3.3_May_1996 3.3 May 1996
3.3_May_1996 3.3_May_1996
User Manual: 3.3_May_1996
Open the PDF directly: View PDF 
.
Page Count: 62
| Download | |
| Open PDF In Browser | View PDF |
The
Analytical Engine
JOURNAL OF THE COMPUTER HISTORY ASSOCIATION OF CALIFORNIA
Volume 3.3
Maiga Wolf, Bookkeeper and Programmer
May 1996
May 1996
The
Volume 3.3
Analytical Engine
JOURNAL OF THE COMPUTER HISTORY ASSOCIATION OF CALIFORNIA
Editorial: THREE YEARS
On April nineteenth, with only the most minor
ceremony, the Computer History Association of
California observed its third anniversary. We look
back on three years - and especially on the most
recent year - of substantial accomplishment and
profound satisfaction. This is a personal, and necessarily abbreviated, report on the current state of
the CHAC's projects and prospects.
Publication: You're reading the tenth issue of the
ANALYTICAL ENGINE, and we look forward
- with you, we hope - to many more. Its present
situation has many positive aspects: Good material
is coming in unsolicited. We have several prospective subjects for interviews, with at least the next
four solidly planned; preservation of oral history is
and will continue to be the ENGINE's single most
important mission. Volunteers have begun to share
editorial responsibility. The ENGINE is poised to
become the first general-circulation trade magazine
of computer history in the United States.
On the other hand: This is ENGINE 3.3, which
says May 1996 on it and should have appeared
then. Since it's September as I write this, you
might reasonably expect to receive issue 4.1; so the
magazine has slipped by two full issues. Most of
the discrepancy has opened up in the last few
months. We, and you, want to get the ENGINE
back on track, but it's proving difficult.
Why? Because, in the three years since our
founding, computer history has attracted much
more general interest - so that the CHAC staff
finds itself handling hundreds of pieces of e-mail,
firing off faxes to far corners, giving lectures and
videotaping them, and trying to stay in control of a
bulging inventory of hardware, software and docs.
Just answering a single day's e-mail for the CHAC
can take two to three hours of one person's time.
We're delighted to be so popular. But somethingmaybe it's living in Silicon Valley - makes us
acutely aware that the CHAC may become a casualty of its own success. This organization has not
grown in step with the interest in what it does.
Later on we'll tell you about some new strategies
to keep the CHAC on top of its own situation.
What else haven't we done? We haven't managed
to publish much history from Southern California,
in spite of making inquiries for about a year. As we
inventory, we learn more about our SDS 930,
which was built in Santa Monica; but we've published very little about the ARPAnet, not much
about the fascinating computer culture that centered on UCLA, and nothing at all about computer
use in the aircraft factories, or the development of
the many experimental "one-off" mainframes. Attracting this material will continue to be a high
priority.
We've published nothing so far about military
computing, certainly a rich topic in California
above all. One or two articles now in preparation
may remedy this before the end of the year.
The sparse selection of powerful desktop publishing software for the MS-Windows platform is irritating. The ENGINE, now produced in Microsoft
Word, is about at the limit of what that very capable word processor will accomplish easily. Our
plans for ad layouts, more and better illustrations,
and color will soon make more advanced page
composition imperative.
The electronic edition of the ENGINE has slipped
badly, for reasons too gory to relate; profuse
apologies to those subscribers who are waiting for
it! 3.1, 3.2 and succeeding issues will appear in
electronic format as soon as we can produce Acrobat .PDF files. Andrew Eisner's generous gift of a
Macintosh SE/30 may help bridge this gap; for the
latest developments, pay an occasional visit to the
ENGINE's Web page, http://www.chac.org/
chengin.html.
Page 2
The Analytical Engine
Collection: Thanks to our newest collaborators
arid benefactors, the Perham Foundation of Los
Altos CA, the CHAC now has a modest amount
of storage space that will serve us well until the
Computer Institute/SFCM is established in San
Francisco - see p. 47. Accordingly, we are once
again acquiring software, docs, and hardware. Read
this issue's Acquisitions column for the juicy
details on, for example, our new HP 2114A.
Recently we have committed significant time,
energy and money to improving utilization of
storage. Edwin V. EI-Kareh, the CHAC's new
Tactical Director, is working energetically to
replace all cardboard cartons with lidded plastic
crates. These not only provide much better protection for artifacts, but they are stackable, meaning
that our storage areas can now be used literally to
the roof. Over the summer we'll also be scrounging steel shelving. Now that the founding of a
computer museum in the Bay Area is a likely prospect, your Association's role as a conservator has
been revitalized.
MUSEUM ACTIVITY
Closest to our hearts, of course, are the two computer museum projects now taking shape in the
San Francisco Bay Area. The Computer Institute
in Berkeley, headed by Fred Davis, Sylvia Paull,
and Andrew Eisner, plans a museum as one part of
a "global center of computer technology" in downtown San Francisco; see Fred's article on p. 47 for
the ambitious details. A second effort, the Computer Museum History Center, is being planned by
Gwen and Gordon Bell of the Computer Museum
in Boston MA; we are promised an update on this
project for ENGINE 3.4.
Personally, we hope that these two projects find a
way to converge if not actually combine. As Erwin
Tomash has wisely suggested, "Of all ways to teach
history, a museum is the most expensive;" and philanthropy in California, as in most places, has long
passed the lighthearted grip-and-grin stage. The Bay
Area's computer museum, to justify its substantial
cost, must offer commensurate value - both perceived and real - in education and in entertainment. Building such a museum, and even making
the best case for it, will require every scrap of energy that all the interested parties can summon up
together.
May 1996
Heinz Nixdorf Museumsforum
Energy, meanwhile, seems abundant at the new
Heinz Nixdorf Museumsforum in Paderborn,
Germany. Curators Ulf Hashagen and Dr. Karlheinz Wiegmann are building on enviable foundations - a realistic budget, a deep fund of
knowledge, and serious floor space in an existing
multistory building. These assets, together with
their obvious dedication and the luck they richly
deserve, will soon produce one of the finest computer museums in Europe. See page 46.
Australian Computer Museum Society
The Australian Computer Museum Society has
circulated a Site Planning report and proposal for
its long-contemplated facility. The authors come to
many of the same conclusions as has the CHAC:
A considerable amount ofcapital and resources, including volunteer labour, would be necessary to estab·
lish and maintain a Computer Museum to attract a
significant public interest.... Because of the planning
time for major developments... .[tlhree to five years
would be good going, but the 2000 Olympic year is a
good target. If established as part ofa major development this strategy will secure us a good site in a prime
location.... We need, in the immediate future, a site
that we can use as a workshop to actively engage in
restoration and development work. We also need
storage space, preferably integral with the workshop.
[ACMS Newsletter #10, April 1996.]
We believe that, like the CHAC, the ACMS is destined to "ride the wave" of public interest to a resounding success. The recent, well-publicized
purchase of a module of Babbage's Difference
Engine by the Power House Museum in Sydney is
a clear indication that computer history has
become a newsworthy topic in Australia. In the
last year the ACMS has raised $Aus10,OOO through
a direct-mail campaign and begun site negotiations
with the Australian Technology Park in the Eveleigh workshops at Redfern, Sydney.
On the other hand, ACMS President Graeme Philipson laments "two major problems ....the continued inability of the Society to find a suitable storage area .... [and] lack of time." Ain't it the same the
whole world over!
May 1996
The Analytical Engine
Page 3
IN MEMORIAM:
DAVID PACKARD
had to be deferred when Packard received a job
offer from General Electric - a prized commodity
in the depths of the Depression.
David Packard, who co-founded one of the world's
most highly regarded businesses and personally
guided it to undisputed leadership in its field, died
in Palo Alto, CA, USA on March 26, 1996. He was
eighty-three.
General Electric, however, was not poised to take
best advantage of Packard's talent. The company
had no interest in electronics, and assigned him to
"uninteresting" work in refrigeration. He quickly
wangled his way into other positions, and found
rewarding work troubleshooting GE's production
of vacuum tubes. The yield of one particular mercury-vapor rectifier, a large and expensive component, was disastrously low. Packard responded by
spending "most of [his] time on the factory floor,"
working closely with line employees to pinpoint
and remedy causes of tube failure. The effort was a
resounding success and Packard's first application
of "management by walking around," which he
soon recognized as key to any kind of technical
manufacturing.
Packard managed the Hewlett-Packard Corporation in a deeply personal and proactive way from
its founding in 1939 to his retirement in 1993. In
the process he set an example for technically adept,
skill-conscious companies throughout the world;
he also helped build a company which, in 1995,
had over 100,000 employees and annual revenue of
$US31.5 billion. HP is routinely cited for its enlightened treatment of personnel, its innovative
management, and its unrelenting pursuit of quality,
all principles which Packard believed fundamental
to corporate survival and growth.
Packard was born in Pueblo, CO, USA, and as a
boy was interested in electricity, general science,
and particularly radio engineering, building a
"fairly sophisticated vacuum-tube receiver"! when
he was twelve and becoming the secretary of his
local radio club in high school. At sixteen he operated his own ham radio station, 9DRV. His other
favorite pursuit was riding, a love and skill that he
kept up almost to the end of his life.
In 1937 Packard briefly visited Palo Alto and held
the first recorded business meeting of his
"proposed business venture" with Bill Hewlett. But
his employment at GE was so important, at a time
when jobs were still brutally scarce, that he returned to upstate New York immediately; and his
bride-to-be, Lucile Salter, traveled there by train (a
four-day journey) so that the two could marry and
have their honeymoon in barely more than a
weekend! Packard would not formally resign from
GE for years to come.
In the fall of 1930 Packard enrolled at Stanford
University, and met his future business partner and
lifelong friend, Bill Hewlett. Packard, Hewlett, Ed
Porter, and Barney Oliver were all encouragedprimarily by the legendary EE professor Fred
Terman - to pursue careers in electronic engineering, a field then so new that it was assumed to be
an ill-defined subset of "radio." Packard also distinguished himself as an athlete, setting rec()rds in
track and earning varsity letters in football and
basketball.
In the meantime, the Packards were at least able to
settle in Palo Alto, thanks to a Stanford fellowship
and stipend arranged by Terman. Packard worked
with other electronic innovators, including the
Varian brothers and Charlie Litton, on advanced
development including vacuum-tube engineering, a
field still of considerable interest to GE. But in
1939 he and Hewlett were at last able to start their
"venture" in the famous garage on Addison Avenue, which would be declared a California Historic
Landmark fifty years later.
Terman's four star pupils, in the spring of 1934,
determined to start "something on their own"
using their cutting-edge technical training, which
Terman had carefully reinforced with hands-on
experience. But the founding of the new enterprise
Some of Hewlett and Packard's earliest projects, in .
the search for income, ranged far afield even for H
and P - a harmonica tuner, a telescope tracking
controller, even a foul-line spotter for a bowling
alley. But the first "Hewlett-Packard product" was
an audio oscillator, well-designed and priced far
below its competition.... actually, and accidentally,
priced below cost. Packard christened it the Model
200A to give the fledgling company an air of expe-
1
The HP Way, David Packard, HarperCollins, 1995,
p.7.
Page 4
The Analytical Engine
rience. Within months the oscillator had been improved, and the resulting Model·200B sold well at
the (corrected and profitable) price of $71.50 each!
Eight 200B's were sold to the Walt Disney Companyforuse on the movie Fantasia. By the end of
1939, HP's first full year in business, the company
recorded a respectable profit and moved to larger
quarters on Page Mill Road.
World War II, justly called the "electronic war,"
saw HP expand its facilities again and put line
workers on double shift. A perpetually over-extended Government was grateful for product quality combined with timely delivery; HP became one
of only three California companies to earn the
coveted Army-Navy "E" award. Although wages
were frozen in wartime, HP's pre-existing bonus
plan - coupled with remarkable gains in productivity - made it possible to pay bonuses of as
much as 85 per cent of base wage. Tangible recognition for improved productivity became another
cardinal principle of HP management, one that
accorded with Packard's deeply philanthropic
nature.
When peace came and controls on employment
were loosened, Packard gave his corporate strategy
unfettered expression. Any company's greatest
asset, he insisted, was a highly skilled, highly motivated, and meticulously trained employee. The
company, therefore, had three primary obligations;
the first was to hire top talent, the second was to
assure congenial and secure conditions of employment and the third was to maintain channels
thro~gh which any employee could easily contribute to the company's success.
These strategic goals were put in place through
tactics that were innovative and even startling for
their time. Elaborate provision for the welfare of a
company's workforce per se was nothing new;
ffiM, to take a well-known example, was renowned
for the scope of support it offered to its employees.
But whereas ffiM intended that the individual
worker should draw strength from a highly prescribed and strongly hierarchical corporate structure, Packard took the opposite tack, trying to restrict the "company" to no more than would serve
to bind its employees into a productive relatio?ship. Designed for agility and constantly prunmg
anything superfluous, Hewlett and Packard's HP
was a company created by engineers for engineers
May 1996
- and it worked; the strength of the company was
the strength of its people.
Every effort was made to keep the company's
structure "flat" and lines of communication short,
so that executives with decision-making power
would have ready access to the company's fund of
ideas. Coffee talks, company picnics and barbecues,
and open offices all contributed to the exchange of
information; at the same time, a support structure
that included flexible scheduling, tuition sharing,
and catastrophic medical coverage made it clear
that HP employees were a valued resource.
The HP Way, as a corporate culture, just about
wrote the prototype specification for Silicon
Valley's high-tech business. Perhaps this model
now goes too far at times, creating personalities
who are better-known than the companies for
which they work. But when we look at HP its~lf,
it's clear that the model worked to near-perfectlOn
at least once. Si monumentum requiris.... look
around and you'll probably see the HP logo!
David Packard would have been famous enough,
certainly, if we had only HP itself to remember
him by. He himself preferred to be remembered.
for his great philanthropic endeavors - the DaVId
and Lucile Packard Foundation, now the largest
charitable private foundation in the United States;
the Monterey Bay Aquarium, which has attracted
millions of visitors to the Central California coast;
Lucile Salter Packard Children's Hospital, one of
the world's most highly regarded pediatric facilities; and visionary projects with a worldwide reach,
such as the Center for the Future of Children and
the Packard Humanities Institute. He was more
than merely a generous man, more than a captain
of industry, and more than a statesman. He was, at
last someone who understood the supreme good
fo~une of his lot in life, and worked hard to give
as he had received.
The Computer History Association of California
extends condolence to Mr. Packard's children,
David W. Packard, Nancy Ann Packard Burnett,
Susan Packard Orr, and Julie E. Packard, and to his
many colleagues and friends throughout the world.
May 1996
The Analytical Engine
Page 5
The LOMAC ADAM (1977) with 50Mbit Control Data hard disk raised for service, terminal, and
Centronics dot-matrix printer.
THE COMPUTER WILL DO
ANYTHING YOU TELL IT TO DO:
An Interview with Maiga Wolf
HLC: See, in accounting, the "assets equals liabilities plus equity" -- that was from about 1300. So,
from 1300 to 1967, huge companies -- international
companies -- functioned without computers in
their accounting departments. Then suddenly,
from 1967 to now, nobody can even write a check
without a computer. What I'm saying here is that
the computer has made a tremendous difference in
everyone's life -- but it seems to me that you
techies think you're the only ones who really appreciate what an incredible breakthrough computers are. I want you to understand how significant it is for the rest of us. I think you should explore how computing changed the life of an ordinary bean-counter.
KC: On January 19th we got a call from Maiga
Wolf. She had an ADAM minicomputer that had
been built by Logical Machines Company in
Sunnyvale, and she was retiring from twenty-plus
years of running her own bookkeeping business;
would the CHAC be interested? The software, oh,
easy! She wrote it herself.
So Hilary Crosby, a certified public accountant;
Edwin El-Kareh, an engineer and CHAC volunteer; George Durfey, an engineer and photographer, and I rose to this challenge.
SYSTEMS MATTER
EVEK: You worked in a large company in the
accounts payable department?
MW: I was brought in to revise their accounts
payable. They had a big problem. They couldn't
pay their bills in time, they couldn't get discounts
in time, and it was one big mess. This was in 1963.
HLC: And they had no automation in the
counts payable?
ac~
Page 6
The Analytical Engine
MW: No, everything was manual. I had 29 people
in accounts payable, processing bills for Ampex. I
went in and revised the procedure, how they were
handling [paperflow.] In the first year I managed to
recapture $275,000 of discounts. In those days that
was a lot of money. [To begin with] we didn't get
all the discounts. It took me a while to get there,
but by the end of the first year we had saved that
much.
In Canada originally I was working for a company,
but on the side I was trouble-shooting. The other
chartered accountants would come to me and say,
"Oh, one of my clients has so many problems, can
you come over and straighten them out? We don't
have time." So I'd work at that site on Saturday or
Sunday. Finally I had so many waiting for me that
I realized I was earning more doing consulting than
from my regular job. So I quit, and began arranging systems to process paperwork, full-time.
HLC: Were you using bookkeeping machines at
that point?
MW: Not bookkeeping machines. There were
Burroughs posting machines and calculators.
May 1996
HLC: Did you have your accounts receivable integrated with your sales and the manual system?
MW: No, no. I had separate girls doing separate
jobs. One girl would do the invoicing, and the
other girl would do the recording in the sales
journal, and another girl would take the invoices
and post to the receivables.
HLC: So each invoice got posted twice - once in
the sales journal, and once as accounts receivable to
the specific customer?
MW: Right, manually. Then in 1967 or '68, I don't
recall the exact date, when we went on the computer at Precision Instruments, it went from the
sales journal right into the accounts receivable.
They had the payroll in a bank, and that was also
computerized, but the bank had lots of problems
there too.
HLC: So, by the late sixties, all the tax withheld
and the deposits and the quarterly reporting was
handled by the bank on a computer, and people
just got their paycheck, and at the end of the year
their W-2?
MW: Well, I'm not a certified public accountant in California. I studied
in Canada to be a chartered accountant, and I went into accounting for
one reason. I was an immigrant after the Second World War and I
could see that I wasn't going to do anything with the positions that
were available for young women in '48, '49 -- I didn't want to be a
nurse, I didn't want to be an office clerk or a teacher -- that was about
my choice, and it just didn't appeal to me. So I decided there was a
little bit more leeway if I went into accounting, and so I studied in
london, Ontario.
First you got your four years at the university. Then you had to be an
apprentice with a chartered accountants' firm, the equivalent of a CPA
firm, for five years, and then you wrote your exam and got your papers.
When I moved to California, I wanted a job to start with, and Ampex
hired me as a consultant in their systems and procedures -- they had
problems, especially in accounts payable. I went in there and started to
work, and things just rolled along, and I never got around to doing my
CPA because nobody asked for it. [laughs.]
Maiga in 1947.
HlC: What about tax returns?
MW: I never did tax returns. After I quit Ampex I wanted to see what the tax laws were in California, so I hired
myself in with a small CPA firm. And I just happened to be there between January and April, with the tax
returns and stuff, and I said, "That's a rat race, I don't want that." And that's why I stuck to this [bookkeeping]
so I never had to bother with taxes. Most of the small businesses also have personal stuff involved, and I don't
want to know anything about the personal.
May 1996
The Analytical Engine
MW: I didn't have to worry about those. That
belonged actually in personnel, so I didn't have
much to do with the payroll at all, except to take
the feeders [reports] and put all that in the general
ledger.
HLC: When was the first time that you encountered the computer?
MW: When I was a controller at PI [Precision
Instruments] in Palo Alto. They decided to install a
computer, one of the great big ones with punch
cards, and my accounting department posted in
through the computer.
KC: Was this an early minicomputer or a
mainframe proper?
MW: It was a mainframe, I think a big Burroughs.
I wasn't familiar with computers at all, so whatever
the company brought in, that was it - I had no say
as to what type. All the departments were
throwing stuff at the programmer, and expected
him to have answers for whatever came into their
heads.
KC: Meaning that the department heads expected
the programmer to know their business and
requirements.
MW: Right, and it didn't work, of course.
KC: Never does.
MW: And so I was looking at what they were
doing, and I realized that nobody told the
programmer what they actually needed [in output]
from that machine. The engineering, and
marketing, and other departments threw
everything at the poor programmer and said "Do
it." And it would come out wrong, or not work at
all, and they had lots of problems. So I decided,
that's foolish, this poor fellow is really not an engineer, or marketing manager, or whatever else
they expect him to be. When my turn came to put
the accounting in, I broke it into separate phases. I
put in - let's say - sales first, and got the sales
journal, then I put in cash receipts, and so on. By
working with one phase after the other, we
brought it off with only slight adjustments.
Everybody [else] was hollering "How come you
don't have anything wrong when we all have
problems?" and they were cussing the programmer!
And I said, well, you want him to bite it off and
chew it up for you, and he doesn't know what you
need. You have to tell him what you need.
Page 7
I worked for Precision Instruments something like
6 or 7 years. That was my start with computers,
and the only experience I had with computers,
until I bought my own.
ALONE, AT LAST
KC: And you went out to do more systems
installation?
MW: No, I began doing bookkeeping for
companies that were too small to keep an in-house
accountant or a full-fledged bookkeeper. I got into
that in a strange way - maybe not so strange,
considering what had happened in Canada.
Someone I had known at PI, not from my own
department, called me up and said "I have two
good friends who want to buy into a business - "
which was a bar and a restaurant in Palo Alto "but the fellow hasn't got any books. Is there any
way you can help?" I'd just quit PI and was sitting
at home, thinking about what to do next; so I said
that 1'd take a look.
I had no equipment at home. My neighbor, who
had a typewriter [that was] from her club, said "I'll
let you use my typewriter." My insurance guy had
just bought a new calculator, so he gave me his old
calculator. At first I wasn't going to buy anything,
because I didn't think I would continue that way. I
set up all the books for the restaurant. It was early
in the year, something like March or May, so it
wasn't too hard to catch up. I took the 1120
[corporate tax return] balances forward and set up
the new year. The restaurant looked like it was
making money, and I told them, "That's what I
can get out of the paperwork." Then, of course,
this fellow who had the bar had a friend who was
in trouble, and.... [Laughter.] It just mushroomed,
so I got an adding machine and a typewriter of my
own, and off I went.
HLC: What made you decide that in order to have
the volume, you needed a computer - what put
that together for you? How did you know that to
service the number of bookkeeping clients that
could give you an income, you needed a computer?
MW: I was on my own and I needed more income.
Right away, I could see that I wasn't getting very
far manually. I was spending hours and hours
adding back and forth across all those pages in the
ledgers and journals, and it was very timeconsummg.
The Analytical Engine
Page 8
May 1996
HLC: There weren't too many women in accounting -- or in computers -- at that time.
MW: Well there weren't too many [women] controllers at that time either. I was a controller for Precision
Instrument.
HLC: Did you think
people there
resented you, a
woman, having that
job?
MW: Oh yeah, very
much. The vice
presidents from
different sections
would come - from
marketing especially
- and say "That girl
in the office there,
she wouldn't sign
that .... " Conrad
Schoebel backed me
up very well. He let
them know, if Maiga
said no, it meant no.
HLC: When you
went to that
convention in
Pasadena, where you
were looking at computers, how many
other women were
there as customers?
Half of the people?
Looking at the books.
MW: No, very few, very few.
HLC: Did it feel strange, walking around there as a customer and a woman?
MW: No, because I've always been working in a man's world. In 1953, accountants were not women, and so
I've been looked at - you know - as an oddball.
HLC: The guy from LOMAC that came to help you, the guy MW: Gary Kench, was very helpful [and] very nice.
HLC: Did he ever say anything like ''You're good at this for a woman ...."
MW: Usually he said something like "How's it going," and I would say "Well, it seems to be working," and he
might say "I knew you would do it." That was about it.
KC: He was the sales guy for the company, right?
MW: Well, he was the salesman for this computer. What else he did for the company I have no idea.
May 1996
The Analytical Engine
Just at that time I read an announcement in the San
Jose paper of the first Business Computer
Conference, in Pasadena, with displays and the
whole business. I got on the plane and went there,
spent five days walking from place to place, from
one computer to the next, and attending any
lecture that was given. I came to the conclusion
that there wasn't one program that would suit me!
And I couldn't afford an in-house programmer,
who at that time would have charged around $40
an hour - not to mention that I'd probably have
to wait two weeks before a freelancer would come,
and I couldn't spare that kind of time either. Then
I discovered this ADAM, and they assured me that
I could program it myself, so I bought it.
KC: Those assurances have been given since there
have been computers, and they haven't always
been true, but in your case it was true.
