197204
197204 197204
User Manual: 197204
Open the PDF directly: View PDF 
.
Page Count: 52
| Download | |
| Open PDF In Browser | View PDF |
:llENCE & TECHNOLOGY
p
April, 1972
Vol. 21, No.4-
CD
COMPUTERIZED OBSERVATORY AT MAUNA KEA, ON THE ISLAND OF HAWAII
"Do What I Mean":
The Programmer's Assistant
Cryptology, The Computer, and Data Privacy
Computerizing a Membership Organization
The Information Industry and Government Policy
The Bad Image That Computers Are Earning
Political Lies:
An Acceptable Level?
W. Teitelman
B. Girsdansky
R. Pol/ert
T. Whitehead
W. G. Gearing,
and others
A. E. Kahn
M.
W.
C.
H.
_ __IIIIIIIJII!"
--- -
-
-
SAN TECH180
1P02121147F6 7301
*N
ECHNICAL SERVICES
00808
180 W SAN CARLOS ST
SAN JOSE CA
95113
•
DO YOU WANT TO
PREVENT MISTAKES BEFORE THEY HAPPEN?
- avoid pitfalls?
- find new paths around old obstacles?
- apply in practical situations the observations and
wisdom of great scientists and wise men?
- stimulate your resourcefulness?
- see new solutions to old problems?
- distinguish between sense and nonsense?
increase your accomplishments?
- improve your capacities?
IF SOl TRY-
The C&A Notebook on
COMMON SENSE. ELEMENTARY AND ADVANCED
devoted to research, development, exposition, and illustration of one of the most important
of all branches of knowledge, i.e. the subject of WHAT IS GENERALLY TRUE AND IMPORTANT =
+
+
+
+
+
+
+
+
THE FIRST SIX ISSUES ARE FREE - see the coupon - THE NEXT 21 ISSUES ARE:
Editor: Edmund C. Berkeley,
author, businessman, actuary,
scientist, computer professional,
first secretary of the Association
for Computing Machinery 1947-53,
editor of Computers and Automation.
RETURNABLE IN 7 DAYS
FOR FULL REFUND, IF
NOT SATISFACTORY WHY NOT TAKE A LOOK?
..... HOW CAN YOU LOSE?
-
7. The Elephant and the Grassy Hillside
8. Ground Rules for Arguments
9. False Premises, Valid Reasoning,
and True Conclusions
10. The Investigation of Common Sense,
Elementary and Advanced
11. Principles of General Science, and
Proverbs
12. Common Sense - Questions for
Consideration
13. Falling 1800 Feet Down a Mountain
14. The Cult of the Expert
15. Preventing Mistakes from Failure
to Understand
16. The Stage of Maturity and Judgment
in any Field of Knowledge
17. Doomsday in St. Pierre, Martinique
- Common Sense vs. Catastrophe
18. The History of the Doasyoulikes
19. Individuality in Human Beings, ...
20. How to be Silly
21. The Three Earthworms
22. The Cochrans vs. Catastrophe
23. Preventing Mistakes from Forgetting
24. What is Common Sense? - An Operational Definition
25. The Subject of "What is Generally
True and Important": Common
Sense, Elementary and Advanced
26. Natural History, Patterns, and
Common Sense
27. Rationalizing and,Common Sense
- - - - - - - - - - - - - - - - - - - - - - - (may be copied on any piece of paper) - - - - - - - - - - - - - - - - - - _
To:
COMPUTERS AND A UTOMA TION
815 Washington St., R5, Newtonville, Mass.
02160
YES, please enter my subscription to the C&A Notebook on Common Sense at $12 a year,
24 issues (newsletter style), and extras.
Please send me (as FREE premiums for subscribing) the first six issues:
1. Right Answers - A Short Guide to Obtaining Them
4. Strategy in Chess
2. The Empty Column
5. The Barrels and the Elephant
3. The Golden Trumpets of Yap Yap
6. The Argument of the Beard
I enclose $
( ) Please bill me
) Please bill my organization
Name ___________________________________ Title_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
Organization ___________________________________________________________
Address ______~------------------------------------------------Signature _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ____ Purchase Order No. ____________
2
+
QUESTIONS
AND
ANSWERS
about "The C&A Notebook on COMMON SENSE, ELEMENTARY AND ADVANCED"
INTERESTING:
Q: Is the Notebook interesting?
NUMBER OF ISSUES.PER YEAR:
A: We think so -- but you can judge for yourself.
You can see the issues, and if not satisfactory,
tell us to discontinue your subscription.
Q: I do not want to miss past issues.
How do I get them?
PAST ISSUES:
A: Some of the issues, like "Falling 1800 Feet
Down a Mountain" and "Doomsday in St. Pierre,
Martinique", are among the most exciting true
stories we know.
A: Every subscriber's subscription starts at
Vol. I, no. 1. Every subscriber eventually
receives all issues. Here is how it works.
The past issues are sent him four at a time,
every week or two, until he has caught up -and thus he does not miss important and interesting issues that never go out of date.
Q: Is the Notebook useful?
USEFUL:
A: It ought to be useful to anybody -- as useful
as common sense. There exists no textbook on
common sense; the Notebook tries to be a good
beginning to common sense, science, and wisdom.
UNDERSTANDABLE:
Q:
How many issues a
A: We promise 24 (newsletter style); we anticipate putting out 30 in each yearly volume.
Q: Is the Notebook exciting?
EXCITING:
Q:
year do you put out?
BOOK:
Can I unders tand the Notebook?
Q: Are you going to publish all the issues
of Volume 1 together as a book?
A: No; they do not fit together into a book.
A: Yes. It is nontechnical -- written in everyday language and using vivid examples.
Q: If I subscribe at the same time to
"Computers and Automation", can I receive
a premium for subscribing to the Notebook?
PREMIUMS:
Q: Do you cover in the Notebook all
parts of common sense, wisdom, and science
in general?
COVERAGE:
A: Yes, the Four Star Reprint -- see the
description below.
A: Yes, we plan to. The main subjects so far
are: systematic prevention of mistakes; avoiding certain fallacies; important principles;
important concepts; illustrative anecdotes; etc.
Q: Can I receive the 6 free issues (No. 1
to 6 of the Notebook) without subscribino?
A: No. Here is what happens. You subscribe.
We send you 8 issues and bill you. You can
send them all back in seven days, and the
bill is canceled. If you do pay the bill, you
receive the 6 free issues, 24 more issues of
Vol. 1, and six issues of Vol. 2, no. 1 to 6.
Q: Will the Notebook save me from
making important mistakes?
MISTAKES:
A: It ought to. One oft~e main purposes of the
Notebook is preventing mistakes.
COST:
Q:
Will the Notebook be worth the cost to me?
A: At about 40 cents per issue (30 issues for
$12), it is hard for you to lose out. EVEN ONE
important mistake prevented, may save you much
time, much trouble, and much money.
Q: If I do not like the Notebook, can
I cancel at any time?
GUARANTEE:
A: Yes. You will receive a refund for the
unmailed portion of ybur subscription.
'* * '* *
A reprint from
CDm~~
FOUR-5TAR REPRINT-------.
SCIENCE AND THE ADVANCED SOCIETY, by C. P. Snow,
Ministry of Technology, Londoh. England (April,
1966 issue of Computers and Automation).
THE INFORMATION REVOLUTION AND THE BILL OF RIGHTS,
by Dr. Jerome B. Wiesner, M.I.T. (May, 1971)
EMPLOYMENT, EDUCATION, AND THE INDUSTRIAL SYSTEM,
by Prof. John Kenneth Galbraith, Harvard Univ.
(Aug. 1965)
COMPUTERS AND THE CONSUMER, by Rtlph Nader
Washington, D.Ci (Oct. 1970)
,-
- -(may be copied on any piece of paper)- - TO: Computers and Automation
815 Washington St., Newtonville, Mass. 02160
)YES, you have convinced me to try the Notebook
on Common Sense, Elementary and Advanced.
Please enter my subscription at $12 a year,
24 issues, newsletter style, and extras.
Please send me Issues 1 to 6 as FREE PREMIUMS
for subscribing.
Please enter my subscription to "Computers
and Automation" at the same time ( ) with
directory $18.50; ( ) without directory
$9.50, and send my FREE premium, the Four
Star Reprint.
I enclose $._ _ __
Please bill me.
Please bill my organization.
RETURNABLE IN SEVEN DAYS FOR FULL REFUND
IF NOT SATISFACTORY
Name,__________________________________________
Title,______________________________________
Organization,________-------------------Address,~---------------------------
Signature_________________P.O. No. ______
COMPUTERS and AUTOMATION for April, 1972
3
Vol. 21, No.4
April, 1972
Editor
Edmund C. Berkeley
computers
and automation
Assistant Editors Barbara L. Chaffee
Linda Ladd Lovett
Neil D. Macdonald
Software Editor Stewart B. Nelson
Advertising
Director
Edmund C. Berkeley
Art Director
RayW. Hass
Publisher's
Assistant
Paul T. Moriarty
Contributing
Editors
John Bennett
Moses M. Berlin
Andrew D. Booth
John W. Carr III
Ned Chapin
Alston S. Householder
Leslie Mezei
Ted Schoeters
Richard E. Sprague
Advisorv
Com in ittee
James J. Cryan
Alston S. Householder
Bernard Quint
Editon'al0.tftces Berkeley Enterprises,lnc.
815 Washington St.,
Newtonville, Mass. 02160
617-332-5453
The magazine of the design, applications, and implications
of information processing systems - and the pursuit of
truth in input, output, and processing.
Computers and Programming
[T A]
8 "00 WHAT I MEAN": THE PROGRAMMER'S
ASSISTANT
by Warren Teitelman, Bolt Beranek and Newman,
Cambridge, Mass.
How to design and implement in a computer, a
"programmer's assistant", so that the sometimes
mistaken and often experimental behavior of a
human programmer is handled with the least
waste of his time and effort.
12 CRYPTOLOGY, THE COMPUTER, AND DATA
[T A]
PRIVACY
by M. B. Girsdansky, IBM Research Division, Yorktown,
Heights, N.Y.
An exploration of how to use the principles and
technology of secret or "hidden" writing, in order to
secure privacy in computerized records through computer programming.
The Computer Industry
Advertising
Contact
THE PUBLISHER
Berkeley Enterprises,lnc.
815 Washington St.,
Newtonville, Mass. 02160
617-332-5453
"Computers and Automation" is published monthly, 12 issues per year, at 815
Washington St., Newtonville, Mass. 02160,
by Berkeley Enterprises Inc.
Printed in
U.S.A.
Second Class Postage paid at
Boston, Mass.
Subscription rates: United States, $9.50 for
one year, $18.00 for two years. Canada:
add 50 cents a year for postage; foreign,
add $3.50 a year for postage.
NOTE: The above rates do not include
our publication "The Computer Directory
and Buyers' Guide"; see "Directory Notice"
on the page stated in the Table of Contents.
If you elect to receive "The Computer Directory and Buyers' Guide", please add
$9.00 per year to your subscription rate.
Please address all mail to:
Berkeley
Enterprises Inc., 815 Washington St.,
Newtonvi lie, Mass. 02160.
Postmaster: Please send all forms 3579
to Berkeley Enterprises Inc., 815 Washington St .. Newtonville, Mass. 02160.
© Copyright 1972, 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.
4
[NT A]
MENT POLICY
by Clay T. Whitehead, Director, Office of Telecommunications Policy, Executive Office of the President,
Washington, D.C.
A survey of regulation, competition, and other new
forms of economic interaction in the field of communications and computers.
[NT A]
21 COMPUTERIZING A MEMBERSHIP
ORGANIZATION
by William R. Pollert, National Association of Manufacturers, Washington, D.C.
How a large membership association planned and
implemented a computer facility using a timesharing system.
35 THE MEANING OF AN INTEGRATED DATA
[T A]
SYSTEM
by W. R. Larson, Western Electric, Corporate Education
Center, Princeton, N.J.
A proposed definition of an integrated data system
for an organization, with special attention given to
the areas of data storage and retrieval and improvement of system performance.
[NT A]
29 THE BAD IMAGE THAT COMPUTERS ARE
EARNING
by H. W. G. Gearing, Worcester, England - and others
Half a dozen examples, in Great Britain and the U.S., of
fouled-up computer output and exasperated customers.
24 THE INFORMATION INDUSTRY AND GOVERN-
COMPUTERS and AUTOMATION for April, 1972
The Computer Industry (continued)
7 On the Legal Side: THE OUTSIDE DIRECTOR
[NT F]
by Milton R. Wessel, Attorney, New York, N.Y.
7 "COMPUTERS ENTER THE BUSING CON[NT F)
TROVERSY" - ADDENDUM
by Robert" L. Glass, Kent, Wash.
[NT F]
34 INTERNAL REVENUE SERVICE: USE OF
COMPUTERS
by William H. Stewart, Jr., University of Alabama
31 IBM'S POWERFUL PARTNER: THE ACCOUNTING [NT F)
PRI NCI PLES BOARD
Samson Science Corp., New York, N.Y.
[NT F)
31 DO YOU WANT TO STOP CRIME?
by William P. Wood, III, Richmond, Va.
[NT F)
32 CDC VS IBM - CORRECTION
by F. O. Parlova, New York, and the Editor
33 Fall Joint Computer Conference: Topics
[NT F]
Edmund C. Berkeley, Editor, Computen, and Automation
20 "The 1972 Computer Directory and Buyers' Guide"
[T GI
The Selection of Personnel - for Computers and Other Purposes
26 PICTORIAL REASONING TESTS - PART 5
by Neil Macdonald, Assistant Editor
27 Pictorial Reasoning Test - C&A No.6
[NT F]
[NT FI
Common Sense, Wisdom, Science in General, and Computers
6 The Old Brain, The New Brain, The Giant Brain
and Common Sense
by Edmund C. Berkeley, Editor
2 The C&A Notebook on Common Sense
3 Questions and Answers About "The C&A Notebook"
[NT E]
[NT GI
[NT G]
Fron t Cover Picture
The front cover ~hows the Mauna.
Kea Observatory, highest in the world
at 13,796 feet above sea level, on the
Island of Hawaii on the top of an
The 88-inch teleinactive volcano.
scope and the dome aperture are
controlled by an IBM 1800 data acquisition and control system. The
system now responds automatically
to star motion and the air idistortion
along the line of sight, and will eventually respond automatically to temperature, humidity, and wind speed.
The Profession of Information Engineer and the Pursuit of Truth
37 DALLAS: WHO, HOW, WHY? - Part II
[NT A]
by Mikhail Sagatelyan, Moscow, USSR
A report published in Leningrad, USSR, by an ace
Soviet reporter about the circumstances of the assassination of President John F. Kennedy and their significance from a Soviet point of view: Part 2.
44 Political Lies: An Acceptable Level?
[NT Al
by Richard M. Nixon, J. W. Fulbright, and others
33 A Concerted Campaign To Deny the American People [NT G]
Essential Knowledge About the Operation
of Their Government
by Prof. Henry Steele Commager, Amherst, Mass.
32 Reducing and Dismantling Science and Research
[NT G)
Institutions, and Social Responsibility
by A. G. Michalitsanos, Cambridge, England
7 Dealing With Today's Problems
[NT F)
by John Skowronski, OPMA, Kansas City, Mo.
34 Some Responsibility for Our Chaotic Society
[NT G]
by S. R. Harrison, Trondheim, Norway
34 Unsettling, Disturbing, Critical ...
[NT FI
Statement of policy by "Computers and Automation"
Departments
49
20
Advertising Index
Calendar of Coming
Events
Correction
Monthly Computer
Census
New Contracts
New Installations
32
48
46
47
Key
[A]
[C]
[E]
[F]
[G]
[NTI
[T]
Article
Monthly Column
Editorial
Forum
The Golden Trumpet
Not Technical
Technical
Computers, Games, and Puzzles
36, 43 Numbles, by Neil Macdonald
49 Problem Corner, by Walter Penney, COP
COMPUTERS and AUTOMATION for April, 1972
[T C)
[T C]
5
C- a
EDITORIAL
The Old Brain, The New Brain,
The Giant Brain, and Common Sense
All computer people from time to time think about
brains - those interesting and complex devices made out
of common chemicals that we human beings and other
animals carry around inside our heads and by means of
which we think.
The prototype brain began in the course of evolution
more than 400 million years ago. As mammals and birds
developed, their brains became very efficient instruments.
These brains accept thousands of simultaneous impressions from one or more senses, and make use of a builtin yet trainable evaluating function. The evaluating function uses observations, context, surroundings, instincts,
learned experiences, and other information in order to
make decisions. The decisions are the ordinary everyday
decisions that are useful and relevant to solving the ageold problems of surviving:
• How to recognize and avoid danger;
• How to find and choose food and shelter;
• How to satisfy other wants, how to mate, etc.
This b rain is what I like to call "the old brain". It is
present, of course, in all human beings. Two of its most
remarkable parts are: the way it deals with streams of
multiple input sensations; and the evaluating function.
Together these parts enable the animal to decide swiftly
and almost automatically what to do next and how fast
he or she should do it. The output of the old brain is
a great quantity of "common sense" behavior - consisting mainly of directions to the body for survival purposes.
At some point more than one million years ago (and
probably not more than 10 million years ago) a branch of
the primates began to develop a new brain. The new
brain has produced language, words to refer to ideas, tool
making, culture, the invention of symbols and writing, the
solving of problems by using symbols, and much more.
The new brain is able to analyze problems into many logical steps, to talk about these steps, to imagine these steps
in different arrangements. Most extraordinary of all, the
new brain possesses the equipment (unlike any other animal except perhaps the dolphin) to deal with the meanings of tens of thousands of words. The output of the
new brain, when brought up in an environment of education and civilization, is intellectual behavior that no other
species of animal can come close to producing.
The origin of the new brain is illuminated in the following quotation from "The Naked Ape" by Desmond
Morris (a fascinating book):
Of all the non-specialists, the monkeys and the
apes are perhaps the most opportunist. As a
group, they have specialized in non-specialization. And among the monkeys and apes, the
naked ape is the most supreme opportunist of
them all. ... All young monkeys are inquisitive
but the intensity of their curiosity tends to fade
6
as they become adult. With us the infantile inquisitiveness is strengthened and stretched into
our mature years. We never stop investigating.
We are never satisfied that we know enough to
get by. Every question we answer leads on to
another question. This has become the greatest
survival trick of our species. For the opportunist
the going may always be rough, but the creature
will be able to adapt quickly to any quick-change
act that the environment decides to put Ion.
The new brain, in spite of its power, has some severe
limitations. It seems to be able to pay attention to only
one idea at a time, while the old brain seems to be able
to pay attention to many aspects of the environment simultaneously and tJ react almost immediately to any suddenly disturbing event, like something noticed out of the
comer of one's eye. The new brain makes far more mistakes than the old brain. The new brain seems to have
rather a small memory - at least, only when human beings make use of external recording can they remember'
more than a small part of what they consciously want to
remember. The old brain in contrast seems to have a
very large amount of information ready for use, especially in emergencies. And the new brain seems not to
have access to all the information stored in the old brain.
The giant brain of course is the computer. The phrase
has been applied to computers since 1949. The giant
brain is strictly an imitation of the new brain. It is not
at all an imitation of the old brain - because it does not
automatically receive thousands of sense impressions per
second; it does not automatically integrate these impressions into representations of the environment; it does not
automatically operate an evaluating function which solves
the problem of living and surviving from moment to moment:
Like the new brain, the giant brain deals with symbols;
it performs operations of logic and arithmetic in enormous
quantities, and at a speed that is on the order of a million
times the speed of human beings.
"Common sense" behavior is what an animal displays
when he is using his old brain: the sense to avoid danger,
to find and choose food, to survive according to the bag
of survival tricks developed by his species.
Computers are enormously busy in these days performing the functions of the new brain. We are just beginning
to apply computers to performing the functions of the old
brain - as in optical character recognition. It is time to
pay more attention to the construction of computers that
will perform adequately the functions of the old brain,
and produce "common sense" behavior.
Editor
COMPUTERS and AUTOMATION for April, 1972
On The Lega I Side:
THE OUTSIDE DIRECTOR
Milton R. Wessel, Attorney
New York, N. Y.
A person who accepts election as an "outside director" of a public company merely for honor or prestige, is making a serious mistake.
During the last decade a host of litigations have
made very clear that a director -- outside or inside
-- has affirmative personal responsibility to a large
number of persons, including stockholders, employees,.
suppliers, customers, and the public generally. Often the standard corporate indemnification arrangement will not help him where he fails to perform that
responsibility, and he must pay money damages and in
extreme cases even suffer criminal conviction. The
fact that the organization involved is a church, hospital, college or other non-profit organization does
not relieve him of even a shred of his responsibility.
What is the director's responsibility? A complete
answer requires a book, and much has been written.
The short, practical answer here is that a director
must act reasonably under the circumstances; he must
act as one would expect a director to act. Section
717 of the New York Business Corporation Law puts it
thus:
"Directors and officers shall discharge the
duties of their respective positions in good faith
and with that degree of diligence, care, and skill
which ordinarily prudent men would exercise under
similar circumstances in like posi tions."
All of this means that the outside director must
pay sufficient attention to the business to satisfy
his responsibility. He cannot successfully defend
against liability by saying, "I didn't know" or "I
didn't attend meetings" or "They wouldn't give me the
financial data I asked for". For example, a director
of a computer company, whose programs and banks of
data are its main asset, should satisfy himself that
they are reasonably protected against fire and fraud.
If he does not and the company is wiped out, he may
be liable to shareholders for negligence. He can't
just sit and wait for management to ask for his approval of a security program.
Nor can a director avoid liability by closing his
eyes to facts about which he should know. If he hears
something at a board meeting which he doesn't like,
he can't say, "Don't discuss that here" or "Don't
tell me about it." He must be able honestly to say
that in his judgment, the company is doing the best
it can under all circumstances. If he cannot, he
must let the other directors know what is wrong in
his opinion. If they do not agree and cannot persuade him to the contrary, he must withdraw promptly;
he cannot sit and wait until he spots trouble just
ahead.
This column deals only with the affirmative duties
of a public director to be a reasonably effective one.
In addition, there are a host of other somewhat more
specific prohibitions which are of an essentially
negative character and which are far more precisely
defined (not to disclose confidences; not to profit
by corporate opportunities; not to engage in short
swing or other improper insider trading, -- in short,
not to deal unfairly with the company), which need
discussion.
The result of these two sets of increased obligations is that informed businessmen and professionals
COMPUTERS and AUTOMATION for April, 1972
are more and more reluctant to accept appointment to
public boards without compensation commensurate to
the degree of responsiblity assumed, or a major stake
in the enterprise, or (more usually) both.
I predict this trend will continue until all directors finally recognize their responsibilities and
boards become true and effective management units,
rather than the dummy operations which characterize
so many newly public companies.
"COMPUTERS ENTER THE BUSING
CONTROVERSY" -
ADDENDUM
Robert L. Glass
26414124th Ave., SE
Kent, Wash. 98031
For the sake of journalistic completeness, I am
sending this note on Seattle's program of busing for
integration as discussed in my article "Computers
Enter the Busing Controversy" which was published
in your July issue.
My statement "Poli tics and public outcry may yet
negate the integration plan which technology has been
able to produce" was, unfortunately, prophetic.
A last-minute local court decision blocked busing
and nearly scuttled Seattle's entire middle school
concept. Schools opened, but busing itself was postponed (nominally) a year.
Citizens Against Mand~tory Busing (CAMS), instigator of the suit which led to the decision, was
jubilant. The School Board, with too little time
left for an effective countermove, could only acknowledge temporary defeat.
Busing advocates in Seattle, and the computing
technologists who constructed the busing implementation plan, can only fall back on the traditional
loser's cry, "Wait 'til next year."
DEALING WITH TODA Y'S PROBLEMS
John Skowronski
Director of Publications
Kansas City Chapter, DPMA
P. O. Box 2425
Kansas City, Mo. 64142
I would like your permission to reprint the article
"The Handwriting on the Wall" (June, 1971, p. 6) which
editorializes a piece entitled "Computerized" by
"Margo". I am also writing to the Boston Globe for
their permission to reprint the "Margo" article.
Data Processing Management Association (D.P.M.A.)
as you probably know is a non-profit organization of
data processing professionals with approximately 260
members in the Kansas City Chapter. Our local newsletter has a monthly distribution of approximately
300.
I would also like to congratulate you for a very
fine magazine that deals with today's problems. Your
articles "The Case of Secret Service Agent Abraham W.
Bolden" (June, 1971, p. 41) and "The Issue is Hypocrisy" (June~ 1971, p. 45) were quite revealing. My
wife was so impressed that she is going to subscribe
personally for your magazine.
7
1100 WHAT I MEAN
II
:
Th e p
tA
rogrammer
5
ssistant
Warren Teitelman
Bolt Beranek and Newman
50 Moulton St.
Cambridge, Mass. 02140
"The programmer's environment influences, and to a large extent determines, what
sort of problem he can tackle, and how far he can go in a given time."
This article deals with the design and actual
implementation in a computer programming system of
"a programmer's assistant". The general function
of the "programmer's assistant" is to make it possible for the human programmer to say to the computer "do what I mean" instead of "do what I say",
and "undo what I just tried - it did not work",
instead of leaving the programmer with the sad consequences of his actual instructions.
In other words, the programmer's assistant deals
with such factors as:
• ease of interaction;
• level of interaction;
• forgiveness for errors (both spelling errors
and errors of thought);
• going back and taking a different path;
• changing one's mind; etc.
and in general, the programmer's environment.
This area of improvement in interactive programming is important. For many applications, the programmer's environment influences, and to a large
extent determines, what sort of problem he can
tackle, and how far he can go in a given time. If
the "environment" is "cooperative" and "helpful",
then the programmer can be more ambitious and productive. If not, he may spend much of his time and
energy performing routine clerical tasks and "fighting the system".
An Analogy
The conceptual role of the programmer's assistant can perhaps best be explained by using an
analogy with a well-equipped chemistry laboratory.
We are attempting to recreate for the programmer the
type of laboratory assistant who works closely with
the chemical scientists. As an example of his
duties, the assistant can watch the scientist perform a sequence of operations, and then be instructed to repeat the operation, perhaps with some rearrangements, omissions, or other modifications, or
to clean up the mess and make things ready for
another run.
This kind of general purpose assistance complements the facilities in the laboratory, but does
not of course substitute for them. Its value is
greatest in those situations which are (1) not
standardized, so that special equipment has not already been built for that purpose, or (2) not sufficiently repetitive, so that new equipment cannot
be designed and built cost-effectively. These
situations usually require the scientist to achieve
8
his goal by executing many small steps,
some similarities or repetitions, while
each point what to do next on the basis
vious results, with occasional, or even
"backing and filling."
with perhaps
deciding at
of the prefrequent,
In the realm of programming, it is easier to
build and discard tools for just a few applications
than in the physical sciences. However, all too
frequently the programmer may not realize he is
going to need an operation repeated, or else the
·operation is not exactly the same in each case, or
else the effort of building the tool (and debugging
it) may not be worth interrupting his train of
thought and action. If it were convenient for him
to specify repeating a previous operation, including possible modifications, without his having
previously prepared for it, he would certainly do
so. Then like our hypothetical scientist, he would
be able to turn his attention to evaluating the resuI ts of his "experiment" so fa r, and deciding what
to do next, while his assistant performs the dog
work.
Undo
One of the important functions of the laboratory
assistant is cleaning up an abortive experiment and
getting things ready for the next run. In the realm
of programming, users prepare for possible disasters
such as a program running wild and chewing up a data
structure, by saving the state of part or all of
their environment before attempting some not yet reliable process. Undoing then consists of backing
up to some previous state and starting over from
there. However, this saving and dumping operation
is usually expensive and time-consuming.
Of course disaster may also strike as a result
of
supposedly innocuous operation. As a result,
the user all too often finds himself in a situation
where he wishes, either idly or desperately, that
he could reverse the effect of a previous operation
or operations. Here we have an advantage over the
physical sciences: a laboratory assistant may not
be able to reverse the effect of mixing two chemicals together, no matter how hard or skillfully he
tries. However, in the programmer's assistant, we
can provide such a capability.
BBN-LiSP
At Bolt, Beranek, and Newman, Cambridge, Mass.,
and at over a dozen other installations around the
United States, the language called BBN-LISP is in
COMPUTERS and AUTOMATION for April, 1972
common use. BBN-LISP as a programming language, is
an implementation of LISP, a language designed for
list processing and symbolic manipulation. BBN-LISP
as a programming system, is the product of, and
vehicle for, a research effort supported by the Advanced Research Projects Agency for improving the
programmer's environment.
Design Philosophy
BBN-LISP contains many user-support facilities.
These include:
• a sophisticated structure editor which can
either be used interactively or as a subroutine;
• a debugging package for inserting conditional programmed interrupts around or
inside of specified procedures;
• a "prettyprint" facility for producing
structured symbolic output;
• a program analysis package which produces a
tree-structured representation of the
flow of control between procedures;
• a concordance listing indicating for each
procedure the procedures that call it,
the procedures that it calls, and the
variables it references, sets, and binds;
etc.
Most on-line programming systems contain similar
features. But the essential difference between
BBN-LISP and other systems is embodied in the philosophy that the user addresses the system through
an (active) intermediary agent, whose task it is
to collect and save information about what the user
and his programs are doing, and to utilize this information to asist the user and his programs. This
intermediary has been named the programmer's assistant (or p.a.).
The Programmer's Assistant
For most interactions with the BBN LISP system,
the programmer's assistant is an invisible interface between the user and LISP: the user types a
request, for example, specifying a function to be
applied to a set of arguments; the indicated operation is then performed, and a resulting value is
printed. The system is then ready for the next request. However, in addition, in BBN-LISP, each input typed by the user, and the value of the corresponding operation, are automatically stored by the
p.a. on a global data structure called the history
list.
The history list contains information associated
with each of the individual "events" that have
occurred in the system, where an event corresponds
to an individual type-in operation. Associated
with each event is the input that initiated it, the
value it yielded, plus other optional information
such as side effects, messages printed by the system
or user programs, information about any errors that
may have occurred during the execution of the event,
etc. As new events occur, existing events are aged,
and the oldest event or events are "forgotten." The length of the history list is set by the user,
and is typically somewhere between 30 and 100
events.
The user can refer to an event on the history
list by its relative event number, for example, -1
refers to ~he most rece~t event, -2 the event before
that, etc. He may refer by an absolute event number, or by a pattern which is then used for searching the history list. For example, the user can
retrieve an event in order to REDO a test case
after making some program changes. Or, having
COMPUTERS and AUTOMATION for April, 1972
typed a request that contains a slight error, the
user may elect to FIX it, rather than retyping the
request in its entirety. The p.a. recognizes such
requests as REDO and FIX as being directed to it,
not the LISP interpreter, and executes them directly. For example,when given a REDO command, the
p.a. retrieves the indicated event, obtains the input from that event, and treats it exactly as though
the user had typed it in directly. Similarly, the
FIX command directs the p.a. to allow the user to
edit the indicated input. When the user has fixed
up this input, it is again processed by the p.a.
exactly as though it had been typed in.
Fifteen Commands
The p.a. currently recognizes about 15 different
commands (and includes a facility for the user to
define additional ones). For example, the USE
command provides a convenient way of specifying
simultaneous substitutions for lexical units and/or
character strings, e.g. USE X FOR Y AND + FOR *.
This permits after-the-fact parameterization of
previous events. The p.a. also enables the user to
treat several events as a single unit, (e.g. REDO
47 THRU 51), and to name an event or group of
events.
All of these capabilities allow, and in fact encourage, the user to construct complex console operations out of simpler ones in much the same fashion as programs are constructed, i.e. simpler operations are checked out first, and then combined
and rearranged into large ones. The important point
to note is that the user does not have to prepare
in advance for possible future use or reuse of an
event. He can operate straightforwardly as in other
systems; yet the information saved by the p.a. enables him to implement his "after-thoughts."
Undoing
Perhaps the most important after-thought operation
made possible by the p.a. is that of undoing the
side-effects of a particular event or events. In
most systems, if the user suspects that a disaster
might result from a particular operation, e.g. an
untested program running wild and chewing up a complex data structure, he would prepare for this contingency by saving the state of part or all of his
environment before attempting the operation. If
anything went wrong, he would then back up and start
over.
Such "accidents" happen all too often in typical
console sessions, and result in the user's either
having to spend what may involve a great deal of
effort in rectinstructing the inadvertently destroyed
information, or alternatively in returning to his
last back-up, in which case the user would have to
redo any useful work performed since that backup.
Instead with the p.a., the user can recover by
simply typing UNDO, and then perform the originally
intended operation.
The existence of UNDO frees the user frqm worrying about such oversights. He can be relaxed and
confident in his console operations, yet still work
rapidly. He can even experiment with various program and data configurations, without necessarily
thinking through all the implications in advance.
One might argue that this would promote sloppy
working habits. However, the same argument can be
and has been leveled against interactive systems
in general. In fact, fteeing the user from having
to anticipate all of th~ consequences of an (experimental) change usually results in his being abl~ to
9
pay more attention to the conceptual difficulties
of the problem he is trying to solve.
Another advantage of undoing as it is implemented
in the programmer's assistant is that it enables
events to be undone selectively.
Finally, since the operation of undoing an event
itself produces side effects, it too is undoable.
The user can often take advantage of this fact, and
employ strategies that use UNDO for desired operation reversals, not simply as a means of recovery
in case of trouble. For example, suppose the user
wishes to interrogate a complex data structure in
each of two states while successively modifying his
programs. He can interrogate the data structure,
change it, interrogate it again, undo the changes,
modify his programs, and then repeat the process
using successive UNDOs to flip back and forth between the two states of the data structure.
1m plementati on
The UNDO capability of the programmer's assistant is implemented by making each function that
is to be undoable save on the history list enough
information to enable reversal of its side effects,
For example, when a list node is about to be
changed, it and its original contents are saved.
For each primitive operation that involves side effects, there are two separate functions, one which
always saves this information, i.e., is always undoable, and one which does not save it.
errors, especially those occurring in type-in, are
of the type that can be corrected without any
knowledge about the purpose of the program or operation in question, e.g. misspellings, certain kinds
of syntax errors, etc.
The p.a. attempts to correct these errors, using
as a guide both the program's environment at the
time of the error, and information gathered by monitoring the user's requests. This form of implicit
assistance provided by the programmer's assistant
is named the DWIM (Qo-~hat-l-Mean) capability.
DWIM is also used to correct other types of conditions not considered errors, but nevertheless
obviously not what the user meant. For example, if
the editor is called on a function that is not defined, rather than typing NOT EDITABLE, the editor
invokes the spelling corrector to try to find what
function the user meant to edit, giving DWIM as
possible candidates a list of user-defined functions. Similarly, the spelling corrector is called
to correct misspelled edit commands, p.a. commands,
names of files, etc. The spelling corrector can
also be called by user programs.
As mentioned above, DWIM also uses information
gathered by monitoring user requests. This is accomplished by having the p.a., for each user request, "notice" the functions and variables being
used, and add them to appropriate spelling lists,
which are then used for comparison with (possibly)
misspelled units. Thus, DWIM "may know" that FACT 'was
the name of a function, and is therefore able to
correct FATC to FACT.
Although the overhead for saving undo information is small, the user may elect to make a particular operation not be undoable if the cumulative
effect of saving the undo information seriously degrades the overall performance of a program, because the operation in question is repeated so
often. The user, by his choice of function specifies which operations are undoable. In a sense,
the user's choice of function acts as a declaration
about frequency of use versus need for undoing.
For those cases where the user does not want certain functions undoable once his program becomes
operational, but does wish to protect himself
against malfunctioning while debugging, the p.a.
provides a facility called TESTMODE. When in
TESTMODE, the undoable version of each function is
executed, regardless of whether the user's program
specifically called that version or not.
As a result of knowing the names of user functions and variables (as well as the names of the
most frequently used system functions and variables),
DWIM seldom fails to correct an error the user feels
it should have. And since the spelling corrector
knows about common typing errors, e.g. transpositions, doubled characters, shift and case mistakes,
etc., DWIM almost never mistaken!y corrects an
error. However, if DWIM did make a mistake, the
user could simply interrupt or abort the computation, UNDO the correction (all DWIM corrections are
undoable), and repair the problem himself. Since
an error had occurred, the user would have had to
intervene anyway, so that DWIM's mistaken correction did not result in extra work for him. It is
this benign quality that makes DWIM so appreciated
by users.
Side Effects
Example
Finally, all operations involving side effects
that are typed-in by the user are automatically
made undoable by the p.a. by substituting the corresponding undoable function name in the expression
before evaluation, as in the earlier example with
the REMOVE PROPERTY operation. This procedure is
feasible because operations that are typed-in rarely
involve iterations or lengthly computations directly
nor is efficiency usually important. However, as
a precaution, if an event occurs during which more
than a user-specified number of pieces of undo information are saved, the p.a. interrupts the operation to ask the user if he wants to continue having
undo information saved.
Automatic Error Correction - The DWIM Facility
The previous discussion has described ways in
which the programmer's assistant is explicitly invoked by the user. The programmer's assistant is
also called when certain error conditions are encountered. A surprisingly large percentage of these
10
Suppose t::e user defines a function FACT of one
argument, N. The value of FACT[N] is to be N factorial.
(LAMBDA (N) (COND
((ZEROP N9 1) ((T (ITIMS N (FACCT 8SUBl N]
(FACT)
~DEFINEQ((FACT
Note that the definition of FACT contains several
mistakes: ITIMES and FACT have been misspelled.
The 9 in N9 was intended to be a right parenthesis, but the teletype shift key was not depressed;
similarly, the 8 in 8SUBl was intended to be a left
parenthesis; and finally, there is an extra left
parenthesis in front of the first T, according to
LISP, on the second line.
~PRETTYPRNT((FACCT]
[1J
=PRETTYPRINT
=FACT
[3J
[2J
COMPUTERS and AUTOMATION for April, 1972
(FACT
[LAMBDA (N)
(COND
« ZEROP N9 1)
«T (ITIMS N (FACCT 8SUBI NJ)
NIL
~FACT(3)
Now, after defining FACT, the user wished to look
at its definition using PRETTYPRINT, which he unfortunately misspells, [IJ. Since there is no function PRETTYPRNT in the system, an "undefined function" error occurs, and DWIM is called. DWIM invokes its spelling corrector, which searches a list
of functions frequently used (by this user) for the
best possible match. Finding one that is extremely
close, DWIM proceeds on the assumption that
PRETTYPRNT meant PRETTYPRINT, notifies the user of
this, [2J and calls PRETTYPRINT.
At this point, PRETTYPRINT would normally print
(FACCT NOT PRINTABLE) and exit, since FACCT has no
definition. Note that this is not a LISP error condition, so that DWIM would not ~called as
described above. However, it is obviously not what
the user meant.
This sort of mistake is corrected by having
PRETTYPRINT itself explicitly invoke the spelling
corrector portion of DWIM whenever given a function
with no satisfactory definition. Thus with the aid
of DWIM, PRETTYPRINT is able to determine that the
user wants to see the definition of the function
FACT, [3J and proceeds accordingly.
~FACT(3)
[ 4J
N9(IN FACT) »--> N)
(IN FACT) (COND -- «T --))) »--> (COND -- (T --))
ITIMS(IN FACT)->ITIMES
[5J
FACCT(IN FACT)->FACT
8SUBl(IN FACT) »--> (SUBI
U.B.A. N9(IN FACT)
FIX?
YES
N9(IN FACT) »--> N)
U.D.F. TON FACT)
FIX?
YES
(COND -- «T --))) »--> (COND
(T --))
ITIMSON FACT) ->ITIMES? ... YES
FACCT(IN FACT) -:>FACT? ... YES
U.B.A. 8SUB1(IN FACT)
FIX?
NO
U.B.A.
(8SUBI BROKEN)
[lJ
[2J
[3]
[4J
[5J
A great deal of effort has been put into making
DWIM 'smart'. Experience with perhaps a dozen different users indicates we have been very successful;
DWIM seldom fails to ~orrect an error the user feels
it should have, and almost never mistakenly corrects
an error. However, it is important to note that even
when DWIM is wrong, almost always no harm is done;
since an error had occurred, the user would have had
to intervene anyway if DWIM took no action. Thus, if
DWIM mistakenly corrects an error, the user simply
interrupts or aborts the computation, UNDOes the DWIM
change using UNDO, and makes the correction he would
have had to make without DWIM. It is this benign
quality of DWIM that makes it a valuable part of the
programmer's assistant.
Conclusion
When a user types a request which contains a misspelling, having to retype it is a minor annoyance
(depending, of course, on amount of typing required
and the ~ser's typing skill). However, if the user
has mentally already performed that task, and is
thinking ahead several steps to what he wants to do
next, then having to go back and retype the operation
represents a disruption of his thought processes, in
addition to being a clerical annoyance. The disruption is even more severe when the user must also repair the damage caused by a faulty operation, instead
of simply UNDOing.
6
~PP
FACT
[ 6J
(FACT
[LAMBDA (N)
(COND
«ZEROP N)
1)
(T (ITIMES N (FACT (SUBI NJ)
NIL
The user now calls his function FACT, [4J. During
its execution, five errors occur, and DWIM is called
fi ve times, [5J. At each poi nt, the error is corrected, a message printed describing the action taken,
and the computation allowed to continue as if no
error had occurred. Following the last correction,
the value of FACT (3) is printed. Finally, the user
prettyprints the new, now correct, definition of
FACT, [6J.
In this particular example, the user was shown
operating in TRUSTING mode, which gives DWIM carte
blanche for all corrections. The user can also operate in CAUTIOUS mode, in which case DWIM will inform him of intended corrections before they are
made, and allow the user to approve or disapprove of
them. For most corrections, if the user does not
respond in a specified interval of time, DWIM automatically proceeds with the correction, so that the
user need intervene only when he does not approve.
Sample output is given below. Note that the user
responded to the first, second, and fifth questions;
DWIM responded for him on the third and fourth.
COMPUTERS and AUTOMATION for April, 1972
We have found that in addition to making life a
lot more pleasant for users, the p.a. has a synergistic effect on user productivity. This effect seems
to be related to the overhead that is involved when
people have to switch tasks or levels:
The DWIM facility greatly facilitates construction of complex programs because it allows
the user to remain thinking about his program
operation at a relatively high level without
having to descend into manipulation of details.
Even more important, however, is that DWIM and thep.a. act to minimize user distractions and diversions
at whatever level the user is operating or chooses to
operate.
We believe that p.a. facilities should be built
into low-level debugging packages, the executive
language of time sharing systems, etc., as well as
other 'high-level' programming languages, for these
facilities provide the user with a significant 'mental mechanical advantage' in attacking problems.
0
References
1. Teitelman, W., Bobrow, D.G., Hartley, A.K., and
Murphy, D.L., "BBN-LISP TENEX Reference Manual", BBN Report, July 1971.
2. Tei telman, W., "Toward a Programming Laboratory",
IJCAI, January 1969.
3. Bobrow, D.G., "The Future of List Processing
Languages", to be published in a forthcoming
issue of the "Communications of the ACM".
11
Cryptology, The Computer, and Data Privacy
M. B. Girsdansky
Scientific Information Dept.
I BM Research Division
P.O. Box 218
Yorktown Heights, N. Y. 10598
"In the modern digital computer, cryptology has gained a valuable ally,
but in ever-growing measure, the computer may find itself in need
of cryptology. "
English writer and wit G. K. Chesterton once observed that a
nighttime view of neon-lit Broadway would be singularly
beautiful-to someone unable to read.
Implicit in the quip was 'realization that any alphabetic (or
syllabic) system of writing in. fact constitutes a cipher to one
ignorant of the system. * An unfamiliar script long concealed the
language and content of the ancient writings known as Minoan
Linear B. It was only in the early 1950s that cipher-breaking
techniques identified the material as Greek of the Late Bron,ze
Age (c. 1300-1000 B. C.), radically revising our notions of early
Mediterranean history.
Such "hidden writing" (the etymological meaning of
"cryptography") is the result of historical accident, but
there are examples of intentional cryptography which
are even older. Deliberately enciphered hieroglyphics
were inscribed more than half a millennium before
Linear B was set down. Closer to our own day-in the
India of the Fourth Century B. C.-ambassadors to
foreign courts were explicitly advised to practice "the
decipherment of secret writings." Over the centuries
since, cryptology-the discipline concerned with the
making and breaking of codes and ciphers-has led a
sometimes raffish, but usually important and frequently
fascinating existence.
Like much else, cryptology has been profoundly
affected by technology. Telegraph, telephone, radio, and
a variety of subsequently developed means of processing
information electronically have presented problems as
well as opportunities. Thus, encoded or enciphered
messages had to be made far more resistant to garbling
* A cipher transforms low·order constituents of a message-such as
individual characters or, more rarely, small groups of characters-according
to some definite strategy. In contrast, a code consists of an agreed-upon
but essentially arhitrary listing of natural-language segments vs. code equivalents, allowing a small group of symbols to represent an entire word,
sentence, or even group of sentences. A volume containing such entries as
DA B L I = "Consign ment two weeks overdue/Customer threatens
cancellation/Expedite or give particulars" is a true code book. The socalled Morse Code is a misnomer; actually tht:- system is a form of simple
substitution cipher.
(Based on a report published in 'IBM Research Reports',
Vol. 7, No.4, 1971, and reprinted with permission)
12
over noisy lines or during poor atmospheric conditions.
New techniques for encrypting data (e. g., such novelities as high-speed transmission and signal-scrambling)
were devised and in turn had to be countered. Increasingly, cryptoJr ' drew upon the insights and skills of
mathematicians, electrical engineers, and others from the
physical sciences.
Of perhaps greatest ultimate significance at the
cryptology/technology interface was development of the
modem digital computer. In it, cryptology has gained
a valuable ally, but in ever-growing measure, the computer may well find itself in need of cryptology.
As the computer is used to store and process data on
individuals, concern has been voiced lest much that is
personal and legitimately private lie open to those with
access to "data banks." In response, cryptologic techniques have been suggested as a logical means of shielding a computer's contents, which might otherwise be
readily called out from the machine. Depending on the
purposes to which the data bank is to be put, some or all
of its contents should be inaccessible and unreadable to
anyone but a given authorized user. Granted, just who is
to be "an authorized user" is largely a question of public
policy. Exploration of technical possibilities is, however,
the province of the scientist and the engineer. In this
area, workers at the IBM Research Division have been
among those seeking effective privacy systems, as well as
probing the weaknesses of systems already proposed.
If, as seems likely, data banks are to be embodied as
centralized computers reached through a network of remote terminals, it would be desirable to have many
features of the system-not only a cipher itself-work
together to provide privacy and security from tampering.
Such an approach has been evolved by IBM's Horst
Feistel, and implemented in an experimental system
developed by William A. Notz and J. Lynn Smith which
features critical hardwar~ designed by Smith.
Perhaps logically prior to system details, however, is the
matter of what type of cipher to use. It has long been
realized that certain classes of ciphers are, by their very
nature, weaker than others. At IBM, Bryant Tuckerman
has gone somewhat further, showing that one popular
COMPUTERS and AUTOMATION for April, 1972
Plaintext
COMEHERE
Key
Ciphertext
EXITEXIT
GLUXLBZX
tttttttt
... ...
...
...
... ...
ABC D E F CHI J K L M N 0 P
A
B
C
D
-.E
F
G
H
-.J
J
K
L
M
N
o
p
Q
R
S
-+T
U
V
II
-+X
y
a I 2 3 4 5 6 7 8 9 10" 12 13 14 15 16 17 18 19 3J 21 22 23 24 25
ABC D E F CHI J K L M N 0 P Q R STU V II X Y Z
... S
Q R
T U V II X Y Z
ABC D E F CHI J K L M N 0 P Q R STU V II X Y Z
BCD E F CHI J K L M N 0 P Q R STU V II X Y Z A
C D E F CHI J K L M N 0 P Q R STU V II X Y Z A B
D E F CHI J K L M N 0 P Q R STU V II X Y Z ABC
E F CHI J K L M N 0 P Q R STU V II X Y Z ABC D
F CHI J K L M N 0 P Q R STU V W x Y Z ABC D E
CHI J K L M N 0 P Q R STU V W x Y Z ABC D E F
H I J K L M N 0 P Q R STU V W x Y Z ABC D E F G
J J K L M N 0 P Q R STU V W x Y Z ABC D E F G H
J K L M N 0 P Q R STU V W x Y Z ABC D E F C H J
K L M N 0 P Q R STU V W x Y Z ABC D E F CHI J
L M N 0 P Q R STU V W x Y Z ABC D E F CHI J K
M N 0 P Q R STU V W x Y Z ABC D E F C H J J K L
N 0 P Q R STU V W x Y Z ABC D E F CHI J K L M
o P Q R STU V W x Y Z ABC D E F G H I J K L M N
P Q.R STU V W x Y Z ABC D E F G H I J K L M N 0
Q R STU V W x Y Z ABC D E F G H I J K L M N 0 P
R STU V W x Y Z ABC D E F G H I J K·L M N 0 P Q
STU V W x Y Z ABC D E F CHI J K L M N 0 P Q R
T U V II X Y Z ABC D E F CHI J K L M N 0 P Q R S
U V II X Y Z ABC D E F G H J J K L M N 0 P Q R S T
V II X Y Z ABC D E F CHI J K L M N 0 P Q R STU
II X Y Z ABC D E F G H I J K L M N 0 P Q R STU V
X Y Z ABC D E F G H I J K L M N 0 P Q R STU V W
Y Z ABC D E F CHI J K L M N 0 P Q R STU V W X
Z ABC D E F G H I J K L M N 0 P Q R STU V W X Y
oA
I B
2 C
3 D
4 E
5 F
6 G
-+7 H
8 J
ABCDEFGHJJKLMNOPQRSTUVWXYZ
BCD E F G H I J K L M N 0 P Q R STU V II X Y Z A
C D E F G H I J K L M N 0 P Q R STU V W X Y Z A B
D E F G H I J K L M N 0 P Q R STU V W X Y Z ABC
E F G H I J K L M N 0 P Q R STU V II X Y Z ABC D
F CHI J K L M N 0 P Q R STU V W X Y Z ABC D E
G H I J K L M N 0 P Q R STU V II X Y Z ABC D E F
H I J K L M N 0 P Q R STU V W X Y Z ABC D E F C
I J X L M N 0 P Q R STU V II X Y Z ABC D E F C H
9 J
J K L M N 0 P Q R STU V W X Y Z ABC D E F CHI
10 K
X L M N 0 P Q R STU V W X Y Z ABC D E F G H I J
"L
L M N 0 P Q R STU V II X Y Z ABC D E F G H I J K
~ M
M N 0 P Q R STU V W X Y Z ABC D E F G H I J K L
g N
N 0 P Q R STU V W X Y Z ABC D E F CHI J K L M
N 0
0 P Q R STU V II X Y Z ABC D E F G H I J K L M N
~ P
P Q R STU V W X Y Z ABC D E F G H I J K L M N 0
~ Q
Q R STU V W X Y Z ABC D E F G H I J K L M N 0 P
"R
R STU V II X Y Z ABC D E F CHI J X L M N 0 P Q
ffi S S T U V W X Y Z ABC D E F G H I J K L M N 0 P Q R
ffi T
T U V II X Y Z ABC D E F CHI J X L M N 0 P Q R S
m U
U V W X Y Z ABC D E F C H ~ J K L M N 0 P Q R S T
21 V
V II X Y Z ABC D E F G H I J K L M N 0 P Q R STU
22 W
II X Y Z ABC D E F G H I J K L M N 0 P Q R STU V
23 X
X Y Z ABC D E F G H I J K L M N 0 P Q R STU V W
~ Y
Y Z ABC D E F G H I J K L M N 0 P Q R STU V II X
3
Z
Z ABC D E F G H I J K L M N 0 P Q R STU V W X Y
Figure 1. "Classic Vigenere" enciphered by establishing a oneto-one correspondence between plaintext and key characters.
The key is repeated in whole or in part when necessary.
Ciphertext characters are then those which stand at the intersections of key rows and plaintext columns.
Figure 2. Vigenere encipherment considered as addition modulo
26. If the letters of the alphabet are given the values A = 0, ... Z
= 25, and combination of a row and a column is defined as
"addition," then the character at the intersection of row and
column is the sum, modulo 26. Here, 21 + 7 = 2, corresponding
to V + H = C.
class of ciphers is, indeed, far weaker than even a skilled
amateur at cryptology might think.
of the plaintext and of the key. For mnemonic ease, it
was formerly customary for the key to consist of an
intelligible word or meaningful phrase, although this
feature constitutes a weakness; incoherent keys, which
reduce the number of clues, are certainly preferable.
VIGENERE-VERNAM CIPHERS
The cipher class studied by Tuckerman occupies a classic
position in traditional cryptology. Although some members of the class are centuries old, one variety was
devised a matter of decades ago to provide on-line
encipherment of teletypewriter text. The class has been
proposed by some as a means of furnishing privacy in
modern data storage and transmission systems. Both
theoretically and by practical demonstration,
Tuckerman has shown that an unauthorized user
("opponent") having only limited material and information with which to work can readily extract the original text of messages enciphered by members of this class
by making use of the speed, capacity, and computational
abilities of the computer.
Representatives of this cipher-class have historically been
kept distinct under the names, "Vigenere cipher" and
"Vernam cipher," but in spite of superficial differences,
their underlying mathematical structures are fundamentally alike. Tuckerman therefore employs the phrase
"V-V systems" to encompass them both.
In its simplest form, the Vigenere cipher is familiar to
even the amateur cryptologist as the "tableau" of alphabets shown in figure 1 and 2. The horizontal and vertical
alphabets set slightly apart from the main block are for
the convenience of the user in specifying the characters
Encipherment of "classic Vigenere" begins by establishing a one-to-one correspondence between the characters
of the plaintext and the characters of the key (see figure
1), the key being partially or completely repeated if
shorter than the plaintext. The cipher character may
then be defined as the character standing at the intersection of a plaintext column and the appropriate keyletter row. If, for example, "EXIT" is used as the key to
encipher the plaintext message, "COME HERE," the
tableau provides a ciphertext of GLUXLBZX: G at the
intersection of "c" column and "E" row; L at the intersection of "0" and "X"; and so on. (Most encipherments omit spaces between words and sentences, since
they can provide valuable hints to the cryptanalyst-the
code-breaker. However, a good enciphering system
should not need this precaution.)
There is an extremely simple rna thema tical model which
can represent such encipherment. If the letters of the
alphabet are assigned numerical values A=O, B=I, . . .
Z=25; if combination of a column and a row is defined
as "addition"; and if the character at the intersection is
defined a~ the "sum" -then encipherment consists of
addition modulo 26. According to such addition, 21 + 7
= 2. Inspection of an appropriately labeled tableau
(figure 2) shows that the corresponding "letter
equation" also holds true: V + H = C. Under this modu-
'.
COMPUTERS and AUTOMATION for April, 1972
13
1)
Ciphertext
BMZX ZJ KCVFMEGRI EWDOCX UCNWZ ZSKBLNTFZ FPNSKKXGRFDTKVKVRNSB TVSJONTDWPFY I ZJ
NDOGIDNKI XOVZHMDNSEVONUNDNTG YEVFCJ PZPFH PDGJ NGDOOPFDNTK E SPL VYTKRR
NPJDFFGRXEETTWMLFMCZ Z VFPPQYZEMDYDEH YRSUJ FDCUZIC ZJ Y BUODDJ PE FUZEEB
LNTFZKHNJ XI
I
c[]
Q
o~o
~.
c:.
~
0
•
..
fA
an
O~"<>-1'
/",,'0
""""-.0
· o·0
~ 2:~
1C
0
Q
.......
r:::=:::J.
'
Z
.6
0
~'-::
I
.1.
'J ' , ~,,- ..........
........
...... £!
o
0
P
4
0
0
-.~
• c:J'
00
N
H
0
'7
8
X
F
0 00 (0 0 0 0
.
• • ·· V"v
..
0
•
·
D
5
V
D
B
0
a
lito
I ~l ;mp~:11 ;;:" lit8T~ I ~
fJ -=#=
[p 00
~
0
<1
9[
Item
D
C
E
F
H
I
Item
C
D
E
H
I
Answer:
J
F
G
J
G
1
In each cell,
6
2
insert the
7
code for your
3
8
4
choice of
9
answer:
5
10
Survey Data: 1. Name
2. Title
3. Organization,______________________________________________________________________________________
4. Address
Good?
Other (121ease Sl2 ecif y)
Excellent? Not your field?
Average?
5. In camp uter programming, are you:
6. In syst ems analysis, are you:
7. In mana ging, are you:
8. What fields (not mentioned above) are you fairly good In (or even expert In)?_--,--________________________
9. What other capacities do you have? (Please don't be bashful -- but be objective) ________________________
10. Any remarks?________________________________________________________________~~~------~----~~
(attach pal2er if needed)
When completed, please send to: Neil Macdonald, Survey Editor, Computers and Automation, 815 Washington Sf., Newtonville, MA 02160
COMPUTERS and AUTOMATION for April, 1972
27
'* * '* *
•
DO YOU LIKE
COMPUTERS AND AUTOMA TION?
•
DO YOU HAVE A FRI END
WHO MIGHT LIKE IT?
•
HOW ABOUT GIVING HIM
(OR YOURSELF) A GIFT
SUBSCRIPTION?
A reprint from
CDrnl?n~a~~J:~
FOUR-5T AR REPRINT-----.
SCIENCE AND THE ADVANCED SOCIETY, by C. P. Snow,
Ministry of Technology, London, England (April, 1966)
THE INFORMATION REVOLUTION AND THE BILL OF RIGHTS,
by Dr. Jerome B. Wiesner, M. I. T.
(May, 1971)
EMPLOYMENT, EDUCATION, AND THE INDUSTRIAL SYSTEM,
by Prof. John Kenneth Galbraith, Harvard Univ. (Aug. 1965)
COMPUTERS AND THE CONSUMER, by Ralph Nader,
Washington, D. C.
(Oct. 1970)
cut here and tuck in flap
r--------------------------,
To: Computers and Automation
I
I
I
I
•
Take advantage of our current offer:
a gift subscription to him (or you)
and for you a copy of our FOUR-STAR ~:,~:,~:,~:, REPRINT
Name
Address
I
I
(
I!! I
evl
-= I
ANY MORE FRIENDS OR ASSOCIATES
WHO MIGHT LIKE SAMPLE COPIES
OF CGA FREE? - just tell us
and we will send them a copy
- so they also may start
liking eGA
B:
I
:
I
I
I
I
I
I
u
I
State
Zip
\
Send 13issues including directory, $18.50 II
in U.S.; add 50 cents for Canada, $3.50for Ig
I~
foreign postage
I~
Send 12 issues without directory, $9.50 in
I;
U.S.; add 50 cents for Canada, $3.50 for
\
foreign postage
I
enclose $
in full payment.
\
)
:
I
I
I
I
I
: Ci ty
II
•
Please enter a subscription for
)
I
I
I
Please send a FOUR STAR
~:,~:,~:,~:, REPRINT to me:
\
I
I
I
Name_______________________________
: Address
\
: City _ _ _ _ _ _ _ _ _ _State _ _ _ _Zip_ _ _ _ _ ;
cut here
...
M
~
e:{
..J
U
t-
en
a:
u::
M
ci
Z
~
~
~
· Ii
'"C
·~
III
Z·
'J:
ot-
:1l
en
0
a: co
w
U
start my subscription to COMPUTERS AND AUTOMATION according to the
instructions checked below.
4/72
One Year (including the Computer Directory and Buyers'
Guide -
13 issues) U.S. only.
0
$18.50
One Year (excluding the Computer Directory and Buyers'
Guide -
12 issues) U.S. only.
0
$9.50
Z
"C
~
'c
g
:::>
c:
~
YES,
'"C
• c:
e:{
~
_ _ _ _ _ _ _ _ _ _ _ Fold here - do not cut
::E
co
~
__ _
eli
'"
~
.z:
__ __ _
~ en
e:{
U5
II
_
'"C
0-
~
__ _
ev
III
I
co
~
~
>
..J
0-
w
a:
en
en
w
l-
>w
...J
w
~
CD
W
Q..
«
en
W
en
+"
E
..-
C
E
w
Z
~
2b
Ci5
:J
co
00.
~
0
Z
0
0
2
0
0.
~
+"
>CD
~
«
0
I::::>
:;{
CD
a:
...J
...J
I-
~
w
(!)
<{
I-
en
0
Q..
W
::::>
Q)
Q)
~
c:
0
en
r.:
J:
en
C'a
~
0
I.t)
t-
U
CD
N
0
co
o
Payment enclosed
fJ
Bill me
«
~
.!E."
.:;:
c:
0
+"
~
Q)
2
o
o
CO
Renewal
g
New subscription
...
J:
~
::I
c:
III
III
To SPEED the processing of your order, please check the one best
descriptor in each of the two categories below. (This information
will be used for statistical purposes only.)
BUSINESS TYPE
+"
0...
~
~
Country if not U.S.: _ _ _ _ _ _ _ _ _ _ __
Signature: _ _ _ _ _ _ _ _ _ __
P.O. No.: _ __
«
"C
~
ev
ev
City: _ _ _ _ _ _ _ _ _ State: _ _ _ _ _ Zip: _ _ __
2
W
a:
u
Address: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
0
c
~
Name: _ _ _ _ _ _ _ _ _ __
Title:._______
Organization : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
IZ
-5
'ro
..J
(Ii
w
en
a:
Q..
a:
01-Computer Manufacturer
02-Aerospace ,I Ail craft
Manufacturer
03-0ther Manufacturing
04-Raw Materials Processing;
(chemical, primary metal,
petroleum, food. etc.!
05-Mining and Construction
06-Computing & Consulting
07-Finance, Insurance. Pub!.,
-and Service Organizations
DB-Transportation Companies
09-Public Utilities
10-Research
II-Wholesale. Retail. Sales,
and Marketing Firms
12-Educational; (College,
University, or School)
13-Government and Military
14-libraries
JOB FUNCTION
___ I-Technical Management; (computer
installation management, program
management, or engineering mgmU
2-Computer Center Personnel;
(methods & procedure analysts.
and operators)
3-Programming Personnel; (system5.
application & research programmers)
4-Professionai: (systems analysts.
mathematicians. operations
researchers. and professors)
5-General Management Executives;
(corporate officers. owners. and
partners)
6-ERgineering Personnel; (systems
engineers. research & development engineers)
7-Research Personnel
B-Students
9-library Subscription
10-Subscription in Company Name
Only
I
e:{
~
ev
...
U
U
The Bad Image That Computers Are Earning
"Shall we as computer professionals correct our programs, so that they function
properly, or wait until we are kicked in the teeth by an enraged public?"
a
1. From Harold W. G. Gearing
Oueenswood House
Stone Drive
Colwall
Malvern, Worcester
England
Many of us here are concerned with the bad image
that computers are getting throughout the world, from
bad data, bad programming, and inadequate training of
data-preparation staff. I would suggest that you
might open a corner of your magazine to hold bad examples. In your June issue, you commented: "Shall
we as computer professionals correct our programs, so
that they function properly, or wait until we are
kicked in the teeth by an enraged public?"
Here are a few to be going on with:
1. A well-known British manufacturer of motor-car
accessories put its accounts-payable onto a computer. The young girls punching the input cards
from the supplier invoices were not properly
trained, and did not appreciate how to distinguish
goods value from returnable outer-packaging deposits. One supplier found that over a period of
several months, items in the Company's details of
payment forms bore little relationship to his
statements for accounts receivable. A day trip of
several hundred miles was necessary, to persuade
the responsible accountant to examine his internal
system.
g
2. A security-transport company computerised its
sales invoices. They show only contract number
and no narrative. Central offices of large users
have to examine contracts each time invoice batches arrive and manually allocate the invoices by
writing on them the source and destination of each
service. The invoicing system appears to be
geared to holding a weekly service charge only;
hence when the service is rendered only monthly,
the contract monthly charge is divided by 4.33,
rounded for entry in the master record; subsequent
multiplication by 4.33 by the standard monthly invoicing program results in an invoice which differs by pennies from the contractual figure, which
in many cases is also being prepaid by a banker's
standing order.
3. A London west-end outfitting company decided to
save money on its sales-ledger system by notifying all its account holders that they were being
transferred to Barclay-card. Many English people
dislike credit cards, and future business is
probably being lost, particularly postal order
business.
In this country the Inland Revenue is modifying
its former ambitions to centralize all income tax
records in a few computer centres. British obstiCOMPUTERS and AUTOMATION for April, 1972
nacy, in continuing to prefer to deal with a local
tax inspector, may be winning the day against the
Big-Brother Computer!
2. From The South Middlesex News, Jan. 18, 1972
Framingham, Mass.
In this age of mushrooming automation, man has
scored a rare victory over a computer -- in this
case a $1.5 million machine described as a "good
worker" which "just couldn't compete with people."
In Sacramento, Calif., in a move that cut teacher
credential processing time by 900 per cent, George
Gustafson fired the machine and switched to human
beings.
"We got rid of a million and a half dollars worth
of computer," said Gustafson, executive secretary of
the State Teacher Preparation and Licensing Commission. "We pulled the plug and sent it back to IBM."
Gustafson Monday described his action as "converting credential processing from a complex and costly
automated system to a streamlined, fully manual operation."
"It was easier to do it by harLd," said Gustafson,
who gave this assessment of the departed computer:
"It was a good worker; it jus t couldn't compete wi th
people."
This giant step backwards, Gustafson said, resulted in cutting the credential period from an average
of 95 days down to the current average of 10.
Gustafson said that as a result of the change he
was able to reduce the staff from 240 persons to
106. Since August, the department has cut 80 employees but all but three were given jobs with other
state agencies.
A side benefit of the switch came during the
summer, when the commission hired more than 50 poor
and minority students to help make the change to
manual processing.
("Man Beats the Machine", by United Press International)
3. From New York Daily News, Jan. 23, 1972
New York, N. Y.
More than 100 state senators and assemblymen
joined Sen. John D. Caemmerer (R-Nassau) yesterday
in sponsoring two bills to curb the power of the
Parking Violations Bureau, which has harassed and
dunned thousands of innocent motorists to pay for
undeserved tickets.
Caemmerer, chairman of the Joint Legislative
Committee on Transportation, said passage of the
29
bills would "prevent another PVB nightmare
New York City-or any other city."
in
The antics of the bureau and its computer, Super
Sleuth, were detailed in a News series after thousands of readers pleaded for help by writing Action
Line, a reader service feature that runs in the
Long Island edition.
Motorists who never had driven in New York City
were being plagued by Super Sleuth to pay parking
tickets. Letters to the bureau, at 475 Park Ave.,
South, brought no reply-only higher fines and
stepped-up threats.
The protests resulted in public hearings by
Caemmerer Dec. 2 at the State Office Building, at
which exasperated motorists told of widespread
bureaucratic bungling and arrogant disinterest by
the bureau.
Caemmerer's bills would require the violations
bureau to provide transcripts of all hearings, requ~re police witnesses if requested and allow defendants to be present at app~als.
Bureau Director Anthony Atlas has often declared
that the abuses of the bureau were being corrected,
but letters to The News from harassed and distressed motoris~did not diminish.
"My mail is worse than ever," Caemmerer said.
At The News, the rate of complaints also is
higher than ever.
("Aim Monkeywrench at Super Sleuth", by Donald
Weinbrenner)
4. From The New York Times, Feb. 20, 1972
New York, N. Y.
In the two weeks since 155,000 maximum base rent
orders have gone out to tenants living in rentcontrolled apartments, 17,379 landlords and tenants
have visited or telephoned the city Office of Rent
Control to protest or challenge the new rents fixed
in the orders.
Most important-at least to the innocent who are
being victimized-the bills would require more
careful vehicle identification and would require
the bureau to answer its mail.
One was an owner who had carefully reported on
the proper form when he increased the rent on an
apartment that had become vacated, only to have it
ignored when the city's computer figured the maximum rent he could collect from it.
"We must deal with the PVB's preposterous attitude toward answering legitimate mailed-in questions," Caemmerer said.
Robert C. Rosenberg, the city official in charge
of the office's computer, traced the error and
partly calmed the landlord.
The most maddening situation facing many motorists was that a parking ticket would be sent to
them in Nassau County, although they had never
driven in New York City. Not only that, it was
marked "second notice."
It turned out that the owner had listed the
apartment as "45" on one form and "4-5" on another.
Computer Confused
In a typical example, a man with a blue Rambler,
license plate XYZ-1234, would get a "second notice"
to pay a ticket issued in the Bronx-where he had
never been-on a yellow Ford with a different plate
number.
Mr. Rosenberg, assistant administrator of the
Housing and Development Administration's maximum
base rent system, said his computer had a onetrack mind and did not realize that "45" was "4-5"
as well as-on some occasions in some landlords'
minds-"S4. "
He then would write a letter-registered-to explain the goof. A month later, he would get a more
threatening letter demanding a higher fine. This
would continue implacably, willi higher fines and
threats to withhold auto registration renewal.
"The computer isn't able to match the listings,"
Mr. Rosenberg said. "A special uni t.-of people, not
computer sections-is resolving these matching
problems. As soon as a problem is resolved, the
computer is notified and a corrected order issued."
So the motorist was forced to pay up-although
innocent-or take a day off from work and drive 50
or 60 miles to straighten it out. In many cases,
the motorist would drive in, be told: "Don't worry
about it," and the next week receive another
threatening letter.
Mr. Rosenberg said landlords and tenants who
discovered apparent errors should file protests at
once. A tally Friday, however, showed that the
17,379 persons, who own or rent 21,250 apartments,
did not wait for urging and have filed protests
already. More will be coming in next week, for
102,000 orders are to go out tomorrow.
Caemmerer's bill would require the cop or meter
maid to put more information on the parking ticket.
In addition to the make of the vehicle, the bill
would call for it,s color, model, body type, plate
number and registration tag number, with its expiration date, accordnig to the Senator.
This should help the Motor Vehicle Department in
Albany identify more carefully the owner of the car
and make sure the ticket reaches the person who deserved it.
In all, more than 1.1 million orders are to go
out.
One set of problems that is giving the computer
pause arises from inconsistent financial data.
When data are not consistent, the computer does
not attempt to reconcile the conflict. It just
omits the apartments to which they refer.
The legislation also would require that all certified mail tothe bureau from motorists must be answered wi thin 30 days or the case must be dismissed.
Until it gets a correction, the computer solves
its problems by tapping out a message on the order:
"Controlled apartments for which you have not
been notified require further research. You will
receive notification on these apartments shortly."
The buteau had never bothered to answer mail.
Letters were stacked up by the thousands.
("17,379 Protest New Rent Orders; Inconsistent
Form Entries and Computer Blamed", by Will Lissner)
30
II
COMPUTERS and AUTOMATION for April, 1972
.~
Do You Want To Stop Cri.me?
William P. Wood, III
Dept. of Information Systems
Virginia Commonwealth Univ.
800 W. Franklin St.
Richmond, Va. 23220
"The last time the three fugitives were seen
after blowing up the University laboratory, they
were driving north towards Canada."
'~he robbers jumped into a waiting car and fled
into the night."
Automobiles Used in Crime
Such are typical daily news reports. What do all
of these reports have in common? Almost every criminal MUST USE AN AUTOMOBILE SHORTLY BEFORE AND/OR
AFTER HIS CRIME.
Automatic Car Identification
Such a fact might be used to great advantage in
law enforcement. A system that would utilize this
fact is as follows:
The principle used is borrowed from the railroad
industry. The railroads use a system named ACI
(Automatic Car Identification) to daily track and
locate railroad cars. A problem the railroads have
had until ACI is the losing of railroad cars. It
is not unusual for a boxcar to be hooked up to the
wrong train, sent across the country, and sidetracked on some remote spur. To remedy this problem, the railroads set up a system of photoelectric
sensing devices along their tracks. These devices
have the ability to read color coded plaques placed
on the side of the cars. These sensing stations
are connected, via telecommunications, to a central
computer complex. This complex stores the location
of the industry's two million cars.
License Plates Read by Sensing Stations
Using a similar system the nation's automobiles
could be kept track of for the purpose of law enforcement. License plates would be required by law
to conform to specifications which would allow them
to be read by the sensing stations. These stations
would be set up at strategic locations along our
nation's highways. They would be tied via telecommunications equipment to centrally located computers manned by law enforcement personnel.
Suggested Systems to Track Criminals
This system could be constructed for either of
two levels of capability. The more limited level
would be that used for only tracking specific vehicles. For example, this would track known criminals. The more capable, and thus more expensive
system, would sense and keep in memory for short
periods, perhaps a few days, the location of all
vehicles on U. S. roads with license plates. Information concerning cars of special interest could
be kept in memory as long as necessary.
Such a system would provide the following advantages:
1. All cars within the area of a crime before
and immediately after could be identified
and their license numbers checked against
suspect's license numbers.
COMPUTERS and AUTOMATION for April, 1972
2. Getaway cars could be traced throughout the
country.
3. Suspects and witnesses would have their
stories verified or discounted.
4. Such a system would be a great deterrent
to crime.
5. Stolen cars could be traced almost immediately.
6. Drug traffic could be traced.
7. Such a system would cut into virtually
every kind of crime: murder, robbery,
rape, drug, arson, kidnappings, bombings,
vandalism.
.Cost of Tracking System and Cost of Crime
The cost of such a system might climb above one
billion dollars, just for the setup cost. The cost
of crime, though, is well in excess of twenty-one
billion dollars per year. Even if the system did
not pay for itself in the dollars and cents terms
of cutting the cost of crime by as much as its own
cost, how can one put a price tag on the misery and
fear generated by this plague on society? Citizens,
fearing for the safety of their families and homes,
flee to the relative safety of the suburbs. Parents
pray that their children will not fall prey to the
evils of drugs and juvenile delinquency. Not only
would this system be a boon to rid our society of
those sinister doers of evil, but it would act as
a tremendous before-the-act deterrent to those
realizing the probability of apprehension.
Cost to Our Personal Privacy
There is one final cost that remains to be considered. This is the cost to our personal privacy
generated by the ever encroaching inroads of computers and electronic gadgetry. Would such a
system shorten the time between now and 1984?
IBM'S POWERFUL PARTNER:
The Accounting Principles Board
(Editorial, reprinted with permission from "Samson
Technology Trends", Dec. 1971, copyright 1971 by and
published by Quantum Science Corporation, 851 Welch
Road, Palo Alto, Calif. 94304)
More than the increased competition from IBM, the
recent ruling from the Accounting Principles Board
(APB) , threatens to decima te the ranks of independent
computer companies that supply plug-compatible peripherals to the end user. The high principled accounting rules are forcing all companies, large or small,
to suddenly be super-honest and as a result, the APB
has forced the majority of the small and medium sized
independent computer companies to their knees, and
sometimes into bankruptcy.
The ruling, instituted last year, requires that a
conditional sale to a third party lessor be treated
on a standard rental accounting basis rather than as
a sale. This ruling has drastically reduced the reported earning of all peripheral suppliers who lease
the majority of their equipment and has made the financing of rental inventory almost totally impossible
for most. In fact, no company which does not have a
sizeable two to four year-old base of rental equip-
31
ment in the field can report profit. Worst of all,
the fastest growing companies are penalized most.
Before this ruling, most peripheral equipment suppliers relied on third party leasing to finance their
sales. Moreover, the third party transactions provided the independents with an excellent profit record
as long as their sales base continued to grow. Third
party sales were recorded as revenue at the full sales
price of the leased equipment rather than only the
rental revenue accruing to the leasing company.
Admittedly, the old treatment unfairly favored the
peripheral equipment supplier because it did not fully
allow for the potential ri sks and wri te-offs resulting
from the return of obsolete equipment. It was nevertheless approved of and certified by the accounting
profession for several years. The ethics of such
misguided support of a false business practice must
seriously be questioned, as well as the sudden application of new principles of accounting that reversed
prior certified profits into losses. The APB, in an
attempt to place the profi tabili ty of the independents
in proper perspective, has now created a situation
that makes sales growth of computer products to the
end user nearly impossible except for large companies
that can finance their own leases.
Many small companies that based their business
plans on old accounting rules have already or will
soon cease existence as independent entities. The
APB now has on its conscience not only dozens of
bankruptcies, but also large losses of hundreds of
thousands of stockholders who had accepted their
views and relied on them to make investment decisions.
What can we look forward to from the accounting profession in 1972 after the pooling-of-assets ruling
in 1970 and third party lease accounting in 1971?
Happy New Year!
CDC VS IBM -
CORRECTION
/. From Frederic O. Par/ova
169 West 88 St.
New York, N. Y. 10024
I read with interest the article on the IBM/
CDC/Greyhbund Litigation (C&A, Feb. '72).
In your introductory remarks you state, " ... and
there is apparently nothing in the court order(s)
which enables the plaintiff, CDC, to obtain access
to the information ...... " May I call your attention to page 49? "Examination and evaluation by
expert IBM, Greyhound, and Control Data personnel
is necessary if the data is to have any meaningful
significance to counsel in preparing their defense."
Also, "IBM shall make the said answers available
only to its counsel and to not more than 15 of its
non-clerical personnel .... CDC and Greyhound shall
comply with this paragraph but with a limitation
of 10 non-clerical employees .... "
II. From the Editor
We regret the error, and appreciate very much
your correction.
CORRECTION
In the March 1972 issue of Computers and Automation,
the following correction should be made: page 27,
"Pictorial Reasoning Test: C&A No.5" - in the
first paragraph under Figure 2, replace the words
"include OK or X as appropriate" by "include V, X,
or Z as appropriate."
32
REDUCING AND DISMANTLING
SCIENCE AND RESEARCH
INSTITUTIONS, AND SOCIAL
RESPONSIBILITY
Andrew G. Michalitsanos
University of Cambridge
Cambridge, Eng/and
The disillusionment with science in the United
States is clearly evident in the continuing decline
in the number of private and industrial research positions and the systematic reduction of research
funds in both governmental and nongovernmental sectors.
The attitude of the Nixon Administration and the
U. S. Congress has been aimed at drastically reducing the volume of scientific research in the United
States, a nation whose growth and prosperity is almost totally dependent on the inventiveness of its
scientists, who can provide new ideas and techniques
for industry.
The Government's reason for this policy is based
on public clamor against "non-relevant science."
It would be interesting indeed to find a person
who could define" relevant scientific research."
Many areas of scientific endeavor were at one time
or another considered not directly applicable to
public needs, but have become almost guiding centers
of industrial applicability many years later. For
example, solid state physics, at one time only of
academic interest, today occupies an important place
in electronic applications of all types. The telephone on the President's desk was constructed with
the aid of "non-relevant scientific research."
The U. S. Government's attitude has a much deeper
psychological effect on the American public in general. The cutting of research funds fosters the belief held by many young citizens today that science
lies at the roots of the country's dilemma since it
has brought the problems as well as the advantages.
The new generation appears to have replaced scientific logic with mysticism, clearly a step backward into the caves.
Pure science has not created the problems, but
rather society's inability to apply scientific discoveries in a resourceful and meaningful manner.
We should recognize that pure scientific research
provides an intellectual store from which we may extract ideas and apply them to our needs. Reducing
funds and decreasing the amount of pure scientific
rese~rch simply makes the available store of informa tion smaller.
I am not advocating indiscriminate spending for
relatively obscure areas of science. However, I can
see little use in dismantling the research institutions that have taken decades to establish.
The Nixon Administration should reassess its
actions and consider what sad effects their decisions will have on the United States and the rest
of the world for decades to come.
(Based on a letter published in the New York Times,
Sept. 24, 1971)
COMPUTERS and AUTOMATION for April, 1972
••
A Concerted Campaign To Deny
The ~merican People Essential
Knowledge About The Operation
of Their Govern ment
Prof. Henry Steele Commager
Amherst College
Amherst, Mass. 01002
The Nixon Administration is at it again. First
it tried to intimidate the television networks. In
vain. Next it attempt~d, for the first time in our
history, to silence the press by the threat of prior
censorship. Again in vain.
Now
on the
branch
United
in the
the Justice Department has launched an attack
constitutional privileges of a co-ordinate
of the Government -- the Congress of the
States and, by implication, every legislature
United States.
The issue is once again the Pentagon Papers. We
might have supposed that with its defeat in The New
York Times case the Government would drop thi s shabby
prosecution. Not at all.
At the instigation of the internal security division of the Justice Department, a Federal grand jury
in Boston has now subpoenaed Dr. Rodberg, legislative
assistant to Senator Gravel, to appear before it.
The Government's brief acknowledges that the thrust
of the subpoena is to prepare the ground for an inquiry into "acts done by Senator Gravel in reading
and inserting into the record the Pentagon Papers."
This may seem like a tempest in a teacup -- especially as the Government has by its decision to publish the Pentagon Papers in toto abandoned its argument that their publication would do "irreparable
injury" to the "security" of the United States -an argument palpably absurd at the time. But in
fact three major principles of our constitutional
system are at stake in this new Justice Department
caper.
First is the hard-won principle of the immunity
of any legislator from just this sort of harassment.
Second is the principle of the separation and the
equality of powers of the three departments of government. Thi rd is the principle of freedom of speech
and of the press.
,
The Constitution provides (Art. I, sec. 6) that
"for any speech or debate in ei ther House, they
[Congressmen] shall not be questioned in any other
place." Neither the purpose nor the meaning of this
clause is obscure. The purpose was to make it forever impossible for any executive authority to punish or intimidate any legislator for what he might
say in the legislative chambers.
These interpretations are now challe~ged by the
Government brief which asserts that Congressional
immunity from "question or debate" does not cover
committee reports, and that it is wholly personal
and cannot be enlarged to embrace legislative staff.
The first of these assertions was disposed of as
recently as 1969 when Chief Justice Warren reiterated the principle that the speech clause covered
committee reports, resolutions, "and such things as
are generally done in a session of the House in relation to the business before it."
COMPUTERS and AUTOMATION for April, 1972
The second assertion raises the larger issue of
the separation and equality of powers in our Government. But the principle laid down in Jefferson's
Manual of Parliamentary Practice and formally adopted by both houses of Congress, has never been successfully challenged, name ly tha t "Congressmen are
at all times exempted from question elsewhere, for
anything said in their own House; that during the
time of privilege neither a member himself, nor his
wife, nor his servants, may be arrested, cited, or
subpoenaed in any court."
The logic of these constitutional and legal immunities is obvious. It is not to confer special
privileges on legislators -- or on judges or Presidents -- but to protect them against harassment and
intimidation which, as Justice Harlan put it, might
"inhibi t and dampen the ardor of all but the most
resolute, or the most irresponsible, in the unflinching discharge of their duties."
The attack on Senator Gravel's immunity is pernicious -- not only constitutionally and politically, but philosophically. It is part and parcel of
what can only be described as a concerted campaign
to deny the American people that knowledge about the
operation of their Government so essential to the
sound functioning of democracy.
It is a direct assault on the Constitution and
the separation of powers; it is an indirect assault
on the principles which the Constitution was designed
to preserve and advance, above all the principle of
freedom of speech and of the press. In this matter
what the father of the Constitution, James Madison,
said in 1794, is still relevant: "If we advert to
the nature of Republican government, we shall find
that the censorial power is in the people over the I
Government, not in the Government over the people."
(Based on a report "A Senator's Immunity" by Henry
Steele Commager published in The New York Times,
October 15, 1971)
FAll JOINT COMPUTER
CONFERENCE: TOPICS
From the Editor
The Fall Joint Computer Conference in its call
for papers has listed the topics which it seeks or
hopes to cover in the program of papers:
User Applications and Requirements
Architecture Trends and Limitations
Terminals
Communications
Complete Systems and Networks
Hardware Advances
Software Development
Measurements, Analysis, and Evaluation
Reliabili ty
Social Issues
New Ideas -- Try Us
Some of the most important topics in the computer field are left out of this list:
Computers and Privacy
Should IBM be Broken Up as a Monopoly?
The Distrust of Computerized Systems by the
Public
The Use of Computers in Dictatorships
Should Computer People Develop into Information Engineers?
Perhaps all these topics arc included silently
in the topic "Social Issues".
33
Internal Revenue Service:
USE OF COMPUTERS
William H. Stewart, Jr.
Asst. Prof. of Political Science
Bureau of Public Administration
Univ.. of Alabama
University, Alabama
(Excerpt from "Computers and Government", Citizens
Information Report No.7, publi shed by Bureau of Public
Administration, University of Alabama, 1972, 54 pp.
"The material in this publication is not copyrighted
in the hope that its use without restriction will be
more widespread.")
Perhaps the most comprehensive use of computers
in American government at any level is that made by
the federal Internal Revenue Service. Automatic
data processing in this agency has resulted in more
accurate identification of potential and delinquent
taxpayers, prevention of duplicate refunding of taxes,
verification of tax computation, billing of outstanding taxes, tax audit, development of statistical information for proposed legislation, and tax administration controls.
These improvements are not simply to the government's benefi t. They help ensure that
the tax burden will be borne equitably and that some
do not escape paying legally owed taxes while others
pay more than they are obliged to pay.
Approximately 76.8 million individual income tax
returns were filed during the 1970 filing period.
The role of computers in the processing of these returns is substantial. The central automatic data
processing installation of the Internal Revenue Service is located at Martinsburg, West Virginia. The
system is so massive that it requires 275 technicians
to feed in the data received from district computer
centers, which handle initial processing and verification of tax returns. Returns are transmitted to
Martinsburg for more extensive analysis and consolidation into the master magnetic tape files. One-half
inch of tape contains a complete three-year running
tax account for each return filer.
The automatic data processing system utilized by
the Internal Revenue Service has been programmed to
look for the most common types of taxpayer "mistakes"
and the classifications of taxpayers who are most
likely to make these errors. The agency's experience
is that the higher the individual or corporate income
the greater is the chance of error. Its data system
program includes a special check for returns over a
certain secret figure. The computer will note and
flag returns which depart markedly from expectations.
An audit may be indicated for taxpayers who list substantially more in the areas of charitable contributions, interest payments, and medical expenses than
they have in previous years and/or exceed the normal
figures expected for individuals in their income
bracket. A combination of incongruencies will increase greatly the likelihood of an audit by mail or
through a personal interview.
At the present time approximately one in44 returns
is audited. This the computer does not do. It verifies arithmetic, prepares tax bills, and indicates
which returns should be audited. The auditing process itself, however, is not automated. Based onthe
number of revenue agents it has, a district Internal
Revenue Service office advises the Martinsburg computer system of the number of return~ it has the manpower to audi t. Given a figure of, for example,
75,000, the data processing center supplies the office
with the names of the 75,000 "best" candidates for
34
further examination based on how far these people
depart from the standard categories.
Federal officials feel that automatic data processing is geared to improving accuracy and equity in
tax collecting. The efficiency of the system makes
possible the collection of greater revenue at smaller
cost. During the initial year in which computerized
tax collection procedures were in operation in the
Southeast, many tax delinquents paid taxes owed without governmental prodding in the belief that these
omissions would be discovered by the government's
computers. Computers have had, therefore, not only
a technological but a psychological effect on tax
collection procedures. Officials of the Internal
Revenue Service also report that embezzling by agency
employees has become harder since the processing of
income tax returns was made the assignment of giant
computers.
SOME RESPONSIBILITY FOR
OUR CHAOTIC SOCIETY
S. R. Harrison, Ph.D.
The Norwegian Institute of Technology
Trondheim, Norway
Good for you and your articles on President Kennedy's assassination. Our country needs people like
you and your staff who are willing to take a stand
and fight for what they believe to be true.
The time is long overdue that our profession had
some life injected into it and begins to accept some
responsibility for our chaotic society. I find too
many highly specialized journals "doing their jobs"
in presenting only restrictive technical publications.
I applaud your efforts.
Unsettling, Disturbing, Critical
Computers and Automation, established 1951 and
therefore the oldest magazine in the field of computers and data processing, believes that the profession of information engineer includes not only
competence in handling information using computers
and other means, but also a broad responsibility,
in a professional and engineering sense, for:
The reliability and social significance of
pertinent input data;
The social value and truth of the output
results.
In the same way, a bridge engineer takes a professional responsibility for the reliability and
significance of the data he uses, and the safety
and efficiency of the bridge he builds, for human
beings to risk their lives on.
Accordingly, Computers and Automation publishes
from time to time articles and other informatjon
related to socially useful input and output of data
systems in a broad sense. To this end we seek to
publish what is unsettling, disturbing, critical
-- but productive of thought and an improved and
safer "house" for all humanity, an earth in which
our children and later generations may have a future, instead of facing extinction.
The professional information engineer needs to
relate his engineering to the most important and
most serious problems in the world tOday: war,
nuclear weapons, pollution, the population explosion, and many more.
COMPUTERS and AUTOMATION for April, 1972
RTR
GGO
NFH
WSW
pez
Qeo
GVR
"fF'D
[J
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 : 2009:03:24 23:25:47-07:00
Create Date : 2009:03:24 23:25:47-07:00
Metadata Date : 2009:03:24 23:25:47-07:00
Format : application/pdf
Document ID : uuid:00ea68d2-6650-4d18-b8b1-88f5eb4f08ef
Instance ID : uuid:3a9c714e-9e10-4dcb-8b53-065cbda62315
Page Layout : SinglePage
Page Mode : UseNone
Page Count : 52
EXIF Metadata provided by EXIF.tools