Skip to main content

Full text of "Interactive Computing newsletters"

See other formats


PRESS 

REVIEW 


The following articles are reprinted solely as items of interest for the independent evaluation by members of 
ATSU. The opinions, statements of fact, and conclusions expressed herein are not those of the Association. 



CRTs Cause of Editors' Cataracts? 


By Nancy French 
And John P. Hebert 

CW Staff 

NEW YORK — Two copy editors 
who work for The New York Times 
were diagnosed as having cataracts af¬ 
ter the newspaper switched to elec¬ 
tronic editing and typesetting — a 
change that put the two in front of 
CRTs all day. 

Both editors, who are in their 30s, 
suspect radiation emitted from the 
CRTs is to blame. 

Cataract victims average 60 years of 
age, according to Dr. Edward Murphy 
Jr., a Boston opthalmologist affiliated 
with the Mar Jachusetts Eye and Ear 
Infirmary who also teaches at Harvard 
Medical School. 

The disease, "which clouds the nor¬ 
mally transparent lens of the eye and 
eventually causes blindness, can be 
corrected surgically by removing the 
lens and then adjusting, vision with 
contact lenses and/or thick glasses. 

Murphy said the disease is so un r 
usual in people as young as 30 or 35 
that when he ,finds such a case he 
usually looks for "some systemic dis¬ 
ease such a£ diabetes." * 

Dr. Milton Zaret of Scarsdale, N.Y., 
an ophthalmologist and authority on 
the effects of radiation on the eye, said 
radiation could be to blame. Zaret was 


called in on the case after the News¬ 
paper Guild of. New York AFL-CIO 
intervened on behalf of the two copy 
editors. 

At first The Times took the position 
that there was no connection between ~ 
use of the CRTs and the two cataract 
diagnoses and that the men probably 
had cataracts before they began work¬ 
ing on the CRTs. But 50 copy editors 
at The Times weren't so sure. 

These editors signed a petition asking 
for an inquiry. 

The Newspaper Guild asked the 
Manhattan Federal District Court to 
grant a temporary restraining order to 
take the two editors off the CRTs until 
new tests could be done. 

Judge Thomas P. Griesa ruled the 
dispute should be settled through ar¬ 
bitration. A ^anel was selected and at 
the moment a representative sample of 
the machines in use at the newspaper 
is being tested for radiation "by an in¬ 
dependent party," according to Harry 
Fisdell, vice-president of the guild. 

Research done to date on the 
cataract-radiation problem is incon¬ 
clusive. A study released May 13'by 
the National Institute of Occupational 
Safety and Health (Niosh) of Cincin¬ 
nati stated that CRTs do not emit 
enough nonionizing radiation to pose 


an occupational hazard. 

The study did acknowledge both 
copy editors have tne type of cataracts 
"compatible with those reported from 
exposure to radiant energy." 

However, the Niosh report has "fatal 
I flaws," according to Zaret. Further, 

I neither qf .the, two men has "any 
l significant medical history that would 
lead to the development of cataracts," 
he said. 

Zaret, who has spent 20 years study¬ 
ing the effects of radiation on the eye, 
said the problems with color TV sets — 
which also contain cathode ray tubes 
—. have been resolved, but CRT ter¬ 
minals are a "totally unexplored area." 

"Deflection coils'' which deflect the 
electronic beam onto the screen "create 
electromagnetic radiation as well as 
nonionizing radiation." These could 
produce cataracts, he said. 

The terminals in use at The Times are 
largely Harris Corp. H-1520 CRTs, 
which create characters using a dot¬ 
matrix display, according to Gordon 
Kilgore, program manager for Harris' 
Composition Systems Division. 

Nearly 1,000 of these terminals are 
now in use all over the world, Kilgore 
said, and no problems like thjs have 
ever been reported before. 


Reprinted with permission of Computerworld, Newton, Mass. 02160, 
August 1, 1977, Copyright 1977. 