MW: I had a good start. Gary Kench, the salesman,
said he would come and get me started. In the
beginning, he spent a couple of hours with me
every day. Meanwhile I was studying the manual
and trying to do the homework he gave me. Then
the next time he came, if my work was okay, we
just progressed from there. After about three or
four weeks he said "You're on your own," and I
took it from there.
KC: It's important to point out that this machine
is not a micro in any sense. This is a minicomputer
with a fourteen-inch disk and a terminal, and it's
built into a desk along with a [dot-matrix] printer.
If you don't mind my asking, how much did the
whole shebang cost when you brought it home
from LOMAC?
MW: Thirty-two thousand dollars.
KC: In 1977.
MW: Yes.
HLC: You had to keep a lot of books to payoff
$32,000 in 1977!
MW: It paid off very easily. As far as money went,
it was no problem. But it was a big risk for me in
the sense that I had no idea if I could handle the
programming or not. There was no other way, so I
took a chance and said that if it didn't work, it
didn't work.
KC: And you had to buy one disk pack per person.
Page 9
MW: Each customer had two [disk packs], one for
backup.
HLC: So you kept a back-up in the same box?
MW: Yeah, they were labeled, number 1, number
2. I would take a back-up after every two hours'
work just in case.
HLC: And how long did it take to do a back-up?
MW: Just a few minutes, I just wanted to copy the
information from the built in disk.
HLC: How much were you paying for those disk
packs?
MW: I have to think, now. At first I had only the
two disks that came with the computer. Then as I
started to work with separate accounts I needed
more disks, and Gary offered me 20 disks at
something like $175.00 each. In my very first year I
bought all those disks, because I knew I was going
to enlarge.
KC: Did that also limit the number of clients you
could take at one time?
MW: No, I could get more disks after that.
HLC: Are they still available?
MW: Yes, come to think, one of the computer
supply catalogues had these large disks, still. Maybe
not the same make - these are Control Data - but
they had some hard disk [packs] still for sale.
KC: Just to recap: You had to computerize in
order to get enough volume. You didn't find any
ready-made software that would meet your needs,
and you knew that you couldn't afford to invest
time and money in custom programming from
someone else; so you decided to do your own
programming. Now, you were programming in
BASIC?
MW: Well, in what LOMAC called English. I'm
no expert on computer languages - I don't know
anything about any other computer except this
one, so I have no idea. They told me it could be
programmed in English, and that's what I was
doing.
HLC: So is your program actually a database
where information is saved centrally, to be accessed
from different perspectives and combined into
reports?
I3Et'--lERAL
PERIOD FROM
o
ACCOUNT HO.
PAR TIC U L A R S
1.1.0
CASH
111
202
2CR
JULY
I
t~
LEt)13ER
1, 1995
PRIOR BALANCE
CURRENT
DEPOSIT
•
13Et--lERAL
PER I OD FROl1
525
3CK
3CK
3CK
3CK
3CK
5GJ
-140
451
528
530
559
893
PAP TIC U LAP S
INSURAHCE-HEALTH
CK.a 553
7.12.95
CK.a 673
7.12.95
CK.a 15184
7.12.95
CK.a 15185
7.12.95
CK.a 15201
7.24.95
DEP.a 995
7.03.95
JULY
GROUP
2
3
4
5
6
7
8
9
~IOHTH
Yo
T.
[I.
TOTAL
BALAIKE
$9,224.00
JULY
31, 1 '395
CURRHIT 110NTH
Y.
T. D. BALAt-ICE
$218.90
$354.91
$1,381. 90
$5,459.53
$123.22
-$424.00
$7,124.55
$7,124.56
RECAP
PR
is
do
and
a.nd
~T
and
TO
$0.00
MOVE
GROUP NO
STANDBY
MOVE
HEAD
GROUP NO
IF
GROUP COUNT
<
2
GO TO
NO TOTALS PRINT
PLF SLC (NPG)
and
a.nd
1,19'35
CGLIC
KAISER FOUNDATIOH
PACIFICARE
EMPLOYERS CHOICE
SAFEGUARD HEALTH PLANS
THE EClUITABLE
RUAF
1 Doe:!'
31, 1995
LEDI3ER
PRIOR BALAt-iCE
ACCOUNT TOTALS
GL
JULY
BANK:
BAY AREA BANK a1537701
DEP.a 995
7.03.95
ACCOUtH t,o.
TO
PSP
a.nd PRINT
and PSP
15
aOlj
PRINT
10 and PSP
15
11
and PRINT
12 and PLF SLC !lPG)
13 and PSP
14 and ~IOIN
STANDBY
o
15 anlj PRINT
3
HEAD
1t:. and PRINT
" TOTALS"
17 and PSP
16
18 and PRS 15_ 2
PM TOTAL
19 and PSP
12
20
and PR$ 1~ 2
MTD TOTAL
21
and PSP
12
22 and ADD
PM TOTAL
MTD TOTAL
23 and PRS 15_ 2
SUM
24 and LABEL
NO TOTALS PRINT
25 arId PLF SLC (NPGl
26 and PLF SLC (NPG)
27 and MOVE
o
PM TOTAL
28 and MOVE
o
MTD TOTAL
29 and MOVE
o
GROUP COUNT
11 _ _ _ _ _ _ _ _ _ _ _ _ I'
II _ _ _ _ _ _ _ _ _ _ _ _ II
II _ _ _ _ _ _ _ _ _ _ _ _ II
..... e r b .
GL
HDG
1 D(. e s
2 and
3 and
4 and
5 and
6 and
7 and
8 and
9
anlj
10 and_
11
and
12 and
13 and
14 and
15 and
16 and
17 and
18 and
19 and
20 and
21
and
22
arId
23
and
and
and
and
and
and
and
and
24
25
26
27
28
29
30
is
a
..... e r b .
MOVE
HEAD
~;AVE HEAD
TITLE PR
PSP
26
PXP
PRINT
"GENERAL LEDGER"
PLF MLC
2
PSP
37
MOVE
FISCAL YEAR BEGIN
PAD
6
CUT
HEAD
4
JOIN
TAIL
HEAD
MOVE
HEAD
YEAR BEGIN
""""'T
HEAD
"~
PRINT
"PERIOD FROM
DATE PR
YEAR BEGIN
PSP
4
PRINT
"TO
DATE PR
END OF MONTH
PLF~.!-.f-----a--p~i' :;,
...
PRINT
"ACCOUNT NO_" . ,
PSP
6
PRINT
"P A R TIC U L A R Sol
PSP
10
PRINT
"PRIOR BALANCE"
PSP
14
PRINT
"CURRENT MONTH"
PSP
14
PRINT
"Y_ T_ D_ BALANCE"
PLF MLC
2
MOVE
SAVE HEAD
HEAD
e
CASH
TRt.
201
202
mL
2CR
2CR
RECEIPTS
ACCTII
121
111
CHECK
TRN
460
461
JNL
3CK
3CK
BATCH
Ace 0 U N T
ACCOUNTS PECEIV. - TRADE
BAY AREA BANK 111637701
REGISTER
ACCTII
111
526
TUL ',-'
REFEREtKE
DEP. II '395
DEP. II '395
•
BATCH
AC C0 UNT
BAY AREA BANK 111637701
INSURANCE-HEALTH
DATE
7.03. '35
7. 03. 95
r]
fY7
p
~
R T
1 (
1'395
U L ~ ~ 5
DATE
12. '35
7.12.'35
·f.
RECAP
.JNL
1
2
~:
4
5
6
7
E:
9
10
11
AMOUNl
-$354.91 •
$354.91
HDG
i : s ..:L ......... e r t:.,
SAVE HEA[I
[Ioes MOVE
HEAD
and TITLE PR
and F'XP
and F'SP
H.
~INL [IESC
and PRINT
and PSP
18
and PRINT
TD MONTH NAME
19"
and PRINT
HI YEAR
and PRINT
and PlF MLC
3
and TRANS HDG
HEA[I
ao.j MOVE
SAVE HEAD
II
12
RECAP
JNL
LIST F'R
i s oS ...... et'"
PAGE HM
"JNL liST"
JOURNAL~;
BEGIN
and lABEL
1 Does
2 an,j
=
general journal, I put any entries that I used to
make corrections. Once I had the journals and
the batch was done, I would make out the
general led er, which is here.
$9,224.00
1 995
P R R TIC U L A R 5
KAISER FOUNDATION
KAISER FOUNDATION·
RECAP
DELETE BATCH
BATCH ENTRY IS a verb.
'NO'
21 and DLF
1 Does IF
BATCH COMPLETED
22 and OlF
do GO TO
INCOMPLETE
23 and OlF MG 38
"~O YOU WANT TO DELETE THIS BATCH?
2 and JNL SELECTION-BATCH
24 and INPUT
ANSWER
'X'
3 and IF
JNL NO
25 and IF
ANSWER
"Y"
do CONTINUE
d. GO TO
INCOMPLETE
JNL NO
TRANS JNL N
4 and MOVE
26 and BATCH DELETION
o
BATCH TOTAL
5 and MOVE
. 27 and MOVE
BATCH START NO
NEXT TRANS NO
6 and LABEL
28 and SAVE
THIS
COMPANY
ADD TO BATCH
29 and MOVE
'YES'
BATCH COMPLETED
7 and TRANS ENTRY
30 and CONTINUE
8 and LABEL
31 and lABEL
BATCH TOTAL PRINT
CLOSE BATCH
9 and BATCH TOTAL PR
32 and TRANS SAVE
10 and CS
33 and MOVE
NEXT TRANS NO
NEXT TRANS NO
11 and LABEL
34 and SAVE
THIS
COMPANY
INCOMPLETE
35 and MOVE
NEXT TRRNS NO
BATCH ";TA~~T NO
12 and BATCH END·OPTIONS·
36 and MOVE
"YES'
BATCH COMPLETED
BE OPTION
=
13 and IF
37 and DLF
do TRANS CORRECTIONS
38 and DLF
2
BE OPTION
14 and Ii
do TRANS DElETI ONS
3
BE OPTION
=
15 and IF
ADD TO BATCH
do GO TO
4
BE OFT ION
16 and IF
There are journals, see, these are batches that I
DELETE BATCH
do GO TO
enter. After each batch I would print out the
5
BE OPTION
=
17 and IF
do BATCH PRINT
journal and screen it, to make sure that
6
BE OPTION
1B and IF
CLOSE BATCH
do GO TO
something wasn't in the wrong account BATCH TOTAL PRINT
19 and GO TO
sometimes your fingers slip, you know. In the
20 and LABEL
.
-~fjL:?OO. ~)()
~c;
JUL "('
REFERENCE
,
CK.II 673
CK.II 673
~1'IC1UtIT
NAt.H(,EI·IEtIT, It,,:.
DEPOSIT
3
LOOP
and
and
and
and
e and
9 and
10 and
11 and
12 and
13 and
4
!:,
6
7
GET
PSP
PXP
PRINT
PSP
PRINT
PSP
PRINT
PLF MLC
GO TO
NEXT
28', -
JOURNALS
"
REF
2
!I_II
2
JNl DESC
2
LOOP
Et'·'lPLD",,.·EE
PAYROLL
LOU I
-
N
DBA
• r
F'E '::OF<~D
t I'
A U Ci U ':::, T
1 '3'::3 '5
~3
SUNtf,")ALE, CR '34087
6. S.
E-'j878
PAY RRTE: 56.50/HR .:os OF: 3.02. '35
DATE t-'IPELt:3.02.'35
~,
TRN
PRL
CK.liO.
~'HrE
4326
8. 15.'35
3000
'338
8.31.
3151
CD
'jS
110NTH TOTRLS
QURRTER T01ALS
YEf.!R T01ALS
GROSS PRY
F. I. T.
5193.'31
5325.00
$0.00
-$33.00
----------$51a. '31
----------$518.31
----------$518. '31
-----------
-----------533.1)0
-----------$33.00
-----------$33.00
-----------
DEDUCTIONS: S-j
F.
I. C.
R.
-$12.02
-$20. 15
----------$32. 17
----------532.17
----------S32. 17
----------
M. C.
".
-$2.81
-$ .... ..
----------
-$7.52
----------$7.52
----------S7.52
----------
I.
10.
r,.
T.
S.
\)tj
-$1.'34
$177. :4
-S.3.25
'£261. 29
-S5. 19
$438.43
-$5.1'3
$438.43
-$5.
$438.
-12,60
----------s2.GO
----------$2.60
----------$2.60
----------
I.
--------
---------------
1'3
--------
WOLF: They claimed, for instance, that I could not have a total for that week, for the month, for the
quarter, and the year print out right then and there. And originally it looked to me as if I couldn'ttime and again I was running into a dead end, and I couldn't get this printed out fom the file. Then I
finally discovered what was wrong with it! You see these references here, the transaction numbers,
they were in the general file, and I was working out of that file. And I would say, go in the payroll
file, then take the next one, do this; take the next one, do this; take the next one, do this - but at the
last reference in the file [the routine] would jump out. It was finished. And I couldn't branch back
into the payroll program to print out these titles.
KIP: lt was a question of recapturing the numbers associated with those transactions.
WOLF: Yes, but right then~ [Laughs.) But I realized, first of all, that I had tran'saction numbers up
to 5 digits built in - and I could even have enlarged that if I needed to. Then you see that on this
sheet, my transaction numbers are ll,400-something, and that was in June. My transaction
numbers would never go beyond five digits, almost certainly; so I put [transaction] number 99,999 at'
the end of the file. Then, when I went into the file, I would say "Get next transaction; if it doesn't
equal 99,999 do this," over and over again, and then "If it equals 99,999, go into this program and
print it out."
KIP: Because you knew that 99,999 was the end of the file WOLF: - and that it would never be a transaction number, yes.
KIP: There's your subroutine, Edwin.
RECAP
PRL PRINT
1 Does
2 and
3 and
4 and
5 and
IS,a
verb,
DLF
DLF MG OS 18
DISPLR\'
"Nc. of COPIPS
COpy COUNT
INPUT
LRBEL
tiET
Fl'f(
~3
6
7
8
9
10
11
12
~-"""13
14
REPEAT
PAGE HOG
"PRL "
BEGIN
PRL TRANS
GET
~EXT
PRL TRRNS
MOVE
PRL DATE
WEEK
MOVE
PRL DATE
WEEK
PRL WK TOTALS 0
PRL WK TOTALS CRLC
PRL TRRNS PR
LABEL
and
and
and
and
and
and
and
and
and
P~'
LOOP
15
16
"
17
18
19
20
21
and
and
do
an d
do
and
and
and
and
GET
IF
GO TO
NEXT
PRL TRANS
PRL DATE
WEE
WEEK TOTRLS
IF
REF
"'39'399'
GO TO
END OF MONTH ~
PRL WK TOTALS CALC
~
PRL TRRNS PR
GO TO
LOOP
LABEL
A
IdEEK TOTALS
22 and MOVE
PRL DATE
WEEK
23 and MOVE
REF
STANDBY
24 and PRL WK. TOTALS PR
25 end PRL WK TOTALS 0
26 and MOVE
WEEK
WEEK PR
27 and BEGIN
STANDBY
PRL TRf
28 and GO TO
LOOP
2'3 and LRBEL
END OF MOfHH
30 and PLF SLC NPG
31 and PRL TOTALS CRLC
32 and PRL TOTALS PR
33 and SUBTRRCT
1
COpy COUN
34 and MOVE
DIF
COpy COUNT
35 and IF
COPy COUNT
<
do CONTINUE
36 and GO TO
REPERT
,::iEI.. jEPAL
LEr:.'I3EF"
.1,
PER I OD FROl1
H,:COllIH
P
rll],
20,
4205
4209
4212
4215
4217
4219
4222
4224
2CK
leI'.
2Cf
':K, II
20:
1
,
316"
CK.II 3165
,;1<.. II
3166
2(1<
CK.II 3167
2CI:
CK.II
BANK
U
L
..,
8. 31.
·~s
JNL NO .
LED G E R
DATE
B A LAN C E
OUTSTANDING ITEMS
3980
ISJ
8.29.95
!:.HLHIIi:!::
-'f.':'juS3.<~1I)
-t411).OI)
-$3U.50
-£t. l74. uO
-£,', :.8B. ::6
-.£"" 'jS2. 54
I
MII'E I
-
WEEKLY
DISCOUI·Wi
OF
AMERICA
REFERENCE
*
-$34'.:., I)U
-$46.54
-tE,:"'{.II.1
AUI3US T
RECONCILIATION
BANK
. TRN
BANI( OF H'1ER I CA
EMPun I'll' tiT DE',IEL(JP"I!::'·l [,EfoT.
REFUND/
P.
:~, CO.,ltK.
8.31. '35
3168
. i. l.
PRIOR 8ALHIKE
f'
8.31,:35
8,31. '35
8,31. '35
8,31. '35
8.31. 35
8.31. '35
II 3161
II 3162
':K.II 3163
'.K,
'.K,
2CI"
r
H P
'3':.15
DAILY REPORT
:1
9'35
~9275-03481
AMOUtH
PARTICULARS
$6, 1 '39. 20
-$1,336.64
DEPOSIT
$I, 174.66
$3,000.00
5'l1).00
HILARY: What about the bank reconciliation? Would this program allow you to clear the checks
when the bank statement came in, and get your outstanding check list?
WOLF: Yes, and this is the program for a selection - it's nearly English. For bank first
~
reconciliation, I would have the selection on the screen; if I wanted BR with an X - cleared - it
would CONTINUE and jump out of the program. If not, number one is from bank one to bank one
transactions, which just meant that it remained uncleared. Number two would go to transaction
deletion. Number three would let me make an adjustment, number four would print it.
HILARY: So that's how you get your clear, clear with error, error adjustment.
RECAP
BANK I PEC
I S
I [Joe,
2
lind
do
:3
.nd
do
4
and
5
6
7
"
BA~~K
do
and
do
"nd
do
.a.nd
IF
2CI(
8.31. 95
B A tl K B A LAN C E
CK.II 3168
I SELECT
BR I
':Otnlt~UE
=~
"X"
IF
BR I
1
BANK I TO BANk I TRANS
IF
BR I
BANK I TRANS DL
IF"
BR I
3
BANK I ADJ
IF
BR I
4
=
BANK I PRINT
REPEAT
-
$1,174,('0
$2,588.25
$4,952. ':.4
4224
l.'er-b,
$345.00
$45. ':.4
$25,»83.
,)1)
0
Page 14
The Analytical Engine
MW: I start with [the screen] called Menu. First of
all, I can change the date, if that was needed. If I
was working in the same month, the date was in
there already. And then I would go in and start
with a batch that I had.
HLC: So you put in a batch of invoices?
MW: So I put a batch in. Every line that went in
got printed out, so that if I needed to check back, I
would see anything wrong on the printout. Then,
when the batch was completed, the program took
over and re-arranged all the accounts and printed
out by accounts into the journal.
HLC: So if you had a stack of purchase orders,
payables, you could put in the name of the
company that you are buying from MW: I have five pieces of information that I put in:
the account number, the date, the reference to the
document, and then the description - vendor's
name or whatever - and the amount.
HLC: And then will your program write the check
to the vendor also?
May 1996
MW: No, you see, my programs were made for
once-a-month processing for the small company.
My customers did the daily work themselves; in
later years some of them even got computers to do
their daily chores. I did payroll for just one
company, but I didn't even bother writing a
program because it was so small - I calculated the
payroll manually, they processed on the computer
from the time cards, and I printed out the payroll
and the employee details. I made an exception for
this one company only because it was so small -about ten people. It didn't pay to computerize the
initial calculations. But I stored the payrolls and
the employee individual information all on the
computer [post facto].
KC: Now, how long did it take you to write all the
modules of this system?
MW: I think I worked about two months on the
very first basic system, or maybe a little more.
That [handled] the smallest account that I had.
When that functioned, I tackled the next bigger
account. To the very basics, I added whatever I
MW: I'm the kind of person who hangs onto things. Everything I have around me is old. I just sold my other
car, a Porsche I drove for 29 years [with] the same motor, and put 300,000 miles on it. For the first, about, ten
years it was the second car, so it never went on big trips .
. HLC: Which model?
MW: It was a 912; actually, my 912 was the first
one in California. When I was still married, the
Porsche manager for the West Coast used to live
across the street; when we went to Europe he
arranged for us to pick one up at the factory in
Stuttgart. But the 912 wasn't shipped to California
yet, so when mine came in, nobody here had seen
one. And I had trouble, because anywhere I parked
I had ten or fifteen people standing around it. But
cops were the worst problem! I don't know where
they came from, but they were right behind me.
And that's unnerving when you have a cop behind
you all the time.
EVEK: A lot of protection for us!
Maiga and her Porsche 912.
MW: A guy called Rolf took care of it from day one
until I sold it, and we overhauled the engine just once, and when I had another 120,000 miles on it [after that]
he said "Well, it's still good for another 100,000 miles." The motor was okay, but of course all the rubber
[gasketing] was getting old ....
May 1996
Page 15
The Analytical Engine
KC: Did you ever try to interest Logical
Machines in the programming you had
done, that they claimed couldn't be
done?
MW: It wasn't the company -- it was one
of the engineers from Lomac who had
gone on his own. He had taken an ADAM
or two and opened up an office where
people could rent time on them. And
when I said "I got the payroll the way I
want it," and showed him the print-out,
he asked me for the program. I said "Oh,
no way." He said, "Well, my clients could
use it," and my answer was "Then send
them to me."
Removing the disk pack.
KC: And did he send them to you, or did
he buy the program?
MW: I wouldn't sell the program. What he was willing to pay for the program probably wasn't very much, and
there were no royalties, so why should I bother? I felt that [the program] was something different, and that I
could give my customers something better. If everybody else started to have that, it wouldn't make any sense.
needed for the next account. Actually, I don't see
any big deal about it. I had the machine do what I
would do manually, followed my manual
procedures exactly and just programmed them in
there.
KC: But instead of working module by module,
you worked company by company, starting with
the smallest company?
MW: Right. So I got as far as printing out journals,
and printing out the general ledger, but I did not
work on receivables and payables and the payroll
until the last. When I started to work on [those
modules,] the people from LOMAC told me that
the kind of payroll printout I wanted couldn't be
done. I said to myself, there's got to be a way.
I struggled for some time. Then I discovered the
reason that they said it couldn't be done, and I
started to work around it. Then the payroll and
payables became very simple for me.
KC: Now how long was it from the time you
wrote the system for your smallest company, to
the time you felt that your program was fully
realized and capable of handling anything you
threw at it?
MW: About a year. I'm almost positive, not more
than that.
KC: And I believe you mentioned earlier that, for
example, by the time your system was finished
every transaction was validated on the fly, and the
system absolutely didn't tolerate rounding errors.
It only took a year from the beginning to get it to
that kind of sophistication?
MW: I had to rush - I needed that machine to
work forme.
SERVICE WITH A SMILE
HLC: How quickly did people get their reports?
What did they get in their reports?
MW: Depending on when they brought their daily
paperwork in, the turnaround would be about
three days. I screened [the data] after it was done
and proofed it myself. This was for ten customers.
KC: So you were giving 3-day turnaround of
reports for all the stuff to ten people? That was a
lot of work.
Page 16
The Analytical Engine
MW: It was! But that's why it paid off that
[computer] fast.
These are my copies, you see.
HLC: So you always had hard copy of everything,
and didn't keep it on the computer.
MW: No. [It would be] deleted out of the
computer. Otherwise there was too much garbage
in there.
HLC: What about the paperless office? [Laughter.]
MW: I found that if I had too much information in
there, it would slow down the processing. And I
needed the time. My customers had no access to
my computer [data,] so they had to look in their
printouts anyway.
HLC: So, from being a manual bookkeeper, you
went onto a computer, and suddenly something
would take too much time because it took a
minute, or five minutes. So rather than keep a lot
of data on your disks, like all these month-to-date
totals through the year MW: I would take time to transfer the information
out of current work into a file for storage, because
the computer had to work, and meanwhile I
couldn't do anything else. Many of my companies
had a lot of entries, between 500 and 1,500 for the
month. So it was worth the time to transfer from
the original file into storage.
HLC: How long would the computer take to close
the month?
MW: I never timed it. This company would take
the longest because they have the most entries. For
the others, it depended on [the number of] entries.
HLC: But how long would it take [the computer
to post the batch]? Was it short enough that you
sat at the console while it did it, or would you go
away from the computer for that time?
MW: Oh, I usually sat there. If I closed the month
on this company - this was the largest that I had
- I would let it run and come [across the room] to
the desk. But for the others I would just sit there,
because it wouldn't take long enough for me to
start something else while it ran.
HLC: I notice you have a television over there.
Would you watch TV?
MW: While I was posting I would. I would prepare
the paperwork before I went to the computer.
