196904
196904 196904
User Manual: 196904
Open the PDF directly: View PDF 
.
Page Count: 64
| Download | |
| Open PDF In Browser | View PDF |
April, 1969
CD
Programming
a
Computer
by '
Tracing
Lines
~~
•
PER 100 I C Al S r-~-'-\"~
lZ
- -l.1--~10 S MARKET S
SAN JOSE CA
95113
104
269
four.
numbers
202/293-3910
WASHINGTON
301/539-5188
BALTIMORE
213/542-1501
LOS ANGElES
415/626-7157
SAN FRANCISCO
that can save you six figures
The concept of saving money through the purchase of used data processing equipment is no
longer the exclusive proper!y of a few EDP
supermanagers. Let I-C-E's professional staff
show you how to take advantage of these savings. Call or write for descriptive literature.
Corporate Headquarters:
1231 25th ST., N. W., WASHINGTON, D. C. 20037
iNTERNATioNAL
CO~pUTER •
EOUIPMENT, INC.
Designate No. 14 on Reader Service Card
goes better
with PDP-10
, C 0 i , :~' U'l' r: ;; J •\I; D
I. :.; 'l 0 .u'\ l' 2<) U
. ~\!~ nII, J 1 ::i 69
Time after time
after time.
You develop and debug programs in hours,
not weeks. No computer
operator between you
and the computer. You
interact with the computer
itself. Through anyone of
64 terminals. Any time.
Anyplace.
You run interactive programs, FORTRAN programs, assembly language
programs without hogging the whole computer.
While you run your program, 60 others can run
theirs, the lab can do
real-time data collection
and processing. All at
once. No waiting.
PDP-10 is the only computer now available that
does real-time, batch
processing, and timesharing simultaneously.
Come alive.
Interact with PDP-1 O.
COMPUTERS· MODULES
Maynard, Mass.
ignate No. 7 on Reader Service Card
Ever wonder what
everyone's doing ... on each job
... each hour .. and how much
it's costing you?
Vol. 18, No. 4 -
April. 1969
Editor
Edmund C. Berkeley
Associate Editor
Sharry Langdale
Assistant Editors
Moses M. Berlin
Linda Ladd Lovett
Neil D. Macdonald
Market Research Director
Advertising Director
Art Directors
I. Prakash
Bernard Lane
Ray W. Hass
Daniel T. Langdale
Contributing Editors
John Bennett
Andrew D. Booth
Dick H. Brandon
John W. Carr III
Ned Chapin
Alston S. Householder
Peter Kugel
Leslie Mezei
Rod E. Packer
Ted Schoeters
Advisory Committee
T. E. Cheatham, Jr.
James J. Cryan
Richard W. Hamming
Alston S. Householder
Victor Paschkis
Fulfillment Manager
William J. McMillan
Advertising Representatives
with PROMPT™
you can stop wondering
... You'll Know!
PROMPT, or Program Monitoring and Planning Techniques, is the new management
control package developed and application-proved by ARIES Corporation. PROMPT
provides detailed computerized reports to all levels of management so that they
know exactly what's happening at any phase of the program cycle ... down to the
most basic task performed by each man, every hour of the day. Effective im~le
mentation of PROMPT will drastically reduce expensive cost, manpower and schedule
overruns since it can accurately pinpoint these problem areas sufficiently in advance to enable rapid correction.
PROMPT is adaptable td any computer operation.
PROMPT assures that all levels of management get the information they
need, when they need it, so that any problem can be quickly and efficiently
corrected.
PROMPT optimizes deployment of available resources, and feeds back his·
torical data for estimating costs and evaluating individual performance.
PROMPT provides maximum return on your management investment while
being easy to use and economical to run.
If you're looking for a better way to plan, monitor and control
your projects, look into PROMPT. For complete information, write or call:
~ ~~!~~s,:£~~~c~TION
McLean, Virginia 22101 • Phone: (703) 893-4400
NEW YORK 10018, Bernard Lane
37 West 39 St., 212·279·7281
CHICAGO 60611, Cole, Mason, and Deming
221 N. LaSalle St., Room 856, 312·641·1253
PASADENA, CALIF. 91105, Douglas C. Lance
562 Bellefontaine St., 213·682·1464
SAN FRANCISCO 94123, Richard C. Alcorn
2152 Union St., 415·922·3006
ELSEWHERE, The Publisher
Berkeley Enterprises, Inc.
815 Washington St., 617·332·5453
Newtonville, Mass. 02160
Editorial Offices
BERKELEY ENTERPRISES, INC.
815 WASHINGTON STREET,
NEWTONVILLE, MASS. 02160
CIRCULATION AUDITED BY
AUDIT BUREAU OF CIRCULATIONS
Computers and Automation is published monthly at
815 Washington St., Newtonville, Mass. 02160, by
Berkeley Enterprises, Inc. Printed in U.S.A. Subscription rates (effective March 1, 1969): United States,
$18.50 for 1 year, $36.00 for 2 years, including annual directory issue - $9.50 for 1 year, $18.00 for
two years without annual directory; Canada, add
50;' a year for postage; Foreign, .add $3.50 a year for
postage. Address ail U.S. subscription mail to: Berkeley Enterprises, Inc., 815 Washington St., Newtonville,
Mass. 02160; address all European subscription mail
to: Box 52, 6354 Vitznau, Switzerland. Second Class
Postage paid at Boston, Mass.
Postmaster: Please sefid all forms 3579 to Berkeley
Enterprises, Inc., 815 Washington St., Newtonville,
Mass. 02160. © Copyright, 1969, by Berkeley Enterprises, Inc.
Change of address: If your address changes, please
send us both your new address and your old address
(as it appears on the magazine address imprint), and
allow three weeks for the change to be made.
Designate No. 9 on Reader Service Card
4
COMPUTERS and AUTOMATION for April, 1969
The magazine of the design, applications, and implications of information processing systems.
Vol. 18, No.4, April, 1969
Special Feature:
Electronic Data Processing and Management
20
GEOPHYSICAL DATA MANAGEMENT -
WHY? AND, HOW?
by Dr. Robert M. White
A proposal for an integrated geophysical data system a national system of data
banks where collected observations can be centrally processed and archived . . . and
then made available to all scientific disciplines.
24
UNLOCKING THE COMPUTER1S PROFIT POTENTIAL
by McKinsey and Company, Inc.
A survey of 36 major companies showing that almost all top management is having
trouble with its computer operations. What problems are managers having? And what
can be done about them?
34
The Implications of the Information Sciences for
INTERGOVERNMENTAL COOPERATION IN COMMUNICATIONS AND
EXCHANGE OF INFORMATION
by William W. Parsons
How a reliable flow of valid information among all levels of management in government
might be achieved - with a warning that current efforts to improve information systems
now in use will not meet future needs.
40
In the front cover picture~
the design of a circuit is being
converted directly into computer language. As the operator traces the drawing~ a MicroMetric digitizing system
converts drawing coordinates
into digital language. For
more information~ see page 52.
TOTAL COMPUTER SERVICE
by Clayton C. Lisle
How the data processing industry could be more responsive to the needs of the managers of businesses.
42
THE SYSTEM FOR DECISION OF PRINCE HENRY THE NAVIGATOR
Anonymous
The "Master Chart" begun in 1415 Prince Henry's remarkably successful system to
manage the information needed to make the decisions which made Portugal master of
half the earth.
NOTICE
*D ON YOUR ADDRESS IMPRINT
MEANS THAT YOUR SUBSCRIPTION INCLUDES THE COMPUTER
DIRECTORY. *N MEANS THAT
YOUR PRESENT SUBSCRIPTION
DOES NOT INCLUDE THE COMPUTER DIRECTORY. SEE PAGE 58.
Regular Features
Editorial
8
The Misdirection of Defense by Edmund C. Berkeley
Departments
and the Social Responsibilities of Computer People,
48
Across the Editor1s Desk - Computing and Data Processing
Newsletter
62
Advertising Index
46
Calendar of Coming Events
C&A Worldwide
61
Computer Census
45
58
Financial and Bu~iness News
I deas: Spotlight
10
The Power of A Computer Rests in What It Can Become, by Allan B. Ellis
Computer Market Report
38
The Computer Leasing Industry -
The Prospects, by I. and U. Prakash
Report from Great Britain, by Ted Schoeters
Multi-Access Forum
11
14
15
16
16
17
17
18
18
18
18
6
The Special Interest Committee on Social Implications of Computers of the Association for
Computing Machinery - Discussion
Censuses of Computers Installed - Discussion
"Machine Language, and Learning It" - Comments, by William F. Sherman
Opportunity for the British and European Computer Industry, by Gordon Hyde
Solving Numbles and Other Puzzles - Discussion
Forecast of Computer Developments, 1968-2000, by Carol Andersen
USASI FORTRAN to Be Extended
ACM Symposium on the Application of Computers to the Problems of Urban Society for Papers
Number of Time-Sharing Vendors, by Alan G. Hammersmith
Computer Market Report - Correction
Who's Who in the Computer Field, 1968-1969 -
COMPUTERS and AUTOMATION for April, 1969
Entries
New Contracts
60
New Installations
7
Call
LeHers to the Editor
59
New Patents
by Raymond R. Skolnick
19
Numbles
by Neil Macdonald
47
Problem Corner
by Walter Penney, CDP
10
Proof Goofs
by Neil Macdonald
5
Letters To The
Editor
Computers: Wonderful
but Dangerous
I missed your fine editorial ("How to
Spoil One's Mind as Well as One's
Computer") in the August issue of
Computers and Automation the first
time around. I did, however, go back
to that issue to read it when I read the
letters in your December issue.
I would like very much to have a
copy of the memorandum which gave
the details on the lies told by our
government.
I congratulate you on a very penetrating editorial and followup. It shows
a great deal of insight and good citizenship on your part. You are providing
a valuable and needed public service
with this kind of editorial. Computers,
like all other important inventions, are
just as dangerous as they are wonderful,
and I feel that we all need to be reminded of this frequently.
B. RUDY GFELLER
Systems & Procedures Analyst
Omaha, Nebr. 68102
Have trouble talking to
general management about
your systems ideas? Now your
problem is solved in •.•
~ Condonood
I
:
Computer
lncyclopodia
By Philip B. Jordain
This new reference work closes the
communications gap between systems men and non-technical management. It translates the technical
jargon of today's computer systems
into terms that both can understand and use. It will help you
speed the mutual development of
systems power that will eventually
decide many business battles.
As your basic reference; it makes
specific information available when
you want it. It adds the muchneeded link between technology
and management. And, as a guide
to "computerese" for management, it offers the understanding
necessary for better systems control.
604 pages, 65/8" x 9", $14.50
At your local bookstore or write
McGraw-Hili Book Company
Dept. 23-COM-469
Designate No. 8 on Reader Service Card
6
Ed. Note - I am .glad that you think
that Computers and Automation should
keep covering "information engineering"
in a broad sense, not a narrow sense. I
fully believe that computer people
should be «information engineers" and
should take as much trouble with the
data coming in, as with the wonderful
machines that they supervise.
City Planning
I have been reading your magazine
with increasing interest, but some difficulty (because of my lack of background). As my grasp of data processing
and systems design grows, with training,
I will enjoy your publication even more.
In this regard, I wonder if you could
answer a question that probably is not
new.
My sphere of experience and training is City Planning. As the computer
age continues to be felt in this field,
more and more planners will need training in "computerese." My main concern
at present is in locating· and contacting
people who have worked with models of
and use simulation programs dealing
with these models, and, generally, people
who apply data processing and analysis
to city planning problems. Do you know
of any people who would fit the above
description? Do you foresee any inclusion of· articles or emphasis on city
planning computer applications in your
magazine?
P. D. CREER, JR., City Planner
Planning Dept.
City of San Antonio
P.O. Box 9066
San Antonio, Tex. 78204
Ed. Note - In the last year, we have
published two articles somewhat related
to city planning: "A Linear Geographical Code for Management Information
Systems" in the April, 1968 issue; and
«Handling Small Area Data with Computers" in the Dec., 1968 issue. In addition, you might wish to contact Doxiadis-System Development Corp. (DSDC), an organization recently formed
jointly by Doxiadis Associates and System Development Corp. The objective
of this new organization is to "help
solve the problems of the American
city by combining expert knowledge in
urban affairs with expert knowledge in
the field of information sciences", and
they may be able to be of some help to
you, or direct you to other sources of inlormation. I believe the new company
can be addressed at the SDC address,
which is 2500 Colorado Ave., Santa
Monica, Calif. 90406.
Games Played by Computers
For a number of years I have been
following the listing in your directory
Talk to the IBM-360-the NCRG. E. - RCA - Etc.
with
THE UNIVERSAL LANGUAGE OF COMPUTERS
Secure a higher future for yourself in Data
Processing with a complete Home Study
COlJrS~ in COBOL Set your own time;
pace and plilce. In il short time you can
master COBOL. Write for information about
our NEW accredited: home study course.
i:lti
oL.lt[]moti[]n
\tl""'oining
in~.
Write Automation Training. Inc. Dept. 14
5701 Waterman' St. Louis. Mo. 63112
Division of Technical Education Corp.
"l.An iH.:credtted membC'r National Home Study COIIIICi(
Designate No. lOon Reader Service Card
issue of Games Played by Computers
and have noticed the improvement that
seems to be taking place gradually in
these programs. I thought you might
be interested in hearing that Behavioral
Science for 1969 has an article by
Eliezer Naddor titled, "GOMOKU
Played by Computers".
This article indicates that there were
two computers, one using a strategy
which depends upon board positions and
the other using an evaluation of the
score of each· empty square determined
bv ~ mathematical formula. In this particular game the second computer won,
but bOUl programs give me the impression of being able to be rated good, or,
possibly, excellent. I am sure you will
be interested in considering these for
inclusion in your next directory issue.
This directory issue has proved extremely valuable in the past in furnishing information about computers which
have been under consideration by secondary schools and colleges for use in their
educational program.
CARL E. HEILMAN
Coordinator Mathematics & Science
Bureau of General & Academic Educ.
Dept. of Public Instruction
Box 911
Harrisburg, Pa. 17126
COMPUTERS and AUTOMATION for April, 1969
t.
NEW PATENTS
Raymond R. Skolnick
Patent Manager
Ford Instrument Co.
Div. of Sperry Rand Corp.
Long Island City, N.Y. 11101
The following is a compilation of
patents pertaining to computers and associated equipment from the "Official
Gazette of the U. S. Patent Office," ,
dates of issue as indicated. Each entry
consists of: patent number / inventor(s)
/ assignee / invention. Printed copies
of patents may be obtained from the
U.S. Commissioner of Patents, Washington, D.C. 20231, at a cost of 50 cents
each.
TELEMATE 300 was designed' and developed to specifications of DACC, a time-sharing computer company that
knows the needs of users.
That's why Telemate 300 is more
compact, more reliable and elegantly
styled.
Top Quality. Competitively Priced
January 7, 1969
3,421,148 / George Aneurin Howells aI'ld
.a
~
Geoffrey Allen Hunt, Aldwych, Londoh,
England / International Standard Electric Corporation, New York, N.Y., a
corporation of Delaware / Data PrOcessing equipment.
3,421,149 / Ernest R. Kretzmer, Holrftdel, Paul Mecklenburg, Fort Lee, DortaId W. Rice, Neptune, and Williath
Ryan, Red Bank, N. J. / Bell Telephone Laboratories, Incorporated, New
York, N. Y., a corporation of New
York / Data processing system having
a bidirectional storage medium.
3,421,150 / Ralph A. Quosig, St. Paul,
and Norman L. Viss, Savage, Minn. /
Sperry Rand Corporation, New York,
N. Y., a corporation of Delaware /
Multi-processor interrupt directory.
3,421,151 / Howard F. Wong, San Diego,
Donald W. Liddell, La Mesa, and William F. Vollmer, Jr., San Diego, Calif.
/ The United States of America as
represented by the Secretary of the
Navy / Coded data translation system,
3,421,152 / W~lliam J. Mahoney, Darien,
Conn. / American Machine & Foundry
Company, a corporation of New Jersey
/ Linear select magnetic memory system and controls therefor.
3,421,153 / William J. Bartik, Jenkintown, Woo Foung Chow, Horsham,
and Edward N. Schwartz, Philadelphia,
Pa. / Sperry Rand Corporation, New
York, N.Y., a corporation of Delaware
/ Thin film magnetic memory with
parametron driver circuits.
3,421,155 / Hans Glock, Germering, Germany / Siemens Aktiengesellschaft, a
corporation of Germany / Magnetic
store.
January 14, 1969
3,422,283 / Donald E. Murray and Walter C. Seelbach, Scottsdale, Ariz. /
Motorola, Inc., Franklin Park, Ill., a
corporation of Illinois / Normal and
associative read out circuit for logic
memory elements.
(Please turn to page 62)
•
•
•
•
Easily portable
Data Rate: 300 Baud
Carrier and on/off lights
Highly selective tuned
oscillator circuit
• Half/full Duplex
• Triple tuned Circuitry
• AGC circuit
FULL WARRANTY
•
'. Acoustic isolation
min. 20 db
• Teletype or EIA
Interface
• Optional: Carrying case,
originate/answer, cable
interface, loud speaker
and --
WRITE FOR BROCHURE
Payroll Systems go on-line
faster with ALLTAxnr
the software package
available in basic COBOL
for all compilers.
ALLTAX calculates payroll
withholding taxes wi th one
standard formula and a table of
variables for each state and city.
It eliminates programming of
individual formulas and substantially reduces program maintenance and memory requirements.
ALLT AX is ,approved by all
states. It's easy to install, completely tested and documented.
ALLTAX is always up-todate. Automatic program maintenance for existing withholding
taxes and new taxes is available
at minimal cost.
Find out why more than 100
companies from coast-to-coast
are using this low-cqst package.
Write today for full information:
r-------:--------,
Management Infonnation Service
P.O. Box 252, Stony Point, N.Y. 10980
Gentlemen:
Please send full details on your
ALLTAX software package.
Name _______________________
Title _____________________
Company __________________
Address ____________________
City _ _ _ _ _ _ _ _ _ _ __
State
Zip
---c
----------------
Management Information Service
Stony Point, N.Y. 10980 • (914) 942-1880
COMPUTERS and AUTOMATION for April, 1969
Designate No. 11 on Reader Service Card
7
EDITORIAL
The Misdirection of Defense - and the Social Responsibilities
of Computer People
Few of the citizens of any nation would I believe disagree
with this proposition:
The main objective of the Defense Department of any
nation is to try to guarantee the successful defense of that
nation against attack.
For there is no doubt that armed. attacks by one nation
against another do occur - one of the most recent examples
being the military invasion of Czechoslovakia by the Soviet
Union and four other nations in August 1968. In Czechoslovakia, the government chose not to resist the overwhelming
force, but to try to adjust to the demands, i.e., surrender.
This was also the choice made by the commanding officer of
the U.S. electronic spy ship Pueblo when the ship was taken
over by North Korean naval vessels either just inside or
outside North Korean waters.
In the case of the Defense Department of the United States,
there is now substantial evidence that its main objective has
shifted - it is only secondarily "the successful defense of the
United States against attack" and is mainly something else.
In fact there is good evidence that the something else is the
serving of the interests of what President Dwight D. Eisenhower identified in 1960 as the "military industrial complex"
and warned Americans against.
What is the military industrial complex? Briefly, it is a
portion or segment of the United States, consisting of industries, regions, lobbies, and people (of many kinds), who make
a great deal of money (profits, income, salaries, wages, research and development grants, pensions, consulting fees, etc.)
from the vast budget of the U.S. Department of Defense,
some $80 billion a year. According to tables in a book The
Depleted Society by Professor Seymour Malman, 73% of this
budget has been paid to 100 companies.
From 1965 to 1967, the main reason the people of the
United States put up with the enormous, rising costs of "defense" was the pair of arguments: "We have to fulfill our
commitments to the government of South Vietnam" (no matter that it was the ninth dictatorship since Ngo Dinh Diem
was shot), and "We can't let our boys down in Vietnam we must give them all they want or need".
But in 1968 it became clear that the war in Vietnam was
not being won. By 1969, over 32,000 Americans had been
killed there; over 150,000 Americans, wounded; over 4000
planes and helicopters had been lost; over $100 billion, spent;
more bombing tonnage had been dropped in Vietnam than
the United States dropped in all the theaters of World War
II; and still no substantial progress. What is the main
trouble? Basically, we cannot tell the difference between Vietnamese on our side and Vietnamese on the other side, and so
our fire power produces hatred for Americans on a large scale.
In 1965 it may have seemed true to many people in the
United States that "defense of the United States" required
winning a land war in Asia more than 9000 miles away from
California.
But it looks now as if the people of the United States no
longer believe that fighting such a war is necessary to our
interests, and they want the war stopped. So the civilian
government of the United States is saying to the Defense Department and the Saigon government, "N 0, with 500,000
American soldiers in Vietnam, you cannot have any more".
8
And a president of the United States has been denied reelection to the presidency because of the war in Vietnam.
As a result, the theory and practice of the U.S. Defense
Department and of the U.S. military industrial complex are
being questioned by thousands of influential persons, including
Senators and Congressmen. Even President Nixon in one of
his campaign speeches promised to bring the war in Vietnam
to a conclusion within six months of his inauguration.
The way in which the military industrial complex operates
is particularly clear in the present pressure from the Defense
Department and associated defense industries to obtain public
approval for the proposed Sentinel, "thin" Anti-Ballistic
Missile System. The proposed system has aroused a great deal
of opposition in the U.S. Congress and in Boston, Chicago,
and elsewhere in locations which are threatened by the proposed anti-missile sites. Clearly these sites will increase the
danger of those areas becoming priority targets in event of, a
nuclear war. In fact, as soon as the first antimissile has been
fired against the first incoming missile, according to a statement by Senator Edward Kennedy, then radio location of the
second incoming missile becomes impossible, because of the
effects of radiation from the nuclear explosion in the high
atmosphere! But does the Defens~ Dppartment honestly and
patriotically admit this flaw? It does not.
Instead, the Pentagon makes use ot an Assistant Secretary
of Defense for Public Affairs and a Chief of Information
Office of the U.S. Army. Both these offices with a total budget. of over $6 million a year have been "programmed" into
the public affairs plan of Lt. Gen. A. D. Starbird for "promoting" the Sentinel Anti-Ballistic Missile system. He is to
provide for "speaking engagements, information kits, exhibits,
films, press releases", etc. In other words, the Pentagon is
using the taxpayers' money to try to persuade the taxpayers
to support a technically illogical project. For example, the
Selectmen of Reading, Mass., are being invited by the Army
on a sightseeing trip to anti-ballistic missile centers.
The military industrial complex (the MIC) by its very nature, evolution-wise, cannot be considered to be really interested in the defense of the United States. Since a large part
of the MIC could not exist competitively in the civilian
market, it must continue to seek large funds from the government, using good arguments if they exist, and any arguments
at all if good arguments do not exist. What it is really interested in is making money from defense contracts. So the real
preferences of the MIC are for billion dollar procurement
programs, which sound meaningful and which can be escalated, even if technologically they are unsound, logically they
are unreasonable, politically they increase the insecurity of the
United States, and financially they threaten the solvency of
the United States and the deepening neglect of our domestic
needs.
Why should computer people be concerned with the interrelation between the defense of the United States and the
military industrial complex?
Computers have been one of the scientific and technological
miracles which have enabled the military industrial complex
(Please turn to page 41)
COMPUTERS and AUTOMATION for April, 1969
b. ~
FR-80 is for
certain birds
People close to computers think they see
the graphic output problems very well. To
render an engineering.drawing from a
digital tape, they might recommend a
Stromberg 4060. For charts and graphs
from tape, a Calcomp 835 is an excellent
choice. 3M's EBR is suggested for forms
generation. An E-K KOM-90 qualifies for
outputs of personnel records.
But a man somewhat above the action can
see that all of these things might be done
by one machine. Instead of four trained
operators, he'd need only one. Instead of
four maintenance contracts, one. Instead of
four rooms and four sets of supplies, one.
Instead of four partly idle, special purpose
systems, a busy all-purpose one.
If you are that man, congratulations.
Here's how you sell your department
managers on a centralized film-recording
system.
Your chief engineer: Tell him his drawings
will be of graphic arts quality - sharper
than any method short of re-drawing,
sharp enough to stand up through
multiple reproductions at E-size.
Your marketing manager: Tell him his
charts, graphs and tables can be in any
format, with any symbols, any line weights,
and of a quality perfectly ~uitable for use
in printed proposals and reports.
Your personnel manager: Tell him he will
be able to do selective, computerprocessed listings on the same pass with
the film recording. Much faster than twostep processing and sharp enough for
multiple reproduction.
Your production manager: Tell him his
parts lists will be keyed to any desired
scheme, placed in any format of your
choosing, selectively ordered, rapidly
g~nerated, and just as sharp as the
_
/engineering drawings.
. FR-80 is the first system to combine a fast ~
computer and the latest developments in
~\
precision CRT imaging on 16 or 35 mm
film. No other electro-optical system can
touch its resolving power (80 line-pairs per
millimeter) and none is so versatile.
FR-80 will accept any tape format and
generate any graphic output.
Write. Even a wise man needs more
information.
Information International, 545 Technology
Square, Cambridge, Mass. 02139,
(617) 868-9810;
1161 West Pico Boulevard, Los Angeles,
Calif. 90064, (213) 478-2571.
[000]
III
INFORMATION INTERNATIONAL
Designate No. 25 on Reader Service Card
IDEAS: SPOTLIGHT
PROOF GOOFS
Neil Macdonald
Assistant Editor
We print here actual proofreading errors in context as
found in actual books; we print them concealed, as puzzles
or problems. The correction that we think should have been
made will be published in our next issue.
If you wish, send us a postcard stating what you think the
correction should be.
We invite our readers to send in actual proofreading errors
they find in books, not newspapers or magazines (for example,
Computers and Automation), where the pressure of near-athand deadlines interferes with due care. Please send us:
( 1) the context for at least twenty lines before the error, then
the error itself, then the context for at least twenty lines after
the error; (2) the full citation of the book including edition
and page of the error (for verification); and (3) on a separate sheet the correction that you propose.
We also invite discussion from our readers of how catching
of proofreading errors could be practically programmed on a
computer.
For more comment on this subject, see the editorial in' the
September 1968 issue of Computers and Automation.
Proof Goof 694
(Find one or more proofreading errors.)
Economic self-sufficiency should be one of the basic aims of
education in a democratic societY'. Gainful employment is a
joyful experience. Honest work has moral value. One's selfesteem is firmly established when the world is willing to pay
for your services; skilled, semi-skilled or unskilled. Every man
must ultimately be inducted into the economic fraternity as
a contributing member. In a responsible society there must be
no such thing as a total drop-out. When a youngster leaves
school he should be able to choose between two exits - one
marked TO WORK, and the other TO MORE EDUCATION. There should be no exit leading to NOWHERE,
and certainly no unguarded exits. It is not just a living we
are responsible for but a life.
What, after all, is education? The dictionary definition is:
"To bring up a child physically or mentally; to educate. Also:
to develop and cultivate mentally or morally; to expand,
. strengthen and discipline the mind, or a facultY', etc. To prepare and fit for any calling or business by systematic instruction; to cultivate; train; instruct. Synonym: Develop, teach,
inform, enlighten, indoctrinate.
It follows, therefore, that it becomes the duty of all who
educate to remove the obstacles - psychological, physical,
emotional, intellectual or environmental - which might undermine this definition.
I am sure that the story is apocryphal, but it is told that
someone asked Michelangelo what method he used for sculpting his Moses. "It is very simple," he is credited as saying.
"You just. take a chunk of marble and chop away everything
that doesn't look like Moses."
The implication is clear: that within any crude piece of
stone (or child) lies a work of art, if you know what to chop
away in otder to reveal it.
10
The Power of A Computer Rests
in What It Can Become
Machines execute procedures and each machine is the
embodiment of the procedure it executes. This is an important relationship that exists for all machines, but people are
just not in the habit of speaking about machines in this way.
It means, of course, that knowing in detail what a particular
machine does - how it works - is enough in theory to know
what procedure it is executing. This is true because when we
say that a machine is the embodiment of the procedure it
executes, we are saying, in effect, that a statement of a procedure describes the machine needed to carry out that procedure. Thus mechanizing means thinking about procedure,
not about hardware, and once we state a procedure explicitly
we should not really be surprised that a machine can be built
to execute it.
Now, a computer is a device whose job it is to accept a
statement of a procedure and to imitate the behavior of the
machine implied by that procedure. This statement of procedure is called a computer program and is usually thought of
as a set of instructions for what the machine is to do. But a
computer program is more like a blueprint which the computer uses to build itself into the particular machine needed
to execute the particular procedure described by the program.
It is as though the computer were armed with pliers and
screwdriver rebuilding itself to conform step-by-step to the
elements of the procedure; and it will then function as that
machine.
A computer without a program will do nothing, whether
or not it is plugged in, because computers are not like other
machines. In a sense the computer is not a machine at all in
its own right, and yet it can become many machines, in fact,
anyone which can be fully described to it. While the power
of most machines is in what they do, the power of the computer rests in what it can become, and the essential idea of
a computer is that it is an incomplete machine ready to be
completed in an infinite number of ways, each way producing
a different machine. Thus, a computer program is at the
same time an explicit statement of a procedure and the blueprint of a machine needed to carry it out, and whether or
not a computer can execute a given procedure depends most
heavily upon how well we understand the components of that
procedure, and how imaginative we are in conceiving procedures in terms of the basic elements of which they are
comprised.
- From Discussion by Allan B. Ellis, Harvard University, p. 32, in Educational Technology: New Myths and
Old Realities, Reprint No.6, Harvard University Program on Technology and Society, Cambridge, Mass.,
[]
1968.
It should also be emphasized that the child will have to
help ",:ith the chopping. Discovering ones self is hard work.
- From Everything But Money by Sam Levenson, pp.
229-230, published by Simon and Schuster, Inc., 630
5th Ave., New York, N.Y. 10020, 1966, 285 pp.
Solution to Proof Goof 693:
Paragraph 5, line 7: Replace "hysterical" with "historical".
[]
COMPUTERS and AUTOMATION for April, 1969
MULTI-ACCESS FORUM
THE SPECIAL INTEREST COMMITTEE ON SOCIAL IMPLICATIONS OF COMPUTERS OF THE
ASSOCIATION FOR COMPUTING MACHINERY
I. To Mr. M. Stuart Lynn, Editor in Chief
Communications of the ACM, IBM Scientific Center,
6900 Fannin St., Houston, Tex. 77025
From: Paul Armer, Robert P. Bigelow, Michael A. Duggan, Roy N. Freed, Herbert R. J. Grosch, Patrick J.
McGovern, Anthony G. Oettinger, Donn B. Parker,
Stanley E. Rothman
President B. A. Galler has recently dissolved the Special
Interest Committee on Social Implications of Computing because of a "lack of interest." We deplore this action, and call
upon our fellow members to join us in resurrecting the only
group in ACM through which a member can show he cares
about the interface of the computer industry with the rest
of the world.
At the Spring Joint Computer Conference in 1967, the then
SICSIC ran a progr~m on the social problems engendered by
computers. Over 100 people came.
And these problems are many. How about Privacy and
Data Banks? How about the problems of Computers and
Communications? And the regulation of the Computer Industry by various branches of government? What about automation and worker displacement?
Growth projections say that by 1980 the Computer Industry will be bigger than the automobile complex. Will the
computer injure the mind of man the way the automobile has
beaten his body? The automobile has changed this nation ...
its living standards, its housing patterns, its work and recreation habits, and, above all, its transportation. The computer
may have a similar effect. If, for example, developments in
computers and communications should cause many people
to work and to learn at home, there might be no need for
huge offices and crowded schools. The implications for society
of such a change are enormous.
ACM claims to be a "profe~sional" society. But the great
difference between a trade and a profession is that the tradesman is interested in his job - his work - and that alonewhile the professional feels that he and his fellow professionals owe a duty to their fellow man to use at least some
of their professional training and talent to improve the status
of mankind. ACM's image as a professional organization is
not helped by shutting down the only open ended group it
has which is concerned with how computers affect our
citizenry.
Under the By-Laws, SICs and SIGs can include nonmembers of ACM. In the area of social implications this is an
asset. The disciplines of sociology, education, business management, law, and medicine, to name a few, all have a function in such a Special Interest Committee. The primary interests of several of the undersigned are in fields other than
computing. And we have friends who are equally concerned
about the impact of this new technology on the fabric of our
COMPUTERS and AUTOMATION for April, 1969
nation, but who may not meet the requisites for ACM
membership.
To reactivate the Special Interest Committee on Social
Implications requires a petition to the ACM Council. We are
starting such a petition. If you want to join us and are willing
to do some work, please write to Robert Bigelow, 39 Grove
Street, Winchester, Massachusetts 01890, or telephone him at
617 -7 42-8300.
II.
To:
From:
Dr. B. A. Galler, President
Association For Computing Machinery
The Computing Center
1000 North University Bldg.
The University of Michigan
Ann Arbor, Michigan 48104
Robert M. Shapiro,
Secretary, SICSIC
Massachusetts Computer Associates Inc.
480 Seventh Ave.
New York, N.Y. 10001
I find it .!post disturbing that the president of the Association for Computing Machinery has dissolved a committee,
the Special Interest Committee on Social Implications of
Computation (SIC:~), without informing the secretary of that
committee, namely myself.
I also find the article printed in Computerworld based on
an interview with Jean Sammet distasteful and misleading.
In particular:
( 1) The article asserts that SIC2 has no mailing list. I
am the secretary and have never been contacted by
Jean Sammet or anyone else in ACM in respect to
the mailing list. In point of fact, there exists a mailing list of over 100 SIC2 members.
(2) The article asserts that SIC2 has done nothing.,
SI C2 has in fact organized round table discussions at
various meetings. SIC2 is at this very moment active
in the New York area. I enclose a paper written
jointly by members of SIG~ with a resolution based
on that paper and unanimously passed at aNew
York SIC:~ meeting held on February 20, 1969.
I formally request that SIG~ be reinstated immediately. I
also request that the ACM Council make an effort to undo
the impression created by the publicity about the dissolution
of the group - an impression to the effect that computer
people are not concerned with social or political issues. Minimally, the ACM Council should recommend that the enclosed
paper, "On the Social Implications of Computers," be published in Communications Of The ACM. It deserves at least
as much space as has been devoted to the "Code of Ethics"
issue.
11
III.
Enclosure in Mr. Robert M. Shapiro's letter: "On the Social Im:plications of Computers"
The responsibility of interpreting and informing the computer industry of the social implications of computing is the
minimum mandate of (SIC)2. Also within its scope should
fall the responsibility of public education, persuasion, within
and without the industry, to further the judicious use of computers in ways that further their social benefit, and to curtail
the use of computers in socially undesirable ways. Investigation and analysis without such advocacy is not only futile but
potentially dangerous, for it could mean that decisions about
the use of computers which have great social impact will be
made by those with no knowledge of the values and limitations of the tool.
The scope of this investigation and advocacy of the socially
beneficial uses and implications of computers is virtually
unlimited. The computer is no more than a tool. Nuclear
energy can be used for generating power or for building
bombs and warships. A computer can be used for medical
research, for guiding spacecraft to the moon or for guiding
nuclear warheads to destroy human life. It is a tool used
not by individuals according to conscience but by society at
large through corporate, educational and governmental institutions.
Having knowledge of the socially destructive potentials and
uses of computers and not sharing that knowledge with society
is a failure of our professional, civic, and moral responsibilities. A doctor or medical association which did not strongly
oppose the improper use of a dangerous drug, by urging the
adoption of laws forbidding its use, for example, would not
be meeting any of its responsibilities. A scientific association
would be meeting its responsibilities to the public by reporting a lack of funds for essential lines of research.
The social implications of computers affect almost every
institutional structure and enterprise in the country today
and computers are having an increasing effect on the personal
lives of every citizen. Because of the decision-making structure in our society, and the newness of the computer industry,
few of our decision makers, corporate, or governmental, are
computer professionals. This all points to a strong need for
computer professionals to attempt to educate and influence
our social decision makers.
It is in the nature of a democratic free enterprise system
that that which is not forbidden will be done if a profit can
be derived from it. It is in the nature of state socialism that
only those enterprises with governmental support will be
undertaken, for only they wiII receive the financial support
required. It is in the nature of bureaucratic institutions that
change will be resisted, and it is in the nature of centralized
authority and institutions to infringe upon individual liberties
and domains without limit unless restrained by the people in
their own behalf.. Because our society is a mixture of all these
things, our approach to analysis of the social implications of
computers and advocacy of their beneficial uses should take
all of these into account.
To mention some of' the social implications of computers
in each of these areas, only briefly consider the following:
financial corporations have found it profitable to install large
data processing systems at the cost of individual customer
service; space research and military projects receive astronomical government grants while educational and medical
research receive only a fraction of these amounts; banks are
very slow to establish computing networks which will ultimately eliminate the need for both money and securities.
And finally, the Government, even while the issue is being
raised in Congress, is proceeding virtually unchecked in its
program to establish mass data banks containing information
on every citizen.
To have an educational or persuasive impact on the uses of
computers and their social effects, individuals and professional
12
organizations must approach the decision-making structures
of the society. These institutions are, in increasing order of
social decision-making power, the people, the corporations
and educational institutions, and the government.
The people can be influenced through public education
campaigns, using the power and facilities of the media.
While this education is vital, it will have little direct impact
on social decision-making. Public education would, however,
help create demand for social decision-making by more
powerful institutions, corporations and government.
Corporate decision-making is influenced to some degree by
public demand, to a greater degree by government control, to
the greatest degree by profit potential. This last and most
important factor, independent of the other two, seems little
susceptible to education or persuasion.
Government decision-making is clearly the most critical in
most social issues, and the use of computers is no exception.
The government itself is a great user of computers. It stimulates corporate use of computers by contracts. It is responsible for restrictive legislation and for encouraging subsidies.
It seems then that the most immediate and effective forum
for education and advocacy on the implications and uses of
computers is the governmental decision-making apparatus.
Since our governmental institutions function, and reach and
implement decisions, through political dynamics, it is hard to
conceive of being concerned with the social implications of
computers without acting in the political arena as educator,
advocate, lobbyist and, if necessary, even partisan.
It is our contention, therefore, that (SIC)2 and ACM
must abandon the misguided concept of professional detachment from political issues and be willing to take and advocate stands on political issues involving the use of computers.
In fact, (SIC) 2 may as well dissolve if it does not recognize
and accept its responsibility in the political arena, for it will
then have no relevance to the social implications of the uses
of computers.
There is no such thing as the "professional neutrality"
which is always invoked to prevent a professional society from
taking public or political stands on social issues. Edward
Teller, who favors a defense policy based on nuclear superiority and the threat of their use, is a "neutralist." Linus Pauling, who recognizes the horror he helped create and wants to
do away with it, is a "political activist." To advocate a
theory of professional neutrality is to exhibit a deep naivete
about the social dynamics of our society and is to take a position supporting the prevailing or establishment position.
Because of the tendency of government to assume any
power not specifically reserved or prohibited by the people,
taking no position on data banks and invasion of privacy is,
in effect, siding with the proponents of mass data banks. They
will be created unless defeated by public opposition because
they are convenient to the government. Not taking a position
-allows the continuance of the government-sponsored myth
that adequate safeguards can be built into a data bank computer system to prevent improper use. The general public
can be confounded by the mystique surrounding a computer:
we can see beyond the technical problems and ask if any
group of people in a less than perfect world could be trusted
with access to such an information system. The decision,
when made, will be a political decision, made by elected representatives and probably along partisan or at least ideological lines; so the position of the professional body must be a
political position.
Similarly, political positions must be taken on the issue of
the programming institutes in the computer industry. We all
know that most of them are frauds and damage both the
industry and the public. Action must be taken by (SIC) 2 on
all three decision-making levels to meet our responsibility to
COMPUTERS and AUTOMATION for April, 1969
police our own industry and profession. A campaign of public
education must be started to warn against fraudulent practices
by these "trade schools." Corporate, governmental and educational users of computers must be convinced to drop all
support either as subsidizers or clients of fraudulent schools.
Finally, (SIC) 2 must press for legislation establishing licensing and regulative bodies under either government or industry control. This last is a political decision and can be accomplished only through political means. Taking no position
aids the continued existence of these fraudulent companies.
On the far greater and more controversial issues of the war
in Vietnam and military uses of computers, we can no
more easily shirk our personal or professional responsibilities
to take a stand, against these enterprises. Arguments to the
contrary assume that professional responsibility is somehow
totally divorced from personal and moral responsibility. It is
properly a subset of personal and moral responsibility and
should conform to and follow from it.
Compartmentalizing the personal man from the professional
man creates a very schizoid, alienating society, which in fact
ours is. The computer is a tool devoid of morality or social
responsibility, but the men who use it are not. Denying professional responsibility to affect decisions on how computers
are to be used makes the professional indistinguishable from
his tool.
The war in Vietnam is immensely destructive in social consequences to our society, as well as to the Vietnamese, in life,
resources, moral energy and political cost. The New Yark
Times has recently reported that the Nixon Administration
has decided that the war must be ended, by compromise if
necessary, as soon as possible because,it is dividing the country
and using too much of our resources - in other words, the
social costs are too great. As individuals we clearly have a
political interest iIi whether the war is allowed to continue.
As professionals we are obliged to take a stand also, both as
a subset of our personal responsibilities and because computers are used so heavily in the war effort.
Again, (SIC) 2 or ACM should act on all three levels of
decision-making. It should take a stand on the social implications of the war in the interest of public education. It
should urge all computer professionals to take professional
stands, including the refusal to use their professional talents
to support the war effort. And it should initiate and support
political action to end the war. The single voice of a professional organization can be more effective than the independent voices of individual members.
The computer professions and the industry are vital to the
war effort and the defense industry in this country. Had warmaking and defense been computerized in the 1930's and
1940's, German computer professionals would have had a
clear moral and professional responsibility not to cooperate
with the Nazis, and would have been justly condemned for
not exercising it. Our case today is little different. Once
embarked upon a policy, however disastrous, the government
has a tendency to continue, especially when so strongly encouraged (in their own interest) by the military-industrial
and defense establishments. Silence in this situation is tantamount to approval and acquiescence. The supporters of the
war can claim the alleged neutralists as their own.
Defense policy, a clearly political issue, relates very closely
to a general question of the society's allocation of resources.
This allocation is affected by the executive and legislative
areas of government, influenced only by political pressure
and action. As professionals engaged in the use of computers
we can and should take positions on how computers are to
be used" which means how the society will allocate its resources. Our role should not be limited to opposing socially
destructive uses of computers but must also include promoting socially beneficial ones. We should be lobbying as a proCOMPUTERS and AUTOMATION for April, 1969
fessional organization for increased resource allocation to
fields like education, research, medicine, social welfare and
urban planning, fields in which computers can play an important and socially beneficial role. Thus we, as computer
professionals, can fulfill both our personal and professional
responsibilities to society, using ourselves and our tools in its
hest interest.
A final point relating to the war and the other general subjects covered here. Other professional organizations of doctors, teachers, linguists, scientists, historians, psychiatrists, and
lawyers have confronted the question of political stands on the
war and other social issues. All have debated professional
neutrality. Many have rejected it in whole or in part and
taken political stands.
The New York Times of February 9, 1969, reported that
"a deep groundswell of discontent is rolling through
scientific communities from Moscow to New York and,
perhaps, even isolated Peking. It was manifest last week
as activists within the American Physical Society tried to
enlist the support of their colleagues in helping to fight
what they called the 'overwhelming' domination of research by the military. . . . they were united in a desire
to shift the emphasis in research from military goals to
pressing social needs. . .. The urged that machinery be
created to help scientists better educate the public to
assess such controversies as those on the A.B.M. and on
biological warfare."
Professional organizations have long taken stands on other
political issues, within and without the field of their purely
technical competence: the A.M.A. on Medicare, abortion,
euthanasia, and marijuana; the American Education Association, on decentralization. In the case of the computer industry, there are few social issues not within our competence
because the computer has pervaded all functions of society.
It is time (SIC) 2 and ACM fulfilled its professional obligation to society to speak out on how computers are used rather
than just how to use them. Only in an organization open to
free exchange of ideas and debate can we provide society
with truly responsible and professional information and service.
RESOLUTION
As professionals in the computer field and members of
(SIC) 2, we have the responsibility, through our professional
association, to oppose the use of our skills for destructive and
anti-social ends. Therefore, we urge that ACM adopt these
proposals as part of its national policy:
1. We oppose the war in Vietnam, U.S. military presence throughout the world, and economic and
political subversion of other nations. Since there is
widespread involvement of our profession in these
endeavors, . we urge all computer professionals to
review the moral consequences of their involvement
in furthering these efforts.
2. We oppose discrimination as practiced in the computer field by direct or indirect means such as educational requirements, arbitrary testing procedures,
and restrictive policies.
3. We oppose the establishment of mass data banks
which pose a threat to our privacy and concentrate
power in the hands of a few.
4. We oppose the economic exploitation of the uninformed by unscrupulous computer schools.
We support the implementation of accrediting standards for the computer educational field.
S. We support the active' encouragement, developmeilt,
and funding of programs for the constructive application of computers toward the solution of the
many problems faced by our society.
13
IV.
From the Editor of "Computers and Automation"
. Several questions arise from the foregoing, on which I
would like to comment not only in the capacity of the editor
of Computers and Automation but also as in the capacity of
one of the first handful of members of the Association for
Computing Machinery when it was founded in 1947, and as
its first secretary, 1947-1953.
First: It seems to me unlikely that the President of the
Association for Computing Machinery by his sole action has
the power under the Constitution to dissolve a Special Interest
Committee. (If the Constitution is now worded in such a
way that he can, without assent from the Executive Committee or the Council, then it seems to me that this power
should be promptly canceled.) Accordingly, SICSIC still
exists, because his action is null and void.
Second: If SICSIC has in fact been dissolved, then it
seems to be desirable that the President or the Council should
forthwith reconstitute it. There is no doubt at all that it is a
vigorous and functioning Special Interest Committee.
Third: Of all the facets of computers and their applica-:
tions, for business, for industry, for science, for the military,
etc., it seems strange indeed that applications for the advantage of society and the social implications of computers should
be placed beyond the pale of professional concern of ACM
members. In fact, such proscription is nonsense. It is at the
same level as the action of the Tennessee legislature many
years ago in passing a law making it illegal to teach the theory
of evolution in schools in Tennessee.
Finally: If any petition is in fact necessary to reestablish
the Special Interest Committee on Social Implications of
Computers, we invite all interested readers of Computers and
Automation to write to Mr. Robert Bigelow, 39 Grove St.,
Winchester, Mass., and enroll on his petition.
CENSUSES OF COMPUTERS INSTALLED
I. From M. L. Melville
Public Relations
NCR (National Cash Registe'r Co.)
Dayton, Ohio 45409
Computers and Automation has compiled an outstanding
record of factual and objective reporting on the data processing industry. For this reason it should be pointed out that
the "Improved" Computer Census published for the first
time in your February issue is grossly erroneous with respect
to the installation figures listed for NCR.
We do not know the composition of the data base from
which these figures were derived, but we suggest that this be
carefully re-examined.
Actually, the census figures previously published by your
magazine gave a generally accurate picture of the numbers
of systems installed by our company to date, although we
are not in a position, of course, to comment on the data
listed for other manufacturers.
II. Report of a telephone call from Norman M .. Bry.
den, Honeywell EDP, Wellesley Hills, Mass., to the
Editor
Mr. Bryden expressed shock over the figures published in
the February issue for Honeywell's computers installed. He
said they ranged from 70% to 20% of the correct figures.
He was not permitted by company policy to state what were
the correct figures, nor could he say for which models the
various percents of understatement applied. He regretted
that company policy prevented him from giving more information. He was dismayed that Computers and Automation referred to the census in the February issue as "an improved" census. He said that figures published in the January and earlier issues of Computers and Automation were
far closer to the correct figures.
III. From a news report (anonymous) in Computerworld for March 5, 1969, Computerworld, 60 Austin
1
St., Newtonville, Mass. 02160
A new computer census, which Computers and Automation
states is more accurate than its previous censuses appeared
in the February issue last week and indicates that the share
of the computer market held by many of the manufacturers
previously had been grossly overstated.
A number of the new figures, however, have been disputed
by knowledgeable industry sources.
In the new census, the figures for Honeywell, for instance,
showed a dramatic change. The population of 120s in the
U.S. was down from 650 to 260, a drop of more than half,
while the successful Honeywell 200 line had apparently
dropped from 800 installations to 448, a 40% drop~ Other
Honeywell systems were equally adversely affected.
Burroughs was hit in the figures for the B300 series, dropping from 370 installations to 183.
Some of the reduction was understandable because the
new figures included only U.S. installations and apparently
were five months old, while the original census had dealt
with worldwide figures and the situation in the middle of
December.
Even so, Honeywell reacted strongly on hearing of the
changes and characterized them as being "completely absurd". A Burroughs spokesman also strongly disagreed with
the new B300 figures ....
IV.
From I. Prakash
D. P. Focus
61 Helen- Drive
Marlboro, Mass.
We have received several comments on the figures reported in our Computer Census published in the February
issue of Computers and Automation.
Our Census is based on hard, factual information, including a listing by name and address for each computer enumer- .
ated in the Census.
Most of the companies do not release figures, but we would
be prepared to change our Census figures on the basis of
hard, factual information - from the companies themselves
or other sources - which includes the name and address of
each computer installation.
COMPUTERS and AUTOMATION for April, 1969
We would much appreciate it if other publishers and reporters would state clearly whether or not they have a listing of the name and address of all (not some) of the installations that they include in a census report.
Our Census is published to assist all our readers and executives in the industry who wish to base their plans, marketing strategies, and other actions on accurate information
which can be verified.
We will continue to do our best to prepare and publish
accurate reports and figures, even if such reports and figures
are not highly' regarded by some in the industry.
All comments are welcome.
v.
From the Editor
We believe it is desirable to publish the best information
we can find in regard to the number of computers installed
and the number of computers on order, in order to fulfill our
efforts in regard to a census of computers.
For a long time we have been dissatisfied with the figures
that we have published in the months through January 1969
in the Computer Census. Our dissatisfaction has shown in
some of the published notes attached in the January computer
census and earlier.
For example, take the note:
(N) - Manufacturer refuses to give any figures on number of orders and installations, and refuses to comment
in any way on those numbers listed here.
To publish a figure marked with a note like (N) attached to
it gives a superficial impression of accuracy that may be really
false, and we do not like to do that. To add such a figure to
a really accurate figure furnished by some other manufacturer
who is frankly telling the truth, is to us even more of a sta-
tistical sin, and has become more and more distasteful to us.
On one occasion we were told by the public relations officer
of a computer manufacturer in California: "When IBM publishes their number of computers installed and their number
on order, then we will, and in the meantime we will give you
no information." What does an editor do?
On another occasion I asked one of the heads of the Institute of Computers and Mechanics in Moscow, U.S.S.R. how
many computers were installed in the Soviet Union; he said
it was his impression that there were about 8000, but that he
knew of no figures available anywhere, and no way of collecting them either. Later, I commented on the 8000 to an
American market research specialist and he said that the
figure should be one third of the 8000. What does an editor
do?
We are glad to change over to a basis whereby we have an
agreement with a competent computer market research organization who will furnish us with computer census information
based on the names and addresses of locations where they
know computers are installed, who take the responsibility for
the correctness of the figures.
We wish to have as little as possible to do with information
from:
• "knowledgeable" industry sources
• "informed circles"
• an unnamed "spokesman"
• somebody who "reacted strongly"
and similar vague, indefinite and faceless informants.
If anybody can give us the names and addresses of U.S. installations where a total of six hundred Honeywell Type 120.
computers are installed, we shall be more than delighted to
increase the total reported in our census to 600, instead of
260. And similarly in all other cases.
0
"MACHINE LANGUAGE, AND LEARNING IT" - COMMENTS
William F. Sherman
MACRO Systems Associates
333 Bayside Dr.
Newport Beach, Calif. 92660
Your editorial of February, 1969 ("Machine Language, and
Learning It") was quite provocative and raised some points
to which members of the programming profession should
respond.
As a professional programmer, I, too, consider machine
language to be more enjoyable and satisfying. There is a
definite element of satisfaction in turning out a good, tightly
coded, systems-oriented, real-ti'me routine. One who has had
the challenging experience of solving a systems problem when
closely bound by the constraints of hardware and time will
certainly admit that this is a true test of the programming
professional. Additionally, the programmer fortunate enough
to be intimately associated with the hardware of a processor
develops a competence and discipline which the programmer
only experienced with higher level languages seems to lack.
There are those who vow never to use anything other than
the highest of high-level languages. These people seem to miss
the same point that the "diehard" machine-language-only
cadres miss, which is that the language selected for a job
is a function of:
a) The nature of the problem to be solved.
b) The environment in which the solving program has
to exist.
c) The utility of the solving program.
Part of the function of the professional programmer/sysCOMPUTERS and AUTOMATION for April, 1969
tems analyst (you will allow me the commonality, I hope) is
to specify a language to solve the problem under discussion.
A programmer "hung-up" on one type, or genre, of programming language can hardly be expected to perform well in this
area.
In similar fashion, data processing shops hung-up on the
use of a single language do themselves a corresponding disservice. The disadvantages incurred by the specification of a
single language only usually seem to outweigh the advantages
obtained.
.
It is my impression that, for the professional exp'erienced
programmer, a machine language or a higher level language
is not all that difficult to learn. He.nce, the recurring and repetitive discussion on this subject has little merit. The professional programmer should be obligated to choose, and be
familiar with, the language necessary to effect the most economical and timely solution to the job at hand.
As a consequence, your discussion of the MOHAC system
has embedded in it the seeds of a basic philosophical discussion which, it seems to me, centers around whether or not a
programmer should be aware of the operating principles of
the equipment he utilizes to solve the problem given to him.
I submit that this knowledge cannot hurt and more than
likely will help the programmer as a professional. Hence any
system or methodology which aids in gaining this end gains
0
my enthusiastic support. ,
15
OPPORTUNITY FOR THE BRITISH AND EUROPEAN COMPUTER INDUSTRY
Gordon Hyde, Scientific Director
Datatrac Ltd.
6 Colling ham Place
London, S.W.S, England
I would like to comment on the IBM anti-trust issue with
particular reference to the European computer scene.
Areas of technology characterized by rapid advance call for
a correspondingly high level' of investment in research and
development. For this reason commercial success and innovative competitiveness go hand-in-hand in such fields. Nowhere is this thesis better supported than in the computer
field. The overwhelming dominance of IBM over the European scene has undoubtedly been one of the major factors in
our failure to develop an adequate home-grown capability,
in the innovative and marketing sectors - although strategic
errors of our own have played a not insignificant role.
The vast dead-weight of punch-card thinking and hardware,
Based on a letter to The Times, Printing House Square, London, England, January 24, 1969.
also a legacy of IBM's commercial success, is also likely to
inhibit commercial exploitation of next generation real-time
systems for some time to come, as far as the conventional
market is concerned.
For these reasons, any change in the balance of power in
the United States computer industry must be reflected in a
determination by the British and European industry to take
advantage of the situation. This will call not only for a more
strategically aware commercial policy, but also a will to get
in first with the next generation of machines.
In this context, we should look closely at the field of smallsized and medium-sized, modular, real-time informationhandling systems, where not only is there a possibility of real
technological advance for relatively low research investment,
but also a hitherto inadequately explored market.
0
SOLVING NUMBLES AND OTHER PUZZLES
I. From Richard Marsh
1330 Mass. Ave. N.W., #822
Washington, D.C. 20005
A Scot once said that golf is an "humblin' game." Paraphrasing, may I say that "Numblin' is humblin.'''
Let me explain.
Due to change of address and failure of postal authorities
(probably assisted by my wife) in forwarding them to me, I
missed the September, October, and November issues of
C&A - just one of those things. In the December issue's
Letters to the Editor I noted references to the Numbles, and
further back in that issue found your December Numble. I
gave it a try, and a couple hours later, most of it spent in
deciding the approach to take, I came up with the solution.
I noted that you invited human or computer programs to
solve such puzzles. Now, though a subscriber to C&A, I'm
essentially a procedures man and wouldn't know a COBOL
statement from one in FORTRAN. But I did think it would
be interesting to try to reconstruct precisely the rules and
logic I had used in solving the Numble. I spent the next
several evenings on this project - somewhere around 16 to
20 hours. (The fact that it takes 10 to 20 times as long to
document a problem than to solve it may explain to some
degree why the software people in general have such a hard
time getting programs documented!)
Anyway, I had a nice set of rules and logic all written out.
I would have mailed it in except that I recognized that a
different problem might require a rule or two I had not yet
included. So, I decided to await the January issue and apply
my rules to the January Numble.
But you, you bum . . . well, the January Numble left me
numb. All those pretty rules and logic I had developed for
the December Numble were useless. I was reduced to a trialand-error routine based on possible values of H, K, and L
which would produce the two T's in the third line. Oh, I
finally got values which would work after only the fifth trial
out of about twenty possibles, but it took over four hours.
Actually it added only one rule to what I had already developed in December, but it was so different in nature that it
looked out of context when placed with my earlier rules.
So, in disgust, I decided to await the February Numble. And
16
what have you done? You've thrown in an entirely different
kind of problem - a simple addition which in a few minutes
can be solved to yield "Bad Luck." But, again, it doesn't follow any of my painstakingly constructed December rules, and
would require another 10 or 20 hours to document.
As I said at the outset, "Numblin' is humblin.''' I have
learned, again, that things are never as simple as they seem.
Complex mathematical equations are far easier to express
than the logic of the third and fourth grade arithmetic which
makes them possible! I think this is an important lesson.
I recognize that in a "conversational" mode a computer
would be handy to come up with a list of possible values for
given letters in a specific problem. It would have saved me a
few minutes of effort in solving particularly the December and
January Numbles. But I'm sure the total cost would be considerably greater than if I did it alone without computer
assistance. (Is this another lesson to be learned - that conversational-computer operations should be carefully screened
to preclude exorbitant costs?)
But back to Numbles.
My alternatives?
• Go to a library or elementary school and study tite
"basic rules of arithmetic again, or
• Swear off Numbles.
Since procedures men are always interested in the least possible effort, I shall choose the latter. Henceforth I will confine
myself to such innocuous pastimes as looking for things like
chances in the February Proof Goof 692.
II. From Neil Macdonald, Assistant Editor
Thank you for your enjoyable letter. I take pleasure in
sending you a copy of our little booklet on Numbles. Don't
swear them off - or swear off them! They're fun; and we
have a program on our DEC PDP-9 computer that does addition Numbles very well, but not yet multiplication Numbles.
We'll try soon to modify it and publish it in C&A.
COMPUTERS and AUTOMATION for April, 1969
III. From Morris Myers, Programmer
Dept. of Chemistry
Univ. of Arkansas
Fayetteville, Ark. 72701
I just discovered Numbles, and have solved Numble 6811.
However, Numble 812 seems to be very difficult to solve. I
am still trying though. Some of my colleagues and I are attempting to write a computer program to solve Numbles.
and I am curious as to whether they ever take the form of
division problems. If so, I would very much like to have one
as an example.
I thoroughly enjoy the relaxation of working Numbles
(already), and look forward to the next issue of C&A.
IV. From Neil Macdonald, Assistant Editor
We are glad you are enjoying Numbles. Yes, they can take
the form of division problems, which are rather easy on the
whole, because of the. large amount of information. We take
pleasure in sending you a copy of a booklet on Numbles and
their solution.
V. From the Editor
One of the reasons we publish Numbles is the fact that one
of the biggest incentives in learning is the solving of interesting problems. The instinct of curiosity that lies back of the
human desire to solve problems is without doubt one of the
elements that has led man as a species of animal to his present dominant position as a form of life on earth.
The instinct of curiosity and the desire to solve problems
might well be the main force which has produced over 200,000
computer programmers and systems analysts in a decade or so,
without benefit of formal training in colleges and schools.
In the pages of Computers and Automation, we hope to
emphasize the area of playing with computers in such fashion
as to lead to learning about them. We intend to publish soon
a d('s~rintion and details of a program for a miniature LISP
(LISt Processing programming language), a LISP that has
only five atoms and only five functions - and show its entire
structure, so that interested readers can play with it.
We invite readers and authors to send us descriptions and
details of small and interesting programs, especially programs that may invite persons into a path by which they
become "addicted" to computers, computer puzzles, computer games, and computer programming.
0
FORECAST OF COMPUTER DEVELOPMENTS, 1968-2000
Carol Andersen
Parsons & Williams
Nyropsgade 43 .
Copenhagen, Denmark
On November 22-24, 1968, 250 computer experts from 22
different countries atten.ded a congress on the organization of
computerized files. The International Federation for Information Processing Societies (IFIPS) was the sponsor, and the
Danish society was the host .for the congress. The high point
of the conference was a forecast of expected computer developments from the present until the year 2000.
The forecast was made using the Delphi technique. Eightyeight of the delegates from 11 countries gave their opinions
on 24 areas of development. Some of the major findings of
the forecast summary are:
1. A 50% reduction of the labor force in present industry is expected by the late 1980's. The reduction will
be partially compensated by shorter working hours
and by absorption of workers by new industries; but
the problem of unemployment is expected to be much
more serious in the future than it is today.
2. In the year 2000, all major industries will be con-
trolled by computers. Small industries will not be
automated to the same extent, since it is not likely
that many will exist by then.
3. The influence on the medical profession by EDP is
expected to be extensive. By 1975, treatment of patients in major hospitals will be controlled by computers and by 1980's a majority of doctors will have
EDP terminals for consultation and will be able to
give reliable diagnosis by computer.
4. The future software will, to a large extent, be built
into the hardware by late 1990's and computers which
learn from their own experience will exist before 1989.
5. In spite of advanced technology, computer prices are
expected to decrease by a factor of 100 by the end
of the 1980's.
The entire survey is published in a book entitled Forecast
1968-2000 of Computer Developments and Applications. Additional information is available from the address above. 0
USASI FORTRAN TO BE EXTE:NDED
X3 Secretary
Business Equipment Manufacturers Assoc.
235 West 42 St.
New York, N.Y. 10017
The USASI X3.4.3 FORTRAN Working Group at its
meeting of January 22, 1969 resolved to begin consideration
of standardizing FORTRAN programming language extensions. (This Working Group developed the existing FORTRAN Standards USAS X3.9-1966 FORTRAN and
USAS X3.1O-1966 Basic FORTRAN). The Group also estabCOMPUTERS and AUTOMATION for April, 1969
lished the principle that any extensions to the FORTRAN
Standards be such as to protect the integrity of existing
FORTRAN source programs written in conformity to the
present FORTRAN Standards.
Inquiries and suggestions should be made to the X3 Secretary at the address above.
0
17
ACM SYMPOSIUM ON THE
APPLICATION OF COMPUTERS TO THE
PROBLEMS OF URBAN SOCIETY
- CALL FOR PAPERS
Jessica Hellwig
Symposium Chairman
Columbia Univ. Computer Center
New York, N.Y. 10027
The fourth annual one-day Symposium on "The Application of Computers to the Problems of Urban Society" will be
held on Friday, October 24, 1969, at the New York Hilton
Hotel. Sponsored by the New York metropolitan chapters of
the Association for Computing Machinery (ACM), this Symposium brings together interested professionals from the computing field and from the urban problem areas, and provides
a forum for the exchange of ideas, experiences, and information.
Papers are invited on computer applications and experiments in: urban information systems; urban planning and
operations research; architecture; pollution, housing, transportation and welfare problems; education; and other areas
germane to computing and urban problems.
Abstracts of about 500 words in length should be submitted
by June 1; final papers will be required by July 1. Abstracts
and requests for further information should be directed to
the address above.
0
WHO'S WHO IN THE COMPUTER
FIELD, 1968-1969 - ENTRIES
Who's Who in the Computer Field 1968-1969 (the
Fifth Edition of our Who's Who), will be published by
Computers and Automation during 1969. The Fourth
Edition, 253 pages, with about 5000 capsule biographies
was published in 1963. The Third Edition, 199 pages,
was published in 1957.
In the Fifth Edition we hope to include upwards of
10,000 capsule biographies including as many persons as
possible who have distinguished themselves in the field
of computers and data processing.
If you wish to be considered for inclusion in the
Who's Who, please complete the following form or provide us with the equivalent information. (If you have already sent us a form some time during the past eight
months, it is' not necessary to send us another one unless
there is a change in information. )
1.
2.
3.
4.
5.
6.
WHO'S WHO ENTRY FORM
(may be copied on any piece of paper)
Name? (Please print) _ _ _ _ _ _ _ _ _ _ _ __
Home Address (with Zip) ? _ _ _ _ _ _ _ _ _ __
Organization? __- - - - - - - - - - - - - - Its Address (with Zip) ? _ _ _ _ _ _ _ _ _ _ _ __
Your Title ? _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
Your Main Interests?
Applications
Mathematics
)
Business
Programming
)
Construction
Sales
)
Design
Systems
)
Logic
other
( )
Management
(Please specify)
NUMBER OF TIME-SHARING VENDORS
Alan G. Hammersmith, Pres.
Time-Sharing Enterprises, Inc.
251 W. DeKalb Pike, Suite C-110
King of Prussia, Pa. 19406
In reference to your desire to find a better estimate of the
number of time-sharing vendors (Feb., 1969 issue, page 17),
you might be interested in the following information.
Since September, 1968, our firm has been publishing the
"Time-Sharing Industry Directory". This publication lists information on the various time-sharing vendors presently offering a commercial remote access service. On Sept. 1 there
were 45 vendors; in Nov. there were 105 vendors; and our
last update on Jan. 1, 1969 lists 117 time-sharing vendors.
7.
8.
9.
Year of B i r t h ? : _ _ - - - - - - - - - - - - - - Education and Degrees ? _ _ _ _ _ _ _ _ _ _ _ __
Year Entered Computer Field ? _ _ _ _ _ _ _ __
10. Occupation ? _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
11. Publications, Honors, Memberships, and other
Distinctions?
------------------
(attach paper if needed)
12. Do you have access to a computer? (
a.
)Yes (
)No
If yes, what kind of computer?
Manufacturer? _________________
Mode 1_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___
b. Where is it installed:
Manufacturer? _________________
Address?
c. Is your access: Batch? ( ) Time-shared? (
Other? ( ) Please explain: _ _ _ _ _ _ _ __
d. Any remarks ? _________________
--------------------
COMPUTER MARKET REPORT
- CORRECTION
Two errors in the Computer Market Report ("The Computer Leasing Industry -- Some Statistics") beginning on
page 39 in our March, 1969 issue need correction. The heading on Table 6, page 40, should read "Geographic Location
of Leasing Companies", not "Leasing Customers" as published.
Likewise, the first sentence in paragraph one on page 69 should
read "The geographic location of the top 175 computer leasing companies is shown in Table 6."
18
13. Associates or friends who should be sent Who's Who
entry forms?
Name and Address
(attach paper if needed)
When completed, please send to:
Who's Who Editor, Computers and Automation,
815 Washington St., Newtonville, Mass. 02160
COMPUTERS and AUTOMATION for April, 1969
C·a
NUMBLES
Number Puzzles for Nimble Minds
- and Computers
Neil Macdonald
Assistant Editor
A "numble" is an arithmetical problem in which: digits
have been replaced by capital letters; and there are two
messages, one which can be read right away and a second
one in the digit cipher. The problem is to solve for the digits.
Each capital letter in the arithmetical problem stands for
just one digit 0 to 9. A digit may be represented by more
than one letter. The second message, which is expressed in
numerical digits, is to be translated (using the same key) into
letters so that it may be read; but the spelling uses puns or is
otherwise irregular, to discourage cryptanalytic methods of
deciphering.
We invite our readers to send us solutions, together with
human programs or computer programs which will produce
the solutions.
Numble 694:
The
HARD
TSI PROXIMITY KEYBOARD
•• • THE ONLY KEYBOARD WITH
100% GUARANTEED RELIABILITY
Tested, approved and accepted, this solid state
keyboard gives you dependability and performance
unparalleled by any other keyboard, including those
labeled "breakthroughs in technology."
HERE'S PROOF OF TSI SUPERIORITY
• RELIABILITY-TSI Proximity Key (patent pending)
utilizes non·contacting proximity transducers to
provide infinite operational life.
• ADAPTABILITY-The TSI Keyboard can be tailored
to meet any specific requirement.
• ECONOMY-No bounce characteristic eliminates
the need for anti·bounce circuitry . . . up to 8
outputs from single key eliminates the need for
diode matrices . . . trouble·free performance
requires practically no maintenance, downtime
or costly repairs.
est
x S T E P
ER0 RE
T L SAO
TSI MINI·lINE KEYBOARD
A S DOT
o
=
L
P
RL S R
SOH SRI E E
02176
58758
62781
93
Constructed to same high standards and ultra· reliability as the
larger keyboard. The TSI Mini is
designed for limited space ap·
plications requiring thin line
construction.
SPECIFY TSI PROXIMITY TRANSDUCERS
Solution to Numble 693
In Numble 693 in our March issue, the digits 0 through
9 are represented by letters as follows:
B=O
=
L=5
1
M=6
E,Y= 2
T=3
O,U=7
N=8
G,K,C
S=4
. J= 9
The full message is: Neglect justly entombs most books.
Our thanks to the following individuals for submitting their
solutions to recent Numbles we have published: A. Sanford
Brown, Dallas, Tex.; Dick Chase, Bloomfield, N.J.; T. P.
Finn, Indianapolis, Ind.; Claude Grenier, Quebec, Canada;
John Lambrecht, Antioch, Tenn.; Morris Myers, Fayetteville,
Ark.; Joseph J. O'Hara, Jr., New Haven, Conn.; L. Rowland,
Columbia, Mo.; D. F. Stevens, Berkeley, Calif.; and Bob
Weden, Edina, Minn.
0
COMPUTERS and AUTOMATION for April, 1969
If you produce key punch, card reader,
paper tape or disc file equipment, you
should be using TSI Proximity TransIducers . . . the most reliable and eco·
nomical method for parity checking,
rack peak detection, displacement sensing and hole detection. Sizes from J{/'
0.0. to X" 0.0.
WE CAN DELIVER UP TO 5,000 UNITS A WEEK
For keyboards write for Bulletin K·9000·A;
for Proximity Transducers Bulletin PT-4000·A;
or telephone.
ITlsl:.:
TRANSDUCERS SYSTEMS, INC.
Easton and Wyandotte Roads Willow Grove, Pa. 19090
(215) 657·0655
Designate No. 12 on Reader Service C.rd
19
GEOPHYSICAL DATA MANAGEMENT
AND HOW?
-
WHY?
Dr. Robert M. White~ Administrator
Environmental Science Services Adm.
U.S. Dept. of Commerce
Rockville~ Md. 20852
(OWe need to consider the cost to the Nation of not having an effective
geophysical data system - what it costs us when a scientist, engineer,
or citizen is unable to get the information he needs to continue his research, to design a building, to plan a dam, to make a business decision
- and what the cumulative effect of this lack of information may be."
Data management seems to be the "in" thing nowadays.
Our technical reports are replete with descriptions of advances in data storage and retrieval. Stock market analyses
give special emphasis to investment opportunities in companies
engaged in the field. Our journals are full of discussions
about the data doomsday which is reportedly about to befall
us. By any measure, data management - with all its implications for science and technology - is a spectacularly growing field. Presumably, then, the question I raise in my title has
been answered, or at least wrestled with, by many people in
many ways.
However, over the past several years, my duties as Administrator of a reasonably large geophysical effort Have brought
me face to face with problems dealing with many kinds of
geophysical data management systems. Naturally, I have arrived at some views about where we are and where we may
be going in this immensely complex area.
Environmental Data Service
The problems of data management are woven through the
fabric of everything we do in ESSA (Environmental Science
Services Administration). They are regarded as so vital that
we have established as one of our five major components an
Environmental Data Service, on a par with the Weather Bureau, the Coast and Geodetic Survey, our Research Laboratories, and the National Environmental Satellite Center. The
Environmental Data Service manages our archival and retrieval systems for the geophysical information for which
ESSA has direct responsibility. But it deals with only one
(Based on an address before the Marine Technology Society,
Washington, D.C., October, 1968)
20
part of the picture: every other element of ESSA is also involved in one form of data management or another. Indeed,
we constitute an organization whose general purpose· is the
collection, processing, dissemination and storage of. geophysical information of all kinds, and for a limitless range of
purposes.
Data management is common to every geophysical scientific
and technical activity. Every scientist or technologist in our
agency is a data manager in one se~se or another. But the
need is far wider than this: almost every economic effort depends to some extent upon the information from our geophysical data.
It probably was inevitable, then, that data management
should give rise to its own cult. There are basic inherent
similarities in the processing, archiving and dissemination of
data of all kinds. There is a commonality of complaints
about the inefficiency of our present systems. New processing,
storage, and retrieval technology form a common technological
base. All of this, however, has led many people to almost ignore the fact that each system is merely a means for carrying
forward a specific program or giving a specific service. There
is a tendency to paper over the difference between programs
and services, and their ultimate purposes, and to subordinate
them to the efficiency of the data management function. It
should be remembered that data management is not, and
should not be, an end in itself.
A Catchall Term
In addition to this kind of thinking, we have a further
complication. Sometimes one feels that data management has
come to mean all things to all people. It is becoming one of
those convenient catchall terms like "the environment". I recently attended a joint colloquium sponsored by the House and
COMPUTERS and AUTOMATION for April, 1969
Senate on a national policy for the environment. I discovered
there are an amazing number of environments. Each of the
Secretarial officers spoke to his concerns. The Secretary of the
Department of Urban Affairs was concerned about the environment, principally social, of the cities. The representative of
the Department of Transportation talked of the environment
of the highways and automobile safety. The Secretary of the
Interior spoke about the conservation of the natural environment. And so it went; everybody was talking about a different environment.
Nowadays, I see the same tendency when we think about
data management. In ESSA, the Director of the Weather Bureau is concerned with the real-time acquisition, communication, processing and dissemination of weather information and
the management of data systems to accomplish that purpose.
The Director of the Coast Survey is concerned with the acquisition of oceanographic data and their processing into the form
of maps and charts and the management of data systems to
accomplish that purpose .. The Director of the Environmental
Satellite Center is concerned with the acquisition and processing of satellite data and the management of data systems to
accomplish that purpose.
Primary S,ystems
The differences between the purposes, the organization and
management of these systems are far greater than the common needs that bind all data systems.' So I think it is logical
to draw a distinction between primary and secondary systems.
In primary systems such as those I have mentioned, there exists relatively good control over input, communication and
processing. There are clear, specific requirements and customers. The fact that data collected by such primary systems
as our weather forecast and warning, or marine mapping and
charting systems may later have wide use for other purposes
is of only peripheral concern to the primary system manager,
is secondary. This, as we shall see, leads to difficult problems
for managers of secondary systems.
Secondary Systems
Geophysical data archival and retrieval systems are secondary systems. Let me emphasize that I use the word "secondary" not to indicate lesser importance: I find many problems
both in design and management in the development of such
systems. Our Environmental Data Service, which is concerned principally with systems for the storage, archival, retrieval and dissemination of historical information - gathered
for a multitude of primary purposes - is a good example of
a set of secondary systems.
In addition to the National Weather Records Center, this
organization operates several geophysical data centers, (geomagnetism, seismology and geodesy), and our Aeronomy and
Space Data Center. It is our representative on the advisory
board of the National Oceanographic Data Center, and it
funds for ESSA's share of that operation.
Lack of Control Over Input and Output
The problems of operating such a set of systems are formidable. By and large, managers have little input control
over the amount, character, accuracy, or format of the information they must archive and process. On the output side,
their mixture of customers and requirements is continually
changing and often indeterminate. By way of example, a decision to build a second canal across Panama, a decision to
operate a new vehicle in air or water - all such decisions can
generate unanticipated demands for data. This condition is
in the very nature of the task, and it will not change.
The principal control which the manager of secondary data
systems has is in the archival and data processing segments of
COMPUTERS and AUTOMATION for April, 1969
the system for which he is responsible. The inherent lack of
control and uncertainty at both the input and output ends of
such secondary systems makes them very difficult to plan,
design and development.
The distinction between primary and secondary kinds of
data systems is far more than semantic. It is a source of confusion. The common threads that bind all data management
activities are without question weaker than the bonds that tie
together a total program or a total data system. The problems
of operating a real-time system for observing and forecasting
ocean data are so fundamentally different from those of operating a system for the archival and retrieving of ocean data,
for example, that lumping them under the general umbrella
of data management serves only to becloud the problems of
both.
.
The Real Problems
The real problems in this area do not lie in our ability to
deal with what is popularly known as the data explosion.
Modern information and data processing technology can cope
with almost any flood of geophysical material we can anticipate, if adequate resources are provided for it. Our thorniest
problems are not those of building bigger and better boxes
into which to feed our information - although these are
needed. Our most difficult problems stem from the lack of
ability to exercise enough influence over the types and formats
of data flowing into the system and arranging quick, responsive retrieval and dissemination to serve our customers.
In designing a storage and retrieval system, we must weigh
three major considerations: (1) the impact on those primary
systems which supply the data; (2) the responsiveness we will
provide the user; and (3) the efficiency of the system itself.
It is clear to me that we have focused too sharply on the last
consideration to the neglect of the first two. This fact may be
leading us toward systems designed primarily to serve their
own ends, and it is certain that such systems will never be
more than a wasted exercise - a burden to those supplying
the data, and a frustration to those who try to use it.
Weaknesses in Present Systems
Most of our current secondary systems take as their task
the absorption of the masses of data collected by routine observational or experimental programs, as well as response to
the retrieval requirements of subsequent users. Our present
systems do not meet the needs. Experience has shown us that
primary data are useful for many purposes after they have
satisfied the first need - purposes whose values may outweigh
in some cases that of the original program.
Measurements of sea-air parameters such as winds, tides,
currents, and temperatures made for environmental prediction
purposes, for example, become valuable to engineers working
in the marine environment, but only after the data have been
systematically identified, summarized, and archived to form
a large enough data base.'
Geophysical data collected for research activities such as
continental drift studies have great economic potential - if
the basic information is made available to those concerned
with assessing and developing the oceans' mineral wealth.
Data collected during surveys associated with the production of navigational charts are also useful in engineering and
exploration activities, if it can be made available in "appropriate forms.
We ought to face up to the fact that at present we have no
satisfactory means in our secondary geophysical data systems
for radically influencing the form, timing, and accuracy of
input material gathered by primary systems. Even in an organization as integrated as ESSA, in which both primary and
21
secondary data systems are under single management, we have
not solved the problem to my satisfaction. There are many
reasons, it seems to me. They range all the way from questions of finance, to utility, to international agreements. We are
perhaps in best shape in the weather data area. But it has
taken a half century to arrive at the present relation between
primary and secondary weather data systems, and even here
there is much to be done.
The Weather Data System
Let me illustrate. The primary weather data system must
respond to real immediate needs for weather forecasts and
warnings. New equipment must continuously be obtained and
installed, new observational locations found. Old ones must be
changed to meet pressing demands; data rates and formats are
designed to serve the needs of the primary system. Priorities
are established without significant reference to the needs of
the secondary weather data system. All of these changes affect
the input to the secondary data system, and the cost and
workload in that system. Ideally, if we had as many resources
as we needed, we could accommodate such impacts. But we
never do have all we require. We are forced to decide whether
we will allow the primary system to move forward and generate adverse impacts on our secondary system. Frequently the
answer is yes, when we are confronted with a pressing requirement of the primary system. The suggestion has been made
that secondary systems should be funded as an overhead on
the primary system. This is only part of the answer, since
it places the secondary system completely at the mercy of the
changes in the primary.
We must begin to develop mechanisms whereby the assessment of changes in the primary systems and the secondary
system can be made, and where the drawbacks and costs can
be compared with the benefits to the primary system.
Determining User Needs
A major problem facing us in the development of secondary geophysical data centers deals with responsiveness of the
system in the face of an inherent uncertainty in the nature of
the user market.
Determining user needs and assessing the degree to which
the data system should be responsive to them is a very difficult
question. Again, ideally we should like a system which would
be sufficiently responsive to the needs of the user to make him
willing to pay for it. I think you are all too familiar with the
difficulties of user requirements and user need studies. Sometimes they take on the aspects of Alice-in-Wonderland adventures.
The cost of satisfying user needs must be related in some
way to benefits. And the question of the cost-to-whom has to
be dealt with. I am convinced that there is a Parkinsonian
law of some kind which says that user needs expand in direct
proportion to the costs somebody else is witling to bear.
The fact that the problem is a difficult one does not mean
we should not tackle it. On the contrary, unless we find some
rationale for confronting and solving it, we are going to have
unresponsive data systems.
It is clear that many user requirements are not known and
cannot in principle be known - because the potential users
may not themselves be able to specify what they will need
next year or ten years from now. More often than not, current
demand reflects only our current response capability - not
what the user really needs or wants. We need to know who
our users are - and what they are trying to do, since they
rna y be unsure of their own exact needs. We need to identify
potential users, customers who are unaware of the data's
availability.
22
Value of Data
Because data have varying degrees of value, ranging from
limited or short-term through indefinite or permanent, we
need user guidelines or priorities to tell us what to process for
high speed, flexible retrieval and what simply to store in the
cheapest way possible. Data essential for defense, intelligence,
and other activities which require a rapid response or frequent
analysis must be stored in sophisticated forms capable of rapid
retrieval. Other types, particularly raw data for which there
are published results or summaries but which still have potential future value, could be stored in a low-cost medium
such as microfilm.
The Cost of Lack of Information
Finally, and most important of all, we ·need to consider the
cost to the Nation of not having an effective geophysical data
system - what it costs us when a scientist, engineer, or citizen is unable to get the information he needs to continue his
research, to design a building, to plan a dam, to make a business decision - and what the cumulative effect of this lack
of information may be. I am convinced that any reasonable
investment in an effective user-oriented geophysical data system would easily be recouped in the "downstream savings" of
the following few decades.
Let's look at our marine data situation today. John Fry,
of the staff of the National Council on Marine Resources and
Engineering Development, has summarized it as follows:
. . . There is a wide diversity of collectors, processors,
and users of marine sciences data employing a wide variety of techniques, many of which are incompatible;
there are numerous uncorrelated data banks in agency
files and data centers, which exhibit varying degrees of
backlogs and user demands. Marine data acquired at
characteristically high cost are not moving expeditiously
from acquisition to an end-product stage in the agencies,
or to data centers with the capabilities to make them
rapidly accessible to industry, universities, and State and
local governments; to regulatory agencies responsible for
monitoring the environment; or to Federal agencies for
use in planning new marine programs.
Marine Data: An Example
To illustrate just one aspect, consider the current dispersal
of facilities and arragements for the archival, retrieval, and
dissemination of marine data. The charter of the National
Oceanographic Data Center charges the agency with archival
and retrieval of only those oceanographic data which were
not already provided for in existing centers. Data from the
ocean's surface up - sea surface temperature, waves, swell,
and the physical measuremeI;lt of the atmosphere above the
sea - are handled by ESSA's National Weather Records
Center in Asheville, N.C. Our organization has also begun a
very limited service concentrating on magnetic, gravity, seismic reflection, and bathymetric data, but this involves mainly
ESSA-generated observations. There is no single focus fQr geomagnetic, gravity, bathymetric, or other ocean survey data.
Not only is there no one place to go to obtain marine data,
there is not even one mechanism which will tell you what
exists. But more important is the problem of collating the
information. In too many areas - of which air-sea interaction is an outstanding example - scientists and engineers
must deal with simultaneously processed data from different
environments, with computers which are in separate centers,
with differing formats, programs, procedures and goals. The
ultimate consequences of this situation are apparent. This
problem must be faced now. And this is by no means isolated;
many of the marine problems plaguing us are only too familiar across the whole spectrum of the geophysical environment.
COMPUTERS and AUTOMATION for April, 1969
I
IE.
M
0
COM
R E
PUTER
L
FOR
E S S
M
0
N
E V
PRICE INCLUDES:
12 bit 2 J.I. sec core store
256 word 400 nsec. Read Only Store
7.2 f.J.sec full word add time
All power supplies
Power fail protection
Chassis and front panel
P.'DI~f
Microprogrammed priority interrupt
Interface for teletype 33 ASR
I/O Bus for 256 channels
1024 words -
$ 6000
4096 words -
$ 7000
I~
r.:..
fS.
WPc
III]
Q]
$ 4900
2048 words -
c."'.
ELBIT COMPUTERS LTD.
9701
N. KENTON AVE. SKOKIE, ILL.
(321) 676-4860
Designate No. 21 on Reader Service Card
It is clear that as we move to probe, understand, explore
and develop our geophysical environment as a whole, we will
increasingly come to require a more comprehensive and better
systematized network of centers capable of providing a wide
range of data. We are going to have to move in the direction
of what I call a national Geo Data System. Such a system
need not be monolithic; it need not be highly centralized; but
it must be an integrated system. It might consist of geophysical "data banks" where collected observations can be
centrally processed and archived, and whose data, analyses,
studies, publications, and services are available to all scientific disciplines. But these "banks" should be integrated into
a common system.
We should seek in designing such a Geo Data System to
set some standards. I would advocate some of these:
1. Provide a "one-stop", rapid-response data-support system for geophysical operations and research capable of answering most user queries on the availability of geophysical
data, facts, figures, and sources. Response time should be
measured in minutes or hours, rather than the current days
or weeks. This service should be accessible on a national
basis, via electronic data links.
2. Provide international geophysical data files of the highest possible quality, through appropriate quality control procedures, while developing a full capability for providing data
evaluation when needed.
3. Provide the techniques and procedures which allow
the best estimates of environmental parameters not specifically
recorded in the observational or experimental record, through
the development of "environmental models".
4. Provide a capability for computer-based data support
to remote operations and research activities, at less cost and
with greater speed and efficiency than is currently associated
with just the acquisition of copies of raw data.
COMPUTERS and AUTOMATION for April, 1969
5. Provide for (a) the inexpensive publication of those
geophysical data most frequently requested by the casual
or unsophisticated user, in a form that discourages misinterpretation; and (b) provide an attractive publication series
to inform the public of the potential of geophysical data
application for long-range planning purposes.
A Changing System
In planning such a system, we must recognize that it will
have to change continually to meet shifting and uncertain
demands. A rigid system designed to give data support today
will surely be obsolete tomorrow, as new areas of interest
and new requirements evolve. Look at our past:
In data collection we have extended our reach from the
immediate environment of man and direct sensing, to the environment of space and the measurement of solar activity and
the earth's cloud cover from a satellite altitude of several
thousand miles.
In data processing we have gone from the accounting machine era to third generation computers in a dozen years;
from the "chunk chunk" of the indestructible IBM 407
Tabulator to electronic speeds in the same length of time.
In data application we have progressed from the problems
of kites and Kitty Hawk to those of satellites and supersonic
transports in less than a lifetime. Users now demand global,
as well as local data, and sometimes both, as in problems
such as the pollution of the air and estuaries. This is because
of man's growing desire to understand and predict environmental change on a geological time scale.
The changes and challenges that will confront us in the
next several decades will surely make these problems pale by
comparison. This fact makes it even more imperative that we
move forward as quickly as possible, making large plans. The
time to eliminate tomorrow's bottleneck in data is now.
0
23
UNLOCKING THE COMPUTER'S PROFIT POTENT,IAL
McKinsey & Company, Inc.*
245 Park Ave.
New York, N.Y. 10017
Profile of the study sample
BY SIZE OF COMPANY
10
9
FOREIGN COMPANIES
_
U.S. COMPANIES
6
Number of companies
6
5
Classified by industry,
the distribution of the
36 companies is as follows:
Airlines
Apparel
Chemicals
Food
Forest products
Insurance
Machinery
Paper
Petroleum
Primary metals
Railroads
Textiles
Transportation
equipment
2
1
8
3
1
3
6
1
3
Annual sales in
millions of dollars
under $200
$200-499
$I,000-I999
over $2,000
BY COMPUTER OUTLAY
_
MORE SUCCESSFUL COMPUTER USERS
a
LESS SUCCESSFUL COMPUTER USERS
2
1
1
$,00-999
Number of companies
7
7
7
4
I
Computer outlays as
estimated % of sales
under 0.24%
0.2'-.49%
I.0-1.99%
over 2.0%
Exhibit 1
*Copyright 1968, McKinsey & Company, Inc. Reprinted with pennission.
24
COMPUTERS and AUTOMATION for April, 1969
((Many otherwise effective top managements are in trouble with their
computer efforts because they have abdicated control to staff specialists.
Only managers. can manage the computer in the best interests of the
business. The companies that take this lesson to heart today will be
the computer profit leaders of tomorrow."
In terms of technical achievement, the computer revolution
in U.S. business is outrunning expectations. In terms of economic payoff on new applications, it is rapidly losing momentum. Such is the evidence presented by a recent study by
McKinsey & Company of computer systems management in
36 major companies. The companies studied represented all
levels of achievement and experience with computers in 13
different industries. Their distribution by industry, sales volume, and relative level of computer expenditure is shown in
Exhibit 1. 1
Diminishing Returns
From a profit standpoint, our findings indicate, computer
efforts in all but a few exceptional companies are in real, if
often unacknowledged, trouble. Faster, costlier, more sophisticated hardware; larger and increasingly costly computer staffs;
increasingly complex and ingenious applications: these are in
evidence everywhere. Less and less in evidence, as these new
applications proliferate, are profitable results. This is the
familiar phenomenon of diminishing returns. But there is one
crucial difference: As yet, the real profit potential of the
computer 'has barely begun to be tapped.
Almost 20 years ago the first computers for business use
made their debut. Five years ago, when we published our first
research report on the computer,2 business was well on the
way to exploiting the awesome clerical and arithmetical talents of its new tool. Today the early goals have for all practical purposes been attained. Most large companies have
successfully mechanized the bulk of their routine clerical and
accounting procedures, and many have moved out into operating applications.
IThe distinction between "more successful" and "less successful" computer users, which is explicit in several exhibits and
implicit in much of the text, requires a word of explanation.
Because of the many variables involved, any absolute standard
of computer success must necessarily be arbitrary. Instead of
setting such a standard, we decided to let "success" be defined
by the range of performance observed in the survey sample itself.
Thus companies identified as "more successful computer users"
are simply those falling in the upper half of this order-ofperfonnance ranking, and the "less successful" companies are
those in the lower half.
2Getting the Most Out of Your Computer, McKinsey & Company, Inc., 1963.
COMPUTERS and AUTOMATION for April, 1969
As a super-clerk, t4e computer has more than paid its way.
For most large organizations, going back to punch cards and
keyboard machines would be as unthinkable as giving up the
typewriter for the quill pen. Yet in these same companies including many that pioneered in the mechanization of paperwork operations - mounting computer expenditures are no
longer matched by rising economic returns.
Failure to Adapt
What has gone wrong? The answer, our findings suggest, lies
in a failure to adapt to new conditions. The rules of the game
have been changing, but management's strategies have not.
There was a time, less than a decade ago, when management could afford to leave the direction of the corporate
computer effort largely in the hands of technical staff people.
That time is past. Yet the identification and selection of neW
computer applications are still predominantly in the hands
of computer specialists, who - despite their professional expertise - are poorly qualified to set the course of the corporate computer effort.
It is not hard to understand how this situation has come
about. Historically, profit-oriented companies have undertaken
computer development work for the sake of a single ultimate
objective: improved financial results. There are just three
ways such results can be reflected in the income statement,
and three general categories of computer applications by which
this can be accomplished directly:
Purpose
1. To reduce general and
administrative expenses
2. To reduce cost of goods sold
3. To increase revenues
Application
Administrative and
accounting uses
Operations control systems
Product innovation and
improved customer service
Improved financial results, of course, can also be achieved
indirectly, through better management information and control. This gives rise to a final purpose and application category:
4. To improve staff work and
management decisions
Information systems and
simulation models
Mainly because of rising clerical costs and the desire to
cut clerical staffs, the practical history of computers in U.S.
25
Where the opportunity lies
BEFORE-TAX
PROFIT
$.20
GENERAL
& ADMINISTRATIVE
Typical breakdown of sales dollar
COST OF
Potential profit impact of 10 percent reduction ...
in general & admini~trative expenses
in cost of goods sold
$.15
Exhibit 2
business to date has been dominated by the first of these four
purposes. A look at current computer development projects
shows that the prime objective in many computer departments
is still the refinement of administrative systems and the reduction of G&A expenses. But this, our study indicates, is an area
where the cream (and some of the milk) has already been
skimmed. It is high time for a change in emphasis, if not a
change of course, in the computer development effort. And
th~ next move is up to management.
Many senior executives are already beginning to recognize
their dilemma. "How can I keep on justifying major computer expenditures when I can't show a dollar saved to date
from our last three applications?" asks the president of a large
consumer goods company. "Maybe I'm a fool to let it worry
me - after all, who tries to find a dollar justification for telephones and typewriters? But I do worry. After all, we know
that what we're doing with telephones and typewriters makes
sense. But that's more than I can say for some of the things
we're doing with the computer, at many times the cost."
The ill-justified expenditures, however, are insignificant
compared to the opportunity costs. Though it has transformed
the administrative and accounting operations of U.S. bm:iness,
the computer has had little impact on most companies' key
operating and management problems. Yet, as Exhibit 2 suggests, this is where its greatest potential lies.
26
In our 1963 report, we noted that no company had yet
come anywhere near exhausting the computer's potential.
Today the gap between technical capability and practical
achievement is still wider, and the stakes have risen sharply.
Until the computer is put to work where the leverage on
profits is high, the penalty of lost opportunities and lost profits
will continue to mount.
Subsequent sections of this report will set forth the dimensions and implications of the issue. We shall outline the developments that have shaped it, explore the current problems
to which it has given rise, and indicate some of the future
opportunities open to companies that take timely action to
resolve it. Finally, we shall offer a few practical guidelines
for the chief executive who recognizes his own vital personal
responsibility for the success and profitability of his company's
computer effort.
The Stakes and the Problem
In 1963, computer manufacturers shipped hardware worth
$1.3 billion to their U.S. customers. By 1967, the value of
computer shipments had risen to $3.9 billion, an increase of
no less than 200 percent in four years. 3 Of every $1 million
that business laid out on new plant and equipment in 1963,
$33,000 went for computers and computer-associated hardware. By 1967, the computer's share had risen to $63,000,
COMPUTERS and AUTOMATION for April, 1969
clearly rising more rapidly than the rental bills. Thus, a company that is paying as little as $125,000 a year to rent equipment of very modest capacity is probably spending upwards
of a third of a million dollars on its total effort. It is a fair
estimate that well over a hundred industrial companies have
rental bills running into seven figures, and there are a handful
whose total computer outlays approach $100 million a year.
Because so much of the total cost is payroll, and because
staffs are dispersed and personnel classifications and accounting conventions differ from company to company, attempts to
formulate "yardsticks" for corporate computer outlays (e.g.,
as a percentage of assets, capital expenditures, administrative
expenses, or sales volume) are likely to end in confusion.
Even if precision in such figures were possible, what a particular company "ought" to be spending on computers will not be
discovered by studying industry averages or the outlays of
individual competitors. At best, such yardsticks will provide
a bench mark from which to start; but the final answer can
only be determined in the light of the company's own situation, strategy, and resources (including the depth and sophistication of its computer experience).
The distribution of costs which go to make up total computer expenditures is, however, fairly consistent among the
companies participating in our study. Exhibit 3 indicates this
distribution. Of every $100,000 of total computer expenditures
about $35,000 goes for hardware; $30,000 for computer operations staff payroll; $15,000 for maintenance programming
and each dollar was buying at least half as much again in .
capacity. Computer spending, both absolutely and as a
proportion of all plant and equipment outlays, is still rising.
Another index to the growth of computers as a factor in
the national economy from 1963 to 1966 can be found in
the published accounts of the largest computer manufacturer,
the International Business Machines Corporation. IBM's ,gross
investment in "factory and office equipment, rental machines
and parts" grew from just under $2 billion in 1963 (double
the 1957 figure) to just over $5 billion in 1966. In its annual
report for 1966, IBM chose for the first time to report separately the value of its ihvestment in machines on rental.
Valued at cost;, that investment had grown from $3.3 billion
to $4.4 billion during the previous 12 months, a one-year
growth of 33 percent.
Cost of People and Supplies
Massive as it looks and rapidly as it is growing, the investment in computer hardware is far from an adequate measure
of business's stake in the computer. For every dollar spent on
equipment, the typical company in our current study spent
close to $2.00 on people and supplies in 1967, as Exhibit 3
indicates; and the payroll component of the total outlay IS
3EDP Industry and Market Report, January 26, 1968.
How computer costs are distributed
RENTAL OR
EQUIPMENT COSTS
Most costs are now fixed
3,%
COMPUTER COSTS
_
FIXED
t=l CONTROLLABLE
OPERATIONS
30%
For every $100 spent on hardware, companies spend $187 on staff
$roo
HARDWARE
RENTAL OR EQUIPMENT COSTS
PAYROLL
COMPUTER OPERATIONS
PROGRAM MAINTENANCE
NEW PROGRAMS
$86
$43
$,8
Exhibit 3
COMPUTERS and AUrOl/ATION for April, 1969
27
(i.e., keeping current systems updated); and the remammg
$20,000 for development programming and other staff time
devoted to new applications.
Development Dollars
These development dollars, the only computer outlays subject to significant short-term management control, are typically a smaller fraction of the total than the company's
annual bill for hardware rentals. Yet their leverage on future
costs and benefits is enormous; in fact, they hold the key to
the company's long-range success or failure with its computer
effort. For unless management segregates these costs and
understands the nature of the resources they buy, the direction
of future computer developments will be in doubt, and the
whole activity will be vulnerable to precipitate, perhaps
emotional, review.
The computer management prQblem as it confronts corporate executives today, then, is a matter of future direction
rather than current effectiveness. The key question is not
"How are we doing?" but "Where are we heading, and why?"
Five years ago this was a less critical issue at the topmanagement level. As long as computer developments were
largely confined to accounting departments there was less
reason for corporate executives to concern themselves with
direction setting: If the controller carried out his function
and kept his costs in line, noone' outside his department
worried very much about how he did it. The situation is
very different today. Now that the conversion of accounting
work to computer processing is virtually complete - as it is
in 30 of the 36 cOI'~panies in our study - the question "What
next?" comes into urgent focus. Many of the alternatives
currently being proposed are complex and costly enough to
require executive approval, but their justification is obscure at
best. When top management, reviewing a proposal, looks in
vain for the promise of profit, it is right to hesitate.
For example, the following three proposals, submitted for
approval during 1967 in one company we studied, would have
consumed 80 percent of the computer staff time available for
development:
1. Design a com/JUter-based «strategic management information system." This was candidly described by its sponsor,
the manager of the Systems & Procedures Department, as
"a basic research project," as indeed it would have been.
Management's information needs had not been determined;
the cost of making information available was uncertain; and
the proposed techniques for putting the manager (assuming
he was interested) in a position to manipulate the information (if it could be provided) had never really been tried out.
2. Design a model of the corporate distribution system, to
be used in both long-range planning and daily management of
operations. Cost data on the present distribution system were
scanty and out of date. Moreover, responsibility for distribution lay with the, marketing vice president, a man who had
made no major changes in distribution policy or practice for
15 years, and had a well-earned reputation for being hostile
to innovation. Perhaps understandably, ,he had not been consulted on the proposal. Yet his support would obviously be
indispensable to its success.
, 3. Design a revised system of sales call reporting. As this
project was envisioned, the computer would analyze salesmen's routes and product and customer profitability; it would
then print out detailed instructions to salesmen each week,
specifying customers to be called on, sequence of calls, target
sales by product, and weekly sales quotas. This project looked
promising, but it had not been evaluated by the very sales
people for whom it was intended. And its assumptions were
based on present account volume, not future potential.
All these proposals were listed, without specific cost or
benefit estimates, in the annual request for budget approval
28
submitted by the Systems & Procedures Department. The
president, when they were presented to him for review, reacted with irritation. Were they the best opportunities available for the application of the computer resources? What
economic results could realistically be expected? No good
answers were available.
Essentially the same questions are raised by any computer
development proposal. They are basically questions of feasibility - a concept often misunderstood and even more often
misapplied where computer projects are concerned. It is this
concept, crucial to soundly based computer development efforts, that we will now briefly consider.
Three Tests of Feasibility
Recently the president of a German chemical company was
asked to examine and approve a proposal for an exciting new
management information system. Featuring a desk-side cathode-ray tube inquiry terminal that would display on demand
any data in the computer files, the system would enable the
president to compare current production figures, by product
and/or by plant, against plan; it would break down current
sales figures in half a dozen different ways; it would display
inventory levels, current labor costs and trends, material costs
- in short, just about any kind of operating data he might
care to request. A few years earlier such a project would
have looked like science fiction; in 1967, its technical feasibility was assured.
Nevertheless, the president turned the proposal down. As
he explained his decision to a McKinsey interviewer, "I care
more about what will happen five years from now than what
happened yesterday. Anyway, I already get all the routine
data I can handle. What would I do with more?'
Proof of a Payoff
The incident is significant because it typifies a trend. Computer technology has made great strides in just the past few
years. Fewer and fewer applications are excluded from consideration because of limits on computer file capacity, internal
speed, or input/output ability; more and more technically
exciting projects are being proposed for management approval.
Particularly when corporate management is unaccustomed
to dealing with the computer department, it takes a certain
amount of hard-boiled skepticism to insist on proof of a payoff. Yet the fact is that technical virtuosity is no guarantee of
problem-solving potential. The most ingenious new proposal
may be merely a fancy new wrapping for an outmoded product. Instant access to data generated by an outmoded cost
accounting system, for example, is at best a dubious blessing.
Back in the days when the computer's full potential, and
hence its full impact on operating systems, was not foreseen,
the overall feasibility of a proposed computer application was
generally equated with its technical feasibility. That being
the case, it made sense to let the computer professionals decide how to use the computer. Today, judging from the findings of our study, this same policy of delegation is being followed in most companies. But it no longer makes the same,
kind of sense. Technical feasibility is only one aspect of overall feasibility. For the great majority of business applications,
it is no longer an important stumbling block.
The concept of "feasibility" really takes in three separate
questions. There is the test of technical feasibility: "Is this
application possible within the limits of available technology
and our own resources?" There is the test of economic feasibility: "Will this application return more dollar value in benefits than it will cost to develop?" And there is the test of operational feasibility: "If the system is successfully developed,will it be successfully used? Will managers adapt to the system, or will they resist or ignore it?"
COMPUTERS and AUTOMATION for April, 1969
Continuous Reassessment
;
1
I
I
Particularly on complex and ambitious computer development projects, these key questions of feasibility can seldom be
answered once and for all at the time the project is proposed.
Continuous reassessment of the technical and economic risks
and payoff probabilities may be vital to keeping such a project
on the right track. But a careful initial assessment can go far
to avert costly misapplication of scarce computer resources.
It is dangerously easy, however, to avoid confronting the
full implications of feasibility until a project is well under
way. Technical feasibility, though less often a question mark
today, is still the test most commonly considered at the start.
The issue of operational feasibility is far too often neglected
until the new application is actually tried out in practice and
perhaps found wanting - the costliest kind of feasibility test.
And economic feasibility - the measure of how much expected dollar returns will exceed expected costs - is frequently given only superficial examination.
Since a company's computer resources are seldom equal to
its computer opportunities, economic feasibility should almost
always be a key criterion in' weighing the merits of technically
feasible projects. Yet it is frequently assessed rather casually,
on the grounds that the important benefits are intangible, and
intangible benefits can't really be evaluated. Actually, of
course, the very difficulty of measuring intangible payoffs is
the best argument for imposing on managers the discipline of
explicit evaluation.
In assessing the cost-benefit balance of a proposed application, computer personnel can, of course, provide the needed
input on costs. The assessment of benefits, however, requires
a full understanding of the operations affected and the policies that govern them - an understanding that only operating
executives can really bring to bear.
To achieve its economic potential, a computer project may
also require substantial operational changes - changes in corporate policies, staff reorganizations, the construction of new
facilities and the phasing out of the old. It will certainly require the support of operating managers and their staffs, and
it may also depend on the cooperation of dealers, suppliers,
and even customers. Consider the case of a hardware distributor who req uires his customers to submit orders on coded
forms designed for the computer. All his customers may want
the faster service promised by the system, but some may balk
if it entails a messy problem of staff retraining.
Corporate computer staffs cannot really judge the necessity
of such changes, much less implement them. At most, they
can advise the operating managers who must make the final
assessment of operational feasibility.
Against. this background, let us look more closely at the
problems and opportunities confronting the companies in our
study.
Past Successes and Present Problems
Ironically, the basic problems currently besetting the man·
agement of the computer effort in most of the companies we
studied have their origin in the successes of the past. In 30 of
the 36 companies, conversions of routine administrative and
accounting operations to computer systems are already complete, or so close to completion that only minor incremental
savings are expected from the mechanization of remaining
manual procedures. Typically, most of the people who accomplished these conversions are still operating and maintaining the systems they helped to install. But others who participated in the early installations now enjoy a different organizational status. They constitute the nucleus of a corporate
computer staff. Instead of reporting to the controller, in some
cases they now report directly to top management.
COMPUTERS and AUTOMATION for April, 1969
For obvious reasons, these computer department staffs are
under pressure to show results in the form of new computer
systems. Technically speaking, they may be superbly equipped
to respond to management's expectations. Typically, they are
highly skilled in computer systems design, and their status as
professionals is unchallengeable. But they are seldom strategically placed (or managerially trained) to assess the economics of operations fully or to judge operational feasibility.
These limitations, although they reflect no discredit on the
corporate computer staff, are raising ever more serious obstacles
to the success of new corporate computer efforts, our findings
indicate.
Another obstacle to future success, also stemming from past
experience, is management's lack of exposure to the feasibility
problem. Back in the days when corporate computer efforts
centered on the conversion of accounting and administrative
systems, management seldom had to concern itself with the
issue of feasibility. With a relatively orderly manual system,
the feasibility question centered on the technical problems of
programming the computer. Economic benefits could be determined with relative ease in terms of clerical payroll reductions: Once a company had learned how to estimate conversion costs realistically, assessing economic feasibility was relatively simple. And operational feasibility was assured when a
single executive, such as the controller, had charge of both
the development and operating phases of the new system.
Today the situation is very different. Applications are not
only more complex, but also more far-reaching in their impact on different operating departments. Feasibility is no
longer an issue that 9perating managers can ignore, for it is
affected by complex economic and operational questions that
the staff specialists are unequipped to answer. Yet many
managers - far too many - are still leaving the whole question of feasibility to the computer professionals. At the same
time, they are neglecting their own responsibility for setting
the direction of the company development efforts.
The background sketched above, then, typically affects the
computer effort today in two ways.
Attitudes of Management
First, today's management practices and attitudes, inherited
from a time \vhen the full scope of the computer's potential
was not foreseen in most companies, are falling short of the
demands of to day's task. Over the past five years, computer
staffs have typically doubled. The department that had 40
people in 1962 has 80 or 85 now, and expects to double again
by 1975. Yet no overhaul of the management practices of
earlier years has taken place. In 14 of the 3 6 companies we
studied, nothing deserving the name of an overall plan for a
full range of computer applications is yet in evidence, and the
economic and operational feasibility of individual projects is
seldom fully explored. Ten companies, including a number
that do have a computer plan, are providing few if any shortterm objectives against which the progress of individual computer projects can be measured.
Range of Computer Projects
Second, the range of computer projects now open to the
company is circumscribed by the limited background of its
computer personnel and the limited initiative of its managers.
Consider a list of four proposed applications recently submitted to the president of a midwestern electric machinery
manufacturer:
1. Put labor records on random-access files, so that production department or machine group efficiency (now the
subject of a weekly report) can be measured daily.
2. Mechanize the follow-up of delinquent accounts receivable. (At present the computer lists delinquent accounts and
29
shows the agc of outstanding debits, but clerks review the list
and handle the follow-up.)
3. Install a data-transmission terminal at the warehouse
receiving dock so that receipts can be recorded immediately
on the computer file when shipments are unloaded. (At present good pieces are counted and the count punched into cards
only after a quality control inspection.)
4. Double the core memory of the computer to permit
multiprocessing of data processing jobs instead of running
them one at a time as at present. (The computer is currently
loaded less than two shifts, five days a week.)
None of these proposals had much relevance to the wellbeing of the corporation. Individual machine shop efficiencies,
for example, had long been appallingly low, and the vice
president for production was convinced that efficient production runs would be impossible until the design department
learned to reduce the catalog of parts used in assemblies. But
no attempt had ever been made to put bills of material on
computer files, and an overall analysis of the catalog of parts
would have to be done manually, an excessively time-consuming job. Accordingly, the president decided to postpone Proposal No.1, making his reasons quite clear to both the computer department manager and the production head. Later,
these two men came up jointly with a project, which was
promptly approved, to transfer bill-of-material descriptions
from a manual file to computer files. With this application
"on stream," production managers are beginning to show real
interest in making more extensive use of the computer, and
the computer staff is gaining a valuable understanding of the
practical problems of production. At least in this area of the
business, the prospects for profitable future computer applications look good.
Operational feasibility was the Achilles heel of Proposal
No.2. The sales vice president firmly opposes automatic
dunning of delinquent accounts. The clerks who now analyze
the delinquent listing are pensioned salesmen who make collections by phone and call personally on seriously delinquent
customers. This system of debt collection will not be changed
until there is a change at the top - and that is unlikely to
happen until 1971 at the earliest.
As for Proposal No.3, the quality control inspection is
essential since high-value components make up 40 percent of
all receipts at the dock, and typically between 3 and 5 percent of all items received are returned to the manufacturer or
held pending billing adjustments. Thus, although a datatransmission terminal would put data on the computer file
two to three days earlier and might avoid some of the interruptions in work flow that now result from reported stockouts, the data would be faulty and would require detailed
subsequent correction and audit.
The Unbridged Gap
Such examples, symptomatic of the unbridged gap between
computer staff and operating management, could probably be
duplicated in most corporations. They are as discouraging to
computer professionals as they are to operating management,
and they doubtless account for the tendency, observable in
many of the companies we studied, for computer staffs to
take refuge in refining the internal operating efficiency of the
computer department itself (as in Proposal No.4).
If computer systems design must be so closely linked to
operating procedures even in apparently simple applications,
it should not be surprising that the more ambitious projects
conceived by computer staffs so seldom meet the tests of economic and operational feasibility. To make better use of computers in the future will require expanding the scope and
capabilities of computer professionals and bringing managers
to a fuller awareness of the computer's vast potential. The
history of computer developments to datc has limited both.
30
The Opportunities: Near and Far-Out
The computer's credentials as a cutter of clerical payrolls
are now beyond dispute. On the evidence of its achievements
in a few exceptional corporations, we believe that the computer can make an equal or greater contribution to corporate
profits by reducing the cost of goods sold.
The more successful companies in our study have recognized this potential and are already beginning to exploit it.
The dominant lesson of their experience so far is that this
second stage of the computer revolution, unlike the first, entails
real operational changes - new, and at first uncomfortable,
ways ()f doing business that will quite possibly encounter resistance within the company.
For the companies moving into operating system applications, moreover, the issue of feasibility has emerged on a new
level of importance. They have found that technical feasibility is often a problem because marketing, production, and
distributioH systems are subject to outside influence and therefore less orderly than accounting systems. Since the benefits
do not derive from reductions m payroll dollars, they have
often found it harder to determine economic feasibility. Most
significantly, the~ have found that the operational feasibility
of a project is vitally dependent on the attitude of operating
managers.
,
I
Teamwork
Teamwork, then, is the key. Where top management provides leadership, and operatmg managers actively and enthusiastically cooperate with professional computer staffs, major economic achievements can result. Even a fairly commonplace computer application such as inventory control requires
such cooperation. Design engineers must give adequate notice
of design changes; sales planners must furnish detailed product sales forecasts; and management must give guidance on
spares requirements and desired customer service levels. But
once developed on the early projects, cooperation between
managers and professional computer staffs becomes an important stimulus to the development of profitable further
applications.
Consider the case of one manufacturer of heavy construction equipment. In this company, whose first computer-based
inventory control system went into operation well over a decade ago, computers now play such an integral role in production planning and control that it is difficult to picture the
company without them. These are some of the jobs now being done by computer:
1. Consolidating sales forecasts from 31 countries. Forecast
data are first consolidated by region, product, and model;
then they are correlated with figures for seven previous years.
Trends are established for each product group, and forecasts
that seem not to "fit" are pulled out for further staff review.
The president and the vice president for sales use these staff
analyses in their annual budgeting discussions with division
heads at corporate headquarters.
4
I
"
I
2. Establishing a quarterly manufacturing plan for each of
13 plants. These plans are updated monthly by reconciling
revised sales forecasts with records of finished goods inventory
and work in process in final assembly. The revised manufacturing program is then exploded into component requirements, and a "net component requirement" analysis is prepared. Extensive manual analysis by production planners is
still required to supplement these computer analyses, but lead
time between customer order and delivery has been reduced,
and the cost of shipping finished goods from depots in surplus
to those with shortages has been cut drastically.
3. Maintaining cost schedules in all plants showing the
economics of make-or-buy decisions. In conjunction with the
"net component requirement" report, these cost schedules
COMPUTERS and AUTOMATION for April, 1969
,
make possible intelligent work-load leveling and allocation
among plants. 'Where there is an option to contract work out,
managers can make their decisions with full knowledge of
both the costs and the effect on specific work centers within
each plant where bottlenecks are predicted.
4. Centra} recording of all engineering changes. Before an
engineering change is put into effect, components in stock are
exhausted first wherever possible. With changes occurring at
a rate of about 2,000 a month, the costs of writing off obsolete stock used to run as high as $1.5 million annually before
the advent of the computer. In the past three years, these
costs have been reduced to approximately $500,000 a year.
5. Maintaining cumulative records on labor efficiency. In
addition to detailed information on direct labor costs and
trends, this system provides production planners with data on
the work content of each component by work center~ These
data have been invaluable for scheduling manpower requirements to meet a varying production schedule, and particularly
in planning the start-up of two new plants, which required
the transfer of hundreds of skilled manufacturing workers.
The complex network of systems which produces such results has been evolving for 12 years now, and its net benefits
to date have been outstanding. Overall, management credits
computers with reducing lead time between order receipt and
delivery by three to five months for U.S. customers, and with
cutting direct labor requirements by 2 percent through improved materials availability and better control of work flow.
Since direct labor costs are approximately $100 million per
year, this fractional saving is significant both in absolute terms
($2 million) and as a percentage of before-tax profits (5
percent) .
I
I
r
~
Evolutionary Development
J
Another example of evolutionary development is offered by
a major consumer goods corporation. This company gives its
product managers and marketing staffs access to a comprehensive, detailed sales history file, in which total U.S. sales over
three years are cross-referenced to show product sales data by
geographic region, type of outlet, timing with relation to promotions, and packaging. This system evolved from an order
entry and billing system that recorded sales solely by customer number.
The direction of this company's development effort was set
early in 1959 by a product manager who foresaw the potential value of a comprehensive marketing information system.
Today, in addition to recording orders centrally, the system
he envisioned is used to schedule production at nine plants
and to coordinate shipments from 13 warehouses. One gauge
of its usefulness is the willingness of marketing men to pay
the salaries of the programmers who prepare on demand
whatever analyses may be needed by marketing managers.
Evolutionary development is typical of systems requiring
audited data bases, since these cannot be built up overnight.
But other systems, equally ambitious, can sometimes be developed quite rapidly where management recognizes that the
data-base approach is not the only, nor necessarily the best,
way to develop advanced computer applications.
A manufacturer of high-style clothing, with national outlets
and multiple plants, decided two years ago that computers,
hitherto used only for accounting purposes, could furnish major help in forecasting sales and establishing preliminary cutting schedules at the beginning of each season. The resulting
computer forecasting model has already proved so successful
in matching production to demand that a project is now under way to put computer forecasting methods to work in
planning purchasing decisions.
Similarly, a number of oil companies have moved quickly
into new fields unrelated to previous computer development
work. Several have successfully undertaken crash programs to
COMPUTERS and AUTOMATION for April, 1969
develop computer-based seismic analysis techniques to assist
in the planning of exploratory drilling, and more than one has
developed a computer model of the crude oil distribution system in order to improve the scheduling of its tanker fleet,
at potentially vast savings.
In a matter of months, one oil company moved to transfer
the production and maintenance records of thousands of domestic oil wells to computer files where they can be correlated
and analyzed. This system enables production decline curves
of wells and fields to be plotted and future production forecast under various alternative secondary recovery programs.
It also calls management's attention to wells that are no
longer producing enough to cover marginal costs. The principal task in developing this computer system was one of data
reduction and file design, and here there was ideal matching
of the talents of the computer systems men and petroleum
engineers. With the engineers' enthusiastic support, the computer staff is now exploring the feasibility of making the same
data accessible to engineers in the field through graphic display units. The obstacles are great, but the potential payoff
from improving the effectiveness of operating engineers, who
control expenditures in the hundreds of millions of dollars per
year, is greater still.
Communicatio'ns
Finally, in industry after industry where such data are
critical, the science of communications is being wedded to the
science of computing to centralize record keeping, planning,
and control in an ever more complex economic environment.
Railroads have "control centers" where up-to-the-minute central records are maintained on the movement of freight and
rolling stock. Retail chains are using teleprinters and central
computer-based dispatch systems to reduce branch-store inventories by cutting the stock-replenishment cycle. A woodproducts company is coordinating production at its nine mills
to match sales orders transmitted by branch offices throughout
the United States directly to a central computer. Banks are
handling branch accounting centrally; it is interesting to note
that one of the main reasons cited for the recent merger of
three large British banks was the opportunity to consolidate
the banks' computers and computer know-how. And virtually
all the major airlines now have their own versions of the seat
reservation system that first proved computers able to control
large communications networks on a commercially feasible
basis.
It is often extremely difficult to assess the overall economic
effects of these advanced computer applications, for the simple
reason that where the corporation would be now without its
computers is well-nigh impossible to determine. But many of
these companies are convinced that they have the computer
to thank for the fact that they are beginning to outdistance
.their competitors.
Leadership
The resources - computers, professional computer systems
men and programmers, management scientists, and communications experts - are available to all. But the team needs
leadership. Advanced computer application concepts, with
potential impact on the central activities of a corporation,
must have sponsors high in the management pyramid to plead
their case. The leadership of enthusiastic managers will gain
the commitment of operating men - and teamwork between
operating men and computer professionals will turn concepts
into practical reality.
If the situation prevailing in the companies we studied is
typical of U.S. industry as a whole, it is a fair guess that
more than half of the proposed computer applications currently awaiting management approval were not originated by
31
operating managers in consultation with computer staffs but
proposed independently by systems and programming professionals. Yet the experience of the more successful computer
users leaves little doubt that operating managers, well motivated and equipped with some knowledge of computer capabilities, are likely to be a better source of ideas for profitable
changes in operations than are computer professionals. The
most profitable applications uncovered in our study had originated with operating executives pondering such ideas as these:
• If only we had a way to test the reliability of the sales
forecasts made by these regional managers of ours, we might
not find ourselves out of manufacturing capacity in Italy at
the same time that we're laying off valuable skilled labor in
Brazil.
• If only we had a way of recording and analyzing all our
customer orders in one place, we ought to be able to allocate
our production better - improve mill efficiencies and raise
the yield from our raw materials.
• If only we could easily check out our historical sales
performance by product, package, and so on, maybe we could
interpret our test marketing results faster and more reliably.
• If only we could play with alternatives on our tanker
deployment, we might use our capacity better - charter in
less and charter out more.
• If only we could project our needs for skilled labor three
months out, we could save the expense of these crash recruiting and training programs.
Two lessons emerge from all the varieties of successful computer experience that we have studied. First, there is a unique
set of feasible and profitable computer applications for each
company. Second, most of these applications are closely related to the key strategic opportunities that the top executives
are really concerned about: marketing and distribution operations in the package goods company; production operations
in the capital equipment concern; facilities planning operations in the chemicals maker; exploration and producing operations in the petroleum company; financial planning in the
conglomerate; and so on. Such applications may be designed
to reduce costs of goods sold, or to increase revenues by
changing operating methods directly. Or, as already noted,
they may seek to improve the staff work and analyses available to decision mahrs.
Implications for Top Management
These lessons, in turn, have important implications for the
top manager. Since each corporation has its own unique pattern of problems and opportunities, there is danger in trying
to duplicate the successes of others: The computer development strategy that" has worked well for one company may not
work at all for its competitor. For the same reason, a company would be unwise to pin all its hopes on vendor-produced
"applications packages" where major development projects
are at stake. Nor can the answers be left to the professionals.
No top executive is going to turn over the operation of his
key departments to specialists with little or no operational
experience.
In almost every industry, at least one company can now be
found that is pioneering in profitable new uses of computers.
In such companies, our findings suggest, the key to success has
been a 'strong thrust of constructive interest from corporate
operating executives who have put their own staffs to work on
computer development projects.
\Ve believe that other companies will follow their lead. Indeed, it may soon be a nearly universal practice to transfer
operating staff to computer development projects, either by
making them members of a project team or by attaching
them for a year or two to the corporate computer staff.
32
Simulation
Another much-discussed area of computer use is management information and control. A few companies have already
succeeded, by means of computer systems that sort out and
speed routine data to the user, in notably improving the quality and quantity of specific information available to operating
managers. Others, as noted earlier, have made profitable use
of the computer in decision making through simulation models
designed to improve decision making by predicting the impact
of alternative actions on economic and operating realities.
Skills in the construction of such models are widespread and
growing, and their results have frequently been noteworthy.
A fertilizer manufacturer has used computer-based simulation to help top management answer such questions as these:
• How much should we plan to manufacture, ship, and
store at the plant location in order to minimize total accumulated costs of production, distribution, and storage over a
one-year period?
I
i
• How much, if at all, could we reduce total costs by
renting additional storage in outlying locations? What would
be the effect on our present production, shipping, and storage
program?
• How large a market area should be served from each of
our warehouse locations?
• Where should new plants be located with respect to
warehousing locations and market areas?
A well-known food products company has constructed and
used a computer-based simulation model enabling it to assess,
under various possible 1970 and 1975 environments: (1) the
relative profitability of different product markets; (2) the desirability of investing in new-market development; (3) the
impact of investment in added plant capacity; and (4) detailed income statements based on these projections.
Again, computer-based risk analysis techniques have demonstrated their value in a wide i'ange of capital investment situations. The industrial chemicals industry is known for the
magnitude of both its investment and its risks. Since industry
capacity directly affects market price, these risks are aggravated by the uncertainty of industry intelligence regarding
competitors' plans for adding new capacity. Risk analysis,
made practical by computers, has proven invaluable for evaluating alternative strategic plans with the help of simulation
models, sometimes even including simulation of alternative
competitive responses by the application of game theory. To
exploit the potential of these and related techniques, an increasing number of corporations are finding it necessary to
supplement the professional skills of computer men by recruiting specialists in the management sciences.
'Vhat is true of simulation models, however, is hardly true
of the so-called total management information systems that
have beguiled some computer theorists in recent years. Much
effort and ingenuity have been devoted to the design and promotion of such systems, and many businessmen are understandably intrigued by their possibilities. Yet in terms of economic payoff and operational feasibility they are as yet illdefined, and certainly they are a long way from practical
realization in business.
Doubtless the computer's information processing capabilities will one day eliminate the need for large staffs occupied
with collecting and interpreting information from various
sources for the use of decision makers. But whether the computer will ever be able to evaluate strategic opportunities or
indicate the proper timing for corporate actions is by no
means assured. Nor are man-machine dialogs via desk-side
consoles likely to become a feature of life in the executive
suite any time in the foreseeable future; top management's
"interface" with the computer is unlikely to be anything more
COMPUTERS and AUTOMATION for April, 1969
I~
1
exotic than a telephone, with a human information specialist
at the other end of the line. What counts, of course, is not
the sophistication of the interface but the responsiveness of
computer-based systems to management's information needs,
and the quality and timeliness of the information they can
provide. Here, without doubt, the potential of the computer
is only beginning to be realized. But "integrated total management information systems" drawing on a single data base,
which have so often been touted as the wave of the future,
are another matter. They have not yet come to pass _. and
it is far from clear that they ever will.
In short, the potential of comprehensive computer-based
information systems and the role of the computer in decision
making are still surrounded by question marks. Research in
these a~eas may be a sound investment for some companies,
even though the costs of experimentation are high. But no
company should embark on a program to develop a major
management information system except to meet a specific,
well-defined need. Even then it should carefully weigh its options - including the option of applying its scarce computer
resources to areas where operating success and economic payoff can be predicted with greater confidence.
Keys to the Future
In embarking on the present study, McKinsey & Company
analysts were not seeking fresh evidence of a gap between potential and performance with respect to the management of
the computer effort. The existence of such a gap has been
obvious for some time to most informed observers. We were
concerned, rather, with determining the present dimensions of
the gap, analyzing its background and causes, and synthesizing
from the practices of the top performers a few succinct management guidelines for maximizing the computer's effectiveness and unlocking its profit potential.
Evidence on the first two points - the performance gap
and its underlying causes - has been reviewed in the earlier
pages, and the general nature of the remedies has been indicated. Against this background, certain lessons emerge for the
senior executive who is dissatisfied with the performance gap
he sees in his own company and is determined to do what he
can to close it.
In the computer field, as in other areas of management, the
usefulness of generalizations from successful experience is
rather sharply limited. It is possible to state some of the principles a company must follow to have a reasonable chance of
success with the computer. But there will always be other
factors - constraints, needs, or opportunities - which are
peculiar to each company and can only be determined in the
light of the individual situation. Hence it is useful to state
general precepts only if their neglect is rather widespread and
the consequences of that neglect are costly. This is the case
in the management of the computer effort today.
The common denominators of successful computer practice,
as seen in the companies we have examined, may be expressed
in terms of three principles: the rule of high expectations, the
rule of diversified staffing, and the rule of top-management
involvement.
The rule of high expectations. In all of the companies that
are realizing outstanding economic results from computer applications, top management is simply unwilling to settle for
anything less. In the less successful companies, many managers exhibit a tendency to keep the computer at arm's length
for fear of exposing their technical inadequacies. This tendency is conspicuously absent among the top computer users.
Departmental and divisional managers in these companies
know that top management will insist on economic resultsand that they will be held personally responsible for achieving those results.
The new president of a capital equ~pment manufacturer,
COMPUTERS and AUTOMATION for April, 1969
who has succeeded in getting a badly stalled computer program in his company moving again, typifies the prevailing
tone of management expecta.tions in the better-performing
companies. Said he: "I ask my department heads to give me
regular formal reports on their current successes and failures
with computers and their future objectives. Right now they're
a bunch of sheep with computers. I aim to convert them into
enthusiasts, so that later I can be jockey, not herdsman."
The rule of diversified staffing. A computer staff whose experience is limited to successful conversion of accounting and
administrative operations is seldom really qualified to design
and install new systems in major operating functions such as
manufacturing and marketing. Computer professionals alone
seldom constitute an adequate corporate support staff.
To make the most of their opportunities for profitable corporation-wide use of the computer, therefore, the top-performing companies take one of two organizational approaches.
Some assign to the corporate computer staff - along with the
usual operations research specialists and other professionalsat least one talented individual with experience in each of the
major functions of the business. Others, relying on the project
approach to computer development, use project teams staffed
by temporary transfers from operating departments. This arrangement, too, encourages good support from all levels of
management.
To head up the computer staff and assume responsibility
for the imple:nentation of development plans, the outstanding
companies have in all cases been careful to pick a manager
who commands, or can quickly learn to command, respect
and confidence throughout the organization. The appointment of the right man to this position is seen as a key contribution that top management can make to the success of the
computer effort. It is also recognized that this individual's
effectiveness depends more on his personal stature and professional skills than on the precise location of his unit in the
corporate hierarchy. We found no evidence, statistical or
otherwise, to suggest that high organizational status assures
effective performance on the part of the corporate computer
staff.
The rule of top-management involvement. If anyone man
can be said to hold the key to the computer's profit potential,
it is probably the chief executive. He has a very definite responsibility for the success of the computer development effort, and it is not a responsibility that he can safely delegate.
At a minimum, the chief executive who wants maximum
results from his company's computer effort must do five
things. First, he must approve objectives, criteria, and priorities for the corporate computer effort, with special attention
to the development program. Second, he must decide on the
organizational arrangements to carry out these policies and
achieve these objectives. Third, he must assign responsibility
for results to the line and functional executives served by the
computer systems - and see to it that they exercise this responsibility. Fourth, he must insist that detailed and thorough
computer systems plans are m.ade an integral part of operating plans and budgets. Fifth, he must follow through to see
that planned results are achieved.
There is nothing novel in any of these recommendations;
they are standard operating practice for most chief executives
in most of their traditional areas of responsibility. Many
otherwise effective top managements, however, are in trouble
with their computer efforts because they have abdicated control to staff specialists - good technicians who have neither
the operational experience to know the jobs that need doing
nor the authority' to get them done right.
Only managers can manage the computer in the best interests of the business. The companies that take this lesson to
heart today will be the computer profit leaders of tomorrow.
D
33
The Implications of the Information Sciences for
INTERGOVERNMENTAL _COOPERATION IN COMMUNICATIONS AND EXCHANGE OF INFORMATION
William W. Parsons
Corporate Vice Pres. for,Administration
System Development Corp.
2500 Colorado Ave.
Santa Monica, Calif. 90406
One morning, not too far in the future, a City Manager or
a Cabinet Officer or a Lieutenant Governor will enter his office, draw his chair up to the desk, activate his computer
terminal, and proceed to have a conversation with his data
base. That, in the current jargon of computerized information processing, is how he will "get the basic facts".
In this way, a few executives in a few large organizations
today are beginning to secure the information they need for
planning, organization, or control. If we look ahead two or
three years, the executive is unlikely to be the only one in his
organization working in this fashion. The technology is too
ready, the advantages of on-line computer use too compelling,
the implementation of interactive data management systems
too easy and too economical for an effective organization to
withhold direct computer access from its star technical and
management performers.
Before going further, let us review some of the history and
current happenings in the area of intergovernmental cooperation in the exchange of information among the various levels
of government.
Exchange of Information in Government: Review
Federal legislation in the 1930's and later to improve the
conditions in our society have led to significant changes in the
pattern of intergovernmental action. Increasingly, federal,
state, and local governments are being brought together to act
as partners in carrying out programs that are designed to meet
public needs. This places a high premium on close cooperation and a steady flow of information. But all levels of government have been slow to change their habits and develop
new methods of working together. This is particularly true
with respect to the development and use of information
systems.
The requirements imposed by the federal grant-in-aid programs beginning in the 1930's inevitably led to an increased
exchange of information flowing through the supervisory and
reporting processes. Prior to this time there had been considerable exchange of information of a census nature, but very
little of this exchange bore directly upon the operating of programs. The one important exception was in the field of tax
information.
(Based on· an address at the Eighth Management Conference, Hawaii, 1968)
34
Tax Information
States have had access to federal tax returns since the beginning of federal taxation. In the early years, states were
able to send agents to Washington to examine returns under
formal agreement. The procedure was formalized by the
Revenue Act of 1926 which opened the federal returns to
state officials at the request of the governors. By the end of
1965, the District of Columbia, and 29 of the 34 states with
broad-based personal income taxes, had agreements with the
Internal Revenue Service for the cooperative exchange of tax
records. In general, the agreements provide for the establishment of mutually acceptable programs, the cooperative exchange of information allowing the federal and state governments to obtain each other's returns, and exchange of other
necessary information to insure effective compliance. The Advisory Commission on Intergovernmental Relations has been
particularly influential in this entire area of cooperative exchange of information.
The development and use of information systems among the
three levels of government is the subject of the recent report
by the Intergovernmental Task Force on Information Systems,
dated April 1, 1968. This report contains recommendations
to improve the flow of information within and among federal,
local, and state governments. The Task Force was arranged
by the Bureau of the Budget, Council of State Governments,
National Association of Counties, National League of Cities,
U.S. Conference of Mayors, International City Managers'
Association, and the Advisory Commission on Intergovernmental Relations. The purpose of the study was to:
Prior to his present position with SDC, William W.
Parsons was the Administrative Assistant Secretary of the
U.S. Treasury. He received his B.S. degree from the Univ.
of Southem Calif., his Master's degree from Syracuse
Univ., and a Doctor of Laws degree from Southeastern
Univ. Mr. Parsons has been active in the American Society for Public Administration, and has served as president
of the Washington chapter, national vice president, and
national president. He is currently serving as a member
of the Advisory Board of the Committee for Economic
Development's Committee for Improvement of Management in Government; he is a founding member of the
National Academy of Public Administration, and is a
member of a number of professional honorary fraternities.
J
COMPUTERS and AUTOMATION for April, 1969
((The technology resources currently available far exceed any needs. But
computers and storage devices, in and of themselves, are of little value.
What counts is how we define our requirements, specify our system,
and implement it. The people in the system are all-important; they
are fundamental from the point of view of design, decision making, and
control."
( 1) Identify impediments to attammg an effective flow
of information within and among governments, and
(2) Recommend actions that could be taken at the federal, state, and local levels of government.
The Flow of Reliable Information
The Task Force concluded that intergovernmental approaches to the solution of public problems require that reliable information flow readily among those who share responsibility so that concerted action may be taken. In general, information systems now in use and current efforts to improve
them are not adapted to satisfy this requirement. A number
of factors impede efficient flow of information. These factors
include: the lack of strong, central coordination at all levels
of government over the development and operation of internal
information systems; and the fragmentation of federal grantin-aid programs available to assist state and local governments in this development.
The Task Force made twenty specific recommendations
under the following headings:
1. "Improving information systems within governr:nents.
2. Improving the exchange of information among gov-
~I,
I
3.
4.
5.
6.
7.
l
ernments. An example is Recommendation 5" which
reads:
"Develop, under the leadership of the U.S. Bureau
of the Budget, a standard 'package' of socioeconomic data to be used as a base by Federal agencies in obtaining information from state and local
governments."
Strengthening information systems at the local level.
Sharing systems knowledge.
Achieving compatibility among systems.
Improving information about federal assistance programs.
Guidelines for action.
improve the statistical services of their educational
agencies. The Office of Education is authorized to
provide grants to cover half the cost of such improvement programs, but no state may be paid more than
$50,000 in anyone fiscal year.
2. The State Technical Services Act of 1965 (PL 89182) provides grants to establish State Technical Information Centers as a means for stimulating indus,:,
trial and economic growth.
3. The Housing Act of 1954, as amended, provides
grants to assist urban development planning programs
in small communities, states, and metropolitan areas.
4. The Law Enforcement Assistance Act of 1965 (PL
89-197) provides grants to states, counties, and cities
to develop new and better methods of crime prevention, law enforcement, and criminal law administration.
Also in many instances expenses for the establishment and
operation of information systems needed to manage grant
programs are recognized as allowable charges to the grant.
Another way in which the Federal Government helps to
improve information systems is by providing direct technical
assistance. For example, the Office of Economic Opportunity
sends teams of systems analysts to states to assist in the development of information systems patterned after a similar system operated by the OEO.
A third way in which the Government assists state and
local governments is by providing federal facilities in the administration of the grant program. To illustrate, the Bureau
of Census has prepared data files on population and housing
in the form of punched cards and computer tapes that can be
processed to provide a user with almost any kind of statistical
summary or small-area tabulation he may desire.
Also, state and local governments are now authorized to
use the federal ADP Service Centers of the General Services
Administration.
Federal Grant-in-Aid Programs
Complex System Leads to Fragmentation
A number of the federal grant-in-aid programs contain
authorizations designed to help state and local governments improve their information systems. Let me cite a few examples:
In spite of these resources, the present system of grant-inaid is much too complex to lead to anything except fragmentation. There are more than 400 separate grant authorizations; each is devoted to specific purposes; and the grants are
administered by more than 20 federal agencies, a fact that
creates major problems of information flow. This leads to
1. The National Defense Education Act (PL 85-864,
Title X, Section 1009) provides grants to states to
COMPUTERS and AUTOMATION for April, 1969
35
further complexity when state and localities seek help in unified information systems. Such proposals not only cross program lines but (obviously) agency lines.
Varied proposals are currently before Congress for improving the situation and I am sure progress will be made, but it
is quite evident that we shall be working on these problems
against great odds for a long time to come apd progress will
undoubtedly be slow.
Several documents are useful in trying to achieve a better
understanding of available federal resources. One is entitled
Catalog of Federal Assistance Programs, produced by the Office of Economic Opportunity. This catalog identifies all the
domestic assistance programs of the Federal Government- 459 of them - and provides a brief description of their
purposes, etc. Another document, issued by the Vice President's office, entitled Handbook for Local Officials, serves as
a guide to federal assistance primarily for local governments.
State and Local Activity
In addition to the extensive activities at the federal level,
developments emerging at the state and local level are stimulating an improvement in the flow of information. For example, efforts are underway for state and local governments
to establish joint service bureaus and/or cooperative agreements among various units of government. In Los Angeles
County a number of small cities are planning the establishment of their own processing center. For some time, there
have been efforts in the state of Iowa to establish a data
processing center to serve all levels of government in that
state. In fact, this particular proposal has been endorsed by
the Council of State Government's Committee on Information Systems.
A Statewide System
Another type of activity is exemplified by the so-called
"California Study," which was undertaken in California under the administration of Governor Brown. This study, entitled The California Statewide Information Systems Study,
was undertaken by the Lockheed Missiles and Space Company, a division of Lockheed Aircraft Corporation in Sunnyvale, California; it is one of the studies demonstrating the applicability of aerospace technology skills to government problems. This study resulted in an extensive report which recommended the establishment of a statewide information system
concept. Simply stated, the basic purpose of the concept is to
augment the information resources of California's public jurisdictions into a single, integrated system serving the information requirements of individual state and local organizations
as well as the needs of the entire state.
It \vas proposed that the State-Wide Information System be
developed as a federation of organizational computer centers
(state and local) tied together by an Information Central and
operating within a framework of compatibility rules. The
state of the arc in information sciences and technology would
permit the implementation of the Lockheed proposal; but the
state of the art of politics has impeded the implementation
considerably, particularly the intergovernmental relationship
aspects. But there has been some progress.
An Area-Wide System
Another type of study, of which the following is only typical, is one made by the firm of Touche, Ross, Bailey and
Smart, proposing an area-wide automated data processing system for the city of Memphis and other local government
organizations in that area. This particular study recommended
a five-year plan for implementation and ultimate automation
of the entire information processes of the city and of its interactions with county and other local governmental units.
36
Probably in hundreds of similar cases progress is being made.
\Vithin two or three years, according to today's best estimates,
it is entirely possible that any well-managed organization will
have 50 to 100 terminals at its headquarters building and
branch locations through the country. Through these terminals
and by means of complex computer programming, managers
in every functional department will put daily information
into, and will receive timely output from, one large central
computer.
A Simple Medium: Terminal and Data System
The medium through which these man/computer conversations are to tal;:e place is deceptively simple. It consists essentially of two parts: one is a computer terminal (the component physically present to the user), which may take the
form either of a teletypewriter or a video screen and keyboard
(voice analyzers are not yet perfected); the other is a general
purpose data system which, stored in the computer, enables
the machine to understand and carry out English-expressed
commands.
If this medium is to be widespread in the near future,
certain prior conditions need to be met:
;
1. That the enabling software exists - as a self-consistent
system - in packaged, off-the-shelf form. (Otherwise
the claim is pure speculation.)
2. That the medium is easy to learn and use (because
no company is likely to invest in a massive programmer training course for its management personnel).
3. That the software system is truly generalized - or
able to handle a wide variety of data for a wide
variety of applications. (Otherwise extension of this
on-line data management capability to all functional
areas of the company would require much more than
two or three years' time.)
4. That it operates under a time-sharing system (for no
company could afford a large private computer for
every user).
5. That a management information system based on the
framework of an interactive generalized data management system offers significant advantages over standard management reporting systems (or managers
would not bother to use the new medium).
Now, the medium does exist in at least some examples.
And the medium opens up a whole new way of thinking and
working with information. When the data are organized in
one central location, anyone who has a terminal and authorized access to the files can find out the information he needs.
\Vith the data "alive" and residing in the computer, the manager who bases a decision on the facts he obtains can review,
manipulate, summarize, and recall facts at will.
Technology Resources Available
The technology includes not only computers, which are
"front and center", but also many other devices, such as
microform applications, memory storage devices, etc. The
technology t:esources currently available far exceed any needs.
They are quite sufficient to meet any requirements for computation, documentation, or communication. A single illustration may help: We now have "micro images", microphotostatic storage devices. One chip, about one inch square,
contains the entire Holy Bible. That can indeed be called
"storage of information." It is certainly a "small testament"
to what is possible.
By far the more important problem, as I view it, is not the
technology, but how we organize to use the resources in terms
of systems and people. Computers and storage devices, in and
COMPUTERS and AUTOMATION for April, 1969
1
I
!
of themselves, are of little value. What counts is how we define our requirements, specify our system, and implement it.
The people in the system are all-important; they are fundamental from the point of view of design, decision making,
and control.
In the information business the word "systems" is a very
common one. To illustrate the information systems concept,
let's consider the short-order restaurant.
A Simple Information System:
A Coffee S,hop
As you will notice the next time you are in a coffee shop,
there is a spindle somewhere between the kitchen where the
cook performs his duties and the counter where the waitress
performs her duties. This spindle is a simple example of
an effective information system. The order blank that the
waitress fills out provides the input. The placement of this
input on the spindle puts the information in memory. The
spindle itself serves as a buffer between the waitress and the
cook. It also provides a queuing device - that is, it lines up
the various orders in sequence. It provides a random access
display whereby the cook and the waitress can both look down
the row of orders and see what comes next and what can be
combined, etc. The spindle provides control and settles arguments as to priorities, and it clearlYI provides a record. These
are all "operations research" terms that are used in the discussion of much more complex and difficult systems.
A Complex Information System:
SAGE
A very complex information system involving the use of immense communications devices and computers is SAGE (Semi- .
Automatic Ground Environment) System. In the SAGE System the computer is used as a device for many of the same
functions that the spindle performs in the short-order restaurant. Information about geography, weather, known plane
flights, etc., is stored in the memory of the computer. Radar
data converted to digital form and fed directly into the computer, are computed and sent to the operators. Operat
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
Producer : Adobe Acrobat 9.1 Paper Capture Plug-in
Modify Date : 2009:03:20 21:04:31-07:00
Create Date : 2009:03:20 21:04:31-07:00
Metadata Date : 2009:03:20 21:04:31-07:00
Format : application/pdf
Document ID : uuid:d7155a39-9264-4313-89b5-b38082fbe4d1
Instance ID : uuid:1f0e111e-7844-498d-8ccf-29797a1413ef
Page Layout : SinglePage
Page Mode : UseNone
Page Count : 64
EXIF Metadata provided by EXIF.tools