Antidote For 
Data Base Disease 


Data base management 
systems without good systems 
management, in actual 
practice, are often unmitigated 
disasters. 


by Kenneth T. Orr 
Contributing Editor 

It is not immediately obvious why 
data base management systems 
should have become so popular to 
the data processing profession. If one 
believes some of the published re¬ 
ports, data base systems are very ex¬ 
pensive; they take up large amounts 
of resources, namely, personnel, com¬ 
puters and time; and they usually re¬ 
quire a great many applications to 
justify their initial costs. 

To be perfectly honest, the primary 
reason data base management sys¬ 
tems are currently in vogue seems to 
be that they sell computer hardware. 
They don't sell them directly, of 
course, since very few of the cost 
projections for implementing a data 
base software package truly cost out 
the ultimate impact of the data base 
software on the operation involved. 
But sooner or later, in order to sup¬ 
port the very growing appetite of the 
data base system for CPU cycles and 
disk storage, their owners are forced, 
much like the proud possessor of a 
baby alligator, to increase their main¬ 
tenance budget or face an unpleasant 
hungry pet. 

And there are other, organization- 
related, problems associated with 
data base management systems. For 
example, if you read the job descrip¬ 
tion of the average data base 
administrator, you begin to suspect 
that the only person in the universe 
that could fill the requirements wears 
a blue jumpsuit with a big red "S" on 
the front. This creates a perplexing 
situation for management. In one 
breath, you read where the data base 
administrator is critical to the installa¬ 
tion of a "true data base environ¬ 
ment," and in the next you are told 
that the job of the data base adminis¬ 
trator is an impossible one. 


By this time, I bet you expect me to 
tell you that data base systems are 
doomed and that it's all a plot by the 
hardware manufacturers to unload 
large amounts of computers on an 
earnest but not too clever public. 
Well, I'm going to surprise you. I like 
data bases, even though I find that 
most organizations starting to devel¬ 
op a data base begin on the wrong 
foot. I think data bases are necessary, 
even critical to long term success of 
information management. But I also 
feel that without good systems man¬ 
agement, in actual practice they are 
often unmitigated disasters. 

For the most part, "data base" is 
simply a term that has been cor¬ 
rupted. Unfortunately, it has come to 
mean some mystical set of computer- 
oriented software which maintains 
large amounts of interrelated data on 
the computer. 

(See Figure 1.) 

Data base is a concept 

But a data base is not software; it is 
really only a concept, a way of look¬ 
ing at data. And rather than being a 
radical new concept, it is a radical old 
one. In fact, if you have a system 
today, you also have a data base. 
However, you probably are not man¬ 
aging it. All of the data that you have 
within the system to produce the cor¬ 
rect output can be thought of as a 
data base. But because of problems in 
definition, we have come to consider 
a data base only as some kind of 
monolithic "monster" file and any set 
of old-fashioned (See Figure 1) files a 
sort of "non-data base." 

As with most good ideas, the data 
base as a collection of the organiza¬ 
tion's data has become confused with 
the software which manipulates the 
data. This is unfortunate, for the 
greater part of any organization's data 
usually resides outside the computer 
altogether. And a true data-based 
systems approach would consider all 
of the data in the organization as part 
of the data base, to be managed, 
updated and maintained. (See Fig¬ 
ure 2.) 

Why the current interest in data 


base anyway? Well, after a couple of 
decades of systems building, our past 
mistakes have begun to catch up with 
us. We now have large numbers of 
systems which are more or less out of 
control. And most of our systems 
have reached the stage of develop¬ 
ment where we can clearly recognize 
the problems involved in any future 
changes. In many places, data base 
management has been seen as the 
answer to all the organization's data 
processing problems, i.e., mainte- 


Figure 1 



a 

a 

a 


The data included under the data base 
management system 


Figure 2 



7 

CZ^ 

□ 




/ 



r.| 

7 


i_i 

u. 

/ 




7 

p. 