May 1996
Then I would know what entries should go in, and
in general what transpired in that month, and the
input was very mechanical. I would listen to public
stations, some lecture or concert, while I was doing
the input.
HLC: So then when you were finished, and it was
sorting and posting the batches, and putting them
into all the different accounts MW: No, the original entry automatically goes
right into the account, and [the program] brings it
from there whenever it's needed.
KC: That was how the transactions were validated,
right? You checked the input against content, as
soon as it went in.
MW: The input was checked for balancing as soon
as it went in. And then when the journal was
printed out - I don't have any journals [now]
because I didn't print them for myself - I would
check to make sure that paper clips were not under
automotive, for example. So the journal gave me
the totals, and each account would give me the
detail. As the journal printed I would proofread it.
The batches were automatically balanced as far as
the numbers are concerned.
HLC: The debits equaled the credits.
MW: Then I would proofread the allocation to
accounts when the journal was printed, and finally,
when I printed the general ledger I would
proofread it a second time. In case of an error I
could go in and change it.
HLC: Oh, you could change it until you closed the
month?
MW: For something that was wrong [in a prior
month] I would make corrections through the
general journal, but within the current month, I
could change anything in the program except the
transaction number - I could change the account
number, the document reference, the date, the
amount or the description, and correct misspellings
anywhere. For instance, I had to hand-pencil
something into my printed copy here - I put in 48
cents, I think this was a payroll correction. And
then here, 56.79 in employee benefits, that went
into the wrong account somehow.
HLC: So then you would change it by hand, key
in the corrections -
May 1996
The Analytical Engine
Page 17
MW: And print out a clean copy for the customer.
But I penciled it in for myself when I found
something wrong.
HLC: So they can't download their information
from this into QuickBooks, or something. They
have to just start over.
HLC: Then after you closed the month, if you had
put this company in concrete pouring sales and the
customer said, "No, no, no, we didn't do concrete
pouring for them; we did drywall," - then would
you reclassify the transaction through a journal
entry?
MW: They have to start over on their own systems
because this doesn't translate into anything.
MW: Yeah, I would change [the entry] from one
account to the other through the journal, but the
current entries could be changed anytime.
WORD OF MOUTH
HLC: When you were getting new business, did
you advertise that you had a computerized system?
MW: I never advertised for business. My new
business came through recommendation from my
other clients.
HLC: Were new clients interested in the fact that
your system was computerized?
MW: Oh, yes. After they saw what I could provide
for them, they all wanted it right away.
HLC: What reports and financial statements could
you give them?
MW: Whatever the company wanted, let's put it
that way.
KC: You've said that any time you found an
apparent disk drive error, which only happened
lately, you traced it to a board on the computer.
Did you ever lose any data off the hard disks?
MW: No, I haven't, not as a defect. When I started,
Gary had me build in all kinds of things I called
"idiot stoppers," that would ask me whether or not
I wanted something. I took Gary's advice and put
them in, but they made the processing too slow.
Once I was on my own I took them all out. Well,
at one point, I had the choice on my menu either
to delete one transaction or to delete the whole
batch; I pushed the wrong button and two hours of
work weill down the drain - I could see it go.
After that I put an "idiot stop" in that place that
said "Do you really want to delete this batch?"
HLC: So you knew, when it came up, that it MW: That it was a batch instead of one
transaction. I also moved the selections further
away, one from the other. Originally they were 2
and 3, and I pushed 3 and I should have pushed 2.
So I moved the selection to delete the batch all the
way to the bottom. When I wanted to delete just
one transaction I wouldn't hit the wrong button.
HLC: You had the ability to change the format?
MW: Most of the time I just changed the programs.
I would get my new customers at the start of a
small business, and usually they would have some
problem with their bookkeeping. There was no
point in overloading them with reports they
couldn't use, I just gave them whatever was
necessary. Then, as their company grew, I would
"grow" the program to match. So I had every
customer on an individual program, customprogrammed for them.
HLC: And you would give them a balance sheet
and an income statement?
MW: They got all the printouts from me.
HLC: Now, you were using this system [from
1977] until October 1995?
MW: I retired as of October 31st, yes.
SUPPORT
HLC: When the computer needed fixing, were you
able to do a lot of that over the telephone?
MW: As a matter of fact, just in the last six months
the battery went, and Raul [the current support
tech] said "Well, since you plan to stop working
and that battery's very expensive, we can fix it. "
And so we bypassed the battery with a little piece
of wood, a piece of Chinese chopstick, and then I
had it plugged into one of those outlets where I
could turn it on and off. With the old battery I
couldn't turn it on with a key, I needed a [new]
battery. But when Raul told me over the phone to
put the piece of chopstick in there, and I bypassed
the battery so it worked.
HLC: So this has an ignition key?
The Analytical Engine
Page 18
MW: This particular [terminal] hasn't got it. We
had to change my [CRT] tube in the last week that
I worked. Raul came out and took the whole
[terminal] from the spare computer and put it on
this. He explained to me that the tube was
supposed to last seven years, and this was the
eighteenth year.
HLC: You got your money's worth.
MW: And so he said, we'll just switch it. The
inside [of the terminal] is now from the spare
c()mputer, and I couldn't find the key. Since it
would have to be turned on with a switch from the
line voltage, instead of with the key and the
battery, the key was not important. He felt there
was no sense trying to chase d()wn a battery in the
last week. It was crazy.
HLC: And the second [ADAM] that you got for
parts, how much did that cost?
MW: Seven hundred dollars.
May 1996
HLC: Instead of $32,000.
MW: Well, it was a used machine and Raul just got
it, I would say, two years ago. He asked me $700
and I said, okay.
EVEK [pulling boards and scrutinizing them:]
Initially these looked almost like DEC PDP
boards, but the backplane doesn't look right, it
looks more like an S-100 type backplane.
KC: 16 Kwords.
EVEK: 16 Kwords?
KC: 16K 16-bit words.
MW: I know nothing about it. If Raul told me
over the phone "Do this, and push that button,
and look at that," then I could do it. But otherwise
I have no idea what it is. Right now it has original
boards in there, some boards from the second
computer, and some of Raul's boards. Last time he
came in, he said "Now we're not going to mess
HLC: Once you paid for the machine, how much business expense did you have?
MW: Not much; what I needed was stuff like paper and ribbons and paper clips. It wasn't much of an expense.
Now, for a number of years I also had a china shop.
HLC: Did you keep your bookkeeping business and your china shop business going at the same time?
MW: In 1985 I decided that, as long as I could still work a little harder than
necessary, I would open the china shop for my retirement. I felt that as I grew
older [the bookkeeping] was taking quite a bit out of me -- it required a lot of
concentration about the things that I had to remember or else. And then no
sooner was there a deadline than my customer would come in two days
beforehand and give me five days' work.
HLC: Right, and you'd work all night.
MW: Oh, yeah. So I thought, well, I can always run a china shop, that's no
problem, so I bought into a little shop and I was running both businesses. But
from '85 to '88 the shopping center that this shop was in went down the tubes; the nice shops went out and
Payless and [some other discounters] moved in, and I was too fancy, I had only good quality china and crystal
and stuff. So this shopping center wasn't doing much for me, and in April [1989] I moved over to Saratoga. In
October we had the [Loma Prieta] earthquake and -- I've got a lot of pictures to show. [Laughs.]
HLC: And you got insurance, right?
MW: Yes, but silly me, being in the business I didn't think rationally. When I got the insurance money was
when I should have packed up and gone home. Instead I thought "Well, gee, so many people have broken
dishes now that it'll be a godsend to them if I can supply them with good dishes." So I took all the money,
refurbished my store, and within two weeks I was back in business. The ladies came in and said "I'm not going
to buy anything breakable any more." That was the attitude. And right after that the economy went down, my
May 1996
The Analytical Engine
around," because I had three more days' work and
the thing conked out on me.
EVEK: Just had to get things finished up to give it
back.
MW: We just patched it up so that I could work on
it to finish up. At least it lasted long enough.
EVEK: It's still astonishing to me that you have
$32,000.00 in that, that's a lot of money.
MW: But if you earn $40,000 to 45,000 a year, in a
couple of years it's [amortized]. It paid for itself.
KC: Handsomely, I would say.
EVEK: And then when you think - I get a new
computer about every two years.
KC: Right, you get a new computer every two
years, but there are computers and computers; and
this goes right to the heart of the old debate that
always ends with the line "It isn't the cost of the
Page 19
hardware, it's the cost of the software." The major
cost here was actually the development time and
the debugging and refinement time of the software.
MW: What I put in there was what I wanted and
needed. Now, truly, once the system was working
and I was working fluently right through, I found a
lot of other things that I still could refine - make
an even shorter verb or something - but it was
working, so who cared? I could see areas where I
could make improvements, but it was giving me
the answers anyway, so I didn't bother spending
the time. By then I had all the customers and I was
too busy to tinker around. Some customers would
ask me "Oh, do you think that we could get some
percentages on this statement," and sure I could get
them, so away we went. But if people didn't want
percentages, I didn't offer them.
KC: Right, you don't give people features unless
they ask for them.
lease came up, they wanted more money, and I wasn't going to work just for the rental, so I packed up and
moved everything here. I still have some $15,000 in china samples sitting in the garage.
HLC: Hope there's not another earthquake.
MW: Well, what are you going to do? I had a big sale and fairly well sold out the regular merchandise, but I
couldn't get rid of the samples. How can you sell one place setting of each pattern?
KC: I have to say at this point that you may be the only person I know who went into retail for relaxation.
MW: I'll tell you, for me it was very simple -- the paperwork was a cinch. I could devote my time to the
merchandise and to the customers.
HLC: Did you yourself work in the store, or did you have employees?
MW: Just myself.
HLC: Just yourself. How many hours a day did you have it open?
MW: It was open from ten to six, and then I'd come home and work [on bookkeeping] until about two o'clock
in the morning. At ten o'clock I'd be back at the store. I worked 16 hours a day, 7 days a week for 7 years.
KC: Hm. Okay.
MW: Well, actually I didn't work 16 hours towards the end. Some of the customers retired, and I didn't take
any new customers, so gradually I went from 10 down to 4.
HLC: In the bookkeeping.
MW: And then after I closed the store I kept the four and didn't add any more. See, I'm not a spring chicken
anymore, I can't keep up. An old lady is not a bullet train. You have to make stops here and there.
Page 20
The Analytical Engine
MW: Now, sometimes I would say "You need this
for your own good," or "I think you need detail of
this or that." But there are all kinds of customers. I
had some who were only interested in the bottom
line, just how much profit was made, that's all they
cared for. They couldn't care less how you got it.
Then I had customers who would look into it but
were not very interested, and then I had some who
wanted to know everything about the general
ledger - how to read it, how to find things that
they wanted to look up. So there are all kinds of
people.
I charged them according to a system figured on
the base work that I needed. Like the one account I
showed you had a lot of entries, so my base charge
would be high. For those who had only 100 or 150
entries, my base charge would be low. In addition I
charged a dollar per thousand on gross income, so
if they started as a small company and kept going
up, my fee automatically would go up [with their
gross]. I also charged $5 per person on the payroll
every month, and when it was time for the
quarterly reports, I didn't have to charge extra that was prepaid already. If the company grew, my
fee automatically went up, and if they had no
income at all, I still had the base.
HLC: That's a good way to charge.
MW: It worked beautifully, and I never had to ask
for an increase.
KC: This was a complete custom system. Who
would sacrifice the hardware platform that it ran
on?
MW: I'm sure that [computer] system is a lot more
capable than what I used it for. If I needed more
information [in the ledgers] then I could have put
in more [in the program] but I didn't need it. I
only put in what I needed.
KC: When I say the capabilities of the system, I'm
not talking about the hardware specifically, I'm
talking about the capabilities of the accounting
software.
MW: Oh, okay. I thought the computer will do
anything you just tell it to do. I don't think there is
any machine that won't do what you want. If you
know what you want, you can put it in there.
May 1996
Quick Take:
RAMAC 40th ANNIVERSARY!
That's right .... the fortieth birthday of RAMAC,
sometimes known as the IBM 305, the world's first
hard disk. The fact deserves some thought - since
in that time a CPU has turned from rack on rack
of firebottles, to TTL, to a single IC; memory
shrank from ferrite core arrays to daughterboards
crammed with RAM chips; II 0 was a Teletype
then and a color flatscreen now. Yet we're only
beginning to think about what might replace the
hard disk, a device as spectacular now as it was in
1956. Three cheers for Rey Johnson and the fine
engineers of IBM San Jose - and many happy
returns.
(If you don't know the stunning story of RAMAC,
you might want to order ENGINE 1.4 and 2.2,
which are still available from our back issues
department at US$6 each. Overseas customers
please add US$3 postage.)
Quick Take:
GETTING FRAMED IN 3D
Did you happen to notice the nicely framed and
glassed Apple One that sold for $22,000 at the 1996
Computer Bowl? Would you like to frame your
own favorite singleboard - or the circuit that
made your reputation? Problem is that an ordinary
downtown framing shop, even a competent one,
will sometimes back off when queried about a box
frame. But the ever-vigilant CHAC has found a
framer who won't flinch. Contact:
F rame-o-Rama
210 Hamilton Avenue
Palo Alto CA 94301 USA
+ 1415-321-3939
and ask to speak to Victoria Miller .... who says she
truly enjoys framing an occasional computer. The
end result is gorgeous, too.
May 1996
The Analytical Engine
THE MAC AND ME:
15 Years of Life with the Macintosh
(Part 2)
by Jef Raskin
ANTI-MICRO ATTITUDES
The computer industry in the middle 70's tended
to ignore or minimize the microcomputers that I
saw as the future of computing. (The term
"personal computer" was to come later.) Nonetheless, I was invited to chair the National Computer Conference session on documentation in
1979 - but this was mostly on the basis of my
presence in the large-computer world.
At first, those who asked to exhibit microcomputers were turned down. By 1978 they were
given a room in the basement. A few of my friends
at the large computer companies asked me why I
was throwing away my career by working for a
microcomputer company.
At one of the National Computer Conferences I
was on a panel where I was expected to uphold the
proposition that microcomputers were useful.
Many mainframers thought and said that micros
were - and would remain - toys. We each gave our
little talks, but I didn't score until the discussion
seSSIOn.
To show the superiority of large computers, one of
the speakers challenged me to some "benchmarks."
The exchange went something like this:
"Anything your little Apple can do, my mainframe
can do, and do it better," he boasted. "For one
thing, microcomputers don't have the speed of a
mainframe!"
"OK," I replied, "name your speed benchmark."
"Invert a 100 by 100 matrix! It will take me about
40 seconds."
"You win," I conceded. "It would take my machine
hours to do it."
The audience gave a bit of applause for the mainframe.
Then it was my turn; "We both have to run across
the hall. The person getting to the other side first,
carrying his computer, wins."
Page 21
There was laughter as people pictured him trying
to pick up and run with his mainframe, larger and
heavier than a refrigerator, and then there was a
solid round of applause as I raised my Apple II
with one hand.
"For my next benchmark, let's discuss power," he
said. "Have each of our machines create an index
to a thousand page book."
I had to concede. My computer couldn't even hold
that much text. This admission got a few guffaws
from the audience.
Then I proposed my second benchmark: "You take
$100 out of your salary every month and I'll take
$100 out of mine. The person who can pay for his
computer first wins."
There was a lot of laughter and applause. "But,"
argued my opponent, "that's not computer
power!"
"A computer," I answered, "has no power at all if
you can't afford it." From the audience reaction, it
was clear I had won the debate.
MACINTOSH PROJECT PRELIMINARIES
Early in 1979, probably in March, I talked with
[Mike] Markkula about my idea for a new computer. He had had an idea for a $500 game
machine, which he called"Annie."
I thought that a game machine, although a good
idea, was not something that 1'd feel comfortable
doing. So I counter-proposed a general-purpose,
low-cost computer based on my own ideas - and
dreams - for an interface. Markkula agreed to it.
I picked "Macintosh" as the name for my project,
since Macs were my favorite apples. I changed the
spelling because I wanted to avoid conflict with the
name of an electronics manufacturer - an attempt
that proved to be in vain.
Most of all, I didn't want to call the project
"Annie," since I felt that the trend in the company
to give new products feminine names was sexist and if you had spoken to the namers you would
agree.
Markkula's "Annie" project would, besides games,
have allowed the user to program in BASIC. But it
was not intended for business, and I thought any
new product should be able to handle a much
wider range of applications.
Page 22
The Analytical Engine
I also said that using a TV set or a third-party
monitor was playing Russian roulette with one of
the most important selling points of a system - how
the screen looked.
With these wants and limitations in mind, Markkula sent me off to do design and cost studies.
Working with my friends at Apple, notably Brian
Howard, I came back with an absolute minimum
selling price of $1,000, far from Markkula's goal.
The machine I designed was based on the 6809 chip
and had a 256 by 256 bit-mapped screen. I came up
with a proportionally-spaced character set that
would display 25 lines with an average of over 80
characters per line on the little display. (To put this
into perspective, the Apple II displayed only 40
upper-case characters per line. The idea of proportional fonts on a display was then unknown at
Apple, though commonplace at PARe.)
My choice of the 6809 was dictated by the tight
price constraint imposed initially by Markkula.
The better 68000, when it first became available a
little later, was $400 - if we bought it in quantity.
That would have made the product have an introductory price of about $3000.
My original concept was biased toward the inexpensive and memory-efficient. I noted that a 256
by 256 display could be addressed in exactly two
bytes, making fast software easier to write - speed is
of the essence in a good interface.
To convey one of the Macintosh design features to
others in the company, I built an Apple II with a
monitor incorporated into the lid. I used it at lectures and demos and it had great appeal wherever I
demonstrated it.
(To this day I don't know why Markkula - to
whom I pitched the idea the strongest - Jobs, and
all the other people in management didn't use my
idea in the II. The Apple II had a pop-off lid, and
we could have sold a replacement lid with an
angled CRT built in.) My very happy experience
with this prototype settled it: the first Mac would
have a built-in display.
FRICTION WITH JOBS
While the company was thinking about manufacturing tens of thousands of computers a year
(another unheard-of idea), I wrote an internal
document called "Computers by the Millions." In
May 1996
it I looked at questions of design, manufacturing,
marketing, and general social and economic impact
of computers in those quantities. Management
found the paper valuable, and would not allow me
to publish it for three years, to avoid letting the
competition know what we were thinking. It was
still years ahead of its time in 1982, when I
published it il). the ACM's SIGPC Bulletin 0101. 5
No.2).
Jobs, unaccountably, did not at all agree with my
views of the future, nor with my distributing them
internally at Apple, even though I was doing so at
Markkula's request. By proposing new strategic
ideas and products independently of Jobs, I began
to get on his "wrong" side. By this time Jobs had
begun to have people who were "in" and those
who were "out;" if you were "in," everything you
did was golden, if you were "out" everything you
did was rotten. By the time Jobs had started NeXT
this had become a major trait of his, according to
Randall Stross's book, "Steve Jobs & the NeXT
Big Thing." My take on this book, and its view of
Jobs, appeared in 1994 as "Hubris of a heavyweight" (IEEE Spectrum, July 1994, pp. 8-9).
But as I began work on the Mac, I didn't recognize
the Jobs phenomenon. Thinking I was still "in," I
kept on trying to get Jobs to go see what PARC
was doing; since I was actually "out," he resisted
the idea strongly.
I , of course, remained oblivious to what was going
on. I thought that he would turn around as soon as
he saw the quality of what I was doing. Besides, we
had been friends, and our disagreements were
purely technical.
PASCAL
Early in 1979, I tried very hard to convince the
company that we should move away from using
BASIC and assembler as our main languages for
applications and system software. After presenting
the case for and against a number of major computer languages, from FORTRAN to APL, I
argued that we should base our work on Pascal. I
hired a clever and inventive ex-student of mine,
Bill Atkinson, who implemented a Pascal developed under Ken Bowles at UCSD. They had it
running on the 6502 processor, the same processor
used in the Apple II, and Atkinson suggested porting it to our product.
May 1996
The Analytical Engine
In the process, Bill had to write graphics routines,
an experience that proved extraordinarily valuable
for Apple. Many in the company had rejected
PASCAL as impossible to put on an Apple II, contradicting several technical memos I had written
showing how it could be done.
As Atkinson later said, "We had a bunch of selftrained amateurs who didn't really understand
modern software development. The system software team actively resented a new language. Once
we had it up enough to demonstrate the word
processor, and Markkula saw that, it was clear
sailing."
I supported Bill's implementation, and then wrote
a PASCAL manual with Brian Howard. Pascal, as I
had predicted, allowed us to hire more professional
programmers, and later became the main development language for "Lisa" and the Mac. At the time,
I personally paid a license fee to UCSD so that
Apple could use their Pascal system. Apple never
reimbursed me, since Jobs insisted that Apple
didn't need and would never use Pascal. Almost all
Mac and Lisa software was written in Pascal
derived from UCSD. I remain amused by the
thought that in some vague sense, it was all owned
by me.
THE MAC BECOMES OFFICIAL
By September 1979, Mike Markkula had - over
Steve Jobs's objections - approved the Macintosh
project. But by going around Jobs I had unknowingly set up a dynamic that made the project far
more difficult politically than I could have anticipated.
From the first, Jobs opposed it, calling the Macintosh the "dumbest idea" he'd ever heard of. He
would often recite a list of imagined advantages
that the Lisa project had over the Mac and put obstacles in the way of my obtaining staff or supplies.
His interference eventually became so overt that
Mike Scott had me move the entire Mac project to
some buildings behind a Texaco gas station across
De Anza Boulevard, so that we would be able to
develop the Mac in peace. Since we were on the
second floor, we called it "Texaco Towers." Later,
when Jobs took over the project, he put up a pirate
flag and claimed that he moved the Mac out of
Apple headquarters so that it would remain pure
and uninfluenced by the stodgy company engi-
Page 23
neers. Tome, the pirate flag really indicated a
pirate within: as I see it, Jobs took over the project
by fiat and lies, and was nearly successful in stealing the credit for having originated it as welL
From the beginning, to keep the project on track,
and so that we would not lose good ideas (and the
reasons for abandoning others) in the press of development, I created a document numbering
system and put the collected documents in the
"Book of Macintosh," which grew to some 400
pages. I wrote most of the book, since I liked to
write - and was the fastest typist in the group - but
the ideas were generated by everybody, and everybody got credit in the text.
Here's one example of the standardized heading
format:
MACINTOSH PROJECT DOCUMENT 18
VERSION 0 DATE: 20 OCTOBER 1979 TITLE:
DELIMITING STRINGS. AUTHOR: JEF
RASKIN.
I asked all the participants to explain the reasons
for their conclusions, their right turns and wrong
turns, as we went along. I wrote most of the
documents late at night at home; we were too busy
during the day to get around to it.
I believe Jobs's opposition was partly due to his
not understanding what I was trying to accomplish, though at the time I incorrectly thought of
him as the supportive friend he had been for so
long. For example, when I insisted on bit-mapping
and square dots, he would retort that W oz had put
a character generator in the Apple II and it didn't
have square dots and its sales were paying my
salary.
THE FAMOUS ANACHRONISM
It has been often said in the computer and general
press that the Mac was a straightforward copy of
the work done at P ARC. It was not, and the idea
does a disservice to the hundreds of people at
Apple who developed the hardware, software,
marketing, and interface concepts. This erroneous
belief turns what was a significant intellectual debt
into the appearance of moral bankruptcy. I can't
blame people for making the mental leap from
hearing that the Mac resembled - and was partially
inspired by - the interfaces at P ARC to guessing
that it was largely "stolen." But at the same time I
cannot forgive those who write on the subject (and
Page 24
The Analytical Engine
May 1996
make the same claim) for not doing their homework. Most of them never saw or used an Alto, a
Dorado, or a Star - the systems developed at
PARC. They simply assume that the earlier
systems were much the same as today's Macs and
Windows machines.
With one button, I reasoned, you could not get
confused about which to use. It took a while, but I
was able to find methods that in every case required the same or fewer operations than those
required by the PARC system; it was faster, easier
to learn and use, and it was far less "modal."
Then there is that apocryphal story about Steve
Jobs visiting PARC, having an "Aha!" experience
and coming back to Apple in full cry to create the
Macintosh project.
Of the methods I invented, the most fundamental
was the idea of pressing and holding a button while
dragging, and using the release of the button to indicate that the operation was complete. This differed from the method - used at PARC and dating
back to the work of [Ivan] Sutherland - of click,
drag, and click again.
Well, he did go, he did see, and he did come back
enthused, but the Macintosh project was well
under way at that time, having been officially
started months earlier. The trip was set up to
convince him of the value of the Macintosh
project. I'm not sure how the story got reversed,
but I later learned that Apple's PR department repeatedly told the false tale to anybody who asked.
