ae
The ANALYTICAL ENGINE
Journal of the Computer History Association of California
ISSN 1071-6351
Volume 1, Number 4, April-June 1994
Kip Crosby, Managing Editor
Jude Thilman, Telecommunications Editor
---------------------------------------------------
CONTENTS
EDITORIAL: _BOOTSTRAP!!_ ...................................... 1
BALLOTS ....................................................... 1
VAPORZINE? .................................................... 1
"JUST LIKE THE GOLDEN GATE BRIDGE" ............................ 2
MINI-ARTICLE: THE "MORE" COMMAND IN UNIX ..................... 16
POMONA ....................................................... 17
TURING, THOU SHOULDST BE WITH US ............................. 20
COLOSSUS RECONSTRUCTED ....................................... 20
IN MEMORIAM: AARON FINERMAN .................................. 22
LIVERMORE UPDATE ............................................. 22
MICRO MUSEUM TOUR AT UC DAVIS PICNIC DAY ..................... 23
ADIOS AMIGA? ................................................. 23
SPOTTER ALERT ................................................ 24
SPOTTER FLASH ................................................ 25
DESPERATE PLEA FOR STORAGE SPACE ............................. 26
DESPERATE PLEA FOR MONEY (AND THINGS) ........................ 26
AND SPEAKING OF MONEY.... ................................... 27
OVERVIEW OF BUREAUCRATIC PROCESSES ........................... 27
HELP FIGHT IRON DEFICIENCY ANEMIA! ........................... 28
APPLE CONTEST ................................................ 29
BOOK REVIEW: FROM DITS TO BITS ............................... 29
ACQUISITIONS ................................................. 32
LETTERS ...................................................... 40
QUERIES ...................................................... 46
PUBLICATIONS RECEIVED ........................................ 58
ADDRESSES OF CORRESPONDING ORGANIZATIONS ..................... 59
THANKS TO.... ............................................... 59
NEXT ISSUE ................................................... 60
GUIDELINES FOR DISTRIBUTION .................................. 60
GUIDELINES FOR SUBMISSION .................................... 61
NINES-CARD ................................................... 62
ADD MONEY, MAIL.... ......................................... 63
The Analytical Engine, Volume 1, Number 4 April 1994 Page 1
-------------------------------------------------
Editorial: BOOTSTRAP!!
-------------------------------------------------
In the last four months the ANALYTICAL ENGINE has done more
-- and more has been done for it -- than ever seemed
possible. Read POMONA and the articles, read SPOTTER FLASH
and the Overview of Bureaucratic Processes, and you'll see.
The Computer History Association of California is doing the
job it was created to do, and in the process, attracting
national and even international attention.
So much has happened in just over a year that, when we look
back, even our own beginnings seem scarcely visible. All the
more important to have clear and unsparing sight of our
future. It's time to invite the cooperation of colleges and
universities, of companies, computer industry workers,
teachers and students, of everyone who sees or _will_ see
that California's history of computing is worth saving. The
CHAC has become greater than we dared hope. It is still too
small, measured against the job that must be done.
Some of you who read this are subscribers. Many more are
not. If you are not, please consider: Subscribing, joining,
donating are the best things you can do, to preserve the
work that you yourself have done, and protect the history
that you love. The time has come to stand and convince the
mighty.
-------------------------------------------------
BALLOTS
-------------------------------------------------
Since mid-March, meaning at Pomona and afterwards, we've
distributed over a thousand subscription blanks with the
National Computing Science Day ballot on the back. Some of
those have come back to us -- not enough, but some -- and
the vote so far has been about six to one in favor.
We are not without regard for the opposing view, which is
roughly "Keep the Feds out of computing," but the Feds have
been in computing since pen touched paper for ENIAC, so
that's an answer to a question we don't think we asked. With
86% favoring the proposal so far, we have a clear mandate to
begin research. Meanwhile, keep those ballots coming!
-------------------------------------------------
VAPORZINE?
-------------------------------------------------
This issue of the ENGINE is 'way late. Read it and you'll
understand why. Nonetheless, it shouldn't have been.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 2
The schedule we try to keep for each issue is first-
fifteenth-first; so that, as an example, the July issue will
close for copy on July first, the e-mail edition will be
uploaded to online subscribers and the request daemon on
July fifteenth, and the paper edition will be mailed on
August first, typically reaching subscribers about a week
later.
There are two ways to make sure this schedule is adhered to.
First, we need volunteers -- preferably volunteers
experienced with Word for Windows or Word for the Macintosh
-- to help with editing, proofing and composition. Distance
from us is no problem since RTF files can be transmitted
rapidly by e-mail. So, if you're a Word user who likes to
proof and format, please consider giving us a hand.
The second, of course, is that we need more articles! Not
only because we like them....why else would we put out a
magazine....but because the more _you_ write, the less _we_
have to write. By contributing to the ENGINE, you assure the
variety, comprehensiveness and interest of what we publish.
And the articles we prefer are short and punchy; it's not a
major undertaking to write one if your references are at
hand. Start one today and know the thrill of seeing _your_
writing in California's premier journal of computer history.
-------------------------------------------------
"JUST LIKE THE GOLDEN GATE BRIDGE":
Rey Johnson and Jack Harker talk about RAMAC
-------------------------------------------------
[INTRO: The story of the modern computer is inseparable from
the story of the hard disk, one of the most successful data
recording technologies of the postwar era. The earliest
commercial hard disk, IBM's 305 RAMAC, is very much a part
of California's computer history -- because it was developed
in IBM's laboratory at 99 Notre Dame Avenue in San Jose.
Reynold B. "Rey" Johnson had worked as an IBM engineer since
1934, developing a mark sensing machine produced as the IBM
805 Test Scoring Machine in 1937. He continued to work on
engineering of input devices and, on January 15, 1952, was
named manager of the San Jose laboratory.
John M. "Jack" Harker joined IBM and the San Jose facility
in May 1952, and became a principal engineer on many random-
access file devices, focusing his research on air-bearing
lubrication and the avoidance of the dreaded head crash. The
work done by these gentlemen answered difficult theoretical
and practical questions and, ultimately, revolutionized the
retrieval of magnetic data. It also confirmed the San Jose
Lab's position as one of IBM's most adventurous -- and most
financially successful -- facilities.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 3
On Saturday, March 5, 1994, Rey Johnson, Jack Harker, Kip
Crosby and Max Elbaum gathered in Rey's study-library-
laboratory for an extended conversation about RAMAC, the Low
Cost File System, and the early challenges of engineering
random-access storage. ]
_KC: In the context of the San Jose labs, let me ask first
-- trying to go back to the perspective of the late '40s,
early '50s -- IBM was, not only in its physical location but
in its mind-set, very much an East Coast company. It had a
formidable European presence through IBM World Trade, and
the data processing needs of Asia were considered to be
minimal. Why did IBM suddenly decide to put a laboratory in
California?_
RJ: Well, in 1951 Tom Watson, Jr., Vince Learson and others
in the IBM top command decided they'd conduct market
research to establish whether computers were business
machines or not. And so they sent a team out -- Dr. Cuthbert
Hurd, I believe, headed the team -- and they visited all the
large companies, [especially] aircraft companies, government
agencies and the military, to establish the size of the
market for computers. Proposing a rental of $15,000 a month,
and assuming it had the qualities of the ENIAC and the
UNIVAC and whatever was necessary, they came out to the
total market for 17 machines. At that point Mr. Tom Watson
made the decision to manufacture 18. But in the process they
discovered that, of the business market, it was really the
aircraft industry that was interested in innovative data
processing. Converting airflow data into usable information
was one of the key problems that required unprecedented
processing power. And since many of the aircraft companies
were located in California and the Pacific Northwest, IBM
was determined to have a laboratory close at hand. I had the
freedom to locate the lab anywhere on the West Coast, but
before I came West, IBM management had pretty much decided
-- or recommended -- that it should be in the San Francisco
area. Now, IBM had a plant in San Jose that was
manufacturing punch cards for the Pacific region, and it had
been a very successful business enterprise since it was
established in 1944. They had a cafeteria, an accounting
system, and a lot of amenities which I -- as a raw engineer,
essentially -- was very pleased to have as resources while
we got started. That's why San Jose was chosen. I had
actually hoped, when I first came out, that we would locate
the laboratory in the Stanford industrial site or somewhere
nearer Stanford.
_KC: At that time the Stanford industrial site had a couple
of other firms that would be important to the computer
evolution -- notably Hewlett Packard and Varian._
The Analytical Engine, Volume 1, Number 4 April 1994 Page 4
RJ: Yes, but remember, I was not a computer engineer. I had
had responsibility for input in IBM. I developed a test
scoring machine (the IBM 805) announced in 1937; and I had
responsibility for keypunch input development in the '40s.