□ 



□/ 

/ 



a 

a 

a 


The data covered under data based thinking 




































































Antidote... continued 


nance, flexibility and increasing infor¬ 
mation requirements. This is often 
wishful thinking; unless you change 
the systems development process to 
make data base thinking an integral 
part, it is unlikely that installing a data 
base management system will do 
much more than make things worse. 

It is more or less a truism that the 
organizations which made the best 
use of the computer were also the 
ones that had the best manual and 
punched card systems to begin with. 
You might be tempted to say that 
breeding counts. Organizations 
which have a strong respect for sys¬ 
tematic operations seem to be able to 
apply that knowledge at each stage 
of development. The same is definite¬ 
ly true of data base management sys¬ 
tems. Those organizations which are 
the most successful with data bases 
are largely those which already have a 
solid systems building process to 
work with. 

But organizations in real trouble are 
among the most likely to see data 
base management as a cheap way 
out of their problems. A good many 
organizations grasp at data base man¬ 
agement systems to save them, much 
as a drowning man grasping at straws. 
But it is precisely in these cases that 
data bases are most likely to fail. As a 
matter of fact, so will anything else, 
except divine intervention. 

I have a strong suspicion that 
somewhere there is at least one orga¬ 
nization making excellent use of 
every one of the major data base 
management packages. And I would 
bet even money that somewhere out 
there/there is an organization having 
monumental problems with every 
single one of the major data base sys¬ 
tems also. There are good and better 
data base systems, but the real secret 
of success in utilizing a data base 
seems to lie in the systems manage¬ 
ment of the organization seeking to 
use data base concepts. 


Part of the difficulty involved in 
using data base concepts is the 
mythology surrounding the area. 
Because of some simple misunder¬ 
standings, many organizations take 
their eyes off the fundamental prob¬ 
lem of systems development and per¬ 
formance in attempting to come to 
grips with the complexities of data 
base administration. This is a mistake. 
For a data base is simply the aggrega¬ 
tion of the data requirements of a 
number of application systems. 
Because a data base exists to aid in 
the production of useful output, and 
because data which is not used can¬ 
not be correct, there is no such thing 
as an integrated data base without in¬ 
tegrated application systems. 

The data processing universe re¬ 
volves not around the data base, as 
some would have you believe, but 
rather around the outputs that our 

application and management systems 
. are called upon to produce. In far too 
many organizations, the building of a 
total corporate data base has become 
the "Bermuda Triangle" where one 
project after another is lost without a 
trace. 

What the world needs is not better 
data base software, though heaven 
knows that would certainly help; no, 
what we need is serious application 
of structured data based thinking. We 
have found that when organizations 
develop correct approaches to defin¬ 
ing systems, they get a good data 
base design as a by-product. Recent¬ 
ly, the heretofore separate worlds of 
data base design and systems theory 
have begun to come together. 

"Data" and "procedure," we find, 
simply represent different aspects of 
the same reality. What we learn in 
one area of systems or data base re¬ 
search can be carried over directly 
into the other. Indeed, when we con¬ 
centrate on doing the right things in 


the right sequence, we are usually 
surprised to find that all sorts of good 
things happen to us. We come up 
with simple workable systems and 
elegant, efficient data base designs, 
too. But the parts of the systems 
problem, procedure and data, exist 
together, not separately. Neither can 
be developed independently, and 
neither can function alone. 

Mirroring reality 

Ultimately, form follows function. 
Our data bases and our systems must 
mirror reality. Because of physical 
limitations, that mirroring will always 
be an approximation of the real 
world. However, if our data bases are 
not organized to make them deal 
with the real world of the organiza¬ 
tion, they will hinder rather than help 
us. 

In the past few years, we have 
begun to see the impact of "data 
structured" systems design at the 
programming level and at the systems 
level. We now see the impact that 
"structuring data" is apt to have 
throughout our organizations. A logi¬ 
cal systems and data base design rep¬ 
resents a major breakthrough. The 
same tools which are helping us at¬ 
tack complexity in programming and 
systems design are now beginning to 
have a significant impact on the de¬ 
sign and development of integrated 
data bases as well. "Structured data 
bases" are a natural outcome of 
"structured" systems design. 

