i.bed- for incorporatin~ overlapped. printing --
into IBM 1130-Commercial FonrrRAN programs.
Communtcation-
to the subprogram which performs a printing operat"ion is
a.chieved. th~ough the FORTRAN WRITE statement rathe~ than
through the CALL statement.
-.
tha~
The advantage of this method _is
limited use can he made ofth~ formatting ability of the
Ji'ORrRAN language.
Headings can- readily be j.ncorporated, and
the layout of the printed page specifica1ly by use of- FOru1AT
-statements.
('
-
'-_ I
o
0 ·"
I
-----...- ..- ...,.
.......... ·......
,-.--~.
-ii-
(
DISCLAIHER
Although each progra.m has 'been__ tested by its contributor, no
_-war:c~nty,
_express or· implicd,is made by the contributor or
- COMMON USERS Group,. _.as to the 'accuracy and functioning -of the
program
..
and
.-
-
related proBram material, nor shall tli"e fact
of
- . _distribution constitute any s:uch warranty, and no responsibj.lity'
is assumed by the contributor or-' COr·1MON USERS Group, in conn-ectlon
therevlith.
o
,-
OVERLAPPED PHINTINGFOR IBM 1130 COMMFA1CIAL
APPLICATIONS USING FORrRl\N v':RIrrE STATE'riffiNT
. I. - Use of Fortran for Conmlercia1
.~ppl:tcattons
on the IBM 1130 •
. - Since the IBM l130.was conceived as a computer to be used for
the' solution of moderate siz.ed engineering and scientific problems,
Fortran is the only compiler .supportedon the system.
inst~llations,
Atmany
it is necessary to use the computer, not only
in this prime role, btlt also in the preparation of. commercial
reports, such as payrolls, labo-r dis'tribution, and time control.
In order to penult the use of Fortran for these applications,
("'-'.
a number of subroutine packages have been developed.
IBtJl' s
--./
o
Commercial Subrouttne Package, Version II, is an example.
are the contributed program packages COHh""T and
I1JE~IJ'
Others
The
purpose of these subroutine packages is to overCOnle the problems
which would
other'~.,ise
cations in Fortran.
be presented in handling 90mmerc.ial appliPrirlcipally
th~se
are the folloi,vinS:
1)
Moving character strings.
2)
Use of floating.dollar signs, check protection, etc.
3)
Elimination of round-off in crossfboting totals
4)
Zone recoBnition
5)
Stacker Selection - -
6)
Overlapped input/output for increas:i.ng processtng
speed.
o
-·s//
,
'--..
_-------------
Use of Fortran - for Cotmnercial A-pE:~~C2.cl,tions on the IBM 1130
cont I d
The first five j.tems' in the It-st 'abov,e exist. ,-on account of' the
fact that Fortran is not intended asa commercia]. dat,B- processing
language.
'l'he sixth -item' occurs on account of the fact that the
Fortran compiler for: the IBlv1 113-0 has been design'gel to conserve-
cor.e space as much' as possible.
-
.
The normal input/output_, subroutines
for use vlith Fortran -use the same_area in core as a buffer area to
contain the image of a card being read or a line being printed. ,It
is th~refore not possible to overlap these opcrattons, although the
interrupt hard\{are provided on the system \'iould other,:iise penni t
~'
(',-..
.
this to be done.
In the design 'of the IBN 1130 Commercial Subroutine
'/
Pa-ckage, it has been recognized that for commercial applications,
pr~cessing speed ~,s~}.unes a greater importance.
Overlap of inp..l t/output
operatiQns h::l.s therefore been achieved at the expense of the ad,d:itional(()
core required for multiple buffers, and for the more
con~licatea
-subroutines Tequired to seryice the interrupts.
II.
Progran1"ning Phi1osop}:ly
Us~ng
Commercial Subroutine Paclzages
The programming philosophy usually 'employed \{hen commercial sub-routine
pack~1ges
1)
are- used
~s
as fo11m'Ts:
Input data 'on cards- is read and stored as a'·card
'image, in EBCDIC form.
i~
2)
A sUBRourrINE subprogram
used for this purpose.
Tests are made
to
deterrnine the type of card
which has been read.
--3":
O~"
-
P~og~~mming
Philosophy Using Commercial Subroutine
Packages
cont'd
"
3)
Data is extracted from the card im~ge, 'and stored
for subsequent calculations,', If 'it is to be used,
in
arithmetic opel'at:io ns, nurr'lerlcal data- is
converted to a code Ylhich permits these operations
to- be performed.
4) "At the conclusion of-the nec_essary calculations,
an j.mage is built up of the' output :Line to be
printed_in the report.
Numeric infonnation is
reconverted to EBCDIC code for inclusion in this
output line image.
(
The output line is printed, using a SUBR01Il'INE-,
c
subprogram.
The use of this philosophy imposes virtually no restriction on
the manner of pr_escntation of data in the inplJt file, or on the
appearance of the final report.
The subprograms' for carel ,
-reading and printing permi't. overlap of these operations. '-lith _
each other. ,Conversion _from card cocle to -EBCPIC is also over-
lapped with card reading.
-III._
Difficul~ies
Presented by the Use of Commercial Subroutine
Packages.
~--,
Al though the use of cormnercial subroutine packages in the manner
described above overcomes the shortcomings of }I'ortran for commc-rcial
-,'
s/a
----~~-~.
_.__ ._--------_...._._.__ ...-... __ .. .. __..__._-------._--_.--_
. l.
-4-
-- to l
cont· d
~
applications, a-number of' additional difficulties -are introduced.
j
1
~
Prinqipally these ax:e the i'ollm,ring:
1)
,i
The introduction of headings into the output
report.
2) -
Loss of traGing capabili ty..
IB1~ 1130 Fortran
has been providedvTith a very useful trace
feature.
This cannot be used \-Then -the overlapped
input/output subroutines of the Commercial
Subroutine Pack[-tge are loaded, as it is not
possible to include the Fortra.n input! output
subroutines- in the sa.me core load.
The Fortran
·"\. !
C
! ;
input:/output subroutines are required- for theuse of the trace feature.
3)
A consid.erable amount of ma.nipulation is often
required to build an image of a line of the output
report.
~onger
h)
This tends to make the coding-of prograxas
and- more prone to errors.
The introduction of masks for editing.
The use
of Commercj:al Subroutine Pacl~age EDIT subroutine
reqtli:r:~s
that non-numeric :Lnformation to be mixed
into a numeric field (such as
$ -.
etc) _be :Lntro-
duced iOn the form of a mask, \olhich is- suhsccluently
.0
o
~.!!f~~}j~~e£i. -E~~E~!'t(:.~..E..:Y:_ th~ Use-=-.
of Com~~~_a~_~u~~::n.rtIne Pa_ckagcs.
--
cont' d
applica.tIons, a· number of JtcJdi tional difficulties are introduced.
Principally these a:re the follouing:
1)
The introduction of headines-- i~to the Ol!tput
report.
2)
j:Joss of tracinr; capability.
IB14 1130- Fortran
has been providccL vlith a very useful trace
feattrre.
This cannot be us eel. \·rhen the overlapped
input/output subrout-inuj of the Commercial
Subrout.ine
PacJ~agc
are loaded., as
j.t
ls not
possible to include the Fortra-n input/output
o
subroutines' in the
sam~~
core load.
The Fortran'
input/output subroutines are requtred for the
use of the trace feature.
3)
A considerabI·e amount of mB,nipu1ation is often required to build an image of a line of t.h.e ou.tput
report.
This tends to make the coding·of programs
--longer and more prone to errors.
1~)
Th8 introduction of masks· for cdi tine.
of COlil.mercial Subroutine
requir~s
Pac}~age
The use
EDIT subroutine
that non-numeriE j.nformat5.on to be mixed
into a numeric field (such
D.S
$ • etc) be :i..ntro-
duced in the form··of a mask, vrhich is subsequently
o
SIS
-----_._"".._----_ ..
__."-_.,,_._.,,...
"""""~-~""~~.-~----~------------
. - - - - - - - - - - -..- - - .
-5" rf~
Diffj"cu:;~.ties
cont t d.L~)
Presented by the Use of Commercial Subroutine Packa.ges
destroyed by the"inclusion' of" the numeric information.
In vie"\-, of the fact· "that alphabetic Ii terals.- canpot
be directly defined in the IBM
~130"
F'ortran".
compiier, this operation is clumsy (Note - Version 11of the IBM 1130 Fortran conipiler includes -the DATA
" statement which overcomes this difficulty).
IV
Use of Fortran
"']RrrE
Statement for Output in Conunercia1
R~R.~ts.
At 'rhe ShavTinigan Engineering Company JJimited offices at Montreal,
Canada, an IBtJI 1130 computer vras" installed in Janu.ary 1967,
pr:!-rnarily to perform engineering calculations.
HOv7evcr~
it was
desired to use the machine for processing the conmlercial reports
which llould be required for the administration of the Company.
_ A study was made "of the available software
fo~
cOITilnercia1 pro-
cessing, and as a result a set -of subroutin"cs "lritten, '·lhj.ch
combined the best features of" the IB!·1 1130 C,ommercial Subroutine
Package, COIvl1"T,,, and IDEAL.
In recognition of the difficul tics vlhich would be presented by
the' use of SUBROurlllE subprograms for printlpg, after the ~9-nner of
the IBM Conunercial Subrout?"ne PacRage, it vias deciclccl ~o adopt a
differerit approach"
A new" subroutine SFI¢( the" subrouti.ne 'fh:tch
handles in Fortrcl,n the :input/ou.tput "functions) "las "Y[1"i t"Len in
"c~
S/6
'.
---".". __..........
"-.~.--"-,,.,,
..- - , , - -
mmn
Ill!!!
TUII!f!M!I!!fW
iTV"
Me,-"';,
-6Use of Fortran vl~~ITE~ st~~ement for Output in ConuUl.~rcial Heports
cont 'd
1130. assembler langua.ge, "having more . limited capabili tie~ than
the _versj.on supplied-as part o~: the sy~tcm, but taking advant-ages
.. of tpe overlap feature of the hardHare.
'l'he ad.vantages of this--
.-:procedure over the tlSe of the Commerc:i.al.Subroutine Pacltage
subprograms are the following:
1).
Head.ings may be introduced' by the use of a FORMAT -
-
'statement, as is done in norfllal processing using
Fortran.
Tracing can be employed.
The amount· of' m.anipulation requj.red· to format
an output. line is reduced, as the FO:ffi·l~T statement
may be used for this purpose.
,. It yTaS not considered necessary to incorporate al1- the features
. - 'Which are supported by the normal SFr¢. subrout:l.ne as supplied
/'\..
-in the system.
Indeed to have done so would have resulted ip.
a su.broutine oc<;!upying a large core space, a consiclerable -portion
of "Thich \vould have been ·wasted. on features vlhich would be rarely used ~n the COlI~mercial area.
The follovTine restrictions were
therefore iI}troduced: .
.
,
1)
Only integer variables and arrays Ihay be-included
in -the output
2).
list_~-
-X'ONg 1'lORD ·-Ir-fJ~EGgRS option must be used.
o
Sf?
----....-.....---.---.--~......... ....-.....-.~~..~~---------------------
..
.
,
'
-1(
Use of }t'ortran vlHEL'E Statenlent for
3)'
Ou~put
in -Comrnerctal -Report~
Card reading is not 'supported, and all ·wr:i.ting
is done on the 1132 printer, regarCUess of thedevic~ cod.e appeari-ng in the \v~ITg staternent.-
Card, type"i'rrj_t-er and console input/Qutput are.
. haru:l~ed by SUBROUTINE E}ubprograms' as in the IBM
Comrnercj al Subroutine Fe,cka-ge.
11.)
Only the follo,\-line; format field specifications
are recognized:
A
. for alphameric information.
The only .-
field widths recognized are Al arid A2
Anythins else is interpreted asAl.
I
X
H
for integer variables
·for spaces
for printirie alph$.betic - information.
form using quotes
ma~
The group repeat, e.g.
6)
Carriage control characters are restricted to
2(lHl,12A2), ie;" not recognized.
the follo\-ling:
Blank
. -Zero
1- 6'
The
also' lYe used.
5)-
Print on next lineSkip 011e line before -printilig
Skip to channels 1 to 6'
o
cont 'd
-8.,.
·r1l>\
~--
Use of Fortran \1RJ:ll'E Statement for Output in Conunercj.al Rep?rts
cont I d
None of these restr:ictions prevent the, use of -the system in
association uith Commerc1al- Subroutine Package,
C_O~JfE,T
_or IDE.A.L.
An attempt. to pi"int a real varia.ble "Till resul tin the outputfie.ld being filled with asterisks..
tracing to j.nteger variables
only~
This restricts the use of
l?eal variables· appeari].1e; as
a- line of asterisks. -It is our experience that this is not a
serious restriction, as it is the values of integer variables
which determine- the logic flmv of a program.
If it is necessary
to determine the value of a real variable for tracing purposes,
then this can be done by converting it to ]-;}3CDIC, for display
using a
vlRrrE
statement temporarily inserted in the program.
, Definition of A~habetic IJiterals Using Fortran HE..A.D Statement
In order
to perrrri t the d.efini tion of alphabetic literal character
strings, a feature _"las introduoed into the re-,\"ri tten SFI¢
-
to permit this to be done using the Fortran READ statement.