And I was aware of the punch card input problems. I had a
record of 53 patents by that time and I developed several
products, including time clock products which were
revolutionary. They therefore gave me complete freedom to
innovate the programs. The programs were to be of my
choosing, as long as they weren't copies of any of the
programs that were in work in the other laboratories, and to
that they added some contract customer work -- adapting IBM
machines to customer requirements.
_KC: So that among other things, you were acting as support
engineers for the installed base of 701s?_
RJ: 701s hadn't been announced. They were in work -- the
first one was finished in the spring of '52. But the 701 was
not a business machine. It wasn't planned that way, and
nobody in IBM really had the insight to see what computation
would do for business -- data processing or computing as we
know it today. At my request Lou Stevens joined me as
assistant manager -- he had been the input/output manager on
the 701 project -- and we worked in the context of card
processing and keypunching, which I had a lot to do with.
There was usually a time card every week for everybody, and
you had to be sure to enter the right information, so you
needed a punched master card to put into the keypunch, for
entering the account number and the rate, _et cetera_. So a
tub file was provided, from which cards were pulled. And
when orders came in from customers, you needed a tub file of
inventory items. If it was groceries you wanted one for
cornflakes, and every item that was ordered by a grocer
wholesale, you needed to have the inventory information --
the cost and the pricing and so on. Thus the tub file was a
central feature of many keypunch departments. IBM at that
point was manufacturing enough punch cards to go around the
world every two weeks, and probably 95 percent of what we
would call machine memory today was in punch cards and
files. We never thought of it as memory, but nevertheless,
that was the memory problem we attacked.
_KC: But the tub file, to my understanding, was a very
laborious way of keeping track of things like inventory.
There were many applications, such as invoicing, for which
the tub file wasn't especially well adapted._
JH: Well, for the systems that were in use it was
surprisingly efficient, because you had these cards with all
the information on the item pre-punched. To assemble an
The Analytical Engine, Volume 1, Number 4 April 1994 Page 5
order you'd pull up a customer card, and you'd pull up a
card for each item, and you'd have a pack that could then be
processed through your keypunch equipment. In terms of
today's technology, that's awkward. But compared to anything
else that was available, it was very efficient.
_KC: For a purely mechanical system, it was probably as good
as you could get._
RJ: The foresight of any of us was pretty bad in terms of
the potential of computers to do the job -- at that point.
Of course there had been years of ongoing work in the
Poughkeepsie laboratory, where Ralph Palmer in particular
had developed a tape processing machine. They had
experimented with sorting tapes, going from tape to tape,
starting and stopping tapes -- that work had proceeded. But
that was still a batch system and it didn't meet the input
problem posed by the tub file. That's the reason I chose the
tub file problem as one of the problems for San Jose.
artin, who was a Cal Tech graduate they'd
succeeded in hiring back East, and Rey chose him to come
back. But most of the people who were hired for the lab were
either out of the West Coast universities, or out of the
aerospace industry.
RJ: On the other hand, it's possible to overplay that. At
one time in IBM I was the only senior engineer -- they
called them inventors -- with a college degree. And my
degree was in educational administration. All the key
inventors that built IBM's data processing were people who
had come from drafting or customer engineering or the
factory -- bright, clever, talented people. They had
developed all our printers, punches, sorters and so on. In
1939 -- this is a bit of background but it'll give you the
feel of IBM at the time -- Thomas Watson Sr. had heard of
someone who was going directly from typing to punching, and
he wanted me to develop an accounting machine that would
work from typed characters, [for which] I could hire anybody
I wanted to. They engaged the patent department to advertise
for scientists or engineers for this work anonymously -- IBM
itself would never advertise for help for its engineering.
It was in essence a blind ad, and a number of people
answered it. I hired an MIT graduate to assist me on that
The Analytical Engine, Volume 1, Number 4 April 1994 Page 6
job, and he and I developed mark sensing. We actually
developed an accounting machine that worked from rather
large typed numbers; it never went into production, but mark
sensing became a very important data input system. Julian
Bigslow was succeeded by another outstanding engineer,
Gordon Slaughter, who went into the service too. When he
came out of the service in 1945, he came to see me. He said
"Rey," -- and he was a personal friend besides my assistant
at that time -- "I have really been sold on electronics and
I think that electronics is the career I would like to
pursue, and I'm not sure that IBM has a career for anybody
in the field of electronics." At that point I'd developed
test scoring and mark sensing, both of which used
amplifiers, and I knew that all the electronic work in IBM,
up to that point, had been done by a single engineer; so I
had to concur. We went down to see Wally McDowell -- who was
lab manager -- and talk about it. As much as we hated to see
Gordon leave, we all three agreed that there wasn't a career
opportunity in IBM for an electronic engineer in 1945.
_KC: By the early '50s, in trying to develop the 701, IBM
found they didn't have that talent. And they converted a
number of mechanical engineers -- they literally gave them
courses in electronics..._
RJ: Frank Towns, I know, built the Harvard ASCC and the
subsequent one [SSEC] that went into Corporate Headquarters
-- he was sent for a short course at Cornell. I also recall
an Engineering managers' meeting with Walter Titus, vice-
president and general manager at Endicott. He said, "I want
to know if any of you know what a binary numeric system is."
And not one of the people of that group knew what a binary
system was -- this in the early '40s. It helps you
understand why I had to run ads to get a graduate student,
and why there wasn't a career in electronics with us as late
as '45. Of course, Ralph Palmer's multiplier was an
application of electronics, to computing in the early
1940's.
_KC: Now the electronic multiplier was the 604?_
RJ: The 603 was first, and 604 was Ralph's successor to the
603. The 604 was a very successful multiplier.
_KC: Right. And many of the people who later became
customers for the 701, 702 and their follow-ons had been
using 603, 604 technology, especially the aircraft
companies?_
RJ: Yes. Some California engineers wired the multiplier
right to accounting machines....
The Analytical Engine, Volume 1, Number 4 April 1994 Page 7
JH: There is an interesting story, about Northrop [Aircraft]
during the Second World War, where a number of customer
engineers -- Bill Woodbury, among others -- actually rewired
accounting machines into a program-controlled machine. It
was a group of about three or four very bright people. I've
never seen that published.
_KC: I have, I believe. This was a bunch of people who were
somewhat familiar with the work being done at the Moore
School. ....You don't think so?_
RJ: There was an electronic data systems group that grew out
of Moore School. But EAM had very little input from Moore
School early on.
_KC: But I'm saying that the people at Northrop might have
been._
JH: They might have. At the time their job was purely
maintaining the equipment and they were very bright people.
The supervisor at Northrop -- Oliphant, I think his name
was, took Bill Woodbury under his wing.
RJ: Woodbury and Toben developed what was known as the
Wooden Wheel. Woodbury was a very important input into the
RAMAC too. He came out and joined our staff about 1954.
Actually, what changed things was the UNIVAC and Van
Neumann's concept. Watson, Sr. staked the pride of the
company on building the Harvard relay calculator.
_KC: The ASCC, the Mark I._
RJ: Yes. The Mark I was all electromechanical. And I think
every engineer in IBM had all the physics they ever needed
for any of those machines. Nothing went beyond electrical
circuits and maybe vacuum tubes.
_KC: And then the SSEC, the one installed at World
Headquarters, was somewhat of a follow-on to the Harvard
machine, and was the last of the big electromechanical
calculators?_
RJ: Yes, it was the last of that architecture. The big
computers IBM built in Kingston for the air force demand --
SAGE -- they were very big. They developed useful technology
but no products.
_KC: Actually, the Air Force provided some of the impetus
behind RAMAC, did it not, through the Material Information
Flow Device Contract?_
The Analytical Engine, Volume 1, Number 4 April 1994 Page 8
RJ: When we were able to size the project, I routinely asked
marketing people to visit the laboratory operation and
discuss our programs. And my main input was a man whose name
I can't recall right now, but he said that for tub files,
"five thousand unit-records is all you need, and anything
beyond that will make it too big." At that point we got this
request to bid for the Air Force Material Flow device, and
John Haanstra and [Arthur] Critchlow, I think, and maybe Lou
Stevens, went out to Dayton. We had quite an in-house
discussion as to whether to use our new concept of disk
files or drums. Drums were the standard at that point, and
IBM being very conservative, it was ordinarily impossible to
get them to try an innovation -- in a customer's machine
particularly. But at the last minute I convinced Wally
McDowell that we should bid the disk files, and they
required fifty thousand inventory items with a hundred
characters each -- that's five million characters, and that
became the specification for the disk file.
_KC: What we would now call a five-megabyte hard disk?_
RJ: Yes.
_KC: Now, when you say that the drum was standard, was this
at about the time of the introduction of the 650?_
RJ: The 650 drum had been in work, yes. But [RAMAC] was
contemporary with the 650 introduction. I had very serious
arguments with Vince Learson, who was responsible for market
planning, and who wanted us to build a component for the
650, rather than the RAMAC as a product. He didn't seem to
want the RAMAC.