Future articles in this series will cov¬ 
er the application of structured sys¬ 
tems design tools to "logical" and 
"physical" data base design. □ 


Editor's note: This article was origi¬ 
nally prepared by Mr. Orr for the 
Second Annual Structured Systems 
Design Conference and was pub¬ 
lished in its copyrighted proceedings. 


Reprinted with permission of INFOSYSTEMS, July 1977, 
Copyright 1977, Hitchcock Publishing Company. 






ON LINE 

WITH ROGER H. 

A stand 
on standards 


Flak 

for 

S. 825 


Biddle 

bites 


TO WASHINGTON 


ALLEN 



A. G. W. (Jack) Biddle, president of the Computer and Com¬ 
munications Industry Association, doesn’t fool around. Perceiv¬ 
ing a problem, he quickly develops a solution. More often 
than not, it's controversial. Take the question of how stan¬ 
dards are set — or, really, aren’t set — in the computer/ 
communications industry. 


Early this year, Biddle complained that standards development 
"has been marred by generally uninspired thinking, selfish 
corporate interests and incessant foot dragging (by) the larger, 
established suppliers." So he suggested a federally chartered 
standards board — for the electronic funds transfer area. 
Chiding the Federal Government for "never really" encoura¬ 
ging data processing standards development, Biddle pointed 
out that: "It has been ten years since the Brooks Bill became 
law; yet there are still no meaningful hardware -- and relative¬ 
ly few software -- standards for ADP equipment." The re¬ 
sult, Biddle scowled: "Untold millions of tax dollars have 
been wasted because of the inability to freely shift peripheral 
equipment and software within the federal ADP inventory. 
There are still no mechanisms in GSA, OMB or NBS to facili¬ 
tate the adoption of ADP standards, and, until Congress pro¬ 
vides these, our government will continue to waste our taxes." 

Now, some seven months later, Biddle's basic suggestion has 
been written into a proposed law: S. 825, the Voluntary Stan¬ 
dards and Accreditation Act of 1977. Sponsored by James 
Abourezk (D-SD), the bill goes far beyond EFT standards. 

And it's getting a lot of flak from the American National 
Standards Institute (Ansi), the Computer and Business Equip¬ 
ment Manufacturers' Association (Cbema) and the American 
Society for Quality Control (ASQC). 


Essentially, the bill would set up an independent, government- 
financed National Standards Management Board and a new, 
NBS-based Institute of Standards and Accreditation (ISA). 

Both, presumably, would replace Cbema and ASQC as the 
Ansi secretariats for standards-setting. The bill vests overall 
standards-making responsibility in the Secretary of Commerce, 
giving the ISA power to set standards if private organizations 
fail to do so. 


Testifying before the Senate Antitrust and Monopoly Subcom¬ 
mittee, Biddle cited many of the problems he criticized when, 
before S. 825 was written, he first proposed a new standards 
procedure. In the computer/communications industry, Biddle 
admitted, "We have numerous standards dealing with how 
a magnetic tape reel should be labeled...as well as standard 
numerical designations for cities." But, he complained, 

"We do not have standard higher level computer languages" 







Pointing 
the finger 


Rebuttal's 

off 

the bull's eye 


that permit program interchangeability. Also absent, he 
argued, are standards for "the efficient high-speed usage 
of computer/communications networks" and for the intercon¬ 
nection of hardware and software units which comprise a 
computer system. 

At fault, Biddle insists, is the voluntary standards-setting 
procedures used now. And he finds five bugaboos: Volun¬ 
tary standards bodies act only in their own self-interests. 
Industry giants dominate the standards process because 
participation in it is too expensive for smaller producers. 
Consumer or general-interest participants, lacking technical 
expertise, are at the mercy of representations made by the sup¬ 
pliers. The "inordinate time required to reach a consensus" 
among participants means that voluntary standards lag techni- 
cological developments. Suppliers resist standards which, 
they fear, might restrict available design options or disclose 
product and marketing strategies. 