The manner of achieving this is sho"1n- by the follO\.,ring example:
DIMENSION IA(13)
RF.AD (99,101) IA
101 FOHIv1Nf (
'Tfirs
-MF.i:)SAGE TO BE STOHED f_)
- These- statemcl)ts cause the EBCDIC equi v$~ent of the ch8.racters
containerl- :i.n the ·li'ORtvl.A.T statement to be stored in IA.
- fication to the compiler is required.
No modi-
Characters are store(1
--
~
-----------,.•- - -..
,--~~,
-.--,--." ... .. ----...
"
-,~~---~----------
':'9-
,.""
\lJ
1
Defini tion of
t~o
Alphabe~:Lc ~?- terals U~ing
per ,vord.
}i'ortran
~E:AD
Ste.:!:-_cment
cant 1 d
If it -is desired to store characters onc' per '\-Tord,
_then apprOl)riate, spaces should be introduced in the f:Leld.
to be used for editing operations.
V.
Perforn1ance of the Subroutine
Using the Disk
Ivlol1:tt~r
System, the subroutine is implemented by
deleting from the dJ.sk the version of SFI¢ as supplied "\-lith the
system.
,..--
The subroutine
PRN'l~Z
must also be deleted.
The sub-
routines SFI¢ and PRNXX, "\-lhich form the overlapped output pacl~age
".
are then substituted.
Note ~hat SFI¢ can be stored as a subtype
3. prograJu, '\-'hich permi t_s j.t to be included in SOCAL overlays.
The subroutine PRN'rl supplied by IBM is also reqLilred as it is
used to drive the 1132 printer.
li10rtran programs should contain
the record
*IOCS(1132 PRII\ffER)
or
*IOCS(1132 PHIN'rER,DISK)
depend.ing :upon '\vhethcr disk- 'input/output is requir-ed.
The
-principal benefit to be derived from using-the method 'is the
simplicity of coding.
Also the processing speed vhi.ch can be
achieved using this subroutine is apparent.ly identical to that
achieved usj.ng the overlapped output subrou.tines suppliccl ,·[i th
I.
I,j·
. -10,:.
Perl'ormance of the Sub rout ine
cont 'd
. the Commercial Stibroutlne-_Package.· The· core requirenlent for·
_SFI¢ anq. PRJr'lX is "Torcls 1.t·70,-· -:Ylhich l.S some\.That in excess- of
the equivalent subprograms supplied rlith the Cornmerci8.~ Sub-
routine ~ackage.
However, tnis is offset- by the slIk'1.1ler core -
requirements of- programs rThich use this method.
As an example,
one of- the sample programs provi~ed '\vi th the· COlU.L'1lcrcial. Subrout;ine
Package was modifi·ed to su:i, t_ the ..Q.Y~rlal'ped input/output sub. routines.
The processings.pe.ed was· the sa.me in both instances.
Despite the smalle-r and simpler main progra.m "Then using the
overlapped Fortran
I/O,
the total .c:ore requi.rement
i-s
somevThat-
larger, as ts shovH1 in the tab-le- below. . For· comparison, the
.
.--
, core requirements and pro.cessing speed using the normal Fortran
input/output subroq.:tj.nes as supplied with the disk monitor system
·are also sho'\vn.
Comparison of Typical Small
Cormllercial AWli.c£\tj-on
csP I/O
Main Program Statements
Main Program Core Requirement
. Total· size of Core· Load
Executio n Time
o
102
94
··978
850
5006
-5298
.1
32
min sec
Stanclard
Fortran I/O.
Overlapped
Fortran r/o
1
32
min sec.-
83
"'
-
800
4858
2
1~5
min sec.
··S:LI
_ _ _ _ _-'"-_ _ _ _ _ _ _ _ _ _ _ _,_..-.-.....,UO:,.....
W""".'L~·
.
,',~,
·1
r",,:','-
f
! f
, '-11-
Performance of the Subroutine
'cont t d
The problem selecte4 "\-Ta's _the" invo~cing problem vrhich is included
as SampJ_e Problem 2 J_n the IBM:. Commercial Subroutine Pa.ckage, ·Version II •
. It is considered that for programs' to produce la-rge and complicated
cc:>mmercial repo:,cts, "\-The~'e the size'of a core load becomes - important,
the total cQre load requlrement of programs using the output method
descriQ.~d
here "\-Tould be smaller than \TGuld be required for the
-
same programs using the Corrnnerctal Subroutine Pac$age;
o
I
,
·
O
I·
S~"
I' W'Z!i¥l!T-l!ff wn llrr-nT1!!fI
2f!P-fW!'..-rZ!t!:r::r::
APPENDIX I
'.
C.
"'\,I
J')
]'ortran Output Houtine, ,,,i th Overlap
Users Guide.
This "is a program to' replace t!l.e IB~1. ,libr~ry program SF'I¢,
.'Whic)1 is called for Fortr.an input/output operations.
Its
'. purpose is to permit overlap of' printing, card reading -and
processing, for commercial
appli~ations.
Ii'or this reason,·
pei~
missible devices, allo1·rable variable types and format field .
specifications have been restricted.
Only integer variables
and arrays, are recognized. *ONE \-IORD INTEGERS option must be
used •
. rlRITE
~ll
writins is done on the 1132.printer, regarcUess of the
devic-e code appearing i-n the \-lRITE statement.
If it is
desired to vTri te on the Console Printer or punch cards then
an appropriate S:tJBROUTINg subprograms must be used.
Data to
'be printed must be stored in EBCDIC or integer form as an
'integer variable or array._
The follovfing F9HMl\T field specifications are recogniz-ed: .
A·
.for alphameric information.
Th~
only
field widths recognized are A1 and
A~.
Anything else is .interpreted as Al, ",ith
no error indication~
I
for integer varlablp.f:
.-13.,
o·
cont'd
x
for spaces
H
for prlntj.ng alphabetic. information.
The form U.stilg quotes e.g.
~_THIS MESSAG~'
Field repeats .e.g.
The group repeat, e.g.
may also be used.
l-2A2
2(lHl, l2A2), vTil1 not be recognized.
The max-imum nur;lbe}" of characters per line is 120, excluding the
carriage control
chara~ter.
The use of E or F field specifications
will cause the field to be filled ,..,i th asterisks ",i thout
further indication of error.
The first character of a line 'fill be treated as a carriage
control character.
The follo'\vine;characters are
blank
print on next line
o
skip one line before printing
1 -
6
skip to channels I to
recogniz~d:
6
-
All other carriage control cha.racters
are treated as blank.
The follo'\'ling errors vTill cause _a program stop ,.;i th the error
code displayed in the accumulator.
EEOl
EE02
.·line exceeds 120
charaC"'~ers
invalio. forina t- type·
-0
"if
j"'"au
o
RFAD
No read statement, accepting information froni an external medium
.is inclucled.
1:;>~
Appropriat.e SUBHOUTINE subproe;rams must
-for input from the card reader or console
keyboar~.
use_d
The
fol101·r1ng stateme:1 t is provided in order to generate alphabetic
literal array-s.
REI\D(99, n). IA
or
READ(99,n)
IA(I)
RE.4D(99, n) (IA( i), I:::K, L)
RFJill(99,n) ((IA(I,J) ,I::K,L) ,J~I.1,N)
(c
",hich transfers H-type fielcls from format statement
,variable or array lA.
'n'
into
The characters \-rill be stored in EBCDIC
form,' 2 characters per viord.
in the FORHAT statement.
Only. H-type fj.eld_s are permit:ted
If the nurnber of
char-acte~s
in a
_field is odd, the rie;htmost character of the last vlord. is filled "lith a blank.
filled.
Characters Hill be moved until the array is
If' the FOH1. 1A.11 statement is exhaust"cd, - control
return to the last open bracket ~
will
The foll?"lin& e"rrors \·rill
C1lasc a pro8ra.m stopvith the error code displayed ·in the accumulator.
EE02
format not II-type
EE03
device code not 99
This rout ine uses the DJN library subroutine IJRl'TTI.
o
Since printing
occurs :tn overlapped mode, a call to ION]) must be made before
a
PAU0E or STOP
statement.
-T"
-15-
READ
" I
cont'd
The inptit/output routine requires ,the use of -X-I,OCS(1132 , PRINII ER).
SinceCOl:l1nunication to the card-reacler, or typc,'rriter is th:rough'
CALL statements to subprograms vlhich load the approprj.ate card
or type,?ri tel"_ subroutines, then CARD, TYPEHRrrEH or h'"EYBOARD
must not apl)ear in the *IOCS re'cord.
Tracing may be used in _conjunction "'~th this overlapped, I/O
routine.
Since real variables are not c~nve::cted by" this routine,
they vTill appear as asterisks in the
~racj_ng
output. _
o
A~PENDIX
II
0Us;i..ng the disk monitor system, the folloi"ing procc_o.ure is
required.
Load the oyerl~pped verslon of SFI¢, ana, PRNXX.
Note that SF1I?; carl be designated· subtyPe 3 (punc!l 3 in column 11
of -l<-S1. 0RE. cardY', and hence "can be includ.ed in saCAJ.) overlays.
1
,-
PRNXX cannot be included In overlays.
Core Rcqu:lrements:.
SFI¢'
PRNXX-
-
456 ,words
II
..
The IBM supplied subroutine PRNTI and lf5tJP from the Com.rnerciaJ,
Subroutine Package are also required.
o
------
....---'------.;....------,--------_._------,
..
II ASM
*LIST SOURCE PROGRAM
LIMITED CAPABILITY I/O SUBROUTINE
;
*
WITH OVERLAP
*
LIBR
----0000
22189580
ENT
SFIO
, 22645100
0076
ENT
SRE-D
0095
229998CO
ENT
S~1RT
0013
22256240
ENT
SIOl
OOlF
22256267
ENT
SIOIX
0026
002F
0033
01C4
22256049
" 22256180
22006517"
176558E9
0000
0001
0003
000'+
0005
0"
01
0
"0
01
1000
66800003
7212
6.AO 1
4COOOo"07
*
SFIO
*
ENJ
510AI
ENT
SlOP
SCOI\1P"
ENT
E"NT
o
PRNT-l
MAINLINE PROGRAM
NOP
*
LDX
12
MDX
STX
Bse
2 18
2 *&1
L *
INITIALISATIO~
SET XR2- TO - LIB F& f
-RETURN TO START OF PROGRAM -
END OF MAINLINE INIT1ALISATION
CALL ROUTINE
_
RETURNS
TO
MAIN
PROGRA~
T~ ~ET
'*
*"
NEXT VARIABLE DESCRIPTION
CALL LD
LOOPS
TEST FOF I/O OF cor·1PLETE
S
ONEl
ARRAY
BSC L RETRN,&
MDX L VARAD,-l
INCREMENT TO NEXT WOR~IN
MDX L LOOPS,-l
ARRAY
sse L DECOD
GO TO DECODE RouTINE
"
RETRN SSC L *
RETURN" TO "'"1A I NL INE PROGRAi'10
* MAINLINE PROGRAM CALLS ONE OF FOLLO~ING "ENT~IES
5101 NOP
ENTRY FO~ SINGLE INTEGER
SET XR2 TO LIBF&lLDX 12 *
LO
2 0
GET AQDRESS OF VARIABLE
UPl
MDX
2 1
COMPUTE RETURN ADDRESS
5AVAD STO L VARAD
LO
ONEl
SET LOOPS TO 1
LPSET STO
LOOPS
SAVE RETURk ADDRES~
STx
~ RETRN&l
GO TO DECODE ROUTINE
"BSC "L DECOD
5IOIX NOP ENTRY FOR SUBSCRIPTED INT.
SEt XR2 TO LI6F&1
LOX - 12 *
GEl ADDRESS OF VARIABLE
LD
2 0
MODIFY BY XRl
A
L 1- JO INS I 0 lEN TR Y
MDX
UP}
ENTRY FOR ARRAY
SIOAI NOP-"
SET XR2 TO LIBF&l
LDX 12 *
GET ADDRESS OF VARIABLE"
LD
2 0
STo L VARAD
LD
2 1
GET NOo OF ELEMENTS
MDX
22
COMPUTE RETURN ADDRESS
MDXLPSET
JOI N 5101. [NTHY
-r.- THIS ENTRY TO" SATISFY LISF SIOF IN
TRACE ROUTINESo JOINS SIOl
*
Nap
SIOF
LDX
12 *
SET XR2 TO LIBF&l
MDX
5101&3
JOIN SIor
ENTRY FOR END OF OUTPUT
scor"ip NOP
LOX 12 *
SET XR2 TO LIBf&l
._~_r~_A_____l_"6_ _ _ _ _S_E T LOOP? T?_~
£~
*
(030
9069
01 4C080011
01 74FFOlll
01 7LI-FF0038
01 4COOOO8D
01 l}cooeo 13
0007
0008
0009
0008
OOOD
OOOF
0011
0
0
0013 0 1000
OCt14 01 66800016
0016 0 (200
0017 0 " 7201
0018 01 Dl tOO0111
OOIA 0 (057
0·018 0 D01C
001C 0 6AF5
0010 01 4COOOOBD
Q01F 0 1000
0020 01 66800022
0022 0 (200
0023 00 84000001
0025 0 " 70Fl
0026 0 1000
00"27 01 66800029
0029 0 (200
002A 01 D4000111
002C 0 (,201
0"020 0 7202
70E(
002E 0
002F
0030
0032
0033
0
01
0
0
1000
66E500032
70F3
1'000 "
0034 01 66800036
0036 0
1810
~-
___"__
._
\,1
I
~
I
'I
I,j
"--~-----""""