_KC: I understand that he was also a very easy man to argue
with._
RJ: He was very strong, but fortunately he had too many
other things to do to oppose our program at that point. And
I was enough of a political animal to recognize that the San
Jose laboratory needed a product, not a component -- because
if it was a component, they would pull that back to
Endicott, rather than do it out here. That proved to be
politically wise, but the RAMAC actually was too
underpowered to carry the load. I mean, its printer was
slow, and its input was punch cards, and it had a hundred
characters of core memory and a drum for a buffer. So it was
overpowered on inventory direct access, but underpowered in
other respects. But they sold a thousand of them.
_KC: We're getting a little bit ahead of ourselves, because
there's something that has to be established here. First of
The Analytical Engine, Volume 1, Number 4 April 1994 Page 9
all, as you mentioned, Vince Learson was never completely
enthusiastic about the machine. Second of all, IBM had
finished the 650 and was beginning to sell 650s to
customers._
RJ: Not when we started. The 650 was developed
contemporaneously. And let me say one other thing that
happened: Vince Learson sent F. J. Wesley out to work with
us, and Wesley really got religion. He wrote a memorandum to
all the corporate people and he said, "IBM has to change
their thinking and planning and their systems. We can no
longer do things on a batch basis, whether it's tape or
cards. The random access is the way to go and we have to
plan our new systems this way." He used all the power of
conviction that he could possibly use. And that was a very,
very important input into the decision to manufacture the
RAMAC in San Jose.
JH: We're skipping over a little bit here. Go back to your
tub file. The first concept of what we were going to do was
the file-to-card machine. The first disk file that was built
and operated essentially replaced the tub file, which was
the original concept. And that was as smart as we were, at
that time, as to what it should be. You're now getting into
sort of the second phase of it, when Rey had assembled a
team of very capable systems planners -- Wesley, and John
Haanstra certainly was brilliant, Murray Lesser. I think at
that time Greg Toben was here, wasn't he?
RJ: Woodbury was there too.
JH: And with this input the concept of a transaction
processing machine evolved, which was the real key of the
RAMAC. It was really the first transaction processing
machine. As you said, unfortunately, it wasn't quite up to
the job.
_KC: It wasn't quite up to the job, but so long as you had
data processing machines that were discrete units, you
always had the problem of some components of the system
being slower than others, and trying to bring them all up to
speed._
JH: But this was really a radical break from corporate
philosophy, because the 701s, 705s were all batch machines.
They mechanized the punch card business, but still operated
as card systems -- in other words, when you ran a payroll,
you would match your input cards against the master file.
The RAMAC and the systems design that went into it was
really an important departure for IBM.
RJ: And there was a lot of competitive action at that point.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 10
There was a unit developed in Los Angeles that provided
random access with tape loops. There was a development at
Rand, I believe, where they had tape strips. And somebody
came back and said that somebody had experimented with
putting all the accounts of the Detroit Edison on the
flywheels of their power station. I was determined to
evaluate every conceivable technology, and so we had one guy
working on wires, one guy was working on plates, one guy was
working on tape strips and drums, and of course one guy on
disks. We evaluated all technologies before we chose disks.
It's been my procedure always to give every competitive
approach a chance to be chosen, before making the choice of
a technology, because so many inventors or engineers choose
technology with an enthusiasm that runs away with them, and
then they're blind to all others.... So I knew, when we
chose disks, that they were the best system.
_KC: In the aid of that, of evaluating competing
technologies, you had introduced a management style at San
Jose which depended on every engineer in the place being
conversant with all projects. That was another way, to my
understanding, in which the San Jose lab was very different
from IBM._
JH: He created a wonderful environment for young engineers.
Every engineer wanted to run his own project, and you really
thought that maybe you had a chance. Everybody tried to come
up with ideas, and Rey created that environment, it was
wonderful.
RJ: I had a statement that I often repeated: "Don't forget,
your most important assignment is to give assistance to
whoever asks for it if you can, and the second in importance
is the job you have to accomplish." One of the employees put
that on a brass plaque and put it in the laboratory. It
became a very important factor, because everybody was
involved in the RAMAC; a small segment called it a bologna
slicer and didn't think it was a particularly good project.
But it was open and good-humored; there wasn't any back-
biting between departments, as there had been in Endicott
over the years, where engineers literally fought for their
projects.
_KC: Again and again we come up with ways in which San Jose
was different from the classic IBM New York laboratories.
It's my understanding that at one point you guys had
Birkenstock -- in his role as the head of product planning --
cutting a memo saying roughly, "I don't know what those
guys are doing out there, but I'm not sure I like it."_
RJ: I never knew him to say that, but he gave me a hard time
on occasion.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 11
JH: It wasn't a transplanted IBM operation. With the
exception of Rey, Hal Martin, and Jim Hood who came out as
assistant lab manager, and Jack Poole who came over from the
card plant to run the administration, there were no IBM'ers.
It was a very free-spirited group that they managed to hire
-- and a remarkably talented one.
RJ: They were good. On the opening day we put an ad in all
the West Coast dailies; we had four hundred applications,
and from that four hundred I chose probably 25. So I had an
excellent choice. The tub file was classified essentially as
an input. Much of my work in IBM was keypunching and mark
sensing and test scoring, work that was keyed to get the
information efficiently into the system. I also had
developed a couple of high-speed printers. I built a high-
speed wire printer....
_KC: For the model 26?_
RJ: The wire printer for the 26 was my design, but after
that I built a line printer -- eighteen hundred lines a
minute, where the code was a rod [rather than the Type 26
code plate -- Ed.] and it was positioned harmonically. I had
a model in my shop which ran quiet enough so that you could
carry on a conversation next to it -- at eighteen hundred
lines a minute --because it was driven by harmonic cams. IBM
decided to make a product out of it. Now I had learned in my
time recording and keypunch days, when I had both product
development and product initiation responsibility, that I
didn't have the temperament to discipline my innovation; and
every Monday morning, my guys used to complain that I'd come
in with an engineering change. So I was very glad to pass
that high-speed wire printer on to Frank Furman, who was one
of the most trusted printer engineers in the business. And
Frank developed that into a product that was announced, and
he changed the drive from my harmonic positioning to an
impact system. This mechanical multi-head wire printer was
the noisiest machine that IBM had ever created. Its failure
as a product was my greatest disappointment at IBM.
_KC: So back to the disk file. Had there been other
research, other theoretical publication about the idea of
recording data on a spinning disk?_
RJ: [Jacob] Rabinow came up with doing that, but his format
of starting and stopping the disk made it unacceptable. He
had a stack of disks, each of which had a sector cut out of
it, and the whole stack arranged like a donut. And then
inside he had a head that had to go around, and when it got
to a disk then this disk would be driven one rotation. In
that sense it was unsatisfactory. And of course there were
Wurlitzers around....
The Analytical Engine, Volume 1, Number 4 April 1994 Page 12
JH: It was a takeoff of a Wurlitzer player.
RJ: We had a very simple technology -- the technology of
drums and tapes. And so the choice of disks just had to be
studied enough to know that surface to surface, you had the
best, most efficient arrangement of data. In the tub file
application of that time, the concept was there'd be one tub
file used by 20 keypunch operators. You could put 20
accesses around this disk array and access each
independently, both as to track and time.
JH: The initial design of the RAMAC stipulated that we had
an 18-degree envelope this access mechanism had to fit in,
because Rey said you've got to be able to get 20 of them
around. We managed to get three working at once on the 650.
_KC: Three heads?_
JH: Three access mechanisms. You could order a 650 RAMAC
with up to three accesses, is my recollection, although I
don't think many people did. If you look at it and wonder
about the configuration of the access, that was a key
parameter that we all had to live with.
_KC: That you had to get 20 sets of carriers, basically, to
the disk at one time?_
RJ: This [_hands over a small section of a disk platter_] is
a surface of a disk segment upon which the recordings have
been developed magnetically -- painted white, and then you
put a magnetic powder on it and develop the bits, and then
you spray a cover. We chose a hundred bits to the inch and
50 tracks to the inch, which was conventional tape and drum
data density. We didn't push the density at all in the
RAMAC.
JH: I think it was 20 tracks to the inch, wasn't it?
RJ: Twenty tracks to the inch. I have this free sample of
Hewlett-Packard's disk. [_He hands over a plastic model of a
Kittyhawk disk, the size of a pocket matchbox._] That's
their twenty-megabyte disk, isn't it? That would be four
times the size of the RAMAC.