Getting more specific, Biddie contends that "voluntary stan¬ 
dards activity...has been carried out under auspices and control 
of the trade association (Cbema) which represents IBM and 
the other systems competitors." He also points out that Ansi 
gave up after five years of working on a standard for intercon¬ 
necting peripheral equipment with host computers. Then, 
after several CCIA member companies spent four years and 
$350K trying to get the project moving again, Biddle says, 

"it had become patently clear that the systems suppliers 
were not working on a good-faith basis toward development 
of a neutral standard that would benefit both users and pro¬ 
viders." (Now, Ansi is circulating a proposed standard for 
public review. But it's the de facto IBM standard. And it's 
getting lots of heat from systems suppliers.) Biddle further 
zings the present system by pointing out that the NBS took 
"little initiative" to promote interconnection standards for 
peripherals despite a ten-year-old legislative mandate (under 
PL 89-306, the Brooks Bill) to do so. 

Interestingly, in their opposition to S. 825, Ansi, Cbema and 
ASQC never refute Biddle's charges directly. Ansi Execu¬ 
tive Vice President Donald I. Peyton says present laws "are 
totally adequate to provide oversight for voluntary (standards) 
activity." Robert W. Shearman, ASQC executive director, 
says "The present voluntary consensus standards system works 
as a fair and equitable arena for commercial product and 
service standards." Improvements don't need more federal 
law, he says. ASQC contends S. 825 would cost three times 
more than the present system. Cbema President Peter F. 
McCloskey argues that voluntary standards practices have 
"operated successfully for over 75 years". To Cbema, 
the bill is "premature and unwarranted." 

S. 825 is due for a Subcommittee vote probably next month 
Then it goes to the full Judiciary Committee. Meanwhile, 

Jack Biddle is perceiving other problems — and solutions. 


Reprinted with permission of Datacomm Advisor, August 1977, 
Copyright 1977. Mr. Allen is the Managing Editor of Datacomm Advisor, 
the monthly newsletter of International Data Corporation, and is also 
a contributing editor for Distributed Data Processing Newsletter. 







Game 

Econometrics Gains Many New Followers, 
But the Accuracy of Forecasts Is Unproven 


By Lindley H. Clark Jr. 

Staff Reporter of The Wall Street Journal 

A century and a half ago, Antoine-Augus¬ 
tin Cournot, a French economist and mathe¬ 
matician, mused about a system of equa¬ 
tions that would explain all the interwork¬ 
ings of the economy and thus help to fore¬ 
cast the future. Alas, he concluded, such a 
system was beyond human powers. 

Today, however, many economists are 
relying heavily on such equations in a popu¬ 
lar (if unproven) process called model build¬ 
ing, or econometrics. Using past relation¬ 
ships among factors affecting the economy 
—population, weather, wars, government 
policies—the economists set up the equa¬ 
tions to try to predict trends. 

Although there isn’t any solid evidence 
that model building produces much better 
predictions than any other forecasting 
method, businessmen in the current uncer¬ 
tain economy are highly curious about the 
future and how to plan for it. And they are 
relying more and more on econometrics. 

A Boom for the Specialists 

The result has been a business boom for 
the small group of economic consulting 
firms using econometrics. Among them: 

—Data Resources Inc. of Lexington, 
Mass., the only publicly held consulting 
firm. It reported net income of $1.5 million 
last year, an increase of 70% from the year 
before. 

-Wharton Econometric Forecasting As¬ 
sociates, a nonprofit company owned by the 
University of Pennsylvania. It has more 
than 200 clients, double the number four 
years ago. It is actively adding to its ser¬ 
vices, according to F. Gerard Adams, trea¬ 
surer, “so that we’ll match Data Resources 
line for line.’’ 