Nearly a decade after the introduction of the Macintosh, Xerox took Apple to court over the issue. I
was briefly invited to be an expert witness, not by
Apple - as I might have expected - but by Xerox.
The Xerox attorneys soon learned that the main
thing I could. testify to was the originality of the
work done on the Macintosh. (The Lisa group did
do what I consider some shameless copying of the
Xerox Star, down to the names for some individual
fonts, but that is a different story arising from the
fact that a lot of key people on the Lisa project had
been hired from Xerox, something that was not
true of people in the Mac group.)
The case did give me a chance to use a Star and an
original Lisa, each for the first time, an experience
that taught me how much further the Mac was
from its predecessors than I had remembered.
INTERFACE INNOVATION
One of the substantive differences in the "look and
feel" of the Mac interface was the one-button
mouse. The one-button paradigm has become so
pervasive that many applications for IBM-compatibles ignore the second button that clutters most
IBM-compatible mice; the third button that was
part of the Engelbart and P ARC mice has also disappeared almost completely from popular use. My
own difficulties with the three-button mouse - and
watching other people have trouble learning it - led
me to rethink the design.
When Larry Tesler came from PARC to join
Apple he was naturally resistant to the one-button
mouse. Larry was comfortable with the threebutton implementation and had long touted its advantages over non-mouse systems. It took considerable effort to convince him, point by point, that
my solution was not only workable (which he and
others doubted at first) but in fact superior.
In any case, the interface we developed was a distinct and new creation, though it shared many
elements with and owed a very real debt to what
had been done at PARCo A major part of that
debt, of course, is that I was able to use PARC's
work as a living demonstration of a highly evolved
graphical interface.
The one-button mouse was not the only major difference between the Mac and the systems at
PARCo Another interface improvement that made
the Mac feel so much easier to use was the way a
user selected something or engaged a menu. At
PARC, menus were relatively static lists of limited
length that the user could summon and dismiss.
Bill Atkinson - later to become 'an Apple Fellow
- proposed that we instead extend my method of
selection and drawing so that just the title of a
menu would be shown, but when you pointed to
it, clicked and held down the mouse button, the
menu would appear! Then you would release the
mouse button when the cursor was pointing at the
desired item. This made menus appear when you
needed them and disappear without apparent
effort. Furthermore, as we both pointed out,
having the menus at an edge of the screen and
having the cursor position confined at the edge
meant that you had to point accurately in only one
dimension, which made the menus easier to use.
The design of Microsoft's Windows and similar
May 1996
The Analytical Engine
interfaces does not have this useful "pin to the
edge" idea.
Atkinson was led, by analogy with my point and
drag methods, to pulldown menus that you can
drag across to your desired item. Probably because
it worked much as typewriter SHIFT keys do and
as a pencil does - you put it down at the
beginning of a line and lift it up at the end - my
method of using a mouse has prevailed.
I extended this idea to drawing lines and to creating
rectangles and other shapes by pointing and sweeping across the diagonal. My "hold and sweep" concept was then applied to making graphical selections. We created a rectangle that surrounds or
touches the items to be selected while the button
was held. The methods I devised are now so universal that some people who worked on the earlier
systems have forgotten how they worked. They
tend to "remember" them working as the Mac does
now. What I remember is the effort it took to
convince my fellow engineers that what I was proposing was better.
I suggested that Apple patent the one-button
mouse and the new way of using such a pointing
device, but Jobs nixed the idea in favor of patenting Atkinson's pull-down menus. Apple missed
this opportunity simply because Jobs didn't want
my name to appear on any Apple patents (though I
have about a dozen of my own). I was still "out."
A more subtle difference between the Mac and the
work at P ARC is this: in the Mac you point to
something and then tell the system what to do
with it. It is the "noun-verb" paradigm that is now
nearly universally recognized as desirable by interface designers. As Bill Buxton of P ARC has reminded me, the Xerox products used a more complex noun-verb-noun method involving a bunch of
function keys Qike the current IBM compatibles).
To quote Buxton, "Both the concept and the operation were quite different ... it is remarkable how
few people who teach and talk about GUIs even
seem to understand the differences to even this
degree of subtlety." Buxton and his colleagues also
published research in the 1990s (on what they term
"kinesthetic feedback") that showed why my clickand-drag paradigm worked so well.
Another fundamental part of the Mac from the
very beginning was the insistence that unifying
software would be built in. Knowing the time con-
Page 25
straints of the real world and the inherent laziness
of all humans, I suspected that if we built in an interface, programmers writing applications would
use it, grudgingly, for their first mock-up as it was
much faster and easier than writing the interface
themselves - standard practice in all products prior
to the Mac.
I knew that writing a rule book would only antagonize the independent spirit of software developers, who are inherently entrepreneurial. They
had to be tricked into using the Mac interface. I
could depend on their time constraints, and the
likelihood that our interface would be far superior
to what they planned, to insure that the details enforced in the software prototype would appear in
the final product.
It worked. When the Macintosh was released, users
found that learning new applications on a Mac required far less effort than the same task on any
competing system. This gave third-party software
developers added incentive to do things in the Macintosh manner, and Mac users have reaped the
benefits.
The success of the Mac led other companies to
copy its interface, and one can now move without
too much difficulty from the Mac to Windows, to
Geoworks, to most workstations, and even to
some mainframe front-ends without retraining and
with barely a glance at a manual or help screens.
My unifying software originally was to be a graphics-and-text editor within which applications could
run as additional commands (via menus), all input
and output being through the interface designed
for the editor. Later, the PARC desktop metaphor
was adopted from the Lisa group, who had adapted
it from the Xerox Alto and Star computers. The
incredible work of the Mac software team designed
and squeezed the necessary code into a "Toolbox"
within a relatively small ROM (Read Only Memory) that we could afford to put into the product.
The interface concepts I wanted to implement required fundamental hardware changes. One example was the way the electronics of keyboards were
designed, not in keyboard layout - which obviously affects the interface - but in the way the keyboard works at the chip level. Before the Mac, and
excepting P ARC which was at that time not a
commercial manufacturer, the makers of commercial keyboards built each key to put out a signal
Page 26
The Analytical Engine
when pressed. By the middle 1970's a special
"encoder" chip took a signal from a key and produced the code for the symbol that key represented. There were usually a few exceptions: the
SHIFT key could be pressed and held and while
other actions took place; the same was often true
of other state-shifters such as the "control" key.
But these exceptions were built into the encoder
chip; what I wanted was a keyboard whose
keystates - whether any keys were up or down would be "known" by the computer. By analogy
with pianos and organs, which can use any combination of keys simultaneously to play what musicians call a chord, this was known as a chord keyboard. I had long believed that this was an essential
step toward improved interfaces and when I first
went to PARC I was delighted to learn that they
had come to the same conclusion.
Burrell Smith, our hardware designer, participated
avidly in these discussions, and often suggested
ways in which hardware changes could help the
interface, sometimes also proposing changes in
software design that could simplify hardware requirements. In each case the interface requirements
took precedence, but this was probably the first
time a commercially successful computer was designed with hardware and software subservient to
the issue of usability.
The Mac succeeded because the initial impetus for
its creation came from a humanitarian impulse,
rather than a hardware dream or a marketing
study.
SELLING JOBS ON THE IDEAS
A popular description of Jobs is that he has a
"reality distortion field." This phrase accurately
described Jobs's ability to convince people that
whatever he was saying at the time was inevitable.
1'd seen him charm otherwise reasonable people
into believing absolute nonsense.
Some of this is helpful when doing something new
in the world, but - as I see it - Jobs lived at the
center of this field and actually believed and acted
not only on vision, but on the basis of his own
falsehoods, sometimes with unpleasant consequences.
It seems to me that Steve Jobs was also mesmerized
by the power of part of his key insight that had
helped make the Apple II a success: 'make it look
May 1996
attractive' became a guiding principle. He contin:
ued to confuse appearances and quality ever after;
years later at NeXT, his first major expenditure
was to hire decorators for the new office complex.
This passion for appearances would have been an
asset, or at worst immaterial, in someone who also
understood the products; but Jobs often did not.
THE PASCAL POSTER
In 1979 he botched the design of a poster that
summarized the structure of the Pascal programming language. Programmers found sets of diagrams created by the originator of the language,
Niklaus Wirth, a handy reference. In writing the
Pascal manual I had discovered several errors in
Wirth's diagrams and also disclosed some simplifications. Diagrams in the manual reflected these
corrections and improvements. I thought that it
would be good advertising, as well as a real benefit
to programmers, to put the entire set into a decorative poster. Color would serve as a key to link
items of the same syntactic type, making relationships among language elements clearer.
Jobs thought it was a great idea, and promptly
hired a prominent graphic artist, Kamifuji, to produce the poster. Jobs asked me for a copy of my
diagrams so that the artist could estimate the
project, telling me that once we had a quote I
would work with the artist. But the next thing I
knew, Jobs proudly came into my office with the
finished work. Thousands had been printed.
The poster was very good looking, with bold
colors on a jet-black background. But some of the
diagrams were no longer correct and the colors had
been chosen purely for esthetic effect, making the
chart unnecessarily hard to use. !'told Jobs that it
was very pretty but wrong; he didn't care, and
blissfully went on to something else. The posters
were sent to stores as advertising posters, but they
couldn't be shipped with the Pascal product as
planned. It was a waste of time and money. This
example is not significant in the history of Apple
per se, but does say a lot about how Jobs thought.
MAC POLITICS
By the end of 1979 it was clear to many people that
unless Jobs had a better understanding of what was
being attempted on both Lisa and the Mac, he
May 1996
The Analytical Engine
would continue to inadvertently sabotage the
former and be antagonistic to the latter.
My friend Bill Atkinson knew a great deal about
what was going on with the Macintosh even
though Jobs had officially forbidden him to work
with members of the project. This meant that Bill
had to keep his involvement with the Mac secret,
lest he lose his "in" status, while he worked on the
Lisa. At the time he was writing his meticulously
crafted QuickDraw graphics system - then called
LisaGraph - for Lisa (Apple knowingly used the
name of the list-structured graphics system I
designed for this central piece of software - without
permission or compensation).
With Bill's connivance and the help of Tom
Whitney - who had given me the title of
"Manager of Advanced Systems" to correspond
with my work on the Mac - and by keeping my
name out of the picture, we at last managed to
convince Jobs to visit PARC.
Jobs later said that after he went to P ARC, he
returned inspired, and launched the Lisa and then
the Macintosh. This story, once promulgated by
Apple's PR department and often repeated in
books, articles, and even by the generally excellent
PBS series on the history of the computer, is inaccurate, to say the least. As Atkinson put it in a
phone call to me, "You were instrumental in
getting Jobs to go to PARC, and that was central
to getting his support for new interfaces." Jobs
pointedly did not invite me on this visit, and excluded me from the conversations about it when he
got back; it would have been very hard for him to
have admitted that I had been right about the value
of the work done at Xerox.
In general, I remained oblivious to the politics
going on at Apple, and concentrated on the design
of the Macintosh. This left almost no room in my
life for anything else, except practicing the piano
and occasionally getting out to fly a model plane. I
bought a house a few blocks from Apple so I could
bicycle in and back on a moment's notice. The
Macintosh project was my life.
CONCEPT AND COSTING
What was the Mac concept like in the early days?
We researched many possibilities. For example, we
considered a bit-mapped LCD display which had a
resolution of 256 X 26 (yes, twenty-six) and a cost
Page 27
to us of about $240. At our usual five-to-one ratio
of parts cost to list price, that part alone would
have been $1200 at retail. A 256 X 256 or larger
display with any technology other than the cathode ray tube (CRT) was then totally out of the
question, since a CRT display cost between $35
and $50.
A drawing done by Brian Howard in 1980 shows a
one-piece box with a built-in CRT, 5 1/4" drive,
keyboard, and joystick. The joystick is in the same
position occupied by the trackball in the later Mac
Portable.
We also worked on a strain-gauge stick almost
identical to the current IBM graphic input device.
Embedded pointing devices have a long history at
Apple; for example, in 1978 W oz came with the
idea building a pair of orthogonal thumb wheels
(one each for vertical and horizontal motion)
under the Apple II keyboard. This was a response
to my request that we build a pointing device into
the box that could be operated without removing
the hands from the keys. This seemingly obvious
good idea reached fruition years later with the
PowerBook series and was probably reinvented
independently by the PowerBook group.
Graphic input was an essential element of the Macintosh from the first. I thought that the mouse in
particular was a clumsy way of going about it - for
one thing, it takes up too much desk space, and for
another you have to find it anew each time you
want to use it.
But Jobs was an adamant mouse-ist, (mainly, I
think, because that's what PARC had), and until
third party vendors supplied trackballs, the mouse
was the only graphic input device available for the
Mac.
TEAM BUILDING AND THE TOOLKIT
One of my basic concepts was of a softwarenu·deus that would be built into ROM, and serve as a
home port to the user, tossed about on the high
and varied. seas of application software. To write
the software, I hired Bud Tribble, who had similar
thoughts. He and the two other "B' s," Brian
Howard and Burrell Smith, were the first Macintosh team.
The Mac Toolkit was initially written by Tribble,
who was in charge of Mac software; it was taken
over by unstoppably hard-working UC computer
Page 28
The Analytical Engine
science dropout Andy Hertzfeld, Bill Atkinson,
Bruce Horn (who, at fourteen, had been one of the
usability testers of Smalltalk at PARC), and others.
Each member of the original "gang of four" came
to the group through a different route. Bud Tribble
was a medical student, a programmer and designer
of genius who I had known at UCSD; he and Bill
Atkinson had been good friends there. Atkinson
pointed out the talents of a man working in repair,
Burrell Smith, and after interviewing Smith, I hired
him as head of hardware design. Brian Howard had
been a friend of mine for years.
There were established hardware designers that I
had tried to bring over to the Mac (and who
wanted to work with me), but Jobs had forbidden
them to join the project. Still, Smith proved a firstrate designer who was open to thinking from a
software and human-interface point of view, and
he was a delight to work with.
I brought MIT anthropology student Joanna
Hoffman on as our marketing person. Her major
contribution to the Mac was to make sure that
design decisions didn't preclude international sales;
this concern was unusual in the then-parochial microcomputer industry. Thus the Mac had, from the
first, the accents, special characters, and diacritical
marks needed in languages other than English. We
had come a long way from the philosophy that
upper-case letters were all you needed. oanna also
introduced me to my future wife.)
a
Steve Clark (another UCSD student I brought to
Apple, and whose Olympic-Ievel-kayaking sister
Candi was later to marry Woz - as I've said, it's a
small valley), and a few others formed the nucleus
of a team easily the equal of the much larger and
better-funded Lisa group. I hired some, such as
programmers (and musicians) Gareth Loy and Bill
Schottstaedt, from SAIL; another, the remarkable
poet Bana Witt, had been a music student of mine
when I taught at the San Francisco Community
Music Center. She later married Bruce Tognazzini,
another Apple employee who worked with me and
was to write and lecture extensively about interface
design. (Being a minister, I had the pleasure of conducting their nuptials.)
Donald Reed, the very image of a bookish intellectual in appearance and manner, worked with me
closely on documentation. Of course we enjoyed
the under-the-table help of Atkinson and others on
May 1996
the Lisa team who believed in what I was trying to
do, and the warm support of the late Tom
Whitney, who had been hired to head engineering
for all of Apple.
THE END OF THE BEGINNING
John Couch, a good manager and insightful computer scientist who was running the Lisa project,
increasingly found Jobs a nuisance, and eventually
managed to get him removed from the Lisa project.
Jobs, at loose ends and hearing rave reports about
the Macintosh, decided to have a hand in it.
Apple's top management helped shunt him to the
Mac project to get him away from Lisa, which was
seen as the company's hope for profitability in the
1980's.
Jobs's attempts to undermine the Mac project now
took the form of destroying my credibility. One of
the more blatant incidents was the "brown bag"
lunch at which I was to describe the Macintosh
project to the company at large. It is discussed in a
confidential memo that I wrote to Mike Markkula
to explain why, though I was seeking someone to
manage the Macintosh project so I could concentrate on technical issues, I didn't want Jobs to be in
charge. The memo specified, in detail and in my
judgment, Jobs's many and egregious failings as a
manager.
I had asked that the memo be kept secret, and
Markkula agreed, though he said that he didn't
think he could do anything to control Jobs. I believed this assurance and, thus, felt betrayed a few
days later when Jobs called me in to his office to
"discuss" the memo. I dimly recall Markkula saying
something about having had to discuss it with Jobs.
But Apple was a very open company, doors were
left unlocked, and people wandered freely into one
another's offices. Any of a number of people might
have seen the memo and made a copy for Jobs, or
he may have noticed it himself. Markkula thinks
that something of this nature is what must have
happened, and it might well be.
The memo reflected the running joke that the way
to get Jobs to agree to something was to tell him
about it, let him reject it, and wait a week; when
he came running to tell you about "his new" idea,
you'd exclaim, "Great, Steve, we'll do it right
away!" In the memo I also made a prediction that
was to prove exact: "Jobs was wrong on his Apple
May 1996
The Analytical Engine
III schedule, wrong on the Lisa schedule, wrong on
the cost and price estimates, and he will be wrong
on Macintosh. He is a prime example of a manager
who takes the credit for his optimistic schedules
and then blames the workers when deadlines are
not met."
The memo also related the incredible brown bag
incident: "Jobs is often irresponsible and inconsiderate. An example is the brown bag seminar I was
scheduled to give on 17 February. In January, he
first cancelled the seminar, but then he agreed that
I was to give it. Two hours before the talk he called
me to say that he was canceling it again. His reason
was: I cancelled it because of the reorganization in
PCs.' However, Jobs did not tell the seminar's organizer about the cancellation, nor did he place any
notices announcing the cancellation."
"At noon, fortunately, I made a last-minute decision to go over to the seminar site, where I discovered a crowd of over 100 employees waiting to
hear the seminar. I announced the cancellation
myself - and then I gave a talk on my current work
and interests at Apple, instead."
I was careful not to mention Macintosh or give
specifics since Jobs had forbidden it, but just
explained the cognitive aspects of the interface and
design principles my group and I had developed; it
was - as everybody knew - a veiled introduction to
the Macintosh project.
The talk was received very enthusiastically. The
morning after the seminar Jobs called me into his
office and told me that I had violated his explicit
instructions and was fired. Ignoring what he said,
since he often spoke without thinking things
through, I told him that I'd come back in the afternoon, after I'd completed something I was
working on, and we'd discuss the matter.
Later I went back and, as I expected, he had decided not to fire me but I was "given" an extended
paid leave from Apple. The leave turned out to be
a very important time in my life. For one thing, I
went to a party at marketer Joanna Hoffman's
house where I met my future wife, Linda Blum. I
found a place to live on ten acres of land high in
the foothills of the Santa Cruz mountains, offering
a magnificent view of Silicon Valley. I rebuilt the
dilapidated old house that stood there, adding a
large music room for my piano and a small flying
field for my model planes.
Page 29
When I came back from my leave I was offered the
position of head of Apple's research division. I had
been offered this before, had accepted, hired a good
group, and seen them whisked away to "put out
fires." In those days Apple didn't know what research meant, and looked at the talented people I
hired as resources wasted if they weren't working
on current products. Besides, there was the matter
of personal integrity. Steve Jobs had become impossible for me to work with.
Most people worked around him or sucked up to
him or were in awe of him. In fact he was no
genius; he resembled a planet shining by reflecting
the light of others. Yet he thought of himself as the
Sun King. He could not abide someone who was
unimpressed by Steve Jobs, yet by his actions he
had lost my respect, and I am incapable of being a
sycophant.
Steve had chutzpah in the extreme; he said that the
Mac would make "a dent in the universe," without
the least idea how big the universe is, or how little
a dent all our activities really make. And you can
also explain Jobs with another Yiddish word,
mensch. It is high praise to say of a person that he
(or, in these enlightened days, she) is a mensch or
"a real mensch." A mensch is cultivated without
losing the common touch, upholds high principles
while remaining practical, is kind and generous
without short-changing himself, and is attentive to
his responsibilities to himself, his family, his business, his associates, his community, and the world.
If you understand the qualities that make a man a
mensch, then you understand a lot about Steve
Jobs. Everything a mensch is, he isn't.
At this point the only alternatives left to me were
to leave or learn to toady to Jobs. I resigned from
Apple and gradually watched my predictions about
the Mac come true. Jobs took until 1984 to get the
project out. Burrell Smith quipped that it was in
"constant time to completion mode" and I was repeatedly told that Jobs did exactly what he said I
would do, make endless mindless changes; I have
many faults, but lack of direction is not one of
them.
Steve was given to imposing absurd requirements
on the project's designers. This was especially
ironic since one of the arguments he used, to convince management that he should be given the Mac
project, was that I was an "academic dreamer and a
perfectionist" who would keep on changing his
Page 30
The Analytical Engine
mind and never bring the project to fruition on
time. My detailed schedule showed release of the
product toward the end of 1982. Jobs's verbal plan
was something like six to eight months shorter.
There is some evidence to back up my perspective:
the next project of comparable scope that I
managed (a workstation for Canon) was completed
on budget and on schedule. The next project Jobs
tried to manage (the NeXT computer) was a disaster in both these regards.
The resultant Mac not only took more time to
come to market, but was a less coherent and - in
some ways - less capable product than what I had
been working towards. The interface was less consistent and harder to use, and there was no way to
get at the hardware bus. Other parts of the Mac
design were improved. Whether my version (as it
would have matured as it approached production)
would have been more commercially successful
than the 128K Mac is an unanswerable question.
The experiment cannot be done, and we will never
know. I feel it would have been a somewhat better
product that would have penetrated the market
faster. I would guess that Jobs would disagree.
DREAM FULFILLED... ALMOST
Three decades ago I dreamed of a computer with
which I could compose music and print it out in
full musical notation, write properly formatted
text in a panoply of fonts, have the ability to mix
text and graphics, and do drawings with precision
and ease.
Today I do all this and more at the tiny and capable Macintosh PowerBook that sits at my desk. It
goes wherever I do. This much of the dream came
true.
My reasons for deciding to abandon teaching for
commerce proved correct. The Macintosh's profitability (as contrasted to Xerox's extensive published record) convinced companies such as Microsoft and IBM that the interface was the controlling
element in most sales. Now a vast majority of the
computers sold have an interface that looks much
like the Mac's.
Because of the Macintosh project, computing has
been made easier and more pleasant for hundreds
of millions of people years before it might have
happened otherwise. I made not a penny for my
work on the Mac, beyond my salary at the time;
May 1996
but I helped change the world in accord with my
own personal vision, and I have seen the effect of
this in my own lifetime. This would be fully satisfying if I didn't know so well that we can do much
better.
The desktop metaphor, used by everybody from
PARe though Apple and Microsoft and extended
almost absurdly by Apple spin-off General Magic,
was a clever way of making the workings of an operating system palatable and learnable. It is far
more fundamentally good to eliminate the need for
an operating system altogether. The current paradigm of using application programs is inherently
wrong from the standpoint of interface design.
This is widely recognized, but the solution offered
is to make them interoperable, which solves some
of the problems but by no means all.
GUIs as presently designed and used are an interface dead end. They can be patched endlessly, but
only a completely different approach can bring a
large jump in usability. The Cat computer, which I
developed for Canon, demonstrated that my alternate approach is implementable and both more
productive and more pleasant than GUIs. Canon
failed to market the product effectively, possibly
because the moribund Electronic Typewriter Division had been selected for the task, and it is now a
dead Cat.
The parts of computer interface design that I am
working on now are not dependent on particular
technologies; any advance in the basics of interface
design will apply - however more powerful computers become, however broad the information
networks of the future spread, and however the
technology is melded into our everyday or even
everymoment lives.
A CRITIQUE OF SOME HISTORIES
Many friends have suggested that I counter the
numerous incorrect accounts of the history of the
Macintosh with a true one of my own. It is difficult, for even though nobody was or could have
been more closely involved with the initial creation of the Macintosh than I, I cannot eliminate
the colors that tinge my memories. It was a very
emotional time, full of strong feelings, massive egos
in conflict, distinctive personalities, and many
rights and wrongs. But I cannot do worse than
some of what has been published.
May 1996
The Analytical Engine
Page 31
There is a strange avoidance of scholarly seeking
after truth. An egregious example of the anti-academic attitude occurs in a book by Robert
Cringely, who writes a delightful column that
appears weekly in Info World. In his book he has
the Mac and Lisa projects being created by Steve
Jobs after Jobs made a visit to PARC in 1980 and
came back inspired.
ideas found their way into consumer PCs." The
people he interviewed were at PARC; their association with Apple began only after the Mac was well
under way. Thus they could only tell him about
the development of the ideas at P ARC and, in the
case of Larry Tesler, about the work on Lisa only
after 1980 - that is after Apple was committed to
the basic direction I wanted the company to take.