_KC: I almost hate to tell you this, gentlemen, but I think
you can get these in forty megabytes capacity...._
JH: The trick was getting a head. That way you could record
magnetically, because we couldn't make disks in those days
as flat as the ones used now. Drums by contrast were all
very precision devices with a few ten-thousandths of run-
The Analytical Engine, Volume 1, Number 4 April 1994 Page 13
out. The heads were adjusted in place away from the drum;
you'd run a head into contact and space it back away,
watching the output signal on an oscilloscope. But the
invention that made the disk file possible was the airhead
to support the magnetic read/write element.
_KC: Right, because there was, to my understanding, the
first problem of -- was a disk like a drum, or was a disk
like a tape? In other words, could you have the heads in
contact with the surface or not?_
JH: If you looked at Rabinow's design -- because you only
spin it once, you could propose putting the head in contact.
But not if it's going to run continuously.
RJ: One of the most important events, for me personally, was
when we put these 150 disks on a shaft. They were machined
out of just ordinary flat aluminum, as flat as we could get
them. When turning they were all quite wobbly. We ran those
disks at a thousand RPMs, and you could put your finger on a
disk and follow it. You could keep your finger on it -- it
didn't get hot, it didn't wear -- and I knew at that point
that we were "in like Flynn". We had to try to get them as
flat as possible, but we were in. Fortunately we'd had some
people with us who had done air bearing work, so we decided
just to put air pressure in the head to keep it away from
the surface, and five three-thousandths of an inch wasn't
too far for this kind of density of recorded spots.
_KC: Now this was not the approach that came later. This was
an approach that actually supplied compressed air to the
disk heads?_
RJ: I think all of the thousand RAMAC computers that went
out had compressed-air head spacing.
_KC: There were compressed air disk systems for other
devices besides RAMAC, were there not? I know somebody built
one for STRETCH._
RJ: STRETCH came along five years later. STRETCH had to have
a full gamut of heads. And the first models were pressure,
but with all that air, the air pumps broke the camel's back.
JH: IBM bid and committed the STRETCH machine, which
required a parallel array of heads because the data rate --
it was a high data rate machine -- all the heads were read
in parallel.
ME: It was a super-computer by the standards of the day.
JH: Yes. And this was the main input, this was how they
The Analytical Engine, Volume 1, Number 4 April 1994 Page 14
would wash data back and forth for computation. It was to be
based on the Advanced Disk File, that became the 1301; and
we ran into trouble making the gliding heads, the self-
acting bearings, work. The only way to meet the commitment
of the STRETCH delivery was with air-fed heads. And it
required a bank of compressors. A good friend, Ralph Golub,
wound up with the development of it, and it was a monster,
but it did the job.
_KC: Let me refer to my notes here briefly. By May of 1953,
you had developed the pressurized recording head, you were
doing experiments with 16-inch disks, and one problem was
that the distance between a head and a moving disk surface
was difficult to measure._
RJ: For us it was difficult to measure. We didn't really
know how to do that at that point.
_KC: So what did you come up with?_
RJ: I think pretty much the recording signal. We had,
indirectly, the recording signal, and so if the signal was
good enough all the way around, that was good enough for us.
And that's the way I remember it, Jack.
JH: You could calibrate the spacing statically.... recognize
that the RAMAC as produced had head spacing of about a
thousandth of an inch. And it being an air-fed head, you
didn't have to have the disk moving, so you could set it up
on a plate, and put a dial indicator on it, and measure the
spacing within the accuracy of what we needed.
_KC: That worked for a pressurized head. When you went over
to a slider...._
JH: Now you're into another generation. You're into the
1301.
RJ: When did the first product come out? And you brought it
out, I guess, with the gliding heads.
JH: 1301. That was shipped sometime around 1960.
_KC: I have June, 1961._
JH: Fine. And that's a whole other development story. Yes,
we did have difficulty measuring a lot of things. It's the
first time we got into optics as measurement.
_KC: Before we get to the 1301, let's discuss two other
RAMAC issues. What about getting the oxide on the disk?_
The Analytical Engine, Volume 1, Number 4 April 1994 Page 15
RJ: We tried painting, spraying, brushing, until one of our
fellows said, "Pour it on and spin it on." And we set up
paper cups which poured a uniform amount for every disk, and
it spun until some of it flew off the edge. That gave us a
very uniform coating. And we had some trouble with
smoothness, and so one of the guys got nylon hosiery and
poured the paint through the nylon hose, so we got rid of
the coarse grains. The spin-coating method was used for
years; it was automated in the factory, but it was still the
same process -- spinning it and pouring the paint on.
_KC: So this was basically a kind of paint; it was a liquid
with an iron-oxide suspension?_
RJ: Yes. It was essentially the same iron oxide that was
used to paint the Golden Gate Bridge -- a ferric, red oxide,
which is there for rustproofing, and for us it was for
magnetic qualities. Jake Hagopian, a very excellent
engineer, had the job of coating. He worked with the paint
people in developing the best mixture that the paint people
thought would work. It worked. And for the density we had,
nobody ever needed to touch the surface.
JH: Wasn't Marcel Vogel the one that finally came up with
the right formulation?
RJ: Marcel Vogel was a chemist. He proved to be
indispensable over many years of developing coatings.
_KC: Now this resulted in disks that did not need to be
polished afterwards, correct?_
RJ: I don't think they were ever polished, the first disks.
JH: I can't tell you when we first started buffing disks. I
don't know whether the first production disks were buffed or
not.
RJ: But Marcel and the paint companies and Jake Hagopian
knew that the mixture had to stick, and it had to be hard,
besides having the magnetic quality.... Here's a story.
[_Shows a disk marked off into successively smaller
segments._] A given amount of information required this much
space [_roughly half the disk_] on the RAMAC, the first
improvement put the same information in there [_an eighth of
the disk_] and that was produced as the 1405. And then each
model that came in, this much information was put in that
[segment] until we got it in there [_points to an area about
the size of a period_] with the later machines. The
improvement came as a result of improving the grain size,
improving the signal processing, improving the head gap and
The Analytical Engine, Volume 1, Number 4 April 1994 Page 16
the head width, with each one of these iterations.
JH: It's been a job of miniaturization, much as with the
semiconductors. It's a very scalable technology. Probably
the greatest gain in density was when we first made the
1301, where we got down into the two-hundred-microinch range
of spacing and the head geometry compatible with that, and
then Winchester got us -- starting out -- to 25-microinch
spacing, those were two major jumps. In signal processing we
went from an NRZI to an MFM encoding, but that was a
comparatively modest increase. In recent years you've seen
signal processing really starting to be used in a classic
communication-equivalent sense....
[To be concluded in October]
-------------------------------------------------
Mini-article: THE "MORE" COMMAND IN UNIX
-------------------------------------------------
by Dan Halbert
halbert@world.std.com
I was a first-year graduate student at UC Berkeley in 1978.
I had been an undergraduate at MIT, and had used the ITS
timesharing systems there, which ran on PDP-10's. ITS put a
"--MORE--" at the bottom of the screen when one typed out
files; you pressed the space bar to continue.
At Berkeley, we'd just gotten our first VAX UNIX system,
though there were already PDP-11 UNIX systems. There was a
very simple program through which one could pipe stdout to
do screen-at-a-time output. It rang the terminal bell after
printing 24 lines, and waited for a carriage return. It was
called "cr3". My guess is that in some version of UNIX,
someone had hacked a page-at-a-time output mode into the tty
output drivers. Using stty, one could already say "cr0",
"cr1", and "cr2", which added different amounts of delay
when printing a carriage return, for the benefit of slow
printing terminals. "cr3" was probably unused, and the page-
at-a-time mode was piggybacked on it. But our version of
UNIX didn't have this "cr3" stty mode; instead we had the
"cr3" program that provided equivalent functionality.
Many of the terminals at Berkeley were Lear-Siegler ADM-3
and ADM-3A "dumb" terminals. Both models (or maybe just the
ADM-3's) rang the terminal bell when the cursor advanced to
near the right margin, as a typewriter bell would.
Unfortunately, they rang the bell on output as well as
keyboard input, which made for incessant beeping. It was
particularly maddening in a room full of terminals. So most
of the bell speakers had been disconnected.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 17
Since "cr3" rang the terminal bell to indicate that a full
page had been output, you couldn't tell when it was waiting
for input on those muted terminals. The problem was
exacerbated by the slow response time of the overloaded UNIX
systems.
So I wrote a simple "cr3"-like program, but had it print "--
More--" instead of ringing the bell. I had it accept space
instead of carriage return to continue, because that was
what I was used to from ITS. I also made it take multiple
filenames, and had it print lines of colons ("::::::::::::")
before and after it printed each filename.
I named the program "more". This was a daring move at the
time, since it was such a long name for a UNIX command, and
was also a real English word.
Subsequently, my friends and fellow graduate students Eric
Shienbrood and Geoff Peck greatly expanded the program,
adding all kinds of command line options and different
possible responses to the "--More--" prompt. It was of
course distributed in the BSD versions of UNIX.