—Chase Econometric Associates of Bala 
Cynwyd, Pa., a subsidiary of New York’s 
Chase Manhattan Bank. It has built its cli¬ 
ent list to more than 600, with some firms 
paying annual fees as high as $50,000. 

—Merrill Lynch Economics, an affiliate 


of the brokerage firm of Merrill Lynch, 
Pierce, Fenner & Smith, today will demon¬ 
strate its new econometric modeling system 
at a luncheon at New York’s World Trade 
Center. “We’re launching ourselves full- 
scale into the econometric area,” Albert H. 
Cox Jr., president of the affiliate, says. 
Among other things, Merrill Lynch will 
make a 400-equation model of the economy 
available to clients through two computer 
time-sharing services 

The Role of Computers 

For better or for worse, computers have 
played a large role in the growth of econo¬ 
metrics. For one thing, they have made it 
possible for econometricians to handle much 
larger models of the economy, solving 
hundreds of equations in a short space of 
time. 

In addition, clients of the major consult¬ 
ing services now often have computer ter¬ 
minals in their offices. By tapping the 
proper keys on the terminal, a client can 
call forth current economic data from a 
computer storage center. Or he can “run” 
the model himself. 

Because no equation can be solved if it 
contains nothing but unknowns, the client, 
like the econometrician, has to make some 
assumptions—about government tax policy, 
for example. The client, in fact, can make a 
variety of assumptions, playing what the 
econometricians call “what-if” games. 

Data Resources Inc., the largest of the 
consulting firms, saw the lure of computers 
from the start, leasing its own equipment. It 
also invested manpower and money in build¬ 
ing up a huge “data bank,” a collection of 
constantly updated economic statistics. 

A Lure for Clients 

As a result, Data Resources has been 
able to sign up many clients who wouldn’t 
have been attracted merely by the economic 
analyses of its president, Otto Eckstein, the 
Harvard economist and former chairman of 
the President’s Council of Economic Advis¬ 


ers. One Data Resources client, for instance, 
is New York’s Citibank, which has its own 
large economics department. 

According to Charles Warden, Data Re¬ 
sources vice president, the firm, founded in 
1969, “turned its first profit in February 
1971, and we haven’t looked back.” When 
the firm first offered stock to the public last 
fall, it reported that its profits had risen 
substantially each year since 1971. 

When the first econometric forecasting 
group was set up at Wharton in 1963, its 
model was solved with mechanical calcula¬ 
tors, not computers. After Data Resources 
started, Wharton put its model on the Data 
Resources computer system-but it soon de¬ 
cided there were disadvantages in having a 
competitor market its model. 

In 1973, Wharton signed a contract with 
Boeing Computer Services, a subsidiary of 
the airplane manufacturer; Boeing now of¬ 
fers Wharton’s models and data to clients. 
The major model of the U.S. economy, 
which is used to make quarterly forecasts, 
has since been supplemented by a long-term 
model, an agricultural model, foreign mod¬ 
els and other services. 

Emphasis on Research 

As befits an affiliate of a major univer¬ 
sity, Wharton has always put strong empha¬ 
sis on pure research. “We aren’t as aggres¬ 
sive as a profit-seeking corporation,” Mr. 
Adams says. “But we operate in, a profit- 
seeking environment, and we act accord¬ 
ingly.” One indication: The firm is steadily 
absorbing more space in the university 
building where it is located; a large confer¬ 
ence room once used to present the quar¬ 
terly forecasts, now holds the desks of five 
economists. 

The models and data of Chase Econome¬ 
tric Associates are made available to clients 
through computer terminals offered by two 
time-sharing firms: ADP Network Services 
of Ann Arbor, Mich., and Interactive Data 
Corp. of Waltham, Mass. “If one system 
should go down, the other one is always 
available as a backup,” the firm tells pro- 





spective clients in a brochure. 

Merrill Lynch Economics will work 
through Interactive Data and National CSS 
Inc., of Norwalk, Conn. Merrill Lynch 
Ecoomics recently hired two men from Data 
Resources to help to beef up its econome¬ 
trics capability—Gary Ciminero and Dick 
Hokenson. 