I wrote to Cringely and pointed out that his
account - like those of several other authors - was
wrong; Jobs had indeed made the visit in 1980
(some say in December of 1979,) but the Mac
project was proposed in the spring and officially
started in September of 1979. In other words, the
project was well underway before the supposedly
pivotal event took place.
Larry was quite resistant to some of the nonPARC ideas that we had developed independently.
He did not at first understand many of the improvements over Xerox's work - such as the onebutton mouse - that I created. He did not work
on the early Macintosh project at all but on Lisa,
which was modeled closely on the Xerox Star,
even to the point of having the same Xerox-created
names for the fonts. The Macintosh proceeded for
years much more independently and (significantly
for the reports that used them as sources) out of
the view of the people interviewed!
Cringely was unabashed. He wrote back: "As for
all the business of what project started when,
whether Lisa started before or after Steve visited
P ARC, whether the Mac had already begun or not,
well I don't think that it really matters very much.
My attempt was to EXPLAIN (I say that at the
front of the book), not to be a historian." How
one can hope to explain what happened, without
even knowing what happened, eludes me.
A PBS special on the history of computers made
the same mistake of attributing the genesis of the
Mac to jobs's visit to PARCo When I wrote to Jon
Palfreman, its producer at WGBH, he replied,
"The part of the program you are referring to
comes at the end of a lengthy segment about the
highly innovative work done at Xerox PARe.
This section was based on extensive interviews
with Alan Kay, Bob Taylor and Larry Tesler. The
purpose was to show that the key concepts of interface design which today are a feature of most
PC's (if you count Windows) were first discussed
at Xerox PARe. When those ideas were embodied
in an affordable machine - the Macintosh - they
began to change the world of personal computing.
I was aware of your key role in the Macintosh
project, and indeed of the contribution of people
who developed Lisa. My aim in this particular
program wasn't to detail the history of Apple, but
to show how the key interface ideas found their
way into consumer PCs. "
Again the false scenario seems so plausible and
story-like that the person in charge does not care to
"detail the history." But it is in that history, and
not only the history ofPARC, that "the interface
The years of thinking and experimentation on the
early Macintosh project have gone unreported,
even though the early work led to the breakthroughs that made the Macintosh and everything
after so much of an improvement over what went
before. Against this reality we have the powerful
mythological image of Jobs going to P ARC,
having an "Aha!" experience and coming back at
full cry to Apple to create a fantastic project.
The fabricated Jobs story is familiar - it parallels
that of Archimedes jumping naked out of his bath
crying "Eureka!," and a dozen other stories. That
there was a little-known computer scientist who
had been working on the concept for over a decade
- who created the project, and then maneuvered
Jobs to go to PARC, so that Jobs would begin to
understand (and thus support) what was already
going on at Apple - is a very different, more complex, and unlikely-sounding story.
Also, the appealing and basically true legend of two
college drop-outs who created the profitable and
excellent Apple II blends in easily with the fiction
that one of the dropouts went on to create the
even more revolutionary Macintosh. It is a less
striking tale that a former college professor and
computer center director with a degree in computer science instigate such a thing - but although
it may not be as good a story, it is what happened.
Page 32
The Analytical Engine
May 1996
Cringelyand Palfreman were not being underhanded, only a bit careless and --- in Cringely's case
- cavalier. In some other cases, authors drew the
wrong conclusion by lacking accurate information.
Jeffrey Young, in his book "Steve Jobs," writes of
the first time that Jobs (along with Atkinson and
others) saw the work at P ARC.
ally at odds with historical fact. Levy retells the
Jobs-at-PARC story. Strangely, he credits me with
having paintings shown at a famous museum; in
fact I have never done any paintings. (An adaptation of this book, published in the February 1994
issue of Popular Science, tells a story that is far more
accurate, - although it still calls me a painter.)
"Atkinson and the others were asking Tesler questions, one after the other." "What impressed me
was that their questions were better than any I had
heard in the seven years I had been at Xerox ...
Their questions showed that they understood the
implications and the subtleties... " But Young did
not ask why their level of instantaneous understanding was so impressive. The reason was that I
had been explaining all this stuff to Atkinson and
Jobs for years; Atkinson (a student of mine who
had worked with me extensively prior to this
meeting) had grasped it very well. Tesler didn't
know about this background, wasn't told, and so
was bowled over.
John Sculley, in his ghost-written book Odyssey,
refers to me as a "programmer" at Apple. I was
never a programmer at Apple, and the rest of what
he says is nearly as inaccurate. He got his misinformation about the history of the Mac primarily
from Jobs, with whom he spent a lot of time. Like
Cringely, Levy, and Palfreman, he chose never to
interview me or even call me - or others who
were there - to check on his facts.
Atkinson couldn't very well say that Raskin had
briefed him and some others, because I was out of
favor with Jobs at the time, and anything I proposed was automatically rejected. Only after much
planning and sleight of hand, during which it appeared that Atkinson supported the PARC trip
and I opposed it, did Jobs agree to go.
There is also the halo effect. During the years that
Steve Jobs was on top at Apple - and before NeXT
showed his fundamental weakness - he was usually
credited with inventing the Macintosh. Later,
when his star was declining - as his company,
NeXT, beat one strategic retreat after another, and
as General Magic, co-founded by Bill Atkinson,
. Andy Hertzfeld, and Marc Porat, was about to
announce its first product - the December 27,
1993 issue of Info World included a story hailing
Bill Atkinson and Andy Hertzfeld as the creators
of the original Macintosh.
Their contributions were essential to the product
and represent some brilliant work, but neither of
them has ever claimed that they created the Macintosh. Again we find the heroes of today falsely
credited with the achievements of others not currently in the limelight.
Steven Levy's history of the Macintosh, Insanely
Great - published to ride the wave of publicity for
the 10th anniversary of the Mac - is also occasion-
My experience with Jeffrey Young was especially
disturbing. I had agreed to the interview with the
understanding that I would see and comment on
the galleys before publication; he never sent them his inaccuracies compounded by a breach of trust.
By way of contrast, Owen Linzmayer's The Mac
Bathroom Reader is more accurate; for example, it
gets the order of events straight.
In any case, Jobs's own view of how things came
about necessarily must have been distorted. For
one thing, we often misattributed ideas deliberately
when speaking to Jobs. If the group admired an
idea by someone Jobs didn't like at the moment,
we gave the credit to someone currently on Jobs's
"good" list. It was also often necessary to use
"reverse psychology" on Jobs; we got a lot of features into the Mac by having someone (usually me)
suggest the opposite. Jobs would then see the
problem in "my" approach and often tell us to turn .
it around.
Another technique was to tell him about something informally. Often he replied that the idea
was dreadful. Then when he "proposed" that same
idea after its merits had settled in, a few days or
weeks later, we'd tell him he was a genius for
having thought ofit.
Thus, Jobs's recollections ofthe history of the Mac
would often be far from what actually went on.
Being very independent, I was often on Jobs's "bad
person" list, so I had to rely heavily on these techniques. You'd think he would have caught on
when one of "his" ideas turned up implemented
later that afternoon or the next day, but he simply
May 1996
The Analytical Engine
believed that his great engineers - and his way of
driving them - could get it done so quickly. He
had little intellectual basis on which to judge the
difficulty of software or hardware tasks, which
often helped us pull the wool over his eyes. How it
all looked to him I cannot say . Eventually his impossible management style became so well known
that Sculley and the board of directors of Apple
had to remove him from all functional duties in
the company. My memo had finally been acted
upon.
JOBS AND THE PIPE ORGAN
Here's another story that doesn't quite fit in anywhere, but gives some insights into the interpersonal dynamics of the time. When I first started
working at Apple, Jobs and I would take long
walks (probably like the much-reported walks he
would later take with John Sculley). I remember
giving him mini-lectures on the philosophy of science or the performance practices of early music.
On one of these walks I shared with him my lifelong ambition to own a pipe organ. I explained
that the valves to the pipes in many organs were
driven electrically, and I hoped to hook up an
Apple II to one which would turn it into a modern
player organ. (I published an article with details in
Byte.) He asked me why I didn't have one; I told
him that it was mainly a matter of space, and that
there was a secondary consideration of cost.
Jobs had a suggestion: if I could find an organ I
could afford, I should buy it and Apple would let
me put it up in the lobby of the new, large building on Bandley Drive, in Cupertino CA. I would
hook it up to an Apple II, which would play it for
visitors. After hours, I could practice on it, or even
give company concerts.
Jobs was excited about the idea and told many
people about the organ that was going to be installed. With this encouragement, I searched for an
organ in earnest. In a few months I got lucky and
found an abbey where the organ was being
replaced. I told Jobs the good news and he congratulated me on the find. I purchased their old
organ, had it crated and moved onto the abbey
lawn (no small task in itself) and called Jobs to tell
him that the crates with the organ would be there
in a day or two.
Page 33
"What organ?" he asked. "The pipe organ we're
putting up in the lobby," I replied, thinking that he
must be distracted to have forgotten. He first said
that he had changed his mind, because it looked
like space would soon be tight; when I suggested
that it was a bit late to change his mind, since I had
already purchased the organ, he retorted that he
had never agreed to have the organ installed at
Apple in the first place.
When I got back I reminded him that he had made
a commitment, and that I had gone to some trouble and expense based on his assurances. He told
me that he had never assured me that he would
give me room for the organ, and refused to speak
in my presence to the people who had heard his
promises. I asked if, until the issue was resolved, I
could store the crates in the still-empty buildings.
The crates were outdoors, this was an imposition
on the abbey, and if it rained, the organ could be
seriously damaged. He simply said no.
I was stuck. The organ was too large to fit into my
house or even a rental storage unit. I called every
organ builder, organ teacher, and church I could
find and, after much desperate work, found a
church in Santa Clara that needed an organ. They,
in turn, found a benefactor to purchase the organ
from me for them. After long negotiations I succeeded in selling it for a fraction of its value.
This was my introduction to the new Steve Jobs,
or perhaps a phase of the old Jobs I hadn't yet seen.
Apple's first employee and his friend of many
years, Dan Kottke, who had traveled with him in
India and worked with him day and night to help
Apple get started, was treated even more shabbily.
In late 1980 Dan was surprised to find out thatin spite of their long friendship and the many uncompensated hours he had put in - he was not
going to get any stock options. Later W oz gave
some stock to Kottke, and to some other deserving
people from the early days of Apple (such as Bill
Fernandez, Chris Espinosa, Randy Wigginton,
Cliff Huston and Dick Huston) all of whom Jobs
had t1,lrned his back on. Woz's was an admirable
act of pure generosity.
As for me, I still don't have a pipe organ.
This is a preliminary version of a portion of a
book in progress. Comments and corrections are
welcomed. Please send them to jefraskin@aol.com.
Page 34
The Analytical Engine
SPEAKING OF ENGINES....
Joel Shurkin has updated his eminent survey text
of computer history, Engines of the Mind, and W.
W. Norton has done us all the favor of publishing
the new edition in paperback. Whether you've
read this book before or not, you may properly
rejoice that it's available once again; one of the few
respectable attempts to pack its topic into a single
volume, Engines of the Mind is compelling, energetic and well-annotated. The new paperback is
priced at $US13.00 and its ISBN is 0-393-31471-5.
Go hound your favorite bookstore.
FTP SITE FOR COMPUTER
CONSERVATION SOCIETY
by Chris P. Burton
The Computer Conservation Society, a special interest group of the British Computer Society in
association with the Science Museum of London
and the Museum of Science and Industry of Manchester, now has an ftp archive site.
All issues of the Society quarterly bulletin,
"Resurrection", are available in several formats, as
well as a small collection of simulators of historic
machines. Many of these simulators have not been
available previously, and more are "in the pipeline", awaiting documentation from their authors.
To access the archive, make an ftp connection to
< ftp://ftp.cs.man.ac.uk/pub/CCS-Archive>.
World Wide Web pages will be announced at a
later date. Meanwhile, readers may be interested in
WWW pages of the work on Colossus at Bletchley
Park, where the CCS has an exhibition room. The
URL is < http://www.cranfield.ac.uk/CCC/
BPark>.
May 1996
THE FRIDEN EC-130:
The World's Second Electronic Desktop
Calculator
(With some notes about the world's first.)
by Nicholas Bodley
BACKGROUND
In late 1963, advances in electronic technology
made it practical to build and market an electronic
desktop calculator. The world's first was the Anita,
made by Sumlock Comptometer (Ltd.?) in
England. The Anita had a beautifully built full
keyboard; each digit place had a vertical row of 10
(or 11) keys, in a row extending toward and away
from the operator. It probably had 10-digit input,
which would imply 100 keys for digit entry. (The
eleventh was a column-clear key, most likely; a
zero didn't need to be explicitly entered.)
The display used shaped-cathode neon glow numerical-indicator tubes, known among the technical community in the USA as "Nixie" tubes; this is
a trademark, probably of Burroughs. The internal
logic used beam-switching decade-counter tubes;
one variety (not necessarily that used in the Anita)
had thirty cathodes. Only one cathode would have
a glow discharge; a series resistor in the anode
circuit kept the voltage below that required for
other cathodes to conduct. Ten cathodes kept the
count;·the other twenty were connected into two
groups of ten, and a two-phase clock stepped the
glow to the next "stable" cathode in two stages by
temporarily "stealing" the glow from the "stable"
cathodes.
The Anita performed all four functions, but
because each digit was costly, it wasn't designed to
display all digits of a full product of a multiplication. This wasn't too bad if the operands were integers; but it did have decimal-point logic, and the
decimal point sometimes appeared in a bizarre
place when mixed operands were multiplied - it
wrapped around the end and popped up in a logical, but peculiar, place. Speed was of no particular
consequence; an "all nines" multiplier imposed no
special wait.
About a year after Friden introduced the EC-130,
Marchant brought out a nice, rather compact
machine called the Cogito, with a display like that
May 1996
The Analytical Engine
of the F riden but with peculiar half-size zeroes that
looked quite odd.
Monroe was the other of the "Mechanical Big
Three" in nonprinting desktop calculators in the
USA, but seemed late in bringing out an electronic
desktop machine; however, they had a brilliantly
designed mechanicga} printing calculator, in two
units cabled together, that was a successful competitor to the EC-130. The model name was PC1421.
THE FRIDEN EC-130
Page 35
ARCHITECTURE
When the EC-130 was designed, ICs were hopelessly expensive; I recall that a Westinghouse DTL930 NAND gate, with perhaps four inputs, cost a
big chunk out of $US50 each. The EC-130 design,
in a militantly cost-conscious setting, was all discrete PNP germanium diodes and transistors. (To
duplicate the internal logic of one of today' s calculator chips with discrete components would - at a
personal, rough guess - result in a box too big for
a desktop, which might draw a few hundred watts,
and cost about $10,000.)
Having sketched in the contemporary competition
to Friden, we can proceed to recollections concerning the design and success of the Friden EC-130, a
very significant product in the history of calculators. The author, Nicholas Bodley
(nbodley@tiac.net as of this writing) was one of the
original eight technicians to be trained at the factory on this calculator. It was fascinating, exciting,
and completely memorable to see this device for
the first time; the EC-130 was beautifully styled
although, in my opinion, some of the prototype's
striking quality was lost in the translation to production dies.
This calculator preceded Hewlett-Packard's desktop machines, and was the first to use RPN, which
Friden called simply "PN". (No, HP wasn't the
first to use RPN, and they have never claimed so.)
The EC-130 had a four-register visible stack functionally very close to that of the HP-48, for example. A stack drop caused zeroes to enter the top. It
even had a "Last x" register, for repeat multiply,
but the contents were not displayed. The Enter
key worked exactly like that on the HP calculators.
EXTERIOR PACKAGING
DISPLAY
In external appearance, the EC-130 was a distinctive, rather low box with a gracefully curved top
cover over the electronics, and a vertical panel
(mostly blank) with an extended glare shield
around the display face. The keyboard extended
across the full width of the machine; it was a 10key, serial-entry type, with various function/ control keys on both sides. The keyboard panel sloped
up toward the vertical front panel. The rear was a
finned die-cast heat sink of generous proportions,
although power consumption was modest. Overall
"footprint" was that of a fairly large electric typewriter, although perhaps longer from front to
back. Friden's chief engineer, the wonderfully capable Robert Ragen - one of the most brilliant
people I have ever met, and a pleasant, rather selfeffacing fellow in the bargain - was responsible for
the remarkably innovative architecture of the EC130, which was produced at a 1963 list price of
roughly $2,100.
Output was displayed on a type 5DEP1, 5-inch
round green-phosphor electrostatic-deflection
CRT, with about 2 kV accelerating voltage. (All
computer CRTs use magnetic deflection.) The
characters were seven-segment, similar in appearance to those now universal in inexpensive LCD
calculators. The beam of the CRT was swept
across the screen to write the strokes that made up
the individual characters; it was a vector scan, not a
raster scan. The deflection waveforms were wondrously complicated, and always the same. (The
slant of the characters was created by just one resistor that cross-coupled the vertical deflection into
the horizontal, without a buffer amplifier; it was a
marvelously simple yet subtle circuit. There was
no visible cross-coupling the other way, and it
didn't seem obvious from. looking at the circuit
how the isolation was done.) Individual digits were
created by unblanking (turning on) the CRT beam
at the appropriate times; the decoding matrix used
a remarkably small number of diodes, probably
Reverse Polish Notation (RPN)
Page 36
The Analytical Engine
about 80. There was a decimal point visible in each
register.
DECIMAL POINTS
Marketing decreed that logic for a floating decimal
point, universal in contemporary calculators,
would be too elaborate for inclusion in a discretecomponent design. The EC-130 did have decimal
points and a decimal-paint-entry key, but the
display had a selectable fixed point, the same for all
registers of the stack. The user had to decide how
many decimal places to work with, then make the
selection through a rotary switch with an edge
wheel knob projecting through a slot; about six
choices were available. I have seen two different
sets of [numbers of places] in different machines.
Internal logic permitted any number of places
within the limits of the machine, but the switch
was the constraint. Someone must have hacked a
freely-selectable decimal selection at some time.
The machine had 13 digits and, through repositioning of the decimal point, could provide all 26 digits
of a product. Overflow and/or truncation naturally occurred in such cases. It could work as a
purely fractional or purely integral machine. Division required 1.050 seconds for an all-nines
quotient; an all-nines multiplier was slightly faster.
KEYBOARD
The keyboard was serial-entry of the type called
10-key.lts unique mechanism was borderline practical as a design. Each keystem was part of a
stamped piece of steel, which included a "blade"
with a rounded bottom edge extending from front
to back. Pressing a key moved this blade down
against ramp-shaped recesses in seven (or eight?)
code bars made of stamped phenolic laminate and
positioned crosswise.
Each code bar had a small magnet attached to it,
which operated a reed switch. As I recall, the code
bars had no return springs but were actuated positively in both directions. One code bar was actuated for any of the ten digits, and reset for other
keys, resulting in the terms "common function"
and "common digit" frequently used by the tech.
folk. There probably was one code bar reserved to
distinguish digits from non-digits. A small, fast
electromagnet with an armature locked the codebars in place - and the keys as well, in some way,
May 1996
- until a time-consuming operation such as multiplication or division was complete. It always was
pulsed, but most operations were quite fast, and it
served as a keyclick noisemaker most of the time.
Some malfunctions could lock the keyboard; it
wasn't rare for a defective machine to have its multiplication key, once pushed, lock down and stay.
All functions and digits were coded, with codes
that specifically allowed minimal component count
in the electronics. The "touch" was quite acceptable despite the oddity of the design.
INTERNAL STORAGE
Like all desktop calculators of its time, the EC-130
required too many digits of internal storage to
allow use of discrete-component flip-flops for data.
Static RAM ICs came quite a few years later, even
after the heyday of serial shift-register ICs. (One
Toshiba machine that came out a few years later,
the so-called TOSBAC, stored in discrete capacitors with refresh circuits; the basic principle was
that of dynamic RAM! Shades of the AtanasoffBerry machine ...)
The Friden machine stored data in a low-cost implementation of the ultrasonic wire delay line, a
truly serial storage. These devices had been used in
expensive systems, with tight control over delay
time and clock frequency, for mass storage of binary data; apparently the idea was to make the
delay time some large multiple of the clock period,
and control it to within a fraction of one period.
F riden took a simpler approach. Once all the digits
had been clocked into the delay line, the timing
chain (cascaded binary counter stages) did a carry
(overflow) out the most-significant flip-flop, and
shut down the clock - so to speak, although the
oscillator continued to run.
The first pulse written onto the line was a dummy
pulse; when it came out of the line again, after a
delay of four milliseconds, it effectively restarted
the clock. Short-term drift was accounted for, and
kept within good bounds. The timing chain had
three fast stages run by the oscillator; these were, as
I recall, jammed to zero by the start pulse, to force
the apparent clock phase to be in step with the
start pulse and succeeding data. (This technique is
probably embedded inside every UART chip.) The
actual clock rate - not the oscillator rate - was
330 kHz.
May 1996
The Analytical Engine
Since pulses on the line were retimed and rewritten
with every "latest" clock frequency, modest medium-to-Iong term drift was of no consequence. A
fairly large timing gap, between the last digit
written and the start pulse, allowed for both longterm droop in clock frequency and mechanical tolerances of the delay line. (The oscillator used a discrete molded inductor and maybe a couple of silvered-mica capacitors, perhaps a Colpitts circuit.
Quartz crystals weren't needed, and ceramic resonators were probably 25 years in the future.)
The line itself was a subassembly on the bottom of
the calculator, a flat spiral of about eight inches
(20cm) radius and roughly a dozen turns. The wire
itself was mild steel; it was carefully selected, but
no exotic alloy was needed. The spiral, supported
by soft silicone rubber sheets with punched holes
and loading slits, sat in a shallow sheet aluminum
tray/chassis with a huge hole on the center.
This type of delay line stores torsional pulses with
a duration of a very few microseconds at most, and
an angular magnitude (probably) substantially less
than one degree of arc. Even given the short duration and small magnitude, stress on the wire was
probably relatively high. Several thousand such
pulses could be launched into this wire, and remain
adequately discrete at the far end. (Whether soliton
phenomena are involved, I don't know, and
probably the original designers didn't, either; I
suspect solitons are a comparatively recent discovery.)
Pulses were launched with magnetostrictive tapes,
probably of pure nickel, and welded carefully to
the exact end of the wire -tangential to the
surface, and at right angles to the length, of the
wire. (If you hold a pencil between your thumb
and forefinger, and let it droop, the tapes are your
fingers and the pencil is the wire.) My recollection
is that each side of the wire used two tapes for engmeenng reasons.
These tapes were passed through the bobbins of
two tiny coils, positioned close to a permanent bias
magnet. The magnet's field made the tapes shorter,
by perhaps a few parts per million, than they were
without the field. Pulses to the coils canceled the
field for one tape, and doubled it for the other. In
probably several hundred nanoseconds, the longitudinal stress pulses traveled to the end of the delay
wire and gave it a sudden twist, followed a microsecond or two later by a relaxation to normal.
Page 37
Stress pulses in the tapes would also travel the
other way, but these superfluous pulses would
reflect from the ends of the tapes, and be absorbed
by sheets of silicone rubber which also functioned
as a support.
Later printing calculator designs (with no display,
just a vertical stack of four lamps to show non-zero
contents) used IC logic and a smaller-diameter coil;
tapes were "single-ended" rather than of the earlier
push-pull type. Apparently they worked quite well
enough. These machines had the Singer logo on
them.
At the other end of the delay wire, a second similar
transducer converted the torque pulses to longitudinal ones; the inverse magnetostrictive effect, fed
into another permanent magnet, developed adequate signal in a second pair of coils.
Pulses coming off the line had a shape reminiscent
of the wavelet sombrero function. (I recognize that
there may be a more formal name for these.) The
pulse takes a negative-going rounded dip, returns
and crosses the zero axis, continues to a peak
maybe three or four times as high as the dip, and
then falls back to a second dip like the original. It
settles quickly afterward. Simple gain stages and a
slicer convert such a pulse to a clean rectangle at
logic level.
INTERNAL DATA REPRESENTATION
The internal representation of the digits was not
BCD; it wasn't really coded at all, but became
radix-one on the wire. Digits one through nine
were represented by strings of pulse count equal to
the digit being represented. (A two was two pulses;
a nine, nine pulses.) Each digit was given its own
time slot in the total data time of 48 llSec. A lack of
pulses occurring at a given time was interpreted as
a zero digit.