Some time later, Don Norman wrote an article for Datamation
entitled "The Trouble with UNIX", in which he complained,
among other things, about the cryptic nature of most UNIX
command names, citing "more" as an example. I never did tell
him that I thought "more" was a great improvement over
"cr3".
I was later amused to see "more" appear in MS-DOS. [Also in
ZCPR3. -- Editors ]
-------------------------------------------------
POMONA
-------------------------------------------------
CHAC had a lot of fun -- and did a lot of work -- at our
first-ever trade show, the Los Angeles Computer Fair at the
Pomona Fairplex on March 18th, 19th and 20th. We attended at
the kind invitation of our friends and computing comrades,
David and Tamara Greelish, co-founders of the Historical
Computer Society of El Paso, Texas.
The three of us were loaded for bear with hardware,
documentation, flyers, spiffy new vinyl booth banners, and
plenty of copies of the ANALYTICAL ENGINE and the HCS
newsletter, _Historically Brewed._ David and Tamara had
brought a vanload of hardware including a Commodore SX-64 --
really one of the most attractive basket-handle luggables
ever -- and a PET, a Radio Shack TRS-80 Model 4, an IMSAI
8080, a Timex-Sinclair 1000, and an Apple Lisa, II+, and Mac
The Analytical Engine, Volume 1, Number 4 April 1994 Page 18
512, as well as a nameless, homebrewed S-100 box. A brown-
case Osborne One and a NEC APC sat forlornly on the booth
floor because we had no more table space, and the IBM PC
never even made it out of the van.... Rather than bring any
of CHAC's hardware, since I flew, we relied on Joe
Schopplein's brand-new and inspiring color photos of our
collection.
Reaction varied but was generally gratifying. The aisle
crowd kept up a good pace, impelled by the hunt for
bargains, but every so often someone would step out of the
current, look around the booth slack-jawed, and say
something like "What's _that_?" The hands-down favorites
were the Lisa, the PET and the IMSAI -- we met about fifty
people who claimed to have built or bought IMSAIs when those
were new. (In classic trade-show fashion, the IMSAI's boot
disk promptly lost a sector after having been dependable for
a million years, but David could at least light the front
panel.) We surprised a few people who said forlornly "Don't
you have a TRS-80?" since they were expecting to see a Model
I, but the Model 4 we _did_ have looks completely different.
Friday overall was a light day and very enjoyable, since it
allowed us to steam the real-time wrinkles out of our
procedures and still have plenty of conversation.
On Friday, too, I had the great pleasure of a long,
undisturbed lunch with Erwin Tomash, a founder of the
Babbage Institute. As might be expected, this became a long,
hard-headed conversation about nonprofit networking,
fundraising, research practices and curatorship. Mr. Tomash
summed up dryly by saying that we'd "set a significant
challenge" for ourselves, but that he wished the CHAC every
bit of the luck that we'll need. His wisdom and experience
are an invaluable gift to our Association and I look forward
to staying in touch with him.
Friday's show traffic tapered off early, but we all knew
that Saturday would be a different story. We made sure that
everything was in place, then headed for our hotel rooms at
pedestrian hours knowing we'd need all the sleep we could
get.... Sure enough, the doors of the Fairplex opened at ten
and the aisles were packed by five after. All in all, we
talked to between eighty and a hundred people who actually
browsed our booth, flipped switches and told entertaining
war stories -- especially, for some reason, about the Jet
Propulsion Lab at Cal Tech. And this one gentleman who had
retired from an aircraft company _did_ think he could point
us toward an IBM 704 -- ! Lots more people just wanted to
know what we were about, and four hundred CHAC FAQ's and
subscription flyers flew off the table on Saturday and
Sunday. A fair number of copies of the ENGINE and
_Historically Brewed_ disappeared too, most of them paid for
The Analytical Engine, Volume 1, Number 4 April 1994 Page 19
although a few simply vanished into the river of people.
In general we were too busy to leave the booth, but I made
an exception when Steve Roberts happened by. As many of you
may know, Steve is the designer/builder of some of the most
advanced human-powered vehicles in the world, and has made a
name for himself riding his creations around the country
while he stays in touch with the world through cutting-edge
wireless communication. He had his latest, BEHEMOTH, with
him at the show -- in true hacker style the name is a far-
fetched acronym standing for "Big Electronic Human-Energized
Machine, Only Too Heavy," which at 580 pounds (including
camping gear) it may be.... I can't really describe BEHEMOTH
here beyond saying that it's an eight-foot-long recumbent
bicycle, equipped with tightly compressed DOS and Macintosh
computers under its front fairing, packet radio at several
frequencies, integrated cellular communication, and
astonishing amounts of custom electronics including a solar-
powered satellite antenna relegated to a trailer. The bike
itself, with three derailleurs, 105 speeds, and hydraulic
power leveling, can "easily" be ridden (by Steve, I mean!)
at ten to twelve miles an hour for many hours. A nice touch
is a water-cooled helmet that, Steve says, "dissipates 75
watts on a warm day." Hey, that's the piece _I_ want! One
paragraph cannot possibly convey the flavor of California's
only (?) local-area network with wire wheels, and we look
forward to receiving an article for the ENGINE from Steve
when his schedule permits it. His absorbing new project, the
Sea Moss Microship (talk about ramming a pun through,) will
use solar arrays to power the electronics of a comparably
decked-out trimaran -- because his goal is still to
"maintain stable publishing and personal relationships while
traveling full-time," and "water doesn't have hills."
By day's end we were parched from talking and burdened with
what my old pal and trade-show vet Carol used to call
"tasteful knee-length headaches" -- mostly because the CD
vendor in the next aisle was drawing attention with
postmodern rock at top volume. Saturday was frankly
punishing, we gave up on some after-hours socializing we'd
been looking forward to, and it's just as well that Sunday
reached the same heights but tapered off by three
o'clock.... All in all the Computer Fair was an invigorating
experience, although a first assessment shows uneven
results; we haven't had a lot of correspondence or contact
from the people we met. On the other hand, no one said
organization-building was quick, and with our appearance at
Pomona, we have begun to bring the Computer History
Association of California to _all_ of California. And the
community of hardware and software developers in the Los
Angeles area -- especially those with ties to the aerospace
industry and to early micro builders -- obviously deserves
further study.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 20
We're looking forward to more fun at a trade show in
Northern California later this year. (Offers of a donated
booth gratefully received at any CHAC address.)
-------------------------------------------------
TURING, THOU SHOULDST BE WITH US
-------------------------------------------------
Forty years ago this spring, Alan Mathison Turing died. Few
can deny any longer that he did it by his own hand.
By the time he died, we must assume, much that was best in
him had been corrupted by reverses of circumstance. It was
his part to spare England and the Allies months, perhaps
years, of agony; yet when he died he was called a traitor.
He was unswervingly loyal to Monarch and Country; yet
shadowy bureaucrats maneuvered to strip him of his freedom,
and took Monarch and Country as their authority. And for no
more than minor indiscretions, he was branded a criminal --
in a way that no one could undo because the charges were
whispered. The fearful tactics of the Star Chamber beat down
one of the finest mathematicians who had ever graced
England.
We will never forget Turing, for we live in his world. From
his ultimate reduction, to the space and the mark and the
stepwise tape, came the vast breadth of the computable, and
from that the power of computing as we know it.
We owe it to his memory to remember, no less, that we live
in a world where the good and the great are hindered -- or
worse -- by intolerance and fear. And it is a more crowded
world than Turing ever knew, with sharper competition and
more perilous contradiction.
We owe it to his memory to lighten the shadows of fear, and
to lessen the burden of intolerance, wherever and whenever
we find the opportunity. We have a command of information
and opinion that can span the globe in seconds; certainly
one of the noblest uses for it is to combat prejudice and
ignorance.
Turing, you should be with us, in silvered eminence,
marveling still at the progress of computation.
-------------------------------------------------
COLOSSUS RECONSTRUCTED
-------------------------------------------------
Construction of an exact replica of COLOSSUS, one of the
earliest programmable electronic digital computers,
The Analytical Engine, Volume 1, Number 4 April 1994 Page 21
headlines the planning and preparation of a Historical
Computer Exhibition at Bletchley Park. Computer-related work
for the Exhibition is being undertaken by the Computer
Conservation Society (CCS), a volunteer specialist group of
the British Computer Society. The Bletchley Park Trust is
making progress in acquiring the Park as an historical
campus, with long-term plans to create active public museums
of Computing, of Cryptography, and of Radio and Radar; the
Park's official opening is currently scheduled for Monday,
July 18th, 1994. If further plans proceed as expected, it is
hoped that the CCS will enjoy substantial facilities for
restoration workshops, as well as archive, library and
research facilities and accession storage.