-----
1'-1
.. "Jr·
l
PAt
0037 0
0038
70·E3
0001 -
00.39 o 00-3A 0
0038 OJ.
003D 00
003F 0
6200
7201
l~4 0 0 0 lll~
C4000002
001+0 01
OOltS 01
0047 00
0049 01
OOl.S o 1
004D 0
OOltE 01
0050 0
0051 01
0053 0
0054 01
0056 0
('+800112
4802
1808
E030
4L} 00_0134
('+000002
irCO 40 0 LtD
-740 1 0 1 1 2
9026
4C28003A
(023
[,(040054
7069
74010112
7066
0057
0058
OOSA
OOSC
005D
005E
C01C·
D4000002
'iAOO0114
72FF
70F(
70SE
0
0042 e
00l+3 0
004l~
0
0
0
00
01
0
0
0
1010
MDX·
JOIN SIOl ENTRY
LPSET
LOOPS S;;S
ARRAY COUNTER
1
END OF CALL ROUTINE
*
HTYPE ROUTINE
*
TRANSPERS·H"';TY")[ CHARACTERS TO I/O BuFFER
*
HTYPE
LDX
2 0
SET XR2 AS CH/\R - ~OUNlER
HLOOP ~iDX
INCRCt COUNT
2 1
BSI L LINE
LD
L 2
BIT lS·OF Xf-~2 TO CARRY
SLA
16
LD
IFfvlT
GET CHAR F RO~1 FOI~i/iAT
C_
BS(
SKIP IF. EVEN Ct-IAR -SRA
8
IF ODD, SHIFT RIGHT
MASKR·
e LEAr~ GARBAGE
AND
B5 I
L lOFIt
MOVE TO. IOBUF
L-D
MOVE FGRrvlA T POINTER IF
L 2
SSC
SECOND CHAR I N---:~\ OF~ D IS
L HODDt"E
MDX L IFMT,l
MOVED
HODD S
TEST FOR LAST CHAR CI
F~'JI DE
BS( L HLOOPt&Z
LEAVE WHE;''-l DIFF -0
END OF TRANSFEf~ IF ODD NO
FvJI DE
LD
Bse L HEND9E
OF CHARACTERS, (,-lOVE FORr-"lA T
rw1DX
POINTER
DECOD
HEND . r\1f)X L I Fi-tiT ,1
MDX
OEeoo
OF
HTYPE
END
ROUT It~E
"*
XTYPE
ROUTINE
*
OVER CHARACTERS IN I /0 BUFFER
*XTYPE SKIPS
LD
FvJ! DE
SET XR2 TO FIELD ~'JIDjH
STO L 2
XLOOP 8SL_ L LINE
TEST FeR LAST CHARACTER
·MDX
2 ..·1
MDX
XLOOP
-MDX
DECOD
END OF ~TYPE ROUTINE
A-TYPE ROUTINE
"*
itTRANSFERS ONE OR TwO CHA-RA CTERS FROM VArtIABl.E
TO I t~o BUFFER
*
ATYPE
BSI L LINE
TRANSFER LEfT CHAR/\( TEi~
lD
VARAD
I
SRA
8
STORE
BSI L IOFIL
LO
FWIOE
S
ONE1
- &
BSC
MOX
CALL
BSI L LINE·
TRANSFER RIGriT Cr-iARAC.T Ei~
LO
1. V-ARAD
AND
MASKR
BS! L IOF!L
(ALL
MDX
ONEI DC
1
STORES FOHi\lAT TYPE
FTYPE BSS
1
STORES FIELD \~ I DTr-I
F ~-J I DE BSS
1
/OOFF
MASKR DC
END OF ATYPE ROUTINE
*INITIALIS!\TION ENTRIES
*
SRED CALLED BY READ STATErvlENT
0
-.
~.
OG5F
0061
0063
006'+
0066
0067
. 0068
0069
006A
1
44000114"
(4800111
1808
4'+000134
COOD
0 - 900A
o _ '~808
0
01
01
0
01
0
0
709D
l;A-00011'+
006E
006F
007L
0072
0073
007Lj.
0075 0
(4800111
E006
'+4000134
7095 .0001
_0001
0001
OOFF
0076· 0
1000
006e
0
0101
001
0
-
*
5RED
Nap
Sc2~_
--..~\,
'~-..
'-19-
0077 01
0079 00
007B·0
-007C 01
007E.30
·0080 0
66800079
C6800000
901E
4C180084
09595100
COJB
_STX
SRA
-'STO
SWRT
0098 O' C07A
0099 0 70EB
!
...
EXIT
. -SET
LD
L
STO
STO
SSC
NOP
LOX
LD
L
12
READ S\vI TCH
COMPUTE RETURN
2 2
L
f
ERROR. DISPLAY IEE03
2 RETRN&l
ADDRESS
-
16
SET 'INITIAL CHARACTER
IOCHR ONE
MULT
LOOPS
SET
COUNT-
SET fvlUL T IPLE F ItL.D_
ARRA~
COUNTER
..
-
COUNT
. I
~_
CALL
*
SET XR2 TO LIBF&l
SET \'1 R1- T E S v!l TCti
ONE
MDX
SAVSW
DC
99
THREE DC
3
EE3
DC
IEE03
_
END
OF
INITIALISATION
*
FIELD ROUTINE
*
TESTS FIELD TYPE
**
.,TRANSFERS TO APPROPRIATE ROUTINE
*
TO MOVE TO I/O BUFFER
FIELD LD
FTYPE
S
. HCODE
TEST FOR HTYPE
BSC L HTYPE.& ....
A
HXCOD
T£S1 -fOR .XTYPE
BSC L XTYPEt&-.S
SXCOD
TES T FOR -SLASH
BSC L ATESTtZ
0063
0003
NINTN
EE03
-
(OD5
906D
4(180039
8010
'+-( 18 0 0 5}
90 16
4C2000AA
44000150
7013
OOAA 01 74000038-·
7004
QOAD 01 44000150
OOAF 01 4:(000011
OOBt p
800A
AlEST
-OOAC 0
as I L
MDX .
PR I NT
DECOD
MDX
LOOPS,O
ATT·S T
L
MDX
aSI L
BSC - L
ATTST A
BSC L
0082 01 4C18005F
0084 0
8005
00 H5 0 1. 4 C1 80 1 79
00B7 01' l{·(0801BF
OOBA a
OOSB 0
-_OOBC 0
·003C
O·
TEST .DEVICE NO.
EE-3
- MDX
6A86
18-10
OnOD 01- D400014F
008F 01 (4000113
009.1' 0 DO 7 7
0092 0 OOA5
0.093 ·01 4(00-0007
0095 0
1000
0096 01 6680"0098
_-00 B9 o·
-.
BSC . L /38.
- 16
ROK· SRA
SAVS", STO L 'READS
LD
2 1
- STO L .. I F~T
008B·O
OOBC 0
-' 0 aA2 0 1
00 A4 ·0 -00A5 01
00A7' 01
OOA9 0
SET XR2 TO LIBF&1
WAiT.
C201
0088 01 0[+00·0112
OO"8A 0
7202
009D 0
009E 0
009F 01
OOA1 0
1.2 O'
LD
0087 b.··
009A 0
- 0098 0
009( 0
12
S
NINTN
BSC· ·.l..._--..--ROK t.&CALL
IOND
DERR
0081 0 3000
-0082 00 4(000038
0-084 6 1810
0085 01 D400010A
*
LGX
LD·
.
A
sse
BSC
MPX
7034
L
l
TEST FOR SCaMP CALLED
PRINT
RETRN
SACOD
ATYPE,& ....
TEST FOR A TYPE
HXCOD
TEST FOR I TYPE _
.ITYPE,&f£TYPt&
INVAL
INVALID FIELD tYPE
HXCOD DC
/0001
HCODE~X~ODE
0003
5XCOD DC
/0003..
/0004
SLASH-X(ODE
0004
SLASH-A(ODE
IOBUF EQU
/003C
FORTRAN I/O 8liFFEH
END OF EIELDFJUTINE·
*
DECOD ROUTINE
*
EXTRACTS FIELD TYPE ~ND WIDTH FROM
0001·
SACOD DC
*
*
----"--
---~---------.-~ .. ---
For~MAT
-
Q
STATE~'1ENT
.
-~-------.---~-~---------.----.--.------ ..
.
-----.-.-.- ....
__._-
-- --_.
__
.•. _" ... _•...•..•. " -
- ---.----.-.~~-~-----.--.-----~-~-----
···r
*
*
'*
*
0080·0 C048
OOBE 0
9054
OOBF 01 4C300101
00C1 01_C4800112
00(3 0 E04(
OOelt 0 DOl+A
DOCS 01e4800112
.00C7 0 ·lBoe
00C80 0045
00(9 01 74010J.12
00e8 0 903F
Ooe( 01 l}C2000D1
ooeE n e040
OOCF 0 D039
0000 0 7030
0001 0 e03C
00D2 0 9034
o0 D3
0 1 t~ e2 0 0 0 E 3
OODS 0
00D6
0007
0008
00D9
00 DA
0
0
0
0
01
OODe 0
e03C
903C
-NOML T L D
OOFO 0 . C>OlC
OOFl 0 ·3000
OOF2 00 I}C000038
·OOF4 01 -C4000074
0106 0
70F2
0107 0
0008
1
STO
LD
BSC
. I-FrvlT
IFMT
READS
L
BSI
LOOPS
NOMLT,Z
RETRN
TYPEC
FTYPE
WI DEC
FWIDE
READS
FIELD,Z
FTYPE
A
REPT,&laND
MOVE TO RIGHT DIGIT
STORES TYPE (ODE
INCREMENT FORMAT·POINTER
TEST FOR FI£LD REPEAT
FIELD ~EPEAT, STORE
NO. OF HEPEATS
TEST FOR REDO CODE
REDO CODE FOUND
RESET FORMAT SCAN P6INTER
TEST FOR READ
READ SWITCH NOT SET
PRINT. AND CHECK FOR
seo~wip CALLED
SET FIELD WIDTH AND TYPE
END OF DECODING
FOR~AT
STo
READ SWITCH sET, TEST TYPE
INVALID TYPE
EE2
ONE
SRA
STo
TRANF LD· . I
STO I
MDX L
SET MUl. r· I PL E FIELD COUNTER
TO NOoOF WORDS ~N H FIELD
1-
MU·L T
IFMT
VARAD
IFMT,.l
Bse
L
(ALL
MULTF 'v1Dx
L
BSe
MDX
L
MULT,-l
READS
FIELDpZ
TRANF
DC
~OT MULTIPLE FIELDt GET
FORMAT TYPE AND DECODE
STORES WIDTH OR REPEAT
HCODE
·WAI T
BSC l· /38
R EPT . LD
L F~I I DE
REDO
Fo0~D
NOt~LT,~
PRINT
.LD
L
Bse L
asc L
SETF LD
STO
LD
STO
ENDDC LD
BSC L·
lD
S
BSC L
INVAL CALL
LD
· LD
PREVIOUSLY
~1ULTFt-Z
WI DEe
· S
Bale
1801
OOF8 0.0010
.. 00 F9 0 1 Cl~ 8.0 0 1 1 2
OOPB 01 D4800111
00 F DOl -, 4 0 1 0 1.1 2
OOFF 01 4C000007
0101 01 ?£.rFFOI09
0103 0 C006
0104 01 ~e20009D
ONE
L
· AND
MASK2
S TO
-. WI DEe
LD
I IFMT·
SRA. .
12
STO
TYPEC
· MDX L IFMT,l
- S
FREPT
BSC L REDTS,Z
LD
WIDEC
STO
MULT
MDX
M-ULTF
REDlS LD
TYPEC
S
REDO
Bse L SETF.Z
LD
I FMl
S
ONE
l}073
OOEl 01 4C000011
00E3 0 C02A
00E4 0 D08E
00E5 0 . (029
- 00E6 0 0080
00E7.0 C022
00E8 01 4C20009D
OOEA ··0 C088
. OOEB· 0 9020
OOEe ·01 4e1800F4
OOEE ·30 09595100
OOF6 0
OOF"! 0
S·
BSC
0039
(030
It C08 00 e 1
OODF 0 I, L}e2000( 1
IF END OF FORMAT· stATEMENT, PRINTS
AND RES-ETS FORt·1AT SCAN. POINTER·
*
IF READ· 99,XXX ,TRANSFERS t~ARACTERS
DECOD LD
MULT
TEST FOR MULT1PLE FIELD
9037
OODD 01 C4000038
IF MuLTIPLE FIE~~, SETS MULTIPLE
FIELD COUNTER
/OOOE;
TRANSFER CHARACTERS TO
VARIA~LE
TO CALL ·TO SEE IF DON£_
REDUCE MULTIPLE FIELD COUNT
TEST·FOR REDO INDICATOR
:5"3./ -
.
I'
.
"
----vJ
_______________ '-21;_-_ _ _ _ __
I;
i
0108,
0109
OlOA
---- 0108 0
OlOC 0
0100 0
.010'E
0001
0001
0001
0009
0005
EE02-
OlOF
'0001
0001
0110 0
0111
0112
0113 0
0001
0001
0001
OfFF
TEMP
BSS
BSS
1
1
MULT
READS BSS
1
·FREPT DC
HCODE DC
DC
EE2
TYPEC BSS
WIDEC BSS
MASK2 DC
/0009
10005
IEE02
1
1
10FFF
VARAD B-SS
I F~~T BSS
ONE
*
**,
**
o
MULTIPLE FIELD STORAGE
o FOR READ 1 FOR PRINT
TEST .FOR FIELD REPEAT
TEST FOR HTYPE FIELD
ERROR DIS~LAY-I~VALID TYPE
TEMPORARY STORAGE fOR TYPE
TEMPORARY STORAGE FOR ~vIDTH
1
11
POINTS TO VARIABLE
~POINTS
TO FORMAT STATEMENT
DC
LIN E . ROU TIN E
CHECKS IF LINE IS FULL
lEE 0 1 IFF UL L', 0 THE I~ It) I S E I~CREMENTS LINE COUNTER
IF START OF LINE. CHECKS PRINTE~ READY, AND CLEARS 1/0 AREA
LINE BSS
1
CHECK FUR START OF LINE
LD
IOCHR
Bse L NOTST,Z
START OF LINE, CHECK
TESf3 LIBF
PRNTl
PRINTER NOT BUSY
DC.