Remember that the timing [counter] chain was
started by the first pulse out of the line after an
"end-around-carry" stopped the counters. This,
combined with good short-term stability of the
clock, meant that the "number" in the counter
chain defined the identity of the digit.
I no longer recall how many internal 13-digit
numbers were kept in the calculator. Four stack
registers were all displayed; two others, not displayed, were a store/recall and an arithmetic register. One was a "last x" register with space to store
The Analytical Engine
Page 38
May 1996
one 13-digit number, and there were Store and
Recall keys.
5
=
11111
6
=
11110
Numbers were written onto the delay line in a
sequence with the "hidden" registers first, then
progressively up the stack; all LSDs were written
before the next digit. Decimal points were not
stored since their location made no difference in
addition and subtraction; multiplication and division decimal settings simply affected where the
result digits were placed.
7 = 11100
= 11000
9 = 10000
8
This code is "wasteful" in that 22 of 32 states are
disallowed, but in the calculator, it permitted a fast
bit rate on the delay line and use of particularly
simple, low-cost logic to count the pulses coming
off the line. This logic is called a Johnson counter
or "switch-tail ring counter;" it's basically as-bit
shift register with a parallel reset for all stages to set
it to zero. The normal and complement serial outputs are connected back to the inputs, but with a
half-twist, so that a zero going out one end shifts in
as a one at the other.
Once in the electronics, the digits took on another
unusual form, but before I explain that I'll pop the
real surprise: There was no adder in this calculator,
or at least not in the conventional sense. You
might well expect some such coding as 8,4,2,1
BCD, or excess-three, or 4,2,2' ,1, with a 4-bit combinatorial adder. But no ....
-
Seri al input
from del ay
line
A
========~
0
(I'm not sure, but I think [A .. D] are
Friden's designations.)
In the electronics, digits were represented by a code
related to the Morse code for each digit. It is and
has been used elsewhere and there are (or were) ICs
that counted in this code. It's a really simple
nonweighted code, which we might call the
baklava code:
= 00001
2 = 00011
= 00111
4 = 01111
3
C - .......... Seri al output
to del ay 1i ne
===(~)
Flip-flop arrays
1
====~
-
L.....,..
o = 00000
,.
B
The code bars in the keyboard defined the ten
digits by this code; "common digit," when asserted,
signified that the code was to be interpreted as a
digit. At the proper time, this digit was read mto
the arithmetic unit in parallel.
ARITHMETIC UNIT
The arithmetic unit contained three counters of
this type, and a simple 5-bit register with parallel
input and output. This totaled to 20 flip-flops, not
an insignificant circuit when built of discrete components. This shows the arrangement of the four 5bit flip-flop arrays, with one block for each group
of five.
In this illustration, the single lines are serial data
paths; the inputs to the A and D counters are logic
pulses from the delay line's read amplifier. These
May 1996
The Analytical Engine
pulses {when enabled by gates, which is usually}
cause the A counter to count up, and the D
counter to count down. {Subtraction!} Likewise,
the C counter feeds pulses serially to the delay
line's write amplifier until it counts down to zero.
Double lines represent 5-bit parallel data paths;
when these are enabled, the contents of a given
counter/register shift to the right, into the "new"
counter/register. This happens every time a digit
comes out of the delay line.
When the calculator is just sitting there, for each
digit, the A counter gets reset; then the serial pulses
from the delay line's read amplifier make it count
up. (The D counter probably counts {down} as
well, but nothing is done with its contents.) At the
end of this time slot, the contents of A shift in parallel into B. The next digit time, the digit in B
shifts into C. The next digit time after that, clock
pulses to the C counter make it count down and
feed serial pulses to the delay line until it counts to
zero.
The D counter's output is usually gated off; it
doesn't go to the B register. (That's why I put the
> in parentheses.)
To do an addition, the A counter is not reset when
it usually would be, and the next digit from the
delay line causes its count to increment "on top of"
the count already there. If it counts past nine, logic
detects the fact, and a carry flip-flop is set. {Extra
logic accounts for carries caused by a carry.} At the
next digit-column time, the logic increments the A
counter by one, to add in the carry. Similar things
take place during a borrow in subtraction.
There is at least one path not shown for parallel
transfers. My memory of this isn't clear, but I
think a bypass around the B register feeds the contents of the A counter directly to the C counter.
This makes a digit return to the line, but one
(register) time earlier than usual. The end result is
that the stack drops. Stack lift involves an extra
delay; I think the D counter serves as a plain register {i.e., not a counter} for that, again using paths
not shown. A given digit would probably take the
path Qine) to A to D to B to C to Qine).
Subtraction, as I said "prematurely", involves decrementing the D counter; it's connected/ defined
as such. I really don't recall the details, but they're
not any great mystery. None of the three counters
is bidirectional.
Page 39
Multiplication involved an extra (time-defined) 13digit register, and might even have involved shifting out multiplier digits from one end, then product digits into the other; the MQ registers of 1970sera computers come to mind. The D counter might
well have controlled the number of addition cycles.
The control logic, while fairly complicated, was no
more so than necessary - every resistor, capacitor,
diode, and transistor counted when determining
cost.
Division, as expected, performed consecutive subtraction (and tallied the number of subtraction
cycles) until underflow, following with a restoring
add-and-shift.
CONTROL
Control was of the state-variable variety. I dimly
recall a 3-bit control counter that advanced when
the next detailed stage of an operation was ready to
take place. There was no explicit diode matrix in
one spot for decoding logic states along with the
contents of the counter; the logic was simply embedded. When square root was added (vff.), the
control counter possibly had to gain another bit.
CIRCUITRY
Most transistors were 2N1305's, germanium PNP,
in TO-5 cans, except where the faster 2N1499's
(TO-1S) were required. The deflection output
stages were 2N2043A's, a popular RF power transistor. Logic diodes {DTL} were 1N662's. The circuitry was highly optimized; collector load resistors and base resistors, all 1/2 watt Allen-Bradley
carbon comps, were individually calculated. Capacitors were chocolate-colored dipped mica,
mostly, and also individually "calculated".
The CRT's high-voltage power supply was an inverter; it had a small transformer like a lowpowered vacuum-tube filament transformer connected "backwards", and a voltage multiplier chain.
The main logic power was quite conventional.
There were roughly 165 logic gates, perhaps 40
flip-flops in all, of which the timing chain had 14.
In total there were 300 transistors.
Of seven circuit boards, each about 12 inches by 5
inches {30x13 em,} six were paired with spacers and
(amazingly) lots of hand-soldered jumpers to join
the two at the edges opposite the connector edge.
One connector served both boards in a pair; the A
Page 40
The Analytical Engine
through D counter!registers occupied either one
board or one pair. Layouts were rather closely
packed. The boards mounted vertically and
plugged downward into edge connectors, in
standard fashion.
The EC-130 no doubt served as an inspiration for
the Hewlett-Packard desktop calculators; I remember how amazed I was to read, in the Hewlett·
Packard Journal of the time, about their scientific
desktop machine, which had no ICs either! That
HP machine, the 9100A, was quite as remarkable
in its own right as was the F riden. 1
MANUFACTURING
Circuit boards were double-sided, with plated
through-holes, and plating was done after drilling
as usual. The pattern of traces was "deposited" by
reverse electroplating in a unique machine developed by the inventor Gilbert Marosi. A copper
negative master for the trace was wrapped around a
drum and clamped. The drum was rotated in close
registry with the drilled and plated copy, while
electrolyte was flushed through a very small gap
between the master and the copy. The liquid went
in clear, and came out blue. The copper was immediately extracted and the electrolyte reused. The
current must have been quite high, perhaps hundreds of amps. Any remaining copper that had disconnected itself electrically was removed with a
quick etch. When this system worked, it worked
quite nicely.
Unfortunately, there was one inherent (and nasty)
contradiction. The process of plating the throughholes created faint anthill-like raised regions
around the holes, which were of no consequence to
the boards themselves, but they shorted out the
Marosi machine. The result was significant damage
to the master; apparently it wasn't easy to stop the
current before the damage was done. (Multihundred-ampere transistors were still in the
future.)
Manufacturing was, one would presume, in a
major quandary. They had no practical alternative
to the Marosi machine; setting up a conventional
photoresist/ etch line wasn't a job of a few days.
They decided to "level" the unpatterned boards by
following the plating step with a surface grind. The
1
See ANALYTICAL ENGINE 2.3, p. 6ff.
May 1996
shop knew more about surface grinding copper
than I do, but not quite enough, and the consequences were horrible. The circuit boards would
work fine until they were shipped to the branch
offices, even sometimes until they reached the customer's premises. Then the failures would begin as
the copper, fatally weakened by grinding, broke
and separated the plating inside the hole from the
trace on the surface.
At this time I was the "depot" service technician
for the Northeastern United States. A torrent of
failures descended on us, about 98% of which were
mechanical intermittents that necessarily resulted
from surface grinding; we had commendably few
actual part failures. We would get a set of boards
in, plop them into a test machine - stock, but
with extenders for all the boards - and start warping the boards. As connections began separating we
saw some wondrously peculiar symptoms, much
more fascinating than those typical of a more conventional architecture.
If we found more than, perhaps, two definite
symptoms, we opened up the pairs and I handsoldered every through-hole on every board with
great attention to technique. I still recall a goldplated transistor lead that was surrounded by a perfectly good-looking fillet of solder. The lead went
through the board to a pad on the opposite side,
which in turn connected to some other circuits;
but the lead was not connected to the pad on the
component side, even though it was surrounded
with a perfect-looking fillet. There must have been
a very thin layer of rosin or other insulator on the
lead, and the component-side pad had broken away
completely from the plating in the hole. Solder
continuity from bottom to top didn't help in this
case.
THE FRIDEN EC-132
After I left depot maintenance on the EC-130,
Friden introduced the EC-132, which offered a different selection of the number of decimal places,
and added square root. Square root was considered
a key feature by Friden; their mechanical Model
SRW did square root using the "fives" algorithm,
and their mechanical masterpiece, the Model SRQ,
did both squaring and square root. Squaring was
trivial on the EC-130; I suspect that the EC-132
also had a squaring key.
The Analytical Engine
May 1996
The internal architecture of the EC-132, as the
model number implies, was an elaboration of the
EC-130's, but the algorithm was probably new to
that machine. More modern calculator architectures aside, at the time of the EC-13x it made sense
to calculate square root by elaborating the division
logic (or algorithm) to perform a division-related
process whose "divisor" is constantly incrementing
in a controlled fashion, instead of remaining static.
The details of square root calculation by this
method become somewhat messy, but the general
scheme for incrementing the "divisor" is worth
describing.
It's been known for a long time that the sums of
the odd integers are the squares of integers, as a few
examples will make clear:
= 1 squared
1 + 3 = 2 squared
1
1+3+5
=
3 squared, etc.
This is the straightforward basis of "direct" square
root calculation; the messy details involve shifts
and preserving existing digits of the root as they
develop, along with the developing "divisor",
which gains another digit every time a root digit is
found. (When I say "direct", I have in mind
Newton's method, which progressively refines an
initial estimate. That algorithm was thoroughly
impractical for a calculator of the EC-13x era, and
is probably not a good one for any calculator. I
could be wrong, but what's taught in school sometimes differs wildly from the Real World.)
In the EC-130, incrementing a register by twos
isn't easy at best; it becomes hairier still when the
developing root digit becomes so large that a 9
must be incremented to an 11. The two-place representation of the current "divisor" digit becomes
especially messy. F riden came up with an alternate
scheme; people aren't generally in a blinding hurry
for their square roots. Division involves repeated
subtraction of a constant divisor; square root, as
noted, increments the "divisor". Friden decided to
subtract 1, first. If no overdraft, then subtract
three; but, to create three, they didn't simply subtract three. They subtracted one, and then one
more in a second cycle. This "one more than that"
scheme (my term), involving a pair of subtractions
for each increment of a root digit, was the key.
Page 41
Each square-rooting cycle of "division" was now
elaborated into a pair of subtraction cycles in each
of which the "divisor" was incremented by one.
Incrementing became much simpler as a result.
So:
Let's assume we want the square root of 16.
16 - 0
=
16 (overdraft not possible)
Subtract one more than (zero): Take
16 - 1 = 15. This has now effectively subtracted 1.
If no overdraft, increment the "divisor", which
now becomes 1. Also tally one count for the root
digit.
Now, do this:
15 - 1
= 14
and subtract "one more than" the 1,
14 - 2 = 12. This pair of subtractions has effectively
subtracted 3. If no overdraft, again increment the
"divisor", which becomes 2, and continue. Increment the tally for the developing root digit.
12 - 2
= 10
then subtract one more than 2:
10 - 3
=7
This pair of subtractions has effectively subtracted
5. There's still no overdraft, so increment once
again, to make the "divisor" become 3. Increment
the root digit again.
7-3
=4
and subtract one more than 3,
4 - 4 = 0 No overdraft yet, so increment the root
digit once again.
There have been four pairs of subtractions; each
time a pair created no overdraft, the tally was incremented, and now stands at four. The "divisor"
also equals four in this instance; the calculation is
complete.
Thus, incrementing by one with proper controls
can calculate by the "consecutive odd integers"
method. This case is simplified; in real life, that
nice zero remainder would probably be ignored,
handy as it looks, because it's a rare case. Most
square root calculations leave a remainder
(although no calculator I know of ever makes that
remainder available to the user; it wouldn't be of
Page 42
The Analytical Engine
much use). In real life, the incrementing process
would continue, the subtraction would cause an
overdraft, and the overdraft would tell the control
logic to get ready to calculate another root digit.
The best information I have is that the EC-132
always computed to a zero result.
In the logic, that repeated "one more than" probably was represented by an extra pulse fed to either
the A or the D counter after the digit in it was
passed on to be rewritten to the delay line. I'm
speculating here.
AFTERMATH
My boss, a decent fellow, finally had me transferred to Friden R&D in Rochester, which is another interesting tale I may tell someday. Somewhat sad; lost opportunity, but also mismanagement.
May 1996
What Was The First
PERSONAL COMPUTER?
An Exploration
by Roy A. Allan
DEFINING TERMS
Recent literature is inconsistent in defining "the
first personal computer" as the term is presently
understood. This has resulted, to an extent, from
limited awareness of some early products with a
claim to the title. Which computer was truly the
first to be personal in the modern sense?
R&D was an instance of youth being wasted on
someone young, and ended (back in 1966, I think)
with early corporate downsizing/R&D cutbacks. It
hit me early! Singer had bought Friden about the
time I first went to work for them, and progressively destroyed a very good company.
This article does not discuss the early personal use
of mainframe computers; it also excludes experimental computers, such as the MIT Memory Test
Computer, and minicomputers, because they were
not oriented to a consumer market. In an article
entitled "Is There Such a Thing as a Personal
Computer?" [7], Lawrence I. Press presented several criteria -such as architecture, software, physical characteristics and marketing - which would
serve to distinguish a personal computer.
One particularly interesting highlight of my time
in Rochester came in 1964, when Friden was unofficially developing an electronic Flexowriter. It retained the basic electric typewriter mechanism, and
even had a bus architecture, something quite advanced at that time. I wasn't directly involved with
the project. It would have considerably extended
the product lifetime of the Flexowriter, but management was, once again, fatally reluctant. The
prototype added to the pain by being built with
multipin connectors that looked great, but turned
out to be miserably unreliable. The resident genius,
Dave Frick, independently conceived of static
RAM chips, and perhaps dynamic ones as well.
This was when the first prototype dual-inline
packages appeared - and it looked so stupid to
standardize on a package that could be inserted
backwards ....
Thus, for the purposes of this article, the term
"personal computer" requires clarification. Up to
the mid-1970's, when the majority of computers
were mainframes shared by many users, a
"personal" computer was defined as being designed
for use by one person. As the technology progressed, our understanding of the term has
changed; we now consider some early hardware
too large and too expensive to qualify as being
"personal," and require that "a PC" should also be
small, easy to use, and of relatively low cost.
Today's personal computer has evolved into a
desktop appliance, available through the consumer
market, that provides ready and affordable computing power to an individual. In this context,
which of the early computers then called
"personal" should we consider for the distinction
of being "the First"?
1957: The IBM Auto-Point
John L. Lentz at IBM developed a small Personal
Automatic Calculator (PAC) in the late 1940's, and
described an engineering model of PAC in December 1954. This project evolved into the IBM 610
computer, described by the reference work "IBM's
May 1996
The Analytical Engine
Early Computers" [1] as being "IBM's first Personal Computer." The Model 610, also called the
Auto-Point, was announced in September 1957,
with a purchase price of $55,000. The computer
system consisted of a floor-standing cabinet - incorporating the electronics, magnetic drum, plugboard, and separate paper-tape readers and punches
- with a keyboard for input and an electric typewriter for output. It was not a stored-program
computer; a programmer entered instructions from
the keyboard, or input them through the papertape readers.
Scientists and engineers used the Auto-Point computer to solve small scientific and engineering
problems. IBM built about 180 units, but was not
then an aggressive competitor in the 'development
of small computer systems. Other organizations,
such as Bendix Aviation, Librascope, Digital
Equipment Corporation (DEC) and the Massachusetts Institute of Technology (MIT), were thus able
to participate in an emerging market and developed, respectively, the G-15, LGP-30, PDP Series
and LINC small computer systems.
1963: MIT LINC
Gordon Bell, a principal in the desigrt of early
DEC PDP-Series minicomputers, has stated that
the MIT LINC was the first personal computer.
MIT developed LINC - an acronym for
"Laboratory INstrument Computer" - to facilitate
the use of computer technology in biomedical
research laboratories [3]. Principal designers were
Wesley Clark and Charles Molnar. MIT demonstrated a prototype in early 1962 and completed
sixteen units by mid-1963, which were assembled
by scientific users to improve their understanding
of the system. LINC had four console modules, an
electronics cabinet and a keyboard. The processor
logic circuits used transistorized circuit modules
from DEC. An oscilloscope module could display a
512-by-512-point image. Memory was magnetic
core with a basic capacity of 1,024 twelve-bit
words; two magnetic-tape drives provided additional storage. Each system cost about $32,000.
Initial software was a text editor, an assembler and
some utilities. A small number of scientific laboratories used the LINC computer in dedicated applicatlOns.
The IBM 610 and the MIT LINC were "personal"
computers developed with reference to established
Page 43
larger architectures, at a cost which confined their
use to major scientific organizations. In the late
1960' s a company called Computer Terminal Corporation (CTC) took the opposite tack by starting
development of a "smart" computer terminal
which would incorporate a microprocessor.
1971: Datapoint 2200
An Invention & Technology article [11] quotes Jack
Frassanito as saying "I invented the personal computer ... " CTC hired Frassanito in 1969 to develop
a computer terminal that would have its own processor and other circuitry which would allow it to
mimic other major computer manufacturers' terminals. The self-contained unit included a 12-line
display, keyboard, two cassette tapes, a Texas Instruments bit-serial processor supported by TTL
logic, and 8K bytes of internal memory. CTC introduced the resulting Datapoint 2200 in June
1970, and shipped the first units in early 1971.
Although CTC designed the unit purely as a terminal with unprecedented capabilities, a number of
commercial users wrote programs and used it as an
early, small, desktop computer. The Datapoint
2200, however, was still not made available at a
price attractive to an individual user. The first microprocessor-based computer affordable for the
general public was the comparatively unheralded
Kenbak-1.
1971: Kenbak-1
An Early Model Personal Computer Contest,
sponsored by the Computer Museum of Boston,
ComputerLand and CW Communications in 1986,
selected the Kenbak-1 as being the first personal
computer. The criterion for selection of the winner
was "interest, significance and date of each model"
[5]. Designed by John V. Blankenbaker and introduced in late 1971, the computer used 130 small
and medium-scale integrated circuits and had a
memory of 256 8-bit bytes; it processed 1000 instructions per second. Input and output were very
limited, with no keyboard or screen. Blankenbaker
sold only 40 units at a price of $750 [4].
The Kenbak-l, with its primitive user interface,
stands in sharp contrast to history's next "first personal computer" - the stunningly innovative, feature-rich and costly Alto, developed by the Xerox
Palo Alto Research Center (P ARC) in California.
Page 44
The Analytical Engine
1973: Xerox Alto
According to Alexander and Smith's "Fumbling
the Future: How Xerox Invented, Then Ignored,
the First Personal Computer" [9], Xerox developed
the Alto computer as a research project in the early
1970's; construction of the prototype was begun in
November 1972 and completed in April 1973. Lead
engineers included Chuck Thacker, Larry Tesler,
Butler Lampson, Peter Deutsch, Bob Metcalfe and
several others. Intended sale price of a system was
about $30,000, and the production cost of early
examples was nearly that high. An Alto computer
system consisted of a main tower case intended to
fit under a desk, cabled to a grayscale bit-mapped
display screen, 8 inches horizontal by 10 inches
vertical, that could display 60 lines of 90 characters
- a full portrait page. Input was by keyboard supplemented by the "mouse" originally designed by
Douglas Engelbart at SRI. The processor was a 16bit custom-made unit and basic memory was 64K
16-bit words, expandable to 256K. One or two 2.5megabyte pack-type hard-disk drives were installed
in the main cabinet. PARC also developed a new
interactive programming language called Smalltalk
which was used to create a windowed graphic environment and the desktop metaphor that was an
unprecedented synthesis of hardware and software.
Furthermore, P ARC surrounded its individual
Altos with a resource pool that included an Ethernet network, shared laser printers, and electronic
mail. These were new human interface concepts
that eventually formed the basis for developments
by Apple Computer, Microsoft and others.
But the Alto, like computers previously described,
and like other contemporary systems including the
National Radio Institute NRI 832 kit, HewlettPackard HP 9830A programmable calculator, EPD
System One computer kit and IBM 5100 portable
computer, still relied on discrete components. Use
of discrete logic kept production labor-intensive
and limited the designer's ability to achieve a low
system price. Technological developments at Intel
Corporation after 1971 were about to surmount
this last obstacle to the affordable small computer.
1971- :Intel
Intel introduced large-scale integrated memory
chips and microprocessors in the early 1970's, providing the basis for the development of low-cost
personal microcomputers. The company released
May 1996
the 4-bit, 4004-based SIM-4 simulator board - the
first commercial product incorporating a microprocessor - in May 1972, followed by the 8008based SIM-8. The Intellec 4 and Intellec 8 development systems, or "blue boxes," followed in
August 1973 [6].
The SIM-4 was not a computer, but a design aid to
facilitate the development of microprocessor applications. Implementation of the microprocessor in
commercial computing really started with the
European Micral.
1973: Micral
The French Micral microcomputer developed by
REE (Recherches et Etudes Electroniques) was the
earliest personal computer to use a microprocessor
[2 & 5]. Truong Trong Thi managed the company
and released the computer, which sold for $1,950,
in January 1973. The unit included an Intel 8008
microprocessor and 256 bytes of RAM, expandable
to one kilobyte.
France had produced the first microcomputer.
Another year would go by before the first North
American microcomputer was advertised in the
March 1974 issue of QST, an amateur radio magazme.
1974-5: Early US Microcomputers
The first personal computer in the USA to use a
microprocessor was the SCELBI-8H, designed by
Nat Wadsworth and Robert Findley of Scelbi
Computer Consulting, Inc. in Milford, CT. Described in advertisements as "The totally new and
the very first - Mini-computer," the 8H used the
Intel 8008 microprocessor and up to 4K bytes of
memory. In kit form it sold for "as low as $440."
The price was right, but Wadsworth had health
problems, and a change in the company's first priority - to publishing - resulted in poor sales of
the computer.
The first "magazine project" microcomputer, called
the Mark-8, was designed by Jonathan A. Titus and
appeared in the July 1974 issue of Radio-Electronics
[10]. It used the Intel 8008 microprocessor and had
256 bytes of memory. Enthusiasts could buy plans
from the magazine for a nominal amount, or a substantially complete kit of parts for $350; sales were
insignificant. Today, just over twenty years later, a
May 1996
The Analytical Engine
Mark-8 is one of the most valuable and sought-after
of all historical micros.
It was the misfortune of both the SCELBI-8H and
the Mark-8 to stumble and fall in the darkness just
before the dawn. Only a few months later, a New
Mexico-based company called MITS - Micro Instrumentation and Telemetry Systems - could not
cope with the orders it received when it introduced
the Altair microcomputer.