In the meantime, CCS Secretary Tony Sale and other CCS
members have mounted an exhibition of wartime code-breaking
memorabilia, including fragments of an original Colossus,
some of the plans for the Colossus rebuild project, a Typex
machine -- the British equivalent of the Enigma -- and a
"bombe" decoder. German hardware on display includes a four-
rotor naval Enigma, a Lorenz SZ42 and a Siemens T52, all
rarely seen before. There is a working radio intercept
station with a pair of National HRO receivers.
This Exhibition, in the Assembly Hall of A Block, is now
open to visitors on alternate weekends, from 10:30 am to
4:00 pm on Saturday and Sunday. The next open weekend is May
21st and 22nd, 1994. Admission is UKPounds 2.00 (UKPounds
1.25 concessions and children), and includes a 90-minute
guided tour of some of the wartime buildings, including the
original 'Hut 6'.
Bletchley Park, a country house and grounds some 50 miles
north of London, was the site of highly secret work
deciphering intercepted German military radio traffic during
the Second World War. Thousands of workers included Alan
Turing, Max Newman, and several other early computer
pioneers. Throughout the war, Bletchley Park produced highly
important strategic and tactical intelligence used by the
Allies -- Churchill's "golden eggs" -- and it has been
claimed that the war in Europe was probably shortened by two
years as a result; but total secrecy was observed by all
those involved. Information on the nature and scale of this
work began to emerge only when COLOSSUS was declassified in
October 1975.
Directions: Bletchley Park is 150 yards from Bletchley
railway station on the line from Euston station, and is
signposted. By car, it is off the B4034 Bletchley to
Brackley road, and about seven miles from the M1. The OS
grid reference of the entrance is SP 863 336. For further
information call the Bletchley Park Trust office at +44 (0)
The Analytical Engine, Volume 1, Number 4 April 1994 Page 22
908 640404, during office hours and on open weekends.
(Thanks to Chris Burton, John Honniball, Tony Sale and
Richard Wendland for information on this fascinating and
inspiring effort. Those of our readers who want to get up to
speed on COLOSSUS and the ULTRA project would enjoy reading:
ULTRA GOES TO WAR
Ronald Lewin
McGraw-Hill, 1978. -- Ed.)
-------------------------------------------------
IN MEMORIAM: AARON FINERMAN
-------------------------------------------------
Aaron Finerman, editor of _Computing Reviews_, died in Boca
Raton, FL, on April 6, 1994. His death followed a short
illness.
Mr. Finerman was a Fellow of the Association for Computing
Machinery and had done much innovative work for both the ACM
and the AFIPS. As Chair of the AFIPS Editorial Committee
during the 1970's, he devoted considerable time and energy
to the inauguration of the Federation's publishing efforts,
including the _Abacus_ and the _Annals in the History of
Computing_; he personally selected Bernie Galler as the
first Editor-in-Chief of the _Annals_.
His tireless attention to, and regard for, the history of
computing will be sorely missed, as will his friendship. We
extend our condolence to his family and colleagues.
-------------------------------------------------
LIVERMORE UPDATE
-------------------------------------------------
The great computers, components and memorabilia at the
Computer Museum of Lawrence Livermore Labs, featured in
January's ENGINE, have been removed from the Almond Avenue
School and placed in storage while a new location is sought.
The Museum's once and future Curator, Ms. Barbara Costella,
has been re-hired by the Laboratory and is coordinating an
effort to refurbish the collection and house it permanently.
We hope that active participation by Laboratory management,
and a lot of willing work by enthusiasts, brings the
attention of a broader community to this very fine
institution. Good luck to all concerned!
The Analytical Engine, Volume 1, Number 4 April 1994 Page 23
-------------------------------------------------
MICRO MUSEUM TOUR AT UC DAVIS PICNIC DAY
-------------------------------------------------
Picnic Day at UC Davis is a time for alumni, families of
current students, prospective students and their families,
and well-wishers to see student-run exhibitions and sporting
events on campus (such as dachshund racing, dog Frisbee,
sheep trials, rugby, water polo, track meet, rodeo,
equestrian dressage, etc.).
On this year's Picnic Day, Saturday, April 16, the CS club
worked with the IEEE student chapter to put on tours of the
University's computer facility. A morning of informal
discussion between students and alumni was followed by
scheduled tours of the computer graphics classrooms,
Internet workstation classroom, and the microcomputer
museum. Students had spent weeks getting a wide variety of
systems up and running, ranging from Ataris (kids loved
'em), to Osborne 1, DG 1 laptop, Kaypro, Radio Shack Models
1 and 2, and many others. Also displayed, but not
operational for lack of time, were a MITS Altair, an IMSAI,
and a Maltron ergonomic keyboard.
The Picnic Day tours were well received by a smattering of
alumni, several prospective students, and hordes of
interested folk. The museum was a highlight of the tour.
Altogether, a definitely repeatable event that generated
several offers to donate more gear for the museum.
We hope to have a permanent exhibit available one of these
days, but are still negotiating for space. At present, the
items up are about ten computers that can remain on display
in a portion of a research laboratory.
-- Dick Walters, advisor, UC Davis Computer Science Club,
walters@cs.ucdavis.edu
-------------------------------------------------
ADIOS AMIGA?
-------------------------------------------------
Commodore International Ltd. of West Chester, PA, USA,
announced on April 29 that it plans to transfer its assets
to trustees and has placed its major subsidiary, Commodore
Electronics Ltd., into voluntary liquidation. The decision
was described by Commodore as "the initial phase of an
orderly voluntary liquidation of both companies." In recent
years the company reported heavy operating losses
exacerbated by poor sales of new products, particularly the
Amiga CD32 video game.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 24
Commodore started 40 years ago as a typewriter repair
company in Toronto (ON), then began producing adding
machines and calculators. In April 1977 it introduced its
first personal computer -- the Personal Electronic
Transactor, or PET -- at the First Annual West Coast
Computer Faire in San Francisco's Civic Auditorium. An
immediate success, the PET allowed Commodore to compete with
Radio Shack and Apple in sale of computers to home users and
small businesses. Later models included the 8016/8032
business computer, the famous VIC-20 and C-64, and the SX-64
color luggable, all introduced in the early 1980s.
The company's shutdown will be lamented particularly by
users of the Amiga, a personal computer optimized for
graphics and sound, which Commodore produced in a profusion
of models since 1985. Amiga partisans claim, with
considerable justice, that theirs is one of the best
available small computers for manipulation of video images
and MIDI soundtracks. At our press time there was some
speculation that a third-party developer would purchase the
rights to this machine and continue production.
-------------------------------------------------
SPOTTER ALERT
-------------------------------------------------
As this issue's SPOTTER FLASH demonstrates, CHAC has begun
to get press -- more, and wider, press than we'd imagined
likely. The national syndication of the New York _Times_
article makes it imperative that our valiant volunteer
spotters be more vigilant than ever.
If you spot any mention of CHAC or the ENGINE in _any_
periodical, please:
* If your copy of the piece is clippable, clip and mail to
the El Cerrito address.
* If you can't spare the physical copy, send the text as
net.mail to cpu@chac.win.net, or photocopy and fax to the El
Cerrito address.
* If you're too busy for that, just send the publication
name, date and page number and we'll do the hunting.
Thanks!
The Analytical Engine, Volume 1, Number 4 April 1994 Page 25
-------------------------------------------------
SPOTTER FLASH
-------------------------------------------------
Mike Malone's column on collectible computers was published
in the New York _Times_ Business section for Sunday, April
17th -- featuring not only CHAC and KC, but our brother-in-
history David Greelish (HCS) and Hewlett-Packard archivist
Karen Lewis. Nicely illustrated with pictures of an Altair
8800 and an HP 65 calculator, it added up to more publicity
than CHAC had seen in its entire life.
This article went into syndication and it appeared in, that
we know of, the San Francisco _Chronicle_, the Orlando (FL)
_Sentinel_, the Everett (WA) _Herald_, and there must have
been others. The phone was ringing off the wall, at least by
our standards, for the next month. This made us rather more
of a national organization than we'd had any notion of
being, but "if you build it...." and we have every intention
of building _on_ this as far as possible. Thanks again,
Mike!
_Forbes ASAP_ for April 11th features a half-page on the
CHAC and its mission, in Kevin Hogan's Tid-Bytes column.
We're pleased and proud on the occasion of our first
national ink, even if we're a bit taken aback by Kevin's
brutally concise description of our strategic goal: "trying
to collect and catalog all significant hardware and software
created, developed or used in" California. If we're trying
to do all _that_ we may need one or two more buildings....or
a disused airport.... Of course, brevity has its points, and
Kevin managed to work in references to the ANALYTICAL
ENGINE, National Computing Science Day, Initiative 1999, and
even our friends the Computer History Association of
Delaware. It's easily the nicest executive summary of our
purposes so far. As for the lead para:
Have an old mainframe gathering dust in the basement?
There's a cabal of fanatical computer hounds ready,
willing and almost able to take it off your hands.