10000
~-
0114
01150
0116 01
011s 20
0119 0
011A- 0
0118 0
OllC 0
011D 0
011E 0
OllF 0
0120 0
0121 01
0123 0
0124 0 .
0125 01
0127 30
0129 0
012A 0
0128 00
0120 01
012F 01
0131 0
0132 0 01330
0001
C039.
l~C200124
176558Fl
0000
70FD
6A06
62(3
C015
D279
7201
MDX
STX
40'+0
TEST3·
2 SAV2&1
- LDX
LD
STRBL STO
2 -61
SAV2
LDX
1/0
AREA
2 1
STRBL
L2
MDX-
*
RESTORE XR2
L INBK
NOTST S
_sse L
CALL
LcD'
~JA
CLEAR
BLNKS
2 IOBUF&61
MDX
MDX
70FD
66000123
7009
900D
4C28012D
09595100
CP07
3000
4C000038
7401014F
'tC800114
EE01
0079
.0 I 5 P L 1\ Y S
LIMIT
LINBK,&Z
IOND
- EEl
CHECK-FOR LINE OVERFLOW
OVERFLOW _ CLEAR INTERRUPTS
AND WAIT WITH ERROR DISPLAY-
IT
SSC
L
LINBK MDX
Bse
~
I
138_
·EXIT
OVERFLOW
INCREMENT
IOCHR,l
NO
LINE
CHARACTER. POINTER
EEl
DC
IEEOl
LIMIT DC
121
MAXo NO~ OF CHARACTERSBLNKS DC
14040
BLANKS
~.
. END OF L Ir'~E ROUT I ;\i-E
-*
IOFIL ROUTINEo ENTERS CHARACTER FROM
ACCUMUCATOR IN FOR~ ooxxINTO IOSUF
IN POSITION GIVEN BY IOCHR
IOFIL BSS
1
*
*
0~34
0135.
'·0136
0.137
0138·
0139
013A
0138
013C
01.3E
0-13F
Ollfl
o1 L,2
0001
0 - 0015
0
C0180
1881
0
8015
0
D006
0
1091
O' C011
01 I~C02 013F
0
1808
01 E(f-OOO 1. 41
0
DOOA
0
C008
STO
LD
SRT ...
HOLD
A
STO
IOADD
AND&l
IOCHR
1
CALCULATE ADDRESS TO
STORE - CHAHAC TEl~
SLT
17
MOVE REMAINDER INTO CARRY
LD·
L
MASKL
AN·D-, C
IF CHARACTER
L
*
BSC
SRA
AND-
SAVE ACCU;"-1ULA. T OF<
AND
STO
'LD
8
TE~Pl
HOLD
~Q.
IS EVEN
o
.- til 0 V E f'vtA S K TOR I GH T S, IDE
DELETE UNWANTED CHARACTER
FRO~ DESTINATION wORD
GtT CHARACTEi-<__ TO__INS[~~4L
..
-_._._-_._------
--._._--_._------...-
... -.....
--~~---
=.!.::.:-------------------,.:I ~.fi
-22-
p~~
I -
Ol'-l3 01 4C020146
·0145 0
1008
- --- O-lA-6 0
E805
0-147 01 [>i+800140
0149 01 4C80013 Lr
oll~B
0001
014C
0001
o ll~D
0
FFOO
014E 0 -003C
014F
0001
- OR
HOLD
TEt-1P 1
MASKL
IOADD
IOCHR
*
.*
*
*
-*
0001
0150
0151
01-52
- __ 0 1S 1+
01-S-6
0157
.0
0 COFD
01 4C080162
00 C400003C
0
901F
01 4C200158
-0_t59 0 COlD
7003
DiSk 0
:0 l_5.8 __01 4(089165
015D 0
100a
015£:-0 E819
01-SF·O-
0173 0 DODB
Oi74 01 '+C800150
0176 0
40FO
0171 0
ODOO·-
0178- 0-- 3000
o
0179 0
COD5
D03C
017A 0
017B 01 (4000074
017D 00 DI~000002
OR,C
IF CHARACTER NO. IS EVEN 8
MOVE CHARACTER TO LEFT
TEMpl
COMBINE
I AND&l
STORE IN -DESTINATION wORD
Bse I IOFIL
RETURN
ASS
1
B$S·
.1
DC
IFFOO
DC
lOBUF
BSS
l'
CHARACIER IN LINE
END OF IOFIL ROUTINE
PRINT ROUTINE
GETS CARRIAGE CONT~OL_CHARACTER AND
SETS UPAPPROF~IATE S~IP
P·RINTS LINEo RESETS CHARAtTER. POINTER
I-F S TAr< T 0 F LIN &f T EST S P R I NT ERR EADY
PR.INT BSS
LD
BSC
LD
SKIP
DOOl
0160 20 176558F1
0161 1 0162
0162 20 f76558Fl
oi63 0 0000
0164 0
70FD
0165 0 (OE9 0166 0
1801
G167" 00 D400003C
0169 01 4C200171
016B 0 eOA7
016C -00 D400003C
016E 0 CO(4
016F 00 D400003D
0171 20 176559E7
0172 0 1810
BSC
SLA
OR
5TO
L
1
IOCHR
TEST2,&
IOBUF
L
S
Bll
NOT-START OF LINE
CARRIAGE CONTROL CHARACTER···-
BSC L
LD
MDX
Bse L
SLA
OR
STO
LIBF
Nur..~, Z
SKP1
IT IS bEROt
L
WAIT -UNTIL SKIP-COMPLETE
-
10000
PRINT LINE
BLNKS -
TES T FOR ZfRO vJC IF COUNT IS ZERO
t-'lAKE COUNT 1 AND
PUT BLANKS IN I03UF & 1
IOBUF&l
PRNXX
PRINT WITH OVERLAP
ONE
IOBUF
STO- L
P1CAL LIBF
sse
PRNTl
TEST2
IOCHR
1
IOBUF
P1CAL,Z
LD
SRA
S TO
~LINE
IT IS A CHANNEL NO
SKIP -AS REQUIRtD
*
CNTRL pC
TEST2 LISF
DC
MDX
OUTLN LD
SRA
STO L
Bse L
LD
STO L
SKIP 1
SKIP
OUTLN,&
8
CNTWD
CNTRL
PRNTl
- 1"6
I
-IOCHR
PRINT
Bll
DC
140FO
BLANK AND ALPHA ZERO
SKPl DC
10000
CONTROL DIGIT fOR SKIP
C NT 'I'D DC
1-3 000
CON TR0 l FUN CT I 6 N * . END OF PR1NT R0 UTI NE
*
ITYPE ROUTINE
*
CONVERTS INTEGER VARIABLE TO DECIMAL
*
TRANSFERS DIGIJS TO 1/0 BUFFER
*
IF FIELD_ WIDTH TOO SMALL, FILLS FIELD
*
wITH *~~*
I TYPE LD
STO
LD
STO
L
I.OCHR
SAVCR
FWIDE
l
~
S A-V E PO SIT ION 0 FCHARACTER ~OINTER
. MARK POSITION OF RIGHT
END OF FIEL.Dt ALSO TLST
---,-
333-'
-
-:;.-
-
.~--
--.....• ,.".,,,._--_._.,,,,....
,,"
..•.,,,,....__.,,--,,.. ,.,,,.,,.
-23PII'
o1 -, F 0 1
0181 0
------0182 0
0183-0
it 4 00 0 lil.
72FF
70FC
COCB
D033
0185 01 C4800111
0187 01 l~CI0018(
0189 0 1810
OlBA 01 9l{·800111
OlBC 0 . D02C
0180 0 (028
018E 0 1890
018{~
I LP
0
POSY
DlVL
FOR SUFFICIENT SPACE
AVAILABLE IN LINE
LD
I
BSC
L
16
GET ABSOLUTE VALUE OF
VARIABLE
IF NEGATIVE,CHANGE SIGN
VARAD
VABS
VABS
DIVIDE BY 10
SRA
S
SiO
LD
SRT
OR
Fl
(OBA
BS 1
LD
I OF I L
IOCHR
9028
S
00B8
S TO
MDX
MDX
LD
Bse
ONE2
I oeM/:<
74000189
0199 0 700E
019A 01 (4800111
019C 01 4(1001A(~
019E 0 COBO
01 AO 01
i~C0801AB
01A2 0
01A3 0
01A4 0
01A5.0
01A6 01
01A8 0
01A9 01
01AS 01
o1-AD 00
OlAF 01
01 B1 0 _
C019
4090
0182 0
l}081
72FF
70FA
(013
001\9
l~C000007
900E
4(30018D
(4000074
D4000002
44000114
FILL
FLP
LD
STO
. BSl
LD
BSI
MDX
r-1DX
L
L
L
FWIDE
2
Q
01BF 01 C4000074
01(1 0
01(2 0
EOOl
70EA-
01(3 0
007F
01(4 0
10'00
Fl
bc'
De
MINUS
De-
TEST FOR NEGATIVE VARIABLE
SIG~
ENTER rv1! NUS SIGN
RESET CHARACTER POINTER
NEXT DIGITe TEST FOR SPACE
AvAILABLE
,,! 0 SPA CE. F ILL FIE LD \\/ I TH
LINE
ASTER
IOFIL
2 r,.l
FLP
BSC . L
CALL
TEN
OlBE
TEST FOR SPACE FOR
DIVL,ooZ
OlBA 0
0060
OOSC
0001
IoeHR
SAVCR
FILL,&
MINUS
IOFIL
FEND
IOCHR
L CALL
SAVCR
~
STORE
MOVE CHARACTER POINTER LEFT
TEST FOR ZERO QUOTIENT
L
SAveR BSS
FEND BSS
VABS. BSS
CONVERT. REMAINDER TO EBCDIC
VABS90
NEXTD
VARAD
RESET,-
BSC
01B3 0
01 B'+ 0
01B5 01 4C000007
01B7
0001
01f}8
0001
0001
01B9
OOOA
I
L
Bse L
LD
BSI
RESET LD
STO
BSC
NEXTD S
COOB
0188 0 . OOFO'
L
·LD
S
9017 .
.0
16
TEN
VABS
16
0028
1090
E828
40;"0
01BD 0
I
o
A82A
01_90 0
0191 0
0192 0
0193 0
0194 0
0195 0
0196 0
0127 01
OlBC 0
VARAD
POSV,'"
STO
SLT
018F 0
019F 0
BS I L L I NE
MDX
2-1
MDX
ILP
LD
IOCHR
STO
FEND
1
RIGHT END OF PREVo FIELD
1
1
ABSOLUTE
10
100FO
10060
1005C
RIGHT END OF T~IS FIELD
VALUE OF VARlp.f~LE
CONVERTS DIGITS ·TO EBCDIC
MINUS SIGN
DC
DC
1
'END Of ITYPE ROUTINE
**
FETYP RouTINE
'*
FOR E TYPE. FILLS FIELD ",lITH .~*-::FETYP LD
L FWIDE
MASK OUT 0 PART OF SPEC~
AND
NOD
. JOIN NO SPACE BRANCH
MDX
FILL&2
NOD
DC
I007F
*
END OF FETYP ROUTINE
PRNTZ NOP
._____ DUMMY ENTRY NDT USED~4._.
ASTER
. ONE2·
o
------
-21~-
't
PI
01C5 0
01C6
01C8
NO
70FE
MDX
OOOl~
*-2
ASS
END
1
ERHORS
IN
ABOVE ASSEtvlBL Y c
o
o
L-._ _ ._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
-25II ASM
*LIST SOURCE PROGRAM
'*
*
0000
176559E7
0000 o 1000
0001 01 (4000003
0003 o 0005
·0004 20 l76558Fl
0005 o 2000
0006 o 003C
0007 1 ·OOOA
0008 01 -4COOOOOA
OOOA 1
0008-
0008 01 4C80000A
603C
PRNXX
PRNXX PRINTS IN OVERLAPPED MODE
THIS SUBROUT l"NE NOT OVERLAID DURING SOCALs
LI8R
ENTPRNXX
NOP
LD
STO
LIBF-
L
PRNTl
DC
12000
DC
IOBUF
EN-DPG
DC
BACK BSC
ENDPG DC
Bse
L
I
lOBUF EQU
000E
,END
NO E-RRORS IN
ABQV E
*BA(Kf71
~.
*ENDPG
IOO3e
SKIP TO CHANNEL 1 IF
END OF PAGE
FORTRAN 1/0 BUFFER
ASSEMBL Y.
o
o
--26I I ASi"1
.
~·LIS·T
a
QQ
09595100
0000
0001
0001 00 7'+000.032
·0003 0
70FD
0004 01 l.C8000-00
0006
ENT
*CALL IOND
*CALL laND
IOND
_ SUBROUTINE NAME
NO PARAMETERS
ALLOWS 1/0 OPERAfIONS TO [ND
PAUSE OR ·STOP IS ENTERED -
'IOND
*
BSS
IOPND MDX
MDX
BACK Bse
1
L
END
.