1975: Altair 8800
The January 1975 issue of Popular Electronics [8]
featured the Altair 8800 developed by Ed Roberts
of MITS - although the unit on the magazine's
cover was only a painted, empty case. It used the
Intel 8080 microprocessor, and basic memory was
only 256 bytes. A lOO-pin bus with 16 slots facilitated expansion of the system with additional
memory and peripherals. Kit price by mail order
was $397. The Altair 8800 was a startling success,
and a flood of orders created severe delivery problems at MITS; in theory, the same computer was
available assembled and tested for $621, but customers who ordered a finished unit waited for
months while the factory caught up.
In retrospect, the Altair's position in the vanguard
of the microcomputer revolution is puzzling. The
layout of the hundred-pin Altair bus (more commonly known today as the "S-100 bus") was deeply
compromised and made the design of third-party
peripherals unnecessarily difficult. MITS' own addon products were sometimes poorly engineered,
like the 4K dynamic memory boards, or sometimes
were announced but never produced. And if the
hardware was attractively priced, the software was
not; a paper tape of Bill Gates and Paul Allen's
"Micro-soft" Altair BASIC interpreter sold for
$150.
Regardless, the combination of the Altair hardware
and the Microsoft BASIC interpreter was the first
runaway success of the "personal computer" industry. More sophisticated products with better promotion, like the Apple II computer introduced in
June 1977 and IBM's PC released in August 1981,
gave substantial computing power to the consumer
in the mass market. Which of these early computers we can honor as "the First" will depend on our
narrow and literal, or broad and inclusive, interpretation of the term "personal computer" -
Page 45
which has obviously meant many things to many
people over the past forty years!
CONCLUSION
The IBM 610 Auto-Point Computer was the earliest personal computer, but did not use the storedprogram concept. The MIT LINC used the storedprogram concept, but was a limited scientific
project with no impact in the general market. The
Kenbak-l, the first low-cost personal computer,
had a very limited interface and enjoyed few sales.
Xerox' Alto introduced many of the concepts
widely used in computers today, but was a
"personal computer" only in a rarefied environment and at prohibitive cost. The French REE
Micral was the first personal microcomputer to use
a microprocessor. The Scelbi-8H and the Mark-8
were the first US microcomputers, but did not
have significant success in the market. The Altair
8800 - with a more powerful microprocessor, an
innovative BASIC interpreter and an affordable
price - was "the first" personal computer to be a
commercial success and start a "personal" technological revolution.
BIBLIOGRAPHY
1. Bashe, Charles J., Johnson, Lyle R., Palmer,
John H. and Pugh, Emerson W. 1986. "IBM's
Early Computers." Cambridge: MIT Press, 1986,
pp. 505-508.
- Describes the Personal Automatic Calculator
(PAC) and the IBM 610 Auto-Point Computer.
2. Buchholz, Werner (Editor). 1988. "Was the First
Microcomputer Built in France?" Annals of the History of Computing, Vol. 10, No.2, 1988, p. 142.
- Discusses the Micral microcomputer.
3. Clark, W. A. and Molnar, C. E. 1965. "A Description of the LINC." Stacy, Ralph W. and
Waxman, Bruce W. {eds.}. Computers in Biomedical
Research. New York: Academic Press, 1965, pp. 3566.
4. Editor. 1989. "Early Small Computers." Annals
of the History of Computing, Vol. 11,No. 1, 1989,
pp.53-54.
- Discusses the IBM 610, Kenbak-1 and Micral
computers.
Page 46
The Analytical Engine
5. Juliussen, Egil and Isaacson, Portia. 1987.
"Computer Industry Almanac." Dallas, Texas:
Computer Industry Almanac, Inc., 1987, pp. 288289.
- Briefly describes the Kenbak, Micral and Altair
computers.
6. Noyce, Robert N. and Hoff, Marcian E. 1981.
"A History of Microprocessor Development at
Inte1." IEEE MICRO, February 1981, pp. 8-21.
- Describes the SIM-4 & 8 and Intellec 4 & 8 development boards.
7. Press, Larry. "Is There Such a Thing as a Personal Computer." Abacus, Vol. 1, No 2, Winter
1984., pp. 69-71.
- Describes "What really distinguishes personal
computers from larger mainframes?"
8. Roberts, H. Edward and Yates, William. 1975.
"Altair 8800 ... The Most Powerful Minicomputer."
Popular Electronics, January 1975, pp. 33-38.
9. Smith, Douglas K. and Alexander, Robert C.
1988. "Fumbling The Future: How Xerox Invented Then Ignored, The First Personal Computer." New York: William Morrow and Co., Inc.,
1988.
- Describes development of the Alto computer.
10. Titus, Jonathan A. 1974. "Computer! - Build
This Minicomputer Yourself." Radio·Electronics,
July 1974, pp. 29-33.
- Describes the Mark-8 microcomputer.
11. Wood, Lamont. 1994. "The Man Who
Invented the PC." American Heritage ofInvention
& Technology, Vol. 10, No.2, Fall 1994, p. 64.
- Describes the Datapoint 2200 "smart" terminal.
May 1996
CHACDONATES
EARLY MICRO TO
NIXDORF MUSEUM
On April 6, in a ceremony at the Santa Clara
Marriott Hotel, CHAC presented a Processor
Technology SOL-20 to the Heinz Nixdorf
Museumsforum fUr Informationstechnik. Participants were Kip Crosby and Edwin EI-Kareh for
CHAC, Joachim Wolf for the Nixdorf Museumsforum, and former Proc Tech designer Lee
Felsenstein.
The donation began with an inquiry to the Smithsonian History of Technology (SHOT) mailing list
by Nixdorf curators Ulf Hashagen and Dr. Karlheinz Wiegmann, explaining that one focus of the
new facility would be "the personal computer and
its development in the USA [and] the history of
pioneers and idealists, who got the development of
PCs under way." They were, at that time, looking
for an Altair 8800, an IMSAI 8080 and a KIM-l; we
were happy to suggest the addition of a SOL-20
which, given its Intel 8080 processor, rugged
design, and 1976 introduction, is certainly
"pioneers' development." Thanks to the generosity
of Al Kossow (see Acquisitions) we even had such
a computer to contribute.
The SOL will travel to Paderborn, DE, for exhibit
in Europe's newest museum of information technology. This complete remodeling of Nixdorf
Computer AG's original headquarters building has
resulted in 18,000 square meters of floor area for
the entire museum, of which roughly half will be
exhibit space. The remainder will be devoted to
conference rooms, a research institute, library and
archive, restaurant and museum store. This project
is underwritten by the estate of the late German
computer pioneer Heinz Nixdorf, through the
Westfalen Foundation; it will be "dedicated to the
historical development and present significance of
information technologies, and their impact on
culture and society."
Sounds like our kind of museum and, frankly, we
hope someday to be among the 200,000 visitors
that the Nixdorf Museumsforum expects to attract
each year. In the meantime, we're happy to send
one of California's earliest micros to this brandnew and truly fine European institution.
May 1996
The Analytical Engine
NEW COMPUTER MUSEUMS
UNDERWAY IN BAY AREA
by Frederic E. Davis
Founder and President,
San Francisco Computer Museum
The Computer Institute Inc., recently incorporated
as a nonprofit organization, has joined with the
CHAC to create the Bay Area's next major
museum of technology: the San Francisco Computer Museum. As envisioned, this Museum will
provide a global center for recording and presenting the scientific, historical, artistic, and cultural
aspects of computing. An accompanying institute
will offer scholars and researchers the opportunity
to extend the bounds of computer technology -a
profoundly important stage in the development of
human civilization.
The San Francisco Bay Area and its environs offer
an ideal setting for this facility. Silicon Valley is the
birthplace of commercially successful personal
computing and remains the world center for computer technology. The City of San Francisco, proposed as the site forthe facility, has been ranked
the number one tourist destination worldwide, is a
major global business center, and serves as a gateway to Pacific Rim countries that are definitive to
the computer industry.
The project has received enthusiastic support from
the San Francisco Redevelopment Agency and
from San Francisco's Mayor Willie Brown. In announcing his support for the museum project,
Mayor Brown said "The time is clearly ripe for
such a museum with its related educational opportunities, and San Francisco is clearly the place for
it. "
Page 47
mouse and graphical user interface in 1968, and the
West Coast Computer Faire, where the Apple II
and many other major computers were launched in
the late 1970s.
Based on the theme, "Computing: Yesterday,
Today, and Tomorrow," this first phase of the San
Francisco Computer Museum will feature exhibits
showcasing computing through history, hands-on
displays of state-of-the-art technology, and exhibits
of the world's premiere computer graphics and
multimedia art. It will be the first museum of computing in the world to have such a comprehensive
focus.
Later Phases:
MUSEUMS, THEATER, INSTITUTE
Later phases of the project have the goal of creating
a larger "World's Fair-style" exhibition that will
become a major tourist attraction for San Francisco. A major facility is planned that will contain
four distinct museums - The Computer Pavilion,
The Computer Jungle, The Computer Gallery, and
The Virtual Museum - as well as a multimedia
theater and auditorium, and the Computer Institute, an educational, cultural, and research resource
for both the general public and the computer industry.
THE COMPUTER PAVIUON
The Computer Pavilion houses a general exhibit
area and a number of company-sponsored pavilions. The general exhibit area traces the history of
computing and displays, where one can see early
computer artifacts and memorabilia. The company-sponsored pavilions provide a look into the
future of computing and are modeled after
"World's Fair" exhibits.
Phase One:
CIVIC CENTER MUSEUM
THE COMPUTER JUNGLE
Long-term plans are to develop the project through
several phases. The first phase, scheduled to start
within the next year pending final approval from
the City, will establish a museum facility in Brooks
Hall in San Francisco's Civic Center Plaza. Brooks
Hall is a 90,000 square foot facility and former
convention site that boasts a notable computer industry heritage; it was the setting for several significant events in computing history, including
Doug Englebart's seminal presentation of the
The Computer Jungle, less structured than the Pavilion, will include an informal exhibit area offering visitors hands-on access to a wide spectrum of
computer technology. Cutting-edge products, and
those still under development, will be showcased in
exhibits underwritten by the companies responsible for the technology. This will be a cross
between an "Exploratorium™ of computers" and
an ongoing computer trade show, without the sales
hype.
Page 48
The Analytical Engine
THE COMPUTER GALLERY
COMMPUTERSEUM
The Computer Gallery will exhibit computer and
multimedia art. Computers are enabling new artistic expressions that are not easy to display in a traditional art gallery. This will be one of the only
facilities worldwide with the technological horsepower to present a broad spectrum of computer art
- from multimedia masterpieces to computer animations and virtual reality simulations.
Grand Opening Season
July 6 - August 19
THE VIRTUAL MUSEUM
The Virtual Museum will extend the Computer
Museum into cyberspace, through the Internet and
other online services, and will also offer special
"exhibits" on CD-ROM and other deliverable media. Providing worldwide access to the Computer
Museum, The Virtual Museum will also serve as a
"community center" without borders, bringing
together computer users, scientists, artists, vendors,
and enthusiasts in real time.
THEATER AND STORE
The San Francisco Computer Museum complex
will contain a multimedia theater and auditorium
and an adjacent museum store. The multimedia
theater will support live presentation of new works
by multimedia artists, as well as previews of new
software and hardware, and other visually intensive
events difficult to present in a standard theater.
Used as an auditorium, the theater will sustain an
ongoing schedule of cultural and scientific presentations by scientists, artists, writers, and others.
The museum store will offer a selection of computer books, software, art, and high-tech souvenirs
to help generate operating revenue for the facility.
THE COMPUTER INSTITUTE
The Computer Institute, which will operate and
administer the San Francisco Computer Museum
facility, will also sponsor resident fellows in the
arts and sciences, who will pursue research into
multimedia and computer-interface design. The
Institute, unlike research labs operated by private
companies, will place in the public domain all
software, research, and design work produced by it
and its fellows.
May 1996
by Kevin Stumpf
Curator, Commercial Computing Museum
Fact: because of 18 people Canada can now boast it
has a computer museum.
I am so proud and happy to write an actual review
of the Commercial Computing Museum's first exhibition season. I mean it actually opened and
people actually came and those people actually
enjoyed themselves.
The exhibits were built around mainframe systems.
The theme was "input- process-output." The slogan
was "A Remarkably Small Museum For Remarkably Big Computers."
Despite the venue and the lack of advertising
(because of the lack of a budget), many people
(adults, youth, and families) took the time to visit.
People came from all over. Many people came
from the Toronto and Hamilton areas (about a 60
minute drive), and we were gratified when people
from New Jersey and New York arrived. Of
course the locals were very supportive, but we
didn't see as many teens as we had hoped. Attendance isn't the only indicator of the success of such
an endeavor, because when visitors spend several
hours browsing, investigating, photographing, and
discussing the exhibits something must have been
done properly.
The Waterloo County Board of Education provided the venue. We operated from July 6 until
August 19 from the gymnasium of University
Heights Secondary School in Waterloo, Ontario.
This was an unusual place to operate a museum,
and no doubt some people didn't take us seriously
and refused to visit.
The Board provided the facility rent-free since our
mandates are both educational in nature. I think
tax payers of the Regional Municipality of Water100 should be pleased that their board of education
acted in such a responsible and creative way to enable Canada's first computer museum to open.
Unlike other attempts at opening a computer
museum in Canada, we first obtained artifacts and
then opened. Nothing would have happened
though without the efforts of many dedicated vol-
May 1996
The Analytical Engine
unteers. No matter how many ancient computers
sat in storage, a group of energetic, helpful, and
interested volunteers were needed to wash them,
prepare them for presentation, move them, and
assist visitors. The Commercial Computing
Museum is blessed with such people.
Page 49
Book Review:
THE MICROPROCESSOR:
A Biography
Michael S. Malone
A much more complete review will be published at
the end of the year. Until then please visit our site
from time to time and stay in touch. Next year's
exhibition is already in the works and there are
many interesting events and fund-raisers on the
way.
New York, NY:
TELOS/Springer-Verlag, 1995
333 pages, $29.95 (cloth)
ISBN 0-387-94342-0
We also had a booth at Comdex Canada exhibiting
"The Art, Science, and History of Computer Control Panels." About 300 people stopped by. This
made July a very busy month. While traveling
back from Comdex I was interviewed for a CBC
radio show that aired on Labor Day.
In the Preface to this book, Mike Malone says that
he called it "A Biography" because "We anthropomorphize a lot of non-human entities in our
world .... people have long referred to their computers by their central processors .... rather than
their brands." Well, maybe, but it's hardly necessary to "anthropomorphize" the microprocessor to
recognize the trait it most shares with humanity a rapacious and half-blind destiny to flow into any
niche that might accommodate it, nudging any
number of apple carts into chaos on the way. I
might once have written this review on a typewriter, but instead I'm in front of a computer so
freighted with microprocessors that I couldn't
count them all. The computer would be bored just
listening to me type, so it's playing a CD on a
drive whose laser is held steady by a....you guessed
it. And I'm drinking coffee that I just heated up
with microwaves whose intensity was controlled
by a .... uh-huh. The microprocessor is a fit subject
for a biographer because it is pervasive - and invasive and transformative - in degrees almost appropriate to a living thing.
.
If you weren't able to visit the COMMPUTERSEUM you can still "picture yourself" there. Please
spend the time to examine the photographs taken
during the grand opening (www.sentex.net/
Nccmuseum). Send us your comments and
suggestions, as well as your mailing address if you
want a copy of the newspaper article about the
museum. It's an honest review of the exhibit and
idea.
Thanks to every member of the CHAC and of the
computer-historical community who gave generously of their support and good wishes during this
long project!
Reviewed by Kip Crosby
Now, of course, microprocessors have only existed
for twenty-five years. "What?" "Huh?" "No way!"
Okay, you're right - not quite yet, because the
silver anniversary of the Intel 4004 won't happen
till November. Yet microprocessors are now so
ubiquitous that you - whoever you are - don't
know how many you own. Even if you count the
ones you can think of (scholastic exercise to say the
least,) you'll miss the ones you don't know about.
They're called "embedded controllers" and you
have more of them than you can imagine. How did
we get here - and in a mere quarter of a century?
And that question, in its startling and sometimes
bewildering complexity , is the one that Malone
attempts to answer in a remarkably entertaining
book.
Page 50
The Analytical Engine
Like any good biography, this one requires unrelenting craft in its writing, since years of life are
condensed into a few hours' reading (talk about
lossy compression!) that still has to cover the high
points, tell an understandable story, pay homage to
the passage of time, and leave us with a coherent
image of the subject. But by taking the storyteller's
approach to the microprocessor, Malone gives
himself a hard row to hoe, because most biographers can rely on a certain baseline of community
that unites subject and reader. Not so this time,
and Malone has to begin by telling us what a microprocessor is, where it originated, how it's produced, how it works and how it fails. Imagine for
comparison the biography of an extraterrestrial,
which would have to include long sections on the
creature's anatomy and internal medicine, its languages, and the energy cycles and atmosphere of its
home planet. (The atmosphere of a microprocessor's "home planet" - the wafer fab - is not much
like what we breathe, being as dust-free as humanly
possible. Soon, Malone assures us, even air will not
suffice as a medium and fabrication will have to be
done in vacuum.) So this book already comprises a
small library.
Groundwork thus laid, it's time for the microprocessor's definitively tangled history, with its
elements of almost Shakespearean treachery (the
disintegration of Shockley Semiconductor and
founding of Fairchild was only a spectacular overture,) and its names worse than Tolstoy's - can
you tell Xilinx, Xidex and Zilog apart, in a hurry?
Only a Silicon Valley native could easily navigate
all this, but Malone traces the industry's incestuous
connections with such agility that we come to understand a great sea-change; the microprocessor,
changing everything, more than anything changed
its own industry, combining irreplaceable minds
into temporary companies, letting designers hop
from tilt-up to tilt-up while they amassed experience that earned fortunes again and again. One of
this book's finest explorations is of intellectual
property in its full flaming weirdness, which has
finally turned litigation from a weapon into a tedious kind of insurance.
These changes, once begun, are not spent; in fact,
they probably haven't reached full acceleration. At
the beginning of this book I grumbled when
Malone called the Intel 8080 the "invention of the
century;" by the end I realized that microprocessors are practically inventing themselves, and the
May 1996
century has four years to go. In the book's last
section, "Dreams of Light," Malone turns his clear
and quick style of inquiry to the concepts and
technologies - like fuzzy logic, neural networks,
optical circuits, and replacements for silicon - that
will drive forward the next consolidation of the
microprocessor revolution. As much as we've seen,
we await wonders that will stun us to silence again.
These few paragraphs are only a sketch of a book
that might as well be called "The Microprocessor:
An Archaeology, Anatomy, History, Ecology, Assessment and Prognosis." It's not a completely even
book, because the range of topics is an awkward fit
in a single binding. The author's talent for explanation, honed by a long and conscientious career as a
journalist, is tested to its limits by the deep wizardry of technical fine points. And if the illustrations are both excellent and necessary, the book's
overall design depends on contrasts that are sometimes distracting. Still, I don't mean to be too loud
about a few little kinks. Building adroitly on the
insider's perspective of Malone's first book, The Big
Score, this book is ambitious, energetic, and compelling; it tries to give a real, rigorous picture of
historical transformation still in wild process. The
Microprocessor: A Biography will appeal to any
reader who wants to understand integrated electronics as primal force.
May 1996
The Analytical Engine
ACQUISITIONS
From smoldering rumor to blazing truth - yes,
the CHAC is accepting hardware again. And
hardware for the ages! We need to know more
about a lot of these computers (and take a longer
look at some of them) but here's a barebone description, for the record, of our accessions in the
last quarter. Share the joy, 0 ye faithful. The
ENGINE's next allocation of time and space will
bring pictures and detailed stories of many of these
machines.
Apple Lisa 2 and Apple Mac XL; thanks to Craig
deRosa. These aren't quite from the computer
store. One of them has a slotted, transparent plexiglass front panel - we suspect made in a p.r0totyping shop, not hacked - and a ten-meg mternal
hard disk ....
Compupro homebrew; thanks to James Birdsall.
A Z80-based S-100 in a rack cleanly fabricated from
plywood and sheet aluminum. Lots of spare boards
including a Hayes modem; lots of software; docs
on everything, neatly alphabetized in manila
folders. The definitive museum-piece micro.
Data General Nova II; thanks to Max Anthony,
and we'll have details later.
GRID Gridcase 3; thanks to Mike Tassano. The
secret agent's laptop, in stealth-black cast alloy.
The orange plasma screen is so sharp that the pixels
are visibly square. Honestly a bit heavy by modern
standards, but it could double as body armor in a
pinch, and it looks so ....serious.
HP 2114A; thanks to Cliff Olson. One of the old,
old, original-original HP rack-mount minis, with a
single-digit prototype serial number. Beautiful in
that faintly patinated way that brings to mind, say,
a classic car. Stay tuned for a feature.
HP terminal and tape drive; thanks to Max
Anthony, and we'll have details later.
Intel MDS; thanks to the Tech Museum of Innovation. This is one of the later Intel development
"blue boxes," with an integral CRT and (we suspect) an i8086, rather than an Intellec. It came with
what look like full docs for Isis.
Page 51
Intertec Superbrain; thanks to the Tech Museum
of Innovation. An early attempt at an all-in-one
business micro, this combined computer/terminal
boasts dual Z80's, dual 8" floppies, a 12" CRT and
32K RAM. (Note to younger readers: In 1979 that
was a lot.) The whole package is clean, imposing
and humongously heavy. Not a bad computer by
the standards of the day, the Superb rain has faded
so far back into history that some people today
think it's a game machine.
Kaypro II; thanks to Craig deRosa. A second example of the well-loved blue-and-silver luggable,
this - like all spares - will make the rounds as an
exhibit computer. Although we haven't yet
checked the two computers side by side for minor
distinctions .... but then, were any two K-twos ever
totally alike?
Osborne One; thanks to Mike Tassano. A pristine
and "late," gray-cased example of the blockbuster
Z80 luggable that every laptop and notebook owes
its soul to. Only a few years after cozying up to a
desktop, micros - thanks to the Ozzie - were free
of even that. A true classic joins the roster.
Processor Technology SOL-20; thanks to Al
Kossow. Actually, two more of the handsome and
gutsy pre-Apple micros with walnut sides. With
AI's enthusiastic permission, we donated one to the
new Nixdorf Museum (see p. 46;) the other will
become an exhibit computer.
Seattle Computer Products 8086; thanks to Charlotte D'Amico. A connoisseur's micro on two
counts. First, this may have been the first production 8086 computer outside Intel. (Article topic!
Article topic!) Second, it was one of these - a close
relative at least - on which Tim Patterson developed QDOS, the precursor of MS-DOS; and,
probably, on which he alpha-tested Microsoft 8086
BASIC in May 1979.
Tektronix 8562; thanks to James Birdsall. This
rarely seen article is a rack-mount development
box based on a ceramic-pack DEC LSI-11. Built at
a time when Tektronix had an almost unrivaled
reputation for quality, the 8562 is of achingly beautiful construction. Since the outside is beige and
has the usual eight corners, we may display it with
the case off. Watch for a feature.
Page 52
The Analytical Engine
Xerox Alto; thanks to Al Kossow. YesllReally!
Absolutely! The CHAC has an Alto, with its
graphical interface, its Smalltalk, its fourteen-inch,
2.5 Mb hard disk .... This computer had stuff no
other computer had in 1974. This computer has
stuff your computer doesn't have now - like a fullpage portrait monitor. And all of this pales next to
the fact that it plays Galaxian. To sit and play at
the Alto, with Butler Lampson's black-covered
manual at elbow, is to enter a major time warp ....
in which the roots of LisaOS, MacOS, Atari GEM,
X-Windows, MS-Windows, etc., are finally laid
bare to the dazzled seeker. One of the few computers that can be totally ingratiating and still send
chills down your back.
May 1996
o
o
o
Figure 1.
Tech Corner:
DE-CRUFfING a POWER SUPPLY FAN
by Kip Crosby and Joan Piker
[Note: The procedure outlined here, if performed
correctly, will clean the power supply generally to
be found in a commodity desktop or tower computer. Neither the authors nor this publication are
responsible for loss or damage suffered as a result
of performing this procedure incorrectly, or performing it on a unit not of the design described
below. Caveat lector. ]
The fan in the power supply of an average desktop
computer leads a ghastly life. Expected to deliver
consistent throughput and constant uptime, it survives on a diet of spiky wall voltage and dust-laden
hot air. The hot air flows through the fan and does
its bit for global warming; much of the dustspongy gray gunk seeded with evil glitter - accumulates on the fan blades or, worse, within the
power supply box itself. Wouldn't it be nice, you
think, to really get in there and suck out all that
cruft? But it's not that easy ....