Most of the other major consulting firms 
so far are holding back from a full commit¬ 
ment to econometrics and computers. But 
business at most firms still seems to be 
good, although they don’t disclose financial 
results. It’s evident, for example, that Town- 
send-Greenspan & Co. expanded its activi¬ 
ties even while its president, Alan Green¬ 
span, was away serving as President Ford’s 
chief economic adviser. 

Results Are Unspectacular 

But Townsend-Greenspan and nearly all 
other consultants now use econometrics to 
some extent. The attractions of the system 
don’t stem from a spectacularly successful 
forecasting record; like other forecasters, 
econometricians have had their troubles in 
recent years. 

Since 1968 the American Statistical Asso¬ 
ciation and the National Bureau of Eco¬ 
nomic Research have been jointly collecting 
forecasts from a panel of business econo¬ 
mists—some who used econometrics and 
some who didn’t. In 1975 Vincent and Jose¬ 
phine Su, two National Bureau economists, 
presented an evaluation of the forecasts in 
the bureau’s publication, Explorations in 
Economic Research. 

The conclusions didn’t offer any comfort 
to anybody. “No method predicts consis¬ 
tently better or worse than other methods, 
and no method predicts consistently better 
in levels or in changes,’’ the researchers 
said. They also compared the two associa¬ 
tions’ forecasts with those produced by the 
Wharton econometric model. The conclu¬ 
sion: There isn’t any clear evidence that the 
associations’ forecasts are any better—or 
any worse. 


“Half Magic” 

As business expands, the econometri¬ 
cians are trying hard to explain to new 
clients that their models aren’t a magic so¬ 
lution to all business problems. Michael K. 
Evans, president of Chase Econometric As¬ 
sociates, says he tells his clients that “my 
forecasts are about 50% my judgment ana 
about 50% model.’’ 

“The unfortunate thing about forecasting 
is that the future remains the future,” Mr. 
Adams of Wharton says. “What we do is 
half magic, half science. Science is con¬ 
cerned with the past. When you get into the 
future, it’s part magic.” 

In analyzing the workings of the economy 
and setting up equations to portray them, 
the econometricians have to rely on the 
ways the economy has worked in the past. 
Unfortunately for forecasters, the economy 
doesn’t always continue to work in precisely 
the way it did in the past. 

Despite the undistinguished forecasting 
record, it is hard to find many economists 
who will knock econometrics. A. Gary Shil¬ 
ling, director of White Weld Economic Ser¬ 
vices, uses econometrics in only a limited 
way. But he stresses that it is good for simu¬ 
lations, for the “what-if” games. An econo¬ 
metric equation give* a logical structure of 
economic relationships. Using such equa¬ 
tions “assures an internal consistency” in 
your forecasting, Mr. Shilling says. 

Econometricians have helped to expand 
the market for their products by bringing 
them doown closer to the level of the people 
actually running business corporations. “In 
the beginning,” Mr. Evans recalls, “all We 
had was a large model of the U.S. economy. 
Now we’re making regional and industry 
forecasts, and we can help companies fore¬ 
cast the markets for their own products.” 

Both Chase and Data Resources now are 
forecasting foreign-exchange rates. Charles 
Warden, vice president of Data Resources, 
says, “Our exchange-rate forecasts are ac¬ 
cepted because they’re useful. They may not 
always be exactly on target, but they point 
to the turns and the magnitude of the 
changes.” 


Reprinted with permission of The Wall Street Journal , 

Dow Jones and Company, August 2, 1977, Copyright 1977. 
All rights reserved. 




Computer Show’s Message: 
‘Be the First on Your Block’ 

By LEE DEMBART 

Special to The New York Time* 


BOSTON, Aug. 25—The computer 
revolution seems endless. Every six 
months, a new product comes along 
that outdates everything before it. The 
latest is the microcomputer, based on 
the same technology as the pocket cal¬ 
culator but capable of putting a power¬ 
ful computer into homes and small 
businesses. 