\ ~;
--------~------~~--------------~--------·~l
_
50~0
IOPND
IO'ID
- ARGUi'vlENT
BEFO~E
ADDf~ESS
ANY INTERRUPTS PENDING
I DEAL /t8,1'
IDEAL488
IDEAL48C
. I DEAL48 D
A IDEAt..48E
I [}EAL'~8 F
I DEALl~8 G
ID[ALl~8H
_IDEALlir 8 I
IDEAL49
I DE!,LLt9 A
NO ERRORS IN. ABOVE ASSEMBlYG
o
o
S37
THE 51: ..
-27-.4--------..------------.;....--w~-S-P-1-2-8-24-l'
/1 "lOB T
// FOR
.** SAMPLE PROBLEM 2
CSP10°
*IOCStl132 PRINTER)
*LIST ALL
*
*
*
NAME .SMPL2·
ONE WORD INTEGERS
EXTENDED PRECISION
( DEMONSTRATE USE OF OVERLAPPED PR1NT SUBROUTINE
C THIS PROGRAM IS IDENTICAL TO COMMERI(\L SUBROUTINE PACKAGE
( SAMPLE PROGRAM NOe 2
C NOT E I NPUT OAT A CARDS CSP 13 96 o· 'TO CS P l'L~030 ARE NOT REQU It~ED
C--ab--THE INPUT IS ~1ADE uP OF A ~1AsTEr~ (t\RD FOLLO~\/~D BY THE THANSACTION
C-----CARDS FOR EACH CUSTOMER. WE WANT TO PRINT AN INVOICE AND PRINT A
C-----NEW MASTER CARD FOR EACH CUStOMERe
DI ME N S ION INC RD( C3 2 ) , I PRNT ( 6) , lOT CD ( 80) f 1ST 0 P ( 5 ) f I ~J t<. ( 1 3 ) , I S Urvl ( 8 ) ,
lIEROR(6)
READ(99,lOl): ISTOP
101 FORMAT(II 5 T 0 PI)'
READ(99.106) IEROR
106 FORMAT(IE R R 0 R
,).
J=2"
1=0
°
L=O
M=O
CALL READ(INCRD,l,80,J)
I F ( J-l ) 22 •.2 ,2
2
IF(NCOMP(I~CRD,1,5,ISTOPtl)
3
(ALL NlONE(INCRD,7Q,5,K)
IF(K-l)
26,4,26
CSP12880
CSP12890
CSP12900
CSP13010
CSP13020
CSP13030
CSP130 l +0
CSP13050
CSP13060
CSP13070
INC RD( 8 1 ) :: 16 4 {~ 8
I NC RD( 82 ) :: 5 L~4
1.
CSP 12 8 /+0
CSPJ.2850
CSPl"2860
CSP1~O
CSPl~~O
(SP13100
CSP13110
3,22,3
4 WRITE(3,102) (INCRD(II),II=1,60)
102 FOR~-1AT('1·',20A1/(1 ',20Al»
REA D ( 9 9 , i 0 3)
10 3 FORMAT ( I
I 'tJ K
t
$ •
CRt)
CALL EDIT(INCRDs61,68,I0K,1,13)
WRI TE ( 3 , 10 4) I 'Ii K
104 FORMAT(///lOX,'QTyl.10X,'NAME',46X,'AMT ' /23X,JPREVIOUS BALANCEI.
1 2 8 X, .1 3 A1 .)
40
5
6
7
8
9
LJNES=8
CAL L ··A 1 DEC ( INC RP, 6 1 ,6 8 , L )
IF(L)
5,5,23
..
CSP13250
CSP13260
CSP13270
CSP13280
CSP13290
CSP13300
CSP13310
CSP1-332J
(S'P13330
CSPl-3340
CAL L M0 VE ( INC RD, 6 1-, 68 tIS U~~ , 1 )
CALL MOVECINCRD,lsgO.IOTCD,l)
CALL READ(INCRD,l,aO,J)
I F ( J-l )
22, 7 , 7
CALL NlONE(INCRD,70,5,K)
IF(K-l)
18,19~8
IF(K-2}
18,9,18
IF{NCOMP(INCRD,1,20,IOTCD,1»
18,10,18
10 READ(99t110) IPRNT
110 fORMAT('
•
0')
_
CALL EDIT(INCRD,49,52,IPRNT,1,6)
REA D ( 9 9 , 1 0 3)
n~ K
·0
-
CALL EDIT(INCRD,41,48,IWK,1,13)
IF(LINES-55) 31,31,17
3 1 \AJ RI T E ( 3 , 1. 0 ~ ) I P RNT. ( INC RD( I I ) , I I :-; 21 , iI- 0 ) , I 'd K
105 FORMAT(7X,6Al,lQX,20Al,24X,13Al)
LINES=LINES+l
,----------------------
5'3'~
---------_._-- -----------.......;.--__
~~
. - ......
--.-~~
.....
- ........ __
.
..
-..
--- --- -- --- - ror'· .:-- -} ....·. rwtr ""TT"·"'''"%, 'flo BffeMiHitM
-----------------------------------~----------
-2'1-
SAMPLE- PROBLEM 2
PAGE 02
CALL AIDEC(INCRDp4lt48,L)
IF (L)
1 2 , 12 s-1 t.
CALL' ADD( INCRDt41 ,48, ISUM~ 1_, 8 t~)-
CSP134-30
CSP1341.1·O
CSP 13L~SO
CSPl.3460
CSP13470
- . CSPJ.3 Lt80
CSP13490
CSP13500
CSP13510
1 F( M )
13 , 6 t 13
CALl. 10NQ-
13
STOP 777
CALL NZONE(INCRD,L,4,Nl)
Nl=O
14
CALL AIDEC(INCRD~LJL,Nl)
IFCNl)
16~16,15
CALL 10ND
STOP '-666
16
CALL DECA1(INCRD,41,48,L)
L=O
. GO TO---11
17 \~RfTE(3tl09) _
109 FORMAT(' l ' ,9X', 'GTY' ,lOX, 'NAME' ,46X,
LINES::1
GO TO 31
_
18
CAL L T YPER ( I E RO Ri 1 -, 5 )
CSP13520
15
CALL
-(;.Sp13 530
CSP 13~LrO
CSP13-S50
CSP13S60
CSP 135-70
'AMT')
CSP13620
TYPER(INCRD~lJ'32)
CSP13630
GO TO 6
CALL DECA1(ISUMjl,8,L)
JF(L)
20,21,20
CAL L IOND - _ __ _ _
STOP 555
--
19
20
" t"dtfe(thttt" ... · 'I.'
CSP1364-0
CSP13650
CSP13660
CSP13670
CSP13680
~,\1 REA D ( 9 9 , 1 0 3) IvJ K --CAL lED I T ( -I SUM, I ; 8",-n", K , 1 t 1 3 )
V
CALL MOVE(ISUM,1,g,IOTCD,61)
CALL TYPER(IOTCD,1,80)
107
~J R I T E ( 3 , 1 0 7 ) H-} K
FORMAT(1123X,'TOTALf,39X,13A1)
CALL fYPER(INCRD,a~,82)
GO "0-1
22 READ(99,1081 IWK
108 FORMAT( 'E N-D.
0 FCAL L TYPER (" I l.A:i< 9 -1 , 1 0 )
~.O B
f)
CALL IONDSTOP 111_
, 23
25
NZQNE ( I NeRD, L, 4, NJ. )
N1=0
CALL AIDEC(INCRD-,L,_L,Nl)
IFCN1)
2~,25,24
CSP13820
CALL _ LON()_ STOP 4'+4
_.
CALLDECA1(INCRD,61968,L)
-CSP13860
CALL
CSP13830 CSP1~8lj·O
CSP13850
CSP1387Q:
CSP1388--Q
CSP13e90
L=O-
GO To 40
2·6
"CSP13900
CSP13910 -
CALL' TYPERiIEROR,l,5)
CALL fYPER(INCRD,1,82)
_ GO TO -1
E'NO
CSP13920
CSP13930
_
CSPI394-0
VAOABLE ALLOCAT·IO!\)S
INCRD::OOSl"
M
00<: B
=
IPR/~T;0057
IOTC-D:::OOA7
K
II
.:: 0 0 CD
102
=0107
=00 CC
STATErvlENT ~ALLOCAT IONS
101 =QOFa 106 =OOFF
ISTOP=OOAC
LINES=OOCE
102
=0111
I vlK =0089 Nl . =OOCF
ISUiVi
=00 C1
IE ROR=oO-C \
110
:::013E
lUS
:::01
L
:53., THL
.-. --------
-'-------------~'
-~--~-----------.__::.c.;;.,;;'~;:-~-~
lS-Ar-.1PL~_OPRAOBL~.·M2
~
-
l-o-'-:-
'=02l~3' 31=0265
19
=0200
20
=02E~21
1 5
~01 SD .~
~
=01C6
11
=0288
12
=02E8
22
4
=0102
'=0292
=0313
13
=0202
=029F
23
=0322
40
.~.
5
li;
24
r
=020(·
PA~E.-l
6
.
=02A3
21.550,'
. ....
=0336
FEATURES SUPPORTED
ONE WORD ~NTEG~RS
-EXTENDED PREC~SION
Ioes
CALLED SUBPROGRAMS
READ
NCQMP'
NZONE
SFIO
-'SIOAI
5-IOIX
INTEGER CONSTANTS
99::'OOD2
- 2=00D3
60=OODC
48=OPE6
lO:::OOFO
.61=00DD
- 55=00E7-111=00F1
-EDl T
SUBse-
16 '+48 =OODlt·
68:::'PODE
21=00E8
441.t·=00F2
CORE REQUIREMENTS-FOR SMPL2
,COMMON
0 VARIABLES
END OF
210
MOVE
PRNTZ
A1DEC
S-TOP -
5440=00D5
13=o"oe F
40=00E9
1911=OOF3
PROGRAM
ADD
IOND
0;-00D68=OOEO
777=OOEA
f638=OOF4
--
DECAl -
1=00D7
20=00E1
4=00EB
1365=OOFS
TYPEr-<
80
l~ 9
666
273
640
COMPILATION
-0
s~o
'~~--------------------.-----------~-----~------------~~-----------.
;l
1/ XEQ
R 47
OB4E (HEX) WORDS AVAILABLE
CALL TRANSFER VECTOR
CARRY
NSIGN
FILL
TYPER
DECAl
IOND
ADD
MOVE
AlDEC
EDIT
1123
1007
1082
OC05
08-5 A .
OB3E
OA7E
OA40
09D2
08AF
. NlONE-
NCOMP
READ
07FD
07AF
0750
lIBF- TRANSFER VECTOR
13AE
135 E
130 E
HOll
o
. PRTY
EBPA
TYPEO
EBPRT
RPACK
SUBIN
ARGS
SWING
SPEED
CAROl
PRNXX
PRNTl
STOP
SCOMP
SIOIX
SUBSC
l1E6
1180
1050
1090
1022
106F
OE06
ODEO
00D2
0(50
0
087C
05A9
053A
SWRT.
SIOAI
SRED
058A'
ESTO
06EO
ElO-
06F6
PRNTZ
SFIO
D I SKZ
o
ILS04
I LS02.
ILSOl
ILSOO
0338
oBAl;
0547
0533
(HEX)
06D8
-0'0514
0
OOF4
1401
141D
142F
1441
IS
THE EXECUTION AODR6l
___--------_--.---_.....,....._-.--~~-- Tr
~¥I
-----------
-------
DAVESMARKE"T
1 997 \-JAS HI NG TON S T" .
NEWTOWN. MASSa 02158
.···0
QTY
NAME
PREV IOUS' BAl.ANC E
8
11
10
8
. 6·
17 .
17 .
17
17
-' 17
17
.25
25
10
.10
- 12
12
12
12
12
12
1.000
4.000
·200
100
5-()
100
'100
100
10
-12
12
12
12
12
12
1,000
~
A.,OPO
200
50
100
100
100
100
10. 12
1,000
4',000
SUGAR .", BAGS
~HICKtN SOUP - CASES
TOMATO SOUP - CASES
SUGAR RETURNED
COOKlES -'CASES
GINGER ALE - CASES
ROOT BEER - CASES
ORANGE ADE ~ CASES
CREME SODA - CASES
CHERRY SODA - CASES
SODA WATER - CASES
DOG FOOD ... CASES
CAT FOOt) - CASES
SOAP ~OWDER - CASES
DETERGENT - CASES
HAM - TINS
HAr·1 - LOAF
. SALAr-11
BOLOGNA
CORNED DEEF
ROAST 8EEF
BREAD - LOAF
ROLLS
MILK - QU~,RTS
MILK - HALF GALS
MILK" GALS
POTATOES --BAGS
TOMATOES - LOOSE
CARROTS - BUNCHES
DETERGENT - CASES
HA~~ - TINS
HA~1 - LOAFSALAr-lI
BOLOGNA
CORNED BEEF
ROAST BEEF
BREAD - LOAF
ROLL~
MILK'" QUARTS
MILK' .. GALS
MILK - HALF GALS
POTATOES - BAGS
TOMATOES - LOOSE·
CARROTS - BUNCHES
DET~RGENT - CASES
HA~w1 - TIKS'
BREAD - LOAF
ROLLS'
AMT
-
$111.29'
$-21002
$38a76
$30.11 $210 02CR '.