An inexpensive power supply is typically cooled
by a muffin fan mounted inside the power supply
box. The fan is covered by a protective grid of
either stamped sheet steel or welded wire. Four
bolts, one at each corner of the fan, run through
the protective grid, through tubes in the corners of
the fan shroud, then through speednut clips that
secure them. (Figure 1 shows a fan frame, with the
fan omitted for clarity, and one speednut in typical
position at the upper left.)
The dread begins when you realize that this entire
sandwich, held together by the bolts, can never be
removed from the tack-welded steel case. You have
to do all your work inside the case, while you
avoid dislodging the speednuts, which would fall in
and be almost impossible to retrieve. (And no, you
can't leave one in there, or you risk shorting out
the power supply circuit board.) This maintenance
is a challenge, but it's possible, and with a little
extra care, it can even be repeated when necessary.
T oals and materials list:
Screwdriver, probably #2 Phillips
Hex driver, probably 1,4"
Mini vacuum cleaner with wand tip
Bench rag
Large paper clip or other short stiff wire
Round toothpicks
Superglue pen
Double-face narrow (1,4 It) tape
Cotton swabs
Isopropyl alcohol
1.) Power-down your computer and unplug it. If
you're really prudent, let it sit for a while so the
capacitors can drain. Go use your other computer.
May 1996
The Analytical Engine
Page 53
in turn. Carefully clean each corner of the fan
frame, and the face of each speednut, with a cotton
swab and alcohol.
6.} Make sure each speednut is aligned with the
tube that its bolt goes through. Touch the edge of
the speednut with the superglue pen (as in figure 2)
to glue it to the fan frame. Hold the fan till the
glue dries.
7.} Using the minivac, or turning the box upside
down and shaking it, remove dust mercilessly. Get
the minivac's wand past the fan and into the box,
to clean the circuit board gently.
Figure 2.
2.) Spread the bench rag on your workspace. Disconnect the power supply's connectors from the
devices and main board, taking note of their orientation as you do. With the hex or Phillips driver,
undo and pull the (probably) four bolts that hold
the power supply against the backplane of the case
- not the four bolts at the immediate corners of
the fan frame. These will be outside the cutouts for
the fan and power cables. In a desktop case, the
supply is also probably held in by two tangs at the
bottom of the computer case that fit into recesses
in the power supply box. Pull the power supply
forward - away from the backplane and toward
the drive connectors .,- about three-quarters of an
inch (20mm); then lift it up and out, and set it on
the rag.
3.} Get all possible dust out of the fan, and off the
perforated sections of the box, with the mini-vac.
Put the box back on the rag with the fan protector
at the top.
4.} Bend the paper clip or wire into a slim hook.
Slip it through the fan protector and snag one
blade of the fan. Remove the four bolts that hold
the fan in place, making sure (as you withdraw the
fourth bolt) that you have a good grip on the fan
with the wire hook. Be careful not to knock the
speednuts out of alignment; they are now loose.
S.} Remove the fan protector. With it gone, you
can probably get a better grip on the fan with your
fingers than with the wire hook. Move the fan
around inside the box so you can see each corner
8.) Put a small piece of double-face tape over each
speednut. Set the box down with the fan facing the
edge of the table, make sure the fan is aligned
properly, and pull the fan back into place to stick
it to the inside of the box. Poke a toothpick into
each bolt hole, through the tape.
9.} Hang the fan protector over the toothpicks.
Replace each toothpick, one at a time, with a bolt.
Re-connect the power supply's connectors and
check any other cables you may have jostled.
You're done! The fan will be significantly more
efficient, and probably quieter. When you want to
do this again, you can just take out the bolts and
unstick the tape, without worrying about the
glued-down speednuts.
Page 54
The Analytical Engine
May 1996
APPLE MACINTOSH
QUERIES
APPLE ONE
Chris Bachmann, c-bachmann@nwu.edu, is trying
to compile a registry of owners of the Apple One
- partly to share increasingly scarce information,
and partly to find out how many boards still survive. Interested parties can reach Chris via e-mail
or care of Bachmann, 225 N. 3rd Street, Wheeling
IL 90090 USA.
APPLE LISA 2/10: MYSTERIES
I recently acquired a Lisa 2/10 from my school,
running an outdated version of MacWorks. I have
a few questions which I have been asking since the
purchase of the machine. If anyone here could lend
a hand, thanks in advance.
1. Under the CRT, there is a small panel listing the
serial number, Applenet #, and Manufactured #.
What do all these fields mean?
2. The hard drive makes a high-pitched squeak
whenever the arm moves out of and into the
parked position (I'm guessing, I know at least that
it's the arm). Is this normal? The disk doesn't seem
to be suffering, and no data is lost. If this needs to
be remedied, what can I do?
3. What is a good way to get that hard-to-reach
crud out of various cavities in the body? Would
Dust-Off or a similar product be a bad idea?
4. The keyboard is covered with that familiar
Apple keyboard crud that accumulates on all keys
of Apple keyboards (maybe others). How can I
remove this?
5. Out of curiosity, the fact that I have an old Mac,
and the fact that the Lisa makes a rotten Mac, I am
reinstalling the Lisa Office System. All I really
need is LisaTerm (which is a term program, right?
I've only seen the Office System once; my apologies if I am using incorrect terms), but does the Lisa
support any sort of networking whatsoever?
Gotta go, thunder ...
Tom Stepleton
ssteplet@artsci.wustl.edu
I have recently rescued a Macintosh 128 from a
trash heap. I have made what repairs I could, but
need some system software to try it out! Please
send me any information you have regarding
where I might get hold of some. I sure would like
to get this little toaster smiling again! Thank you
for your time.
Stephen Jones
swj0001@jove.acs.unt.edu
UNT
Box 7519
Denton TX 76203
(817) 243-5242
BOSTON IV
I'm looking for the company, or any info, on a
Boston IV home computer system. All I know is
that it had a 6502 processor. Can you help?
Thanks ....
Chuck
chuckS@psln1.psln.com
COMPUTER AUTOMATION
Does anyone at CHAC know of a place from
which I could purchase programming manuals for
certain obsolete minicomputers (without the associated hardware)? My interest is in one of the lesser
known manufacturers, Computer Automation,
Inc. They were probably forced out of the minicomputer manufacturing business by the advent of
the microprocessor, but used to be in Irvine, California. The models I am interested in are the
Alpha-16 and the LSI-4 minicomputers. Thanks,
Roy Campbell
75537.2422@compuserve.com
May 1996
The Analytical Engine
Page 55
IBM
DATA GENERAL NOVA 3
I recently acquired a Data General Nova 3 minicomputer. Anyone with memories or information
on this beast? I'd like to know more of its history.
Is it true that the Nova design was initially rejected
by DEC?
Edwin R. Parsons
edwin@peninsula.apana.org.au
I have a bunch of IBM unit record equipment including a large number of plug boards card sorters
with tubes and printers. When it was taken out of
service a few years ago it still ran. Any idea of who
might be interested in it and how much it would
be worth.
Thanks,
Bob Swartz
rs@interaccess.com
EAGLE PC-2: DOCS AND S/W WANTED
A rthur Bauman has recently inherited an Eagle PC2, an Intel 8088-powered California IBM PC clone
that can run either MS-DOS or CP/M-86. He
would appreciate hearing from anyone who has
docs or software for this machine, especially a
user's manual. If you can help, please write to him
at 124 Orchard Avenue, Mountain View CA
94043.
IMSAI
Do you know of anyone selling any IMSAI 8080's?
I am VERY interested in obtaining one with
floppy drives. Being the computer in the movie
War Games makes it quite interesting.
Thanks.
Nick
conartis@tiac.net
80 MICRO MAGS OFFERED
I have a complete set of 80Micro Magazines. The
magazine for the Radio Shack TRS-80 Model I, II,
III etc. The collection includes Issue 1 (1-80)
through Issue 74 (3-86) With a special edition
published Jan 1983. Wayne Green was the publisher.
I am looking for a good home for the magazines,
please contact me if you would like them or if you
know of an organization which would like them.
Regards,
Rick Hoover
rick_h@community.net
ENIACDOCS?
Does anyone know where I can find technical
documentation for the ENIAC computer? I would
be very interested in detailed descriptions of the
hardware and instruction set. Ultimately, I would
like to write a ENIAC simulator for the PC and
run some of the original programs if I could get
hold of them.
Jim Stewart
jkmicro@dsp.com
[8080's are relatively common because they were
gutsy and performed well, so people kept them.
The going rate for a functional one seems to be
about $300. Be warned that you mayor may not
want the floppy drives, which earned themselves
the name of "Pizza Oven;" you might be just as
well off with paper tape.]
MINIX/286
I have a couple of old 286s witch I wish to put to
slow redundant tasks. And to do this of course I
can not use DOS. I am looking for someone
willing to sell/ donate/pay-me-to-take a useable
copy of Minix (I figure everyone must be running
Linux primarily nowadays) that would run on a
286 machine .....
E-mail if you can help.
T.Camp
camp@industrial.com
Page 56
The Analytical Engine
NCR 1-9020
I have acquired an old (1978) NCR 1-9020 rack
mount computer, with ribbon printer, and 20
14inch floppies.
All the hardware seems to be OK, but the
SYSGEN tapes are corrupt. I have been unable to
locate a source for the tape which is required to
boot up the system.
Hope someone has an idea or suggestion. Any help
would be appreciated. Thank you,
Joe Mohnike
jmohnike@valleynet.com
OSBORNE ONE: DISKS, DOCS, MAGS
I'm looking for Osborne 1 disks, and any of the
Osborne 1 magazines that still might be out there.
Can anyone help? Thanks,
Bill Johnson
bjohnson@moa.com
P. O. Box 579781
Modesto CA 95357 USA
: DISKS, DOCS, or OFFER
I have an Osborne I in mint condition; looking for
software and manuals.
Alternatively, open to some sort of counter offer.
Susan Rosen
RRosen@postoffice.worldnet.att.net
SHUGART 14" DRIVE INTERFACE:
ANYTHING?
Does anyone have any information about the history of the interface used with the Shugart SA4000
14" hard drives, and information about any other
drives that were compatible with this interface?
Also, did the SAI000 interface, which, if I recall
correctly, is the predecessor to, and (about?) the
same as, the ST412/ST506 interface, precede the
SA4000 series hard drives? Can anyone provide
any information on the SAI000 hard drives?
I'm trying to figure out it the SA4000 interface was
meant to be an improvement upon the SAlOOO
interface, as it has a higher data transfer rate (7.11
May 1996
MBit/s vs. 5 MBit/s), and in that it uses NRZ
(non-return-to-zero) pulses for the data, bears some
resemblance to ESDI (although 10 to 15 MBitl s, it
also uses NRZ for the data, and uses MFM encoding, like the SA4000 and SAlOOO/ST506).
Also, out of curiosity, are there any others out
there, aside from those of us still using PERQ-1
graphics workstations, who are still using these 14"
Shugart hard drives (or remember using them)? I'm
sure that there must (hopefully) be some people
out there who are still using these drives with Z80,
S-100 bus, based CP/M systems as well, as there
was a Discus M26 Winchester Disk System controller used with these S-100 based systems. The
User's Manual for the Discus M26 (written by
George Morrow) is rather nice, containing schematics and source code. Is anyone still using the
M26?
Thanks very much in advance for any information
that anyone can provide about any of the above!
R. D. Davis
http://www.access.digex.netrrdd
WANG 320SE
I have a ~ 1967 Wang Model 320SE multi-user electronic calculator unit. It is approximately
60x20x13cm and weighs about l5kg. It has four
connectors for external keyboard-display units. I
would like to find more information about this
beast. Do you know where I might look? Any information would be welcome.
Thanks,
Scott Coburn
scott@bnl.gov
XENIX
Hey, HELP. I am looking for a copy of Xenix-286
Ver. 3.2 (2.3). I need the boot disk and utility disks.
I am surfin' the net for any help I can get. Thanks
Dennis Rapp
rapp@i-link.net
May 1996
The Analytical Engine
ZILOG Z8000: ONLINE DOCS?
Does anyone know where to find some online
documentation on the Z8000 family of CPUs? I
wrote email to someone that said Olivetti made a
computer based on one of the Z8000s, does anyone
out there know what it's called? I'm also trying to
track down a cross-assembler for the Z8000 (the
Z8002, specifically, which can only address 64K),
and I think there's a commercial table assembler
available that has the Z8000 in it, but it's $200 and
I was hoping to find a freeware or shareware
program that might have it. Any info would be
greatly appreciated!
Gary
garyd@haus.efn.org
PUBLICATIONS RECEIVED
In the interest of getting ENGINE 3.3 out the door
during the present century, listing of publications
is postponed until the archive ("heap") can be reviewed and reorganized. We hope to finish an updated list in time for 3.4.
Page 57
The Computer Journal, P. O. Box 3900, Citrus
Heights CA 95611. Dave Baldwin, editor.
Computer Preservation Society (Inc.), Ferrymead
Historic Park, 269 Bridle Path Road, Christchurch
8002, New Zealand. Abraham Orchard, secretary;
abe@voyager.co.nz.
Computer Technology Archive, Box 4376,
Stanford CA 94309. Bill vanCleemput, director.
East Bay FOG, 5497 Taft Avenue, Oakland CA
94618. Tom Lewis, president.
Hewlett-Packard Journal, Hewlett-Packard
Company, Box 51827, Palo Alto CA 94303-0724.
Richard P. Dolan, editor.
International Association of Calculator Collectors,
14561 Livingston Street, Tustin CA 92680-2618.
Guy Ball, Bruce L. Flamm, directors.
IEEE Computer Society, 10662 Los Vaqueros
Circle, Los Alamitos CA 92640. Bob Carlson,
director.
Lambda Software Publishing, 149 West Hilliard
Lane, Eugene OR 97404. David A. J. McGlone,
editor and publisher.
Lexikon Services, Box 1328, Elverta CA 95843.
lexikon2@aol.com. Mark Greenia, director.
ADDRESSES OF
CORRESPONDING
ORGANIZATIONS
Perham Foundation, 101 First Street #394, Los
Altos CA 94022. Don Koijane, president.
Amateur Computer Group of New Jersey
(ACGNJ), P. O. Box 135, Scotch Plains NJ 07076.
Joe Kennedy, president.
Santa Clara Valley Historical Association, 525
Alma Street, Palo Alto CA 94301. John
McLaughlin, director.
Australian Computer Museum Society, PO Box
103, KILLARA 2071, NSW, Australia. Michael
Chevallier, secretary.
Charles Babbage Institute, 103 Walter Library, 117
Pleasant Street SE, Minneapolis MN 55455. Bruce
Bruemmer, archivist.
Commercial Computing Museum (formerly Unusual Systems,) 220 Samuel Street, Kitchener ON
N2H lR6, Canada. Kevin Stumpf, president.
Computer Conservation Society, 15 Northampton
Road, Bromham, Beds. MK43 8QB, UK. Tony
Sale, secretary.
The Computer Museum, 300 Congress Street,
Boston MA 02210. No contact at present; use
www.tcm.org.
Page 58
The Analytical Engine
THANKS TO....
Aaron Alpar for his donation.
Charlotte D'Amico and Dan Lahey for handdelivering the Seattle Computer Products 8086.
Fred Davis for the Dell Dimension 486-100 now
serving as our full-time fax machine, among other
things.
George Durfey for his help with the Maiga Wolf
interview.
Andrew Eisner for the Macintosh SE/30; Frank
Freeman for the SE video board; and Gavin
Carothers for setting it all up with System 7.1.
Tom Ellis for ongoing support of MS-FrontPage
and MS-Windows NT; and for configuration of
our brand-new Ethernet network (yessssss!.~
Edwin Vivian EI-Kareh for his fomidable contributions as your Association's Tactical Director.
Don Koijane for his collaboration and support on
the lecture circuit.
Bob Ragen for fact-checking Nick Bodley's Friden
article.
Erich Schienke for fine-tuning this issue's graphics.
NEXT ISSUE / COVER ART
To begin with, it's the silver anniversary of the
microprocessor and that's worth commemorating ....there's an interview in the works with a
workstation pioneer.... maybe an excerpt from a
major new computer history title .... and a review of
Hafner and Lyon's new book on the origins of the
Internet. In your hands as soon as we can get it
there.
Cover: Maiga Wolf at work in London, Ontario,
1949.
May 1996
GUIDELINES FOR
DISTRIBUTION
The ANALYTICAL ENGINE is intellectual
shareware. Distribution of complete, verbatim
copies through online posting, Internet mail or
news, fax, postal service or photocopying is encouraged by the Computer History Association of
California.
Excerpting or brief quotation for fair use, including
review or example, is also permitted, with one exception: Any material copyright to or by a third
party and reprinted in the ANALYTICAL
ENGINE by permission shall not be used in another periodical or context, unless the permission
of the copyright holder is separately secured for
the new use.
Alterations, abridgments or hacks of the ANALYTICAL ENGINE which change the intent or
meaning of original content; or which contrive to
bring income to any person or organization other
than the Computer History Association of California; or which contrive to injure the Computer
History Association of California, its officers, contributors, volunteers or members; are PROHIBITED. Reproduction of the ANALYTICAL
ENGINE without its subscription coupon is
abridgment in this sense.
GUIDELINES FOR SUBMISSION
The ANALYTICAL ENGINE solicits manuscripts of 750 to 2500 words on the general topic of
the history of computing in, or with significant
reference to, the State of California. Articles
should focus on one interesting or illuminating
episode and should be written for a technically literate general audience. Submissions are welcome
from both members and non-members of the
CHACo Article deadlines are: July 15 for the November issue, October 15 for the February issue,
January 15 for the May issue, and April 15 for the
August issue.
Each author may publish a maximum of one
signed article per year. This restriction does not
apply to letters, queries, book reviews or interviews. Thank you for cooperating to protect diversity of voices and topics. Previously published material will be republished only in clearly attributed
May 1996
The Analytical Engine
quotations or citations; or when its publication in
the ANALYTICAL ENGINE will bring it to the
attention of a significantly broader audience; or
when the original publication is materially obsolete
or inaccessible.
Decision of the editors is final but copyright of all
published material will remain with the author.
The preferred document file format is Microsoft
Word for DOS or Windows, but almost any DOS
or Macintosh word processor file will be acceptable. Submit manuscripts on DOS 5.25" or 3.5", or
Mac HD (1.4) diskettes. Alternatively, please send
your article as ASCII or ISO Internet maiL Please
avoid submitting on paper unless absolutely
necessary.
The ANALYTICAL ENGINE
Volume 3, Number 3, May 1996
ISSN 1071-6351
newsletter of the Computer History Association of
California, is published four times a year - in February, May, August and November - at Palo Alto,
California.
This magazine is available both on-line and on
paper. Basic, domestic subscriptions are $25 electronic and $35 paper, with $25 deductible as a
charitable donation. For information on institutional, international, and low·income subscriptions, contact the Association at:
41S9-C El Camino Way
Palo Alto, CA 94306-4010 USA
Internet: engine@chac.org
WWW: http://www.chac.orgl
Page 59
NINES-CARD
WHO INVENTED "SOFTWARE?"
by Herb Kanner
In 1958, I managed by sheer chutzpah to wangle
myself an assistant professorship at the Institute for
Computer Research, University of Chicago. My
graduate degree was in physics, and all that I knew
about computers was that I had written some
programs for an IBM 650 in assenibly language, in
Perlis's IT, which was perhaps the first high-level
language on earth, and in the Bell Labs interpreter,
which simulated a three-address machine.
The Institute was building a transistorized computer (one of the first, if not THE first) on an
Atomic Energy Commission contract. So here
were these guys soldering transistors, capacitors,
and resistors .onto printed circuit boards and
quaintly referring to the stuff as "hardware." This I
found very amusing; my concept of hardware was
something you bought in a hardware store.
Because I was responsible for the initial programs
for this machine, e.g. a symbolic assembler, I
thought it would be a cute idea to put a sign on my
office door that read "Software Department."
I'm sure that the term "software" did not radiate
from my usage. But I'm also fairly sure that it was
not floating around the literature at that time. So,
at worst, I made an independent invention. It
might be interesting to try to track down the other
origins of the term.
CONTENTS
Editorial: THREE YEARS ..•....•.•.....••...•.••..••••...•.•••••.•.••.•.•.•...•.•...•••••...•...••.................•....•..••.•.................... 1
IN MEMORIAM: DAVID PACKARD ...•...........•.•....••.....•.....••......••..•....•.•.•..•.••..•.........•.•...•.........•.......... 3
THE COMPUTER WILL DO ANYTHING YOU TELL IT TO DO:
An Interview with Maiga Wolf by Hilary Crosby ••..••..•....•.....•.•.••.•.•..•.••..••••.........•....•............................... 5
Quick Take: RAMAC 40 th ANNIVERSARY! ........................................................................................... 20
Quick Take: GETTING FRA.MED IN 3D ••••...•..•..•....•.•.•.•..•••....•.•..........•........•.....••••.•.•..•.•...........•.....•... 20
THE MAC AND ME:
15 Years of Life with the Macintosh (Part 2) by Jef Raskin •......•...•.•.....•.•..•.•....•................•..........••....•..... 21
SPEAKING OF ENGINES ..•....•.....••...•..••...•......•.......•...•..•..•.•••...•..........•..................•..•.......................... 34
FTP SITE FOR COMPUTER CONSERVATION SOCIETY, by Chris P~ Burton ..•............................ 34
THE FRIDEN EC-130: The World's Second Electronic Desktop Calculator, by Nicholas Bodley ...•...... 34
What Was The First PERSONAL COMPUTER? An Exploration, by Roy A1lan ..................................... 42
CHAC DONATES EARLY MICRO TO NIXDORF MUSEUM ............................................................ 46
NEW COMPUTER MUSEUMS UNDERWAY IN BAY AREA, by Frederic E. Davis ........................... 47
COMMPUTERSEUM Grand Opening Season July 6 - August 19, by Kevin Stumpf......•..................•.. 48
Book Review: THE MICROPROCESSOR: A Biography, by Michael S. Malone ...•.•..•.••...•.•......•........... 49
ACQUISITIONS ••.••.•.•••••.•.••••••.•.•......•••••.•.•••..•......•.•.•..•.•...•••.•..........•.........•.•....••.•.•....••................:......... 51
Tech Comer: DE-CRUFTING a POWER SUPPLY FAN, by Kip Crosby and Joan Piker ...................... 52
QUERIES .........••••....•.................••....••....••............•••....•....•.•.....•.•....•.....•....•.•...................•••...•.•..••............. 54
PUBUCATIONS RECEIVED .••..•.•...•.•....•.•....•••••.•••..•.••••.•.•...•..•......••..........•.....•....•....•.....•.................... 57
ADDRESSES OF CORRESPONDING ORGANIZATIONS ................................................................. 57
THANKS TO .•..•••••.•••••....•.••.•.••.•.•.•.•.•.•.•...•.•....•.•..•.•.•..•.•.•.•.•••.•...•••..•.•....•...•.•....••....••...•••....•....•.....•........ 58
NEXT ISSUE / COVER ART ••.•.•.•••..••••.•.•..•.•. ~ ......................................................................................... 58
GUIDEUNES FOR DISTRIBUTION ..................................................................................................... 58
GUIDEUNES FOR SUBMISSION ..•.•.•.....•.•.••.•.•..•.••••.•.•.•.•.•....•...•...•.•..•.....••...•............•.•.•.•.•.•.•.....•...... 58
NINES-CARD: WHO INVENTED "SOFTWARE?," by Herb Kanner ................................................... 59
SUBSCRIBE!
US$7.00
UK£S.OO
12.S0DM
¥1000
Lee Felsenstein and Joachim Wolf examine the SOL-20 (page 47)
and. ...
The Computer That Could Do Anything • David Packard, 1912-1996
The Earliest PC's • Friden Calculators • Raskin's Mac and Me, part 2
Malone on Microprocessors • San Francisco Computer Museum
Dealing with Fan Cruft • Nixdorf Museum in Paderborn • and more!
The Computer History Association of California
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.3 Linearized : No XMP Toolkit : Adobe XMP Core 4.2.1-c043 52.372728, 2009/01/18-15:56:37 Create Date : 2014:11:03 18:58:07-08:00 Modify Date : 2014:11:03 18:15:13-07:00 Metadata Date : 2014:11:03 18:15:13-07:00 Producer : Adobe Acrobat 9.55 Paper Capture Plug-in Format : application/pdf Document ID : uuid:9ba670fd-f2dc-344c-938e-5e9c48f0e0bd Instance ID : uuid:4fe50c04-39c9-e94b-8f5e-8df5a5bff4ba Page Layout : SinglePage Page Mode : UseNone Page Count : 62EXIF Metadata provided by EXIF.tools