....well, sometimes the terrible truth must be told. And
what a wealth of meaning there is in that "almost." Thanks,
Kevin! (To anyone packing a mainframe for shipment to our El
Cerrito address; please call or e-mail first.)
The Analytical Engine, Volume 1, Number 4 April 1994 Page 26
-------------------------------------------------
DESPERATE PLEA FOR STORAGE SPACE
-------------------------------------------------
Look -- we were desperate for space when we _started_ this.
You can imagine what we are now. And there are three
fundamental reasons:
1) Small computers. As you'll see from this issue's
_Acquisitions_ column, we're assembling a truly fine and
representative collection of California micros -- from the
BYT-8, SOL-20 and Apple One to a Mac 128 and an Atari ST.
The best we can say about these is that they're out of
harm's way. They aren't all in one place, they aren't
particularly accessible, and they couldn't be exhibited
without a lot of logistical hassle; and for this second-rate
arrangement we pay a significant storage bill every month.
2) Medium-size computers. We now have two truly wonderful
minis, and they're sitting in somebody's living room in San
Francisco. Luckily for us, it's a big living room, and the
person in residence is one of CHAC's staunchest backers. But
this obviously can't be considered more than a stopgap!
3) Documentation -- to such an extent that even we're
surprised. Back in the days when computers were scarce and
expensive, they arrived with many more manuals than they do
today, and HP systems in particular travel with _walls_ of
docs. (The manuals for the 9100 programmable calculator are
bigger than the device itself.) Add in the docs for
commercial software; event-related books like the Computer
Faire symposia; magazines like _User's Guide_, _Dr. Dobb's
Journal_, and _Softalk_; newsletters like FOGHORN and _Cider
Press_ and BAMDUA.... We can see a looming need for,
literally, thousands of linear feet of shelf space. For the
moment we'd be very happy with a couple of hundred.
Since August we've engaged in several negotiations for long-
term storage, and they've all come and gone. We keep hoping
that some public-spirited person will tap us on the shoulder
and offer us a secure, dry room in a warehouse. Hellooooo
out there....
-------------------------------------------------
DESPERATE PLEA FOR MONEY (And Things)
-------------------------------------------------
Time for us to have a wish list! Until very recently our
accession technique has amounted to not much, but we need
some guidelines for inventory, or we won't even know what's
in what box. In pursuit of that we could definitely use the
following:
The Analytical Engine, Volume 1, Number 4 April 1994 Page 27
Mylar archival bags (comic-book bags)
Open-back document shelving boxes
File boxes or drawers for all diskette sizes -- c'mon,
you've got those old eight-inchers somewhere....
Secondhand steel shelving (but with all the bits, please)
Volunteer time for shelving and labeling
Volunteer hardware repair
Reference manuals for California hardware/software
...._and of course cash always welcome!!_ Unlike some of
this other stuff, it ships easily, too.
-------------------------------------------------
AND SPEAKING OF MONEY....
-------------------------------------------------
At the end of April we did a paper mass mailing, looking for
new ENGINE subs. We wanted to reach as many people as we
could, so we used the Internet to ask our friends for
mailing addresses. A gratifying number of you responded and,
of those, quite a few said how _glad_ you'd be to join our
mailing list.
However, some of you have not subscribed yet. No doubt this
is a simple oversight on your part. After all, if you got
that letter, you know how eager we are to have you as a
member. And if you know that, naturally you feel encouraged
to become one. The inertia involved in writing and mailing a
check is trivial by comparison to your enthusiasm.
Tell us that sleepless hours over a purring LaserJet(tm)
were not spent in vain. If you got that letter, PLEASE
SUBSCRIBE TODAY. And we'd be glad to have your sub even if
you didn't.
-------------------------------------------------
OVERVIEW OF BUREAUCRATIC PROCESSES
-------------------------------------------------
This was the quarter when everything broke loose. Success
upon success, long-awaited, hotly pursued success.
CALIFORNIA TAX-EXEMPT STATUS
-------------------------------------------------
This, and everything contingent on it, was delayed by a
mixup over a check, but a physical visit to the Franchise
Tax Board's San Francisco office did wonders to clarify the
situation. On February 28, the CHAC was granted tax-exempt
status as a charitable and educational organization, under
The Analytical Engine, Volume 1, Number 4 April 1994 Page 28
section 23701(d) of California's Revenue and Taxation Code.
This in hand was half the battle won, since with our
California exempt status assured, we could go after the much
more complex Federal equivalent.
CALIFORNIA CORPORATE CERTIFICATION
-------------------------------------------------
On March 24th, the CHAC was certified as a nonprofit public
benefit corporation by California's Secretary of State, with
appropriate notification to the state's Attorney General. We
responded on April first by filing the required Statement by
Domestic Nonprofit Corporation, and your Association is now
not only the Right Thing, but the Real Thing.
FEDERAL TAX-EXEMPT STATUS
-------------------------------------------------
All that done, we were in position to file IRS Form 1023 and
petition for Federal tax-exempt status. This was enough of a
grind to make the ENGINE late, since the form itself is ten
pages, and with the required attachments, continuations and
samples, ours went into the envelope at a handsome seventy
pages or about a quarter-inch thick. CHAC director Bruce
Rice cautions that we may be called in (or at least called
up) if the IRS wants more complete answers, but for better
or worse, it's in the mail....
NONPROFIT MAILING PERMIT
-------------------------------------------------
With California exempt status and corporate certification,
we can apply for a nonprofit second-class mailing permit,
which would potentially cut the ENGINE's mailing costs by
almost eighty per cent. Application forms and instructions
are on their way to our El Cerrito office.
CREDIT CARD PURCHASE
-------------------------------------------------
Not with a ! but a whimper.... No, you still can't subscribe
or renew with a credit card. But there may be a glimmer of
light in the distance because we're about to try our luck at
yet another (our fourth) bank. Stay tuned.
-------------------------------------------------
HELP FIGHT IRON DEFICIENCY ANEMIA!
-------------------------------------------------
Attention, those who know and love old iron! The ANALYTICAL
ENGINE is not getting enough articles about early computer
use in California! There are even those who have had the
temerity to grumble about our supposed "bias" towards IBM
The Analytical Engine, Volume 1, Number 4 April 1994 Page 29
and/or Intel, when there is no such bias for or against, we
are completely non-partisan but we can only publish what is
submitted. Therefore we, in turn, propose the following
article topics in hopes that they will jog someone's
interest or memory:
Logistics Research and the ALWAC
BDDDA, the G-15 or other Bendix adventures
BINAC or MADDIDA at Northrop in Hawthorne
Computer Research: CADAC, 105, or 107
CA/DIC at UC Berkeley
Consolidated Engineering 36-101
JOHNNIAC at RAND
MINAC at Cal Tech
The RAYDAC experience at Point Mugu
SWAC at UCLA
UNIVAC, STRETCH, CDC or Cray at LLNL
and _any_ early computer use at the aircraft factories!
We would be _exceptionally pleased_ to receive articles, or
proposals therefor, on any of these topics. Deadline for
Volume 2, Number 1 is July first, and the issue will appear
in mid-July; preferred article length is 750 to 2000 words
and submissions should be sent in machine-readable form.
(ASCII E-mail is fine.)
-------------------------------------------------
APPLE CONTEST
-------------------------------------------------
In the January ENGINE we asked "What was the text, in tiny
letters, that ran around the edge of the picture frame in
the original Apple logo?" The answer, which is "A mind
forever voyaging through strange seas of thought -- alone,"
comes to us from Gregory Nakshin of Staten Island, NY, who
laments that he has "all the manuals for an Apple One, but
no Apple One." Hey, the opposite predicament could be almost
as frustrating. Gregory receives our congratulations and an
extra issue on his ENGINE sub.
-------------------------------------------------
Book Review: FROM DITS TO BITS
by Herman Lukoff (1923-1979)
-------------------------------------------------
Robotics Press, 1979
ISBN: 89661-002-0
Reviewed by Brian Deith, University of Wisconsin
From Dits to Bits could as justly be titled "The Journey of
an Electrical Engineer." The book is autobiographical to a
The Analytical Engine, Volume 1, Number 4 April 1994 Page 30
charming extent; rather than presenting a cold, hard series
of dates and places, Lukoff lets us see computer history in
the making, as he did.
The journey begins in Lukoff's youth, as a radio hobbyist in
Philadelphia in the early 1930's. He discusses what it was
like to be involved in radio, detailing his triumphs and
disappointments:
'Dad, Dad,' I yelled, 'I'm getting Italy on my
radio!' My illusions were dispelled a half hour
later when the announcer said, 'This is WDAS in
Philadelphia, your foreign language broadcast
station.'