$45(121
$52037'- $52037
$52.37
$52c37
$52037
$.52037
$101026
$101.26
$72.89
$72.89
$36.75
$33.75
$33.75
$33.75
$33.75
$33075
$150.00
$150.0~i
$57.4~
$57.42
$57 ~lr2
$11.23
$11.23
$11.23
$72089
$3.6075
$33.7.5
$33.75
$331_75
$33075
$33.7"5
$150.00
$150.00
$57042
$57 a'. 2
$57.42
$11 c 23'
$110'23
$11e23
$72089
$36.75
$150000
$150000
··C
. II·)
--_. --.~..----- . .. "-'-._---_....... --..~.,
Qry
200
100
50
. 100
100
100
10.
1-·2
12
12
12
12
12
f,ooo
4.000
200
_ 100
100
100
100
100
10
12
'-------:-------. ~
. NAME
(\1ILK - QUARTS
MIL K HALF G/,LS
MILK
GALS
POTATOES - BAGS
rOr-1ATOES .,.. LQOSE
CARROTS - BUNCHES
DETERGENT - CASES
HA~1 - TINS
HAM'- LOAF
SALA\11
BOLOGNA
CORNED BEEF
ROAST BEEF
BREAD ... LOAF
ROLLS
MILK
QUARTS
NO
fVLl L K •• HALF- GAL 5
MILK'" HALF GALS
POTATOES'" BAGS
TO~·1ATOES
tl>
CARROTS DETERGE~T
HA~1 - T l-NS
LaOS E
BU~CHES
~ CASES
-
AMT
~
.;
$57.1+2
$57042
$57 ~42.
$11.023
$11023
$11023
$72089
$·36 () 75
$33c75
$33,675
$33(175
$336)75
·$33075
$150'000
$150000
$ 5 -, () '~2
$5704?
$57,,4·2
$11e23
$11~23'
$11023
'$72.89
$36075
TOTAL
c
._ _ _ _ _ _.________--:-:=:__
5-'1',3
1.
__
~ "_C"~"-:~"'.:T"~
STANDISH MOTORS
10.WATER STREET
PL¥.¥lOOTHt MASS.02296
1
t·' .
.
,4
I
r~·.
~/-
aTY
20 .
6
20
50
50
100 ..
NAME
PREVIOUS BALANCE
AIR C(EANERS - CASES
GREASE
BARRELS
TIRES
650 X 13
TIRES
150 X 14.
TIRES
800 X 14
GASOLINE CAPS
AMT
$2,356036
$200 (l03.
- $1650 2l~
$260.038
-
·$90.0053
$1,-·012000
$99.68
TOTAL
~--~-~----------
__.__£
J..jJ..j _TI
.1
o
~~
. w·GJ
1130 COMMERCIAL SUBROUTINES
~---------
ABSTRACT:
Version II of the IBM 1130 Commercial Subroutines will be
the primary emphasis of this session. - The general problem of
commercial programming in pure FORTRAN and the history of
partial solutions available at this date will be discussed.
Version II of 1130 CSP will be compar~d to Version I and other
available commercial routines. Performance in execution time
and core requirements as well as current limitations will b~
presented.
,"
,
-----_._----
,
-------
I:
,.
G:
-. . -~------ ...
--~~.---------~----.---.--- ...
c/~·
~
------·--------by
s. F.
----- --.-.
__-------
Seroussi
Thomas J. Watson
. Research Center
P. O. Box 218
Yorkto\vn Hts •• N. Y.
A series- of programs has been \vritten to-al1ow-180.0.
user s to fully utilize the capabilities of the mM 2841- 2311 disk
storage system within the frame \vork of the Time-Sharing
Executive -- TSX, Version.3, Mod 1.
A comparison is made between the 2311 and the 2310
disks in programming techniques arid timing.
o
Physical Characteristics
2310
Capacrtr - Oxide coated disk with 200 primary, 3 alternate
2 track cylinder s capable of storing 521, 304 16 bit words
(1,042, 304 bytes) of which about 5I2K words are available to
the user the rest being used as alternates to defective tracks.
Each track is divided into four sectors.
A sector is the basic
. addressable unit \vith a capacity of 321 v/ords one of which is
usually used for sector addressing and error checkout.
o
Titning - The disk rotates
a revolution taking 40 ms.
a~
a speed of 1500 rpm;
Word transfer rate is 36 K words/ sec.
l
Cylinder to cylinder access' is 15 ms in single steps
or 20 for double cylinder steps and a delay of 20 ms must be
added to allow the carriage to stabilize itself. Average access
time 530 ms.
Bet"',veen sectors the shortest time available is 235!lsec
(plus
27~sec
for each word not used).
As an exarnple, if we
write 310 v/ords in a s ector and then continue to ·\-vrite on the
next sectors the delay betvveen writes will be a minimum of
235 + 27 x 11 or over half a millisecond. 1
Programming - The following hard\va~ (rr..achine
o
languaJ e) operations are available:
1
11odt1s AI, A2, and A3 ..
SJ./7
z. ,
CE Mode
READ TO MEMORY
READ - CHECK
INITIALIZE ·READINITIALIZE WRITE
CONTROL
(SEEK FORWARD OR BACKWARD)
SENSE
These operations must be executed one at a time with no
chaining allowed and ,vith several restrictions imposed such
as the need for a 20 millisecond delay between seeks or an error
will result.
Only ~ sector can be processed per I/O c0-mmand.
Interrupts - Sensing the DSW for each disk unit provides
the user with a series of indicators.
The occurrence of an
Operation Complete, causes an interrupt in the 1800.
o
All other
indicators 'Nill only be detected tog ether \vith Operation Complete.
Data Status Word
Bit
o
1
2
3
4
5
6
7
8
9
10
11
12
13
14-15
Interrupt
Any Error
OperatiQn Complete
Disk Not Ready
Disk Busy
Carriag e Horne
Parity Check Error
Storage Protect Error
Data Error
Write Select Error
Data Overrun
Not Used
CE Not Ready
CE Busy
Not Used
Sector Count
o
\
3.
Two or znore indicators '\vill be on when an interrupt occurs.
Number of Drives Per System- Standard up to three.
RPQ three
mo~e.
Typical I/O Comznand:
IOCCI
o
XIO
IOCCI
"DC
TABLE
DATA TABLE
DC
14DOI
WRITE ON DRIVE 1,
SECTOR 1
321
COUNT
TABLE DC
BSS
321
A previous XIO would have moved the arm to the proper
cylinder.
o
4.
2311
Capacity - 12 oxide
coat~d
disks with the ten inside
surfaces used :for recording data up to 7. 5 million 8 bit bytes.
A pack is divided into 200 primary, three alternate
cylinders and ten tracks (one for each recording surface) of
3694 words each.
Each track can be subdivided into records
of variable length as defined by the user.
The home address
. which identifies the track and the first record on the track,
RO , usually used to define the condition of the track (and its
alternate if needed), are generated with special IDM provided
programs with a standard format.
A track would have the
()
follo\ving configuration:
2311 Track Format
Each of the above areas contains the follo\ving in.formation:
Index Home
Marker Address
Oato Record R I
Home Address
0\
5S0
5.
o
Home Ad~ress Flag Byte:
The flag byte in the home
address is transferred automatically to the using system by
perforrn.ing a read home addres s operation on the track.
The
bit significance is:
lit
Function or Setting
o
Zero
Zero
Flag
Byte
2
Zero
3
Zero
4
Zero
S
Zero
6
Track Condition
o indicotes operative track
7
Track Use
o indicates primary track
;,..
1 indicotes defective track
1 indicates alternate track
Flag:
Byte 0 of the count area is generated by the 2841
as each record i~ '\vritten.
o
It is not sent from the CPU.
Bit
7=1 indicates an alternate track.
Bit
o
Function
0 for even-count records (RO I R2, R4, Rt/
1 for odd-count records (Rl , R3, RS····)
Used by the 2841 to ensure that all address
markers (and records) ore present. The 2841
signals a missing address marker when two
consecutive, identical bits are encountered
(unless on index point intervenes).
Used only with record overflow feature.
o for all non-overflow records and for the
Flog
last
segment of an overflow record.
Byte
1 for each segment, except for the lad segment
of on overflow record.
2
Zero
3
Zero
4
Zero
S
6
Track condition
7
Track Use
Zero
0 indicates operative trock
1 indicates defective track
o Indicates primary track
1 Indicates a hemat. track
Bits 6 and 7 are transmitted to the flog bytes of
a II records on the treck from the flog byte
,.r the
Home Addr~ss of thot track by the 2841 •
SSt
------_.... _---... -._•..
_--_
__
..•.._....... .•....... _....".".----,..._,..
'''.''-.'''.-~--.".--~~~~~------~---------~------------
6 •.
5
6
__------________
~-------------------y
-J
Coont Area
index
Key Area
Data Area
HA
~~~~~______~_______R_l______~________~
rn
~
~
Address
Marker
Address
Marker
• May Not be Present
Record RI - Rn Format
o
Storage
Track Copocity in Bytes
When RO is Used as
Units
23tr
Oo~o
When RO is ~sed for
Sp'!cifj~d
Applicotion Data
By IRM Programmi"9 Systems.
3625
Records (except for lost record)
Without Key
61 + 537
512 °L
3694
Number of Bytes per Record when
(KL = Key Le.1 g th.)
DL = Dotu L~ngth
Lost Record
Bytes Required by Data Records
Track Capacity
With Key
Without Key
81 + 537 (K + D )
l
L
512
.D
.with Key
20 + (K l + Dl )
L
Number of Equal Length Records Per 2311 Track
Record RO vsed as specifIed by
IBM Progrommin.l Systems.
I
2
3
4
5
6
7
8
9
10
II
12
13
Without Ke)'
3625 1740 1131 830 651
532
447 384
334
295
263 236 213
W' h K ~ I neTudes key bytes +
.t
ey t doto bytes - I<'L + DL
3605 1720 1111 811
512
428
315
275
244
632
364
14
15
i?3
177 162 149
16
17
18
19
20
138 127 li8
2'7 194 174 158 143 130 119
loa
99
Record gaps are of a length determined by the size of the
record.
Track capacity then, varies with record length as
shown in the above figure.
Timing - The disk rotates at 2400 rpm with a revolution time of 25 ms.
Access time fron1 cylinder to adjacent
o
7.
o
cylinder is 25 ms; maximum access time (from cylinder 000
to 200) is 135 ms and average time for' a random access ~s
Once the cylinder is accessed, an average of 12.5 ms
75 ms.
is needed to reach the desired record.
No delay is encountered
in going from track to track because the access mechanism
includes one read/write head per track.
Progranlming - A series of commands each taking 3
1800 words, are available to the user which allow for manipulation of both the control unit and the 2311 drives.
o
These commands
can be classified mto four groups as recognized by the selector
Output Forward (Write, Control)
channel:
Input Forward (Read, Sense)
Branchi...'1.g (Transfer in Channel)
Test I/O
There are forty possible command codes(l) SOIne of which allow
the user to operate on any of the elements contained in a track
or in a whole cylinder.
cylinder boundaries.
There is no capability to cross
This has to be programmed for and is
very simple.
Interrupts - Before considering the interrupt capabilities
o
of the 2841-2311 system let us look at a simple diagram of one
of these devices connected to an 1800.
1. See Appendix I, pp L.2 for table of comrnands
SS3
-~
----- ...
.......-.- ....- ............
-r
---.-.~~--
8.
1
I
1800
1826-M2
Selector Channel
REQIN"
~R_E--..;;:Q__O.;...U..-.;.;T__""j')8 41
2311
2311
From this diagram, it is evident that there are two interfaces
of importance; (1) Selector Channell Control Unit (2841) and
(2) Control UnitlIIO Devices (2311).
o
Selector Channell Control Unit
(a) If more than one control unit is attached to the
selector channel, SENSE ILSW will deternline 'which one of
them has interrupted.
The machine configuration under
consideration has one control unit so there is no need to
execute this instruction.
.
(b) To determine the status of the selector channe.1, a
SENSE CHANNEL STATUS \VORD can be executed.
similar step to sensing DSW on the 2310.
It is a
It provides the pro-
grammer with the following information in the accumulator:
BIT
o
STATUS
MEANlliG'
o
Not Operational
Device Addressed not ready
or not connect ed
1
Unit Status Pending
Control.Unit and Device Status
Waiting for Further Action
2
Program. Control
Interrupt
.Requested by Programmer
3
Program Check
Bad Parity on IOCC - Request
Out of Bounds
4
Data Check
Parity Error - Write to
Protected Ar ea
5
Interface Check
Hard\vare Error
6
Incorrect Length
Requested Length and Length
in Disk Unequal
7
Channel Busy
Busy - Do not Request Another
I/O or Card Check will Occur
8
Unit Operational
Unit Up - Actually In Use
The fir st four conditions generate an interrupt.
Further
information can be obtained from the status of the control unit
and devices.
Control Unit/ Devices
Up to eight 23lls can be attached to a single control
Wlit.
To obtain their status, a SENSE UNIT STATUS can be
executed.
If Unit STATUS PENDING occurred, it is necessary
to do this sense in order to clear the interrupt.
The 16 bits
10.
0
of information placed in the accunlulator are:
BITS
STATUS
MEANING
0-3
Address of Control Unit
4 - 7
Device Address (0 - 7)
8
Attention
Not Used
9
Status Modifie,r
With 11, Control Unit Busy
10
Control Unit End
Control Unit Available
11
Busy
With 9
12
Channel End (GEl )
Indicate Status
13
Device ,End (DE)
Of Operation
14
Unit Check
Error
15
End Exception
End of File Detected (A
Record with Zero Count)
The above bits can be on in s everal
the operation in progress.
C;
com~inations
depending upon
In general, CE - DE together and
alone, indicate a successful end of operation.
the other hand, indicates an error.
Unit Check on
Three more sense operations
are available to assist the programmer in determing the error
source:
(a)
SENSE CCvV ADDRESS - will give the address plus
three of the last commands used.
(b)
SEI'~SE
BYTE COUNTER - will give the residual
0'
byte count.
(c)
SENSE BYTES - will transfer six bytes of informa-
tion concerning the unit check from the 2341 to the CPU at
~1'l
1
n
le:r'Y
'111
o
address specified by the user.
The first two bytes contain
most of the needed information as shown below:
DesignatIon
Significance of "I"
Trock
Overrun
Indicates thot writing has not been completed
by the time the index point is detected.
2
Cylinder
End
Indicates that Cylinder End has been detected,
but the CCW Command Chain has not been
completed.
·3
Invalid
Sequence
Indicates that on attempt has bp.en made to
execute on invalid sequence of CCWs or that
two Set File Mask commands appear in the
some command chain.
Valid commpnd sequences are defined in the
Write and Erase command descriptions. Command Reject (Byte 0 bit 0) is also set when on
Invalid sequence is detected.
4
No Record
Found
Indicates that while executing a chain of CCWs
the 2841 has d~tected two I ndex Points witholJt
completing on intervening command to read or
write or search the Data Area, Read Home
Address, or Read RO. It is also set in
conjunction with Missing Address Marker if
there is na data on the track. No Record Found
is never set if the Multi-Track bit in the command (8it 0) is on. No Record Found will be
posted if the address marker in front of the last
physical record on the track is not detected.
5
Fil.
Indicates that a command was issued contrary to
the file mask. The Comrr.and Reject oit is also
set by this condition, if the operation violates
the write portion of the file mcuk.
Byte lit
o
o
o
Command
Reject
Intervent ion Indicates that the specified file is not physIcally
Required
attachcd to the system :;,r, if physi colly attached
to the system, it Is not available for use because
the file mator is not on, a cover interlock is
open, etc.
o
2
Bus Out
Parity
Check
o
3
Equipment
Check
Indicates thot on unusual condition is detected
in the control or storage unit. Conditions
covered by this bit are defined by Sense 8yte 2
(See Appendix B).
o
4
Data Check
Indicates that a dota check error hos been detected in the inFormation received by the 2841
from the storage unit.
o
5
Overrvn
Indicates that a chained CCW was issued but It
was received too late to be propP.riy executed;
or that a byte was received too late (during
... ,;,~ operation) to be tronsferred properly; or
the channel did not respand fast enough during
a r~ad or s('arch.
o
Indicates that the 2841 has detected a parity
error during the transier of a command or data
from the channel to the 2841. A parity error
detected during command transfer signals a
Parity Check.
Whon writing, the remaining portion of the
record area Is filled with zeros and the overrun
check is generated. When rcoding or searching,
the remaining portion of the record is ignored.
o
6
Trock
Condition
Checlc
,
!
o
7
o
o
Indicates that t~e 2841 has received an invalid
operation code, on invalid sequence of commands, on invalid Seek AddreH. The write
portion of the file mask hos been violatcd.
(See Set File Mosk.)
Seek Check
Protected
6
Mhslng
Address
Morlcer
A mining Addrt!u Marker, Which may indicate
a missing rccord is dctected Juring the eKCCUtion of command or chain of commands whicf'l
operates on successive Count Areas on a track.
The condItion detected is two successive records
on a track with equal bit conditions in bit 0 of
the Flog bytes, with nl) interv~ning Index Point.
MIssing Address Marker is set in coniu"ction
with No Record Found if there is no doto on the
track.
7
Overflow
Incompl.te
This bit is used with the Record Overflow
special feature. It Is set with other Indicators
to signal conditIons as follows:
Indicates defective track.
A Track Condition check is generated under the
following conditions:
t. If on overflow record is being reod, written,
or searched which overflows to a defective
track, the interrvpt occurs after the lost
byte on the previous trac k has been operated
on and before the first byte for the defective
track is requestcd from or sent to the channel.
In this case overflow incomplete Is also set.
2. If a single track command other thOR a
Search HA, Read HA, or Read RO is executed
on a defective track.
3. If a multiple track command or an overflow
operatiOf'1 attempts to switch from on alternate or defective track after an operation
has been executed.
Indicates that the file has been unable to complete a Seek becl]use:
1. The Seek address is outside the valid address
boundaries of the storage device. Unused
seek address bytes must be a valid address
for the device selected. Command Reject Is
also set.
2. less than six seek address bytes were sent.
Command Reject Is also $et.
3. The equipment foiled, which resulted In the
access mechanism going to either the inner
or outer stop.
1
Condition
Sets Overflow
Incomplete and
Other Indication
Overflow to a
defective trock
Track Condition
(Byte 0, bit 6)
Overflow from on .
alternate track
Track Condition
(Byte 0, bit 6)
Data check in data
oreo of overflow
record other than last
segment.
Data check
(Byte 0, bit 4j
Ove rflow to File
Protected boundary
File Protected
(Byte 1, bit S)
Overflow to wrong
track (Head number
unequol)
Seek Check
(Byte 0, bit 7)
Data Check Indicates that cyclic check error has been dein the Count tected in a Count Area read from the storage
FifJid
device. Data check (bit 4) II' byte 0 Is also
turned on.
Sense Bytes 0 and 1
SS7
12.
o
,-
TyPical I/O Command:
SIO
CCW
SEEK
READ HOME ADDRESS; CYL 180-HEAD 8
XIO
SIO
START I/O
DC
CCW
START I/O - IOCe
DC
"/9502
AREA CODE 18 - UNIT 2
A.L~D
DC
5
SEEK
DC
/4007
FOR RECORD SPECll"'IED
AT SEEI{ ADDRESS
DC
SEEK
DC
6
READ HO}Y1E
DC
/OOlA
ADDRESS INTO
DC
HOME
THIS ADDRESS
DC
0
DC
180
CYLINDER
DC
8
HEAD
HOME BSS
3
CHAIN
"f"'\
~;
13.
Sofnvare
The available software for the 2310 is well known to
most users.
In brief, there are programs to initialize the
disk cartridges, test them for errors, label and define length
of LET and FLET tables, and dump their contents when needed.
Both FORTRAN and ASSEMBLY LANGUAGE routines are
available to read and write to almost any area on the 2310
. disks.
System routines will define and label at your request
buffer areas and will assign sector addresses.
All the user
has to do is punch a define file card, or a STO REDAT A card
c
and the system will do the rest.
Moreover, all of this is
usually done with re-entrant subroutines enabling the user to.
be free of worries about using a device from different interrupt
levels.
The only penalty paid, as the sample program vlill
show, is time and overhead costs.
All 2310 routines eventually
call an I/O subroutine, DISKN which has been charg ed wi~h
the responsibility of controlling all
II 0 to the
servicing the resulting interrupts.
It should be noted that
2310 sand
FORTRAN calls are not overlapped, 1. e., control is returned
'to the user after the operation has been completed successfully
or not.
o
The software available for the 2311 as provided
SS~
·.-----............ ...... ..
"
,
..
~.,.- -~~~----------------~,
14.
is restricted to diagnostic programs and a disk initialization
program.
TSX does not support the 2841-2311 but programs
have been written "which allow the 1800 FORTRAN and
ASSEMBLY LANGUAGE user to fully utilize the 2841-2311
capabilities..
These routines are at the present time being
submitted as Type TV
programs~l)
Work in switching
T SX to the 2311 is also being contemplated.
The structure
. of these routines is similar to those used to service the 23l0s
but three differences must be mentioned:
(1)
The routines are not re- entrant.
Given the speed
of data transrnission, it is possible to rnask any interrupts
while the 2311 routine is generating a chain of commands
prior to executing the I/ O.
The time delay is comparable to
the time it would take to save and restore all needed parameters
if made re-entrant.
(2)
The I/O can be overlapped.
The return to the
user is made irnmediately as soon as the I/O is initiated.
The
user has the option to examine the final interrupt to determine
if the operation was successful o.r not or to ignore this fact
'comp1etely and continue with his program.
(3)
The different I/Os are requested through calling
sequences.. There is no special WRITE or READ as L~ the
1. See Appendix II for further explanation and exarnplese
M,~\ij;-;hi'i1" *,""wiri,Wfiitjfi
,. \Wrt' ftbttfttH#ii""#*#iNWWiTij'fj""'
- TW"fWfrwr@MwtiV'.epw .
!1PeRlr
en
ar
15.
2310, and most of the parameters in FORTRAN must be fixed
integers.
There are 10 calls that can be used to generate a
sequence of in~structions that will perform a certain I/O on the
;
2311 and two auxiliary calls; one that allows the user to write
his own command chain and the other a busy test routine.
In
machine language, a th-ird special call allows the user to get
to a table of
param~ters
and indicators saved by the 2311 ISS.
As in the case of the 2310, the 2311 calling sequences execute
all I/ a through a subroutine, DISKZ.
Having provided a brief sketch of the software available
for both 2310 and 2311 disk storage devices, it is possible now
to look at the programs run to obtain some timing comparisons.
The philosophy used was very simple; write and read back a
fixed length record to both disk units under similar conditions
and record the lapsed time between start and finish.
TIMING
Time
Routine
0
WRITE-READ I/O
2310
2311
• 3-22 sec.
Language
FORTRAN-TSX
- TTh1El
3200 words
7.188 sec.
TTh1E2
3200 words
L 609 sec ..
TIME3
3200 words
• 192 sec.
A.SM-TSX
$TIM:E
320 words
.262 sec.
ASM-ABS
TIME4
320 words
.. 918 sec.
ASM-TSX
ASM-ABS
16
It should be pointed out that a data area for use of all the se
routines was set up in FLET and '\lnprotected by means of
DWRAD.
This time was not included in the co:mputationso
Also J the time shown for
$TIME doe s not corre spond to a
single write of 320 words but to five repeated attempts to
write the same record.
T:his was caused by an error in one
of the routine s used and show s the timing effect of retrying
a'n operation that failed with a recoverable error (in this
case unrecove rable).
It is evident that the use of FORTRAN for disk I/O instead of Assembly Language is costly for the 231 o but that
in the ca se of the 2311 storage device s, at the pre sent time,
the difference is quite tolerableo
o
I~
'<)
PHYSICAL CHARACTERISTICS
2310
1. 02
2311
STORAGE CAPACITY (MILLION BYTES)
7.5
530.00
HIGH SPEED ACCESSIBILITY (milliseconds average)
·72.00
DATA TRANSFER (KILO BYTES)
3
75.0
156.0
MULTIPLE UNIT GROWTH
8
~
....~
203
Cl
Z
NUMBER OF CYLINDERS
20.3
10
~
2
NUMBER OF TRACKS/ CYLINDER
<
5. 1
CYLINDER CAPACITY (TIIOUSAND BYTES)
At
At
SECTOR
640
r-4
•
~
37.0
BASIC ADDRESSABLE UNIT
RECORD
MAXDAUM CAPACITY OF BASIC UNIT (BYTES)
3900
r-4
....•
o
o
~ ~
.........
·0
APPENDIX II
110 1
The 2841 and 2311 Subroutine Package
The 2841-2311 subroutine 5 were written to allow FORTRAN
and/or Assembly Language
users to avail themselves of
the se high speed storage device s.
The aim. wa s to or ganize the routine s
of the 2310 support programs.
To accomplish
along the line s
this~
a set of
three subprograms were developed:
1)
The interrupt service subroutine (ISS), DISKZ,
that ha.ndle s all I/O reque sts through its subroutine
~-"
~)
entry point and take s care of all re sulting interrupts inforilling the user J if he care s to know, of
the status of the last I/O executed.
2)
Ten calling sequence s which operate upon all or
any elernent of a record, track or cylinder (see
Table I).
3.
One additional subroutine with three entry points:
(a) One entry point, DISIf prog:-ams in the inJ?ut -stream is defined by
a JOB card.
The automatlc loadlng and executlng of the next stacked
program is initiated by anyone of the following:
C)
·EXECUT'
ON
CONFIGURATION
0\
6YM60l TAGLE
J\ElOCA1A~LE
~U~~OVTJ"JE~
o
~~~~~~~
ftELCADfD
~REACH
P~O(7RAM
CO#\E
ftEGtDENT
0, ',
"
Fig.
4
61/
.
-
-
._-
-
.EXECUTION CONFIGURATION
STUDENT VEft:"CN
•
5YMGOL TABLE
RELOAOfO
...--------------FO~ EACH·
P~C~~AM
F'XfO
'ftELOCATABL£ '
SUGttOUTtNES
o
Fig. 5
--
---
-~~
-
-
- - - - - - ---- ---- -
-----
-
--- -----.--
-- ...... -.-...
-.----.-.--~-----
1
o
1.
Execution of a CALL LINK or CALL EXIT statement:
These statements functionally terminate a FORTRAN program,,The primary difference between the s,tatements is conceptual.
CALL EXIT terminates a job and CALL LINK is used together with
COMMON to link overlay phases of ~ program.
The overlay of a program can be profitably used to- conserve
core in large programs. For example, all program variables
can be put in COMMON and initialized in a phase which is then_
overlaid by the main program. Figure 6 illustrates the source
deck arrangement for this example.
2.
Readin~
of a JOB card:
An attempted READ of a JOB card by the FORTRAN program will be
treated as an end-of-file condition for the input to the program.
This will cause the monitor to branch to the specified statement
if an ON ENDFILE statement is used or to terminate the job and
cause loading of the next program. The ON ENDFILE statement is
described later.
c
3.
Execution of an END statement:
For those who forget a CALL EXIT.
4.
An ,Input Data Error
An input error will be considered a job terminating error-unless
an ON DATA ERROR statement is used.
This statement is further
described under the FORTRAN compiler.
5.
A Console IInterruptl:
Pressing INSTANT STOP, RESET, INSERT, RELEASE, and START at any
time during execution will cause the current job to be aborted
and the next program found and loaded.
6.
Lin€ Count or Times Overflow:
These will be described below.
A· separate termination message is typed by the monitor for each of these
end-of-job conditions.' The JOB card is also typed £:or each job.
~/3
-
------. .
·~-······1
i
'EXAMPLE OF- CALL LINK
ENO
CALL
£~lr
- COMC'1CN A,0, ••• ....--......
$"00
'~J'1~AltZAr'Of'J
cCMMOru A,S, ...
$'-'00
I---o-'
o.
",,'!!':mt'lIf'ij'NU""lf'lP'""trliT
o
The PDQ system has two major drawbacks, however. Source language
error detection is minimal, and is unsatisfactory for use in a
teaching system.
In addition, the subroutines must be separately
loaded with each object deck. This constant loading becomes quite
expensive when a large number of short programs_are to be run.
TheC.S.C.H. FORTRAN maintains the speed and language attributes of
PDQ while reducing these disadvantages. The three principal components
of the system are
1.
FORTRAN Precompiler
. Since the. level of error detection r-equired could not be
provided in the compiler, a separate precQmpiler was
written which provides a high level of-error detection.
The total time -- both student and machipe -- required to
produce a working program is greatly reduced in this manner,
since the previous compile-execute-debug phas~necessary to
find one error is replaced with a precompiler phase which
detects multiple errors.
-
2.
o
FORTRAN Compiler
Some additional features have been added to the PDQ language
and some modification and deletion of existing statements
made.
statements have been included that provide communication with the batch processing execution monitor.
Subroutines and Batch Processing Monitor
A monitor has been added to the subroutine library which
allows batch execution with only one loading of the subroutines for any number of stacked execution jobs. Job
card-control, FORTRAN calls to the moni~ori output line
counts, program timing, program linking facilitie~, and
continuous no-halt execution of programs are features provided by the monitor.
Both_the compiler and precompiler operate in a batch processing mode
under control of JOB cards. This means that any number of programs
can pe processed to execution in three phaser~ ~i th only one -loading
of the precompiler, compiler, and subroutines. The procedure is
indicated in Figures 1, 2, and 3.
The entire system is card oriented, and no typewriter input or- listing
of source. statements is allowed.
It is also possible to suppress type--writer output during execution (by using a monitor option). in order to
achieve faster throughput.
c
616
FORTRAN Compiler
o
The compiler is essentially the PDQ FORTRAN "compiler with the followirig
mc:;>difiqations <;1esigned to'increase language and batch compiling capabili·ties: .
1.
"The compiler has a batch compiling option which suppresses
the halt between jobs. A card with eighty asterisks is
punched following the trailer cards for an object deck.
This permits easy separation of object decks when batch
compiling with no stops.
'2.
A JOB control card must precede each source deck. The
JOB control card will be recognized by-the compiler and
typed to provide a record of'jobs. The JOB card also
contains compile options that specify execution subroutine
configuration.
.
3.
All source input to the compiler is from cards~ Also, no
typed listing 'of the source deck or symbol table is allowed,
although the symbol table can be punched alphamerically.
4.
Subroutines cannot be compiled into the object deck.
5.
Almost all error detection has been eliminated. A program
with no precompiler errors will compile correctly, however.
6.
The following statements have been added to the PDQ language:
-
a•
CALL nnnnn
where nnnnn is a five digit memory address. This generates
a BTM instruction to the address specified. The operand is
the return address to the-calling program. This statement
_
is used for communication with assembly language subroutines.
h.
IF (SENSE LIGHT n) S1' S2
SENSE LIGHT n ON
SENSE LIGHT n OFF
where n is a single digit number and S1 and 52 are statement
numbers.
The IF statement is similar to the IF SENSE SWITCH
statement except that i t tests an internal logical indicator.
The test does not turn off the IIsenselightll. The other two
statements ar~ use~ to set tbe ten indica~ors.
Since the senselights -are internal indieators and require no
core storage,their use will require less object code and
symbol table space than the use of arithmetic variables as
indicators ..
,
(
c.
CHECKPOINT nnnn
This statement is the same as the STOP statement except
that the program does not halt after typing the s2ecified
number. This has been found to pe-a useful-logic flow
debugging aid without the slow execution typeout disadvantages-of TRACE.
d.-
ON ENDFILE GO TO Sl
where Sl is a statement number. This stat~ment specifies_
the number of the statement to be executed when the last
data card has been read. - This condition is defined by the
last card indicator or the reading 0.£ a JOB card. The
statement can be ,placed anywh,ere in the program but must
be -executed before the end of file condition occurs. If
this statement is missing, the reading of a JOB card will
terminate the program and cause execution of the following
stacked program.
e.
ON DATA ERROR GO TO Sl
This is similar to the ON ENDFILE statement and specifies
a program entry point if an input conversion error is
de±ected. An input error will terminate the job if this
statement does not precede the occurance of the error.
o
f. ' CALL EXIT and CALL LINK
These are described under the batch processing monitor.
The PAUSE and CONlROL statements have been deleted from the language and
the syntax of Procedure Statements has been changed.
FORTRAN precompiler
The FORTRAN precompiler is a one pass error checking program which
executes at approximately the same speed as the compiler. JOB card
recognitio~ and batch processing capabilities are included.
Some of
the major features of the precompiler are
1.
As far as is known, all non-logical checkstop errors are
detected. Execution errors sUch-as a'calculated sUbscript
going out of bounds.are, of course, undetected.
2.
A large number of separate diagnostics (approximately eighty)
are provided. Errors are identified by a two digit number
which is typed out along with the statement containing the error.
In
addi1;.ion to controlling the initiation and termination of programs
in a batch execution mode, the monitor, also provides optional monitor D "
functions during the execution of a program,,_ The features provided ar' ,
1.
output-Line-Count Monitoring:
The number of output lines allowed i,s specified- on the
JOB card. If this limit is exceeded, the job is aborted.
2.
Suppression of Typed Output:
A TYPE or PRINT statement will execute as a PUNCH stat~ment.
This results in considerable throughput increai:;e in a batch
processing environment. JOB cards are_punched when encountered in order to easily-separate output when listing
the cards from a series of jobs. Because of the construction
of the JOB card, a 407 board maybe easily wired to cause
ejection to a new page when a JOB card is encountered during
listing.
3.
Suppression of STOP statement:
In order to maintain the no-stop, batch processing concept,
the halt of a STOP statement may be suppressed.
4.
Execution Time Monitoring:
~\
V
Since the FORTRAN compiler produces very little in-line code
and practically all operations are performed by a few basic
execution subroutines, it is possible to provide software
execution timing capabilities. This is done by placing code
in fifteen basic subroutines which will increment a software
timer by an amount proportional to the length of time required
to execute that subroutine. ~his will increase the execution
time 0 f such a subro"u tine by 240 l<"-: S if the time monitor is
-disabled and by 1040/ls if the monitor is operative.
The
total increase in execution time has been found to be less than
3% if _the time feature is not used and approximately 15% if the
timer is used.
The software time count has been found to be
accurate to within 10% of the actual t~me with one-tenth of a
minute accuracy.
The maximum time limit is specified on the JOB card and overflow
is considered a terminal condition. The timer feature is primarilyused in closed-shop, batch processing operations where
quite lengthy jobs are often run-.
All of the above options are enabled or suppressed by sense switch settings
when the subroutines are loaded.
( "1_!
"
-
r
______
~~
_
________·__ ,,_··__ ··__ ·" ____
H_~_·_··
-
o
3.
Multiple errors in one statement can be detected. This
reduces the number of precompil-er passes necessary to
produce an error free program.
4.
Error checking is to a higher level than is required by
the compiler in order to produce valid code. For instance, .
key words are generally identified by only one or two letters
in the compiler but are checked for six letter-s in the p+,ecompiler.
Some of the major errors detected are
1.
Possibly undefined symbol
Any variable which haa not been defined in a previous statement
(in terms of card order, not logical flow) will be typed out as
a possible undefined symbol When it is used. This procedure
catches most undefined symbols. statement numbers which are
- undefined at the end of the program will be typed out.
o
-·2.
Branch to FORMAT statement
3.
Duplicate statement Numbers
4.
All DO Loop Errors
These include overlapped DO loops, modification of-DO parameters within the loop, end of theloop i~ previous statement,
etc.
5.
AII-- Subscripting Errors
These include contradictions involving the number of subscripts
-en a variable and complete checking of subscript structure.
6.
All FORMAT Errors
7.
All Procedure Errors
8. --
I/O statement References non-FORMAT statement
9.
Exponentiation -of Fixed Quantity
IO~
All Syntax Errors: (misplaced commas, etc.) in-Other Statements
In addition, some logical errors and some potential errors such as program
possibly too large are detected.
T1ie precompiler symbol table is the same size -as that of the compiler
so-that compiler symbol table overflow conditions can be detected.
- "';,.C
from
\
,
__I
The following special implementation restrictions exist but-, aside
the firs~, they hav~ no practical implications in our environment:
1.
-A FORMAT statement must precede the first I/O statement
to reference it.
2.
Maximums of three continuation cards, five negted DO's
and five levels of exponentiation are allowed.
3.
A maximum of twenty symbols can be undefined at any time.
The important known flaws in the precompiler are
1.
There is no way to detect all undefined symbols in a genera.l
program. The likelihood of undeflned symbols at execution
causing checkstops is reduced, however, because the monitor
fills in the unused core with fixed point constants. This
- produces invalid results in the case of an undefined symbol
but preve-hts the subroutines from being destroyed by the
Transmit Field instr~ction.
2.
The appearance of a variable i~a DIMENSION statement causes
that variable to be considered defined.
3.
Arguments of functions are incompletely checked bu~ this will ~;.
not produce checkstop errors. Mixed mode expressions in the
argument will be det~cted.
~
------------
......
~--....l
•
I~~~,~
"
W' L.-.
o
ABsrRACT
The following describes and addition to the programming system for the
IBM 1800 computer. The expanded system will support up to 16 remote teletype
terminals being used in a time-sharing environment primarily to solve realtime information processing problems.
INI'RODUCTION
1800 Hardware
The 1800 CPU makes extensive use of hardware interruyt levels and
data channels to operate its standard data processing I/O equipment. In
addition the 1800 can have analog and digital I/O capable of communicating
directly with almost any kind of equipment. The terminal system uses 2
words of digital input and 1 word of digital output (16 bits/word) to
control all 16 terminals simultaneously. No data channels or additional
hardware are employed.
TSX Programming System
The 1800 programming S,ystem provides many conveniences.
of two types, pr~cess and non-process.
Programs are
Process programs can be initiated by externally generated interrupts
or can be queued by any pror-;ram for execution. They can be a part of the
S,1stem skeleton which is in core at all times or they can be kept on disk
in core-image form. Process programs have highest priority and generally
use some analog or digital I/O.
Non-process programs ordinarily do not use the process (analog and
digital) I/O. They are usually stacked jobs of a conventional type to be
run under the non-process monitor.
When time-sharing, the system does the following:
(1).
Rursnon-process programs for background (job shop).
(2).
Periodically
(3).
Pern'.!i:tB externally or internally generated interrupts to
load and execute process programs at any time.
chec~the
queue for process programs.
Both (2) and (3) require that the non-process job be saved and restored when
the process work is finished.
The Terminal System
The terminals extend the time-sharing capabilities of the system by
allowing up to 16 users to communicate with the s.ystem simultaneously.
0··"
I,
Terminal activities \nc1ude:
(1).
Queing process programs for execution.
.-..........
'.'-'.----.~--~-
- 2 -
(2).
Communicating with programs during execution.
(3) •
Programming in NUtran (conversational fortran).
_ ;i;'
'e"""",'
Maqy other activities are planned.
Most of the terminal system capabilities are achieved by simply making
the standard TSX functions more accessible. The only programmin~ efforts
unique to the terminal system are the communications controller (simulated
by an in-skeleton program), and the time-slicing of terminal service programs.
iI.M\NIOTES
l,cc.:}·i:·j(;LOGY DEPT.
Pl)i:~:.)Ut;~ i.iNIVEHSITY
C f\ L \. )r·~ F T ~: t>. ~l P ! J S
HAMMOND, IN 46323
r)r·j
'r'\i\)
COMi:)UTi:f~
c
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.3 Linearized : No XMP Toolkit : Adobe XMP Core 4.2.1-c043 52.372728, 2009/01/18-15:56:37 Create Date : 2014:03:05 14:51:40-08:00 Modify Date : 2014:03:05 14:26:36-08:00 Metadata Date : 2014:03:05 14:26:36-08:00 Producer : Adobe Acrobat 9.55 Paper Capture Plug-in Format : application/pdf Document ID : uuid:ad9a94dd-2b92-7e4c-8be4-d3233a2166f4 Instance ID : uuid:cd29de91-c4eb-7243-ac06-ded96bc64730 Page Layout : SinglePage Page Mode : UseNone Page Count : 624EXIF Metadata provided by EXIF.tools