Several thousand people, many of 
them students or businessmen, turned 
out here today.for “Computermania,” 
a .major exposition of microcomputers 
carrying price tags of $300 to $3,500. 
The displays were extensive, the ex¬ 
hibitors excited and the computer en¬ 
thusiasts eager to study everything 
new. 

Limited Only to Imagination 

But no one could say for sure why 
people might need a computer at home.. 
“For fun” seemed the most honest an¬ 
swer. Some manufacturers said a home 
computer could balance a checkbook, 
although a $15 calculator could do the 
same thing. The personal computer 
seems to be a spectacular toy in search 
of a use. 

“Sometimes it is difficult/to explain 
to somebody what they need it for just 
as it would have been difficult to ex¬ 
plain to someone in 1850 what they 
needed an automobile for,” said Dave 
Armitage, president of Computer 
Power Inc. of Warwick, R.I., who was 
demonstrating a Sol terminal computer. 
“The uses of a machine like this are 
limited only to the user’s imagination.” 

At another booth, Steve Jobs was 
demonstrating the Apple II computer, 
which is the size of a portable type¬ 
writer and hooks up to a regular televi¬ 
sion set. It plays games', displays color 
graphics or does sophisticated mathe¬ 
matics. Mr. Jobs suggested that ama¬ 
teur radio operators could use the 
$1,300 device to figure frequency skips 
and that investors could use it to ch 
and that investors could use it to chart 
stock prices or do commodity spreads. 

A Keyboard and Power Chip 

But, Mr. Jobs agreed, “most people 
are buying computers not to do some¬ 
thing practical but to find out about 
computers. Jt will be a consumer prod¬ 
uct, but it isn’t now. The programs are¬ 
n’t here yet.” 

On the outside, all of the personal 
computers have a keyboard, much like 


those of a typewriter, and a-display 
screen or a capability for attaching to 
one. Inside, their key ^element is a 
micro-processor chip that is a computer 
on a piece of silicon a quarter-inch 
square. The chip has the power of the 
original Eniac computer of 1949, which 
took up a city block. 

Three years ago, there wfere no 
microcomputers. People who wanted 
computers at home had to rent or buy. 
a terminal and plug in to a large com¬ 
puter somewhere, frequently at a uni*? 
versity, where they would be one of 
a hundred other users in a time-sharing 
network. 

“Compared to time- sharing, the 
microc omputer is very, very powerful,” 
said^Gordon Stitt- who was demon¬ 
strating the IMS AT system, 10,000 of 
which have been sold in the last year 
and a half. 

The only thing that big computers 
do better than microcomputers is a lot 
of arithmetic. By computer standards, 
microprocessors are slow at math, 
capable of doing several hundred addi¬ 
tions a second as against hundreds of 
thousands by the big ones. 

Small Business Market 

The personal computer industry, 
which is estimated to total $30 million 
a year (up from $5 million two years 
ago) thinks its biggest market today 
is not the home hobbyist but thesmall 
business with sales of $250,000 a year 
or less. Up to now, computer time was 
too expensive for such enterprises. 

“It used to be hobbyists,” said Alan 
Hald, own&r- of the Byte Shop of 
Tempe, Ariz., which calls itself the af¬ 
fordable computer store. “Now it is 
more personal business users.” 

Ervin Fraser, a data processing man¬ 
ager in Boston, said became to the 
show today to see what innovations 
there had been. “You go away for three 
months and come back, you find things 
you never heard of before,” he said. 

A 19-year-old junior at the Massa¬ 
chusetts Institute of Technology, Man¬ 
uel Ulloa, said he planned to buy a 
personal computer because, “you can 
take it in your room and turn the lights 
out.” 

“If you invert a matrix on a big com¬ 
puter, that’s nothing,” Mr. Ulloa said. 
“But if you add two plus two on your 
own computer, that’s something.” 


Reprinted with permission of The New York Times, 
August 26, 1977, Copyright 1977.