Lukoff's education took him from the schools of Philadelphia
to the University of Pennsylvania's Moore School of
Engineering, birthplace of ENIAC. During his senior year he
was invited to participate in Project PX, an ENIAC
forerunner; his task was to construct test equipment
necessary for the construction of the computer. His
supervisor on the project was J. Presper Eckert, then a
graduate student at the school, later to be renowned as a
computer pioneer.
Lukoff's involvement allows him to bring to light facts not
well known. He reproduces a memo by Dr. John Mauchly, dated
August of 1942, proposing the use of electronic circuits for
high speed computation. Perhaps more surprising is Eckert's
memo of January 29, 1944, describing how to store data on
magnetic disks. Lukoff points out that Eckert's memo
specifically suggests storing _programs_ on disks, predating
Dr. John von Neumann's similar -- and more widely credited --
suggestion. (The text is given in the appendix of the
book.)
After a leave from the Moore School to serve a tour of duty
in the Navy, Lukoff returns in time to join the EDVAC
project. Again he provides views of "inner workings" that
could not be duplicated by others. The EDVAC engineers
developed mercury delay-line memory to implement Eckert's
idea of storing instructions; Lukoff's task was to devise a
control system to compensate for temperature changes in the
mercury column, which would also change its length. Here he
goes beyond the technical details of the memory to provide
some of the more human aspects of developing and presenting
computational equipment:
The next day we were there at 10 a.m. to greet the
thousands of attendees. No sooner was the
equipment turned on than unusual things started to
happen. The mercury memory would suddenly fill up
with all kinds of extraneous pulses. When cleared,
The Analytical Engine, Volume 1, Number 4 April 1994 Page 31
it would refill. I noticed that the pulses were
changing rhythmically every two seconds. We
wondered what could be causing that phenomenon.
Onlookers pressed for explanation, which was
embarrassing, to say the least. As I pondered the
problem, the corner of my eye caught a rotating
Army Signal Corps radar antenna halfway across the
hall. Every time it rotated in our direction, the
mercury memory pattern changed. That had to be it!
I approached the radar operator, and asked him if
the antenna was radiating power. Sure enough, it
was. It did not make any difference to him whether
the radio frequency power was on or off, so I
convinced him to leave it off. That cured our
problem...
One day, Lukoff received a call from home -- an invitation
for him to join a company, the Electronic Control Company,
that Dr. Mauchly and Eckert were forming to build the
Universal Automatic Computer later known as UNIVAC.
Lukoff joined the company and continued to work with memory,
this time with CRT tubes. He explored the possibilities of
storing and reading an electrostatic charge from the face of
an ordinary CRT screen. This would provide faster access to
memory than the delay line did, as the cathode-ray beam
could be deflected more quickly than the acoustic signal
could be read from the column of mercury. (Paradoxically
enough, radio was again to be Lukoff's undoing. After some
detective work, he discovered a local radio station atop a
nearby building was causing distortion in his memory.
Shielding solved the problem.)
Development of electrostatic memory was shortly suspended
(to be resumed later) to free up resources for Eckert and
Mauchly's next computer, BINAC. This device, built for
Northrop Aircraft as a missile-guidance computer, is
sometimes called the first commercial stored-program
computer built in the United States; it utilized mercury
memory, as the technology was more developed and considered
more reliable.
Though Lukoff did not know it at the time, BINAC was in a
neck-and-neck race for the honor of being the first
commercial stored-program machine. Manchester University's
Mark I (MADM) ran its first stored program on June 21, 1948,
while BINAC was still undergoing construction; but BINAC was
accepted in August 1949, while the commercial version of the
Mark I was delivered to the customer on July 7, 1951. BINAC,
with 512 31-bit words and unprecedented processing speed,
was significantly faster, more powerful and more flexible
than MADM with 128 40-bit words. (BINAC, however, was
The Analytical Engine, Volume 1, Number 4 April 1994 Page 32
fragile, and the Mark I eventually made a better transition
to commercial use.) The UNIVAC I, Eckert and Mauchly's first
commercial success, was delivered to the U. S. Bureau of the
Census on March 30, 1951; forty were ultimately sold, and
their reliability owed much to Lukoff's innovations in
computer manufacturing.
Obvious throughout this book, Lukoff's engineering abilities
were practical and irrepressible, and not confined to
working hours. He shares the detail of some "incidental"
projects, like the Bubelator, which was of great
significance to his household if not to history in general.
He was minding his son while his wife was off shopping, and
quickly discovered that every half hour, he had to stop what
he was doing to rock his son's carriage. Within an hour, he
had an electric baby carriage rocker built and running. His
wife, returning, was incredulous, the baby was delighted,
and (since the baby's nickname was Bub) this crucially
useful machine was christened the Bub-elator!
Lukoff became the head of Sperry Univac's engineering
department, supervised the creation of UNIVAC-LARC and many
other UNIVAC models, established Sperry's semiconductor
manufacturing capability, and was a pioneer in computer-
aided design. He recounts his career with humor and
perspective too often lacking in works of technical history.
Enjoyable reading, insights and facts not well known, and
personal human interest are well blended here to produce a
truly valuable source in the field.
-------------------------------------------------
ACQUISITIONS
-------------------------------------------------
APPLE MACINTOSH 128
-------------------------------------------------
Bruce and Mary Yow
The computer for the rest of us! How could a computer
collector live without one? The original Mac was one of the
few computers designed within deliberate limits -- even the
standard 128K RAM was a concession wrung out of Steve Jobs,
who thought half that was plenty -- yet it gradually opened
up almost unlimited possibility. The Mac's combination of a
windowing interface, easy operation, small footprint,
affordability, and tongue-in-cheek humor made it a winner
from Day One.
In the last ten years, Macs have grown steadily more
elaborate, and it's refreshing to go back to the roots
represented by this humble yet ingratiating box. This one is
historically pristine because it never suffered the semi-
The Analytical Engine, Volume 1, Number 4 April 1994 Page 33
universal memory upgrade; in fact it seems that this case
hasn't been cracked.
However, we will need to crack it someday. It brings up a
diagnostic code that means one of the RAM chips is bad, and
we even know which chip, but opening the case of a Mac is a
task best left to those who've done it before. Anyone who
would care to donate parts and labor for the repair will
earn our considerable gratitude.
And thank you, Bruce and Mary Yow, for your donation of --
just maybe -- California's quintessential computer.
MORROW MICRO DECISION
-------------------------------------------------
Neil Abbey and Arthur Colton
That early admirer of user-friendliness, George Morrow, was
always pushing the envelope. In particular he liked to build
computers that a new owner could unpack, put in place, hook
up, boot up, and use to get some work done. This is a
sequence we take for granted now; twelve years ago, when the
Z-80A was a hot chip and internal floppy drives were
uncommon, it was much more the exception than the rule. But
Morrows worked, and this one does today.
The Micro Decision was also unusual in the scope of its
standard configuration. Its monitor was optional, but
everything else was bundled: floppy drives, ports, CP/M 2.2,
a word processor, spreadsheet, database, three languages,
and even a menuing system with a CP/M tutorial. Sitting at
this machine, watching its futuristic Lear-Siegler
turtleback monitor, evokes the dream of a computer for
Everyman and Everywoman. In 1982 that was still a distant
dream; if by now it has nearly come true, Morrow is one of
the pioneers we have to thank.
And thank you, Neil and Arthur, for your donation of this
computer, monitor, docs, full software, backups, a Silver-
Reed daisywheel printer and half a dozen ribbons! Hmm, maybe
I'll use WordStar for a few of those letters....
LOBO MAX
-------------------------------------------------
Steve Scharf
Lobo International of Goleta, CA -- down by Santa Barbara --
was primarily known as a maker of extremely rugged disk
drives. This machine came to us with a dual 8" floppy that
is a tank! I mean, think carefully before you pick it up!
Around 1982-83, though, Lobo took a side trip into computer
The Analytical Engine, Volume 1, Number 4 April 1994 Page 34
development and produced this handsome MAX-80 that, like the
drives, went just a bit further than the standard spec of
the day. Its Z-80B processor runs at 5 MHz when many
manufacturers thought that 4 was quite quick enough; it
addresses 128K RAM accessible through bank-switching,
instead of the more usual 64K. The systems manual says
proudly that the disk controller can talk to four "3 1/2 or
5 1/4-inch drives" _and_ four 8" drives _and_ a hard disk.
The video is remappable and so are the keys....
Like the TRS-80 Mod I or Atari ST, this is a "computer-
keyboard" that only needs to be attached to drives and a
monitor. Its dove-grey case and sharp-edged styling make an
aggressively attractive machine that seems packed with
muscle. Ours arrived with CP/M Plus and a full suite of apps
including Pascal, and, incidentally, some of the most
voluminous and detailed docs ever seen for a micro. Thanks a
lot, Steve; this one's a showpiece.
EXIDY SORCERER
-------------------------------------------------
Klaus Krause
The idiosyncratic and colorful Sorcerer takes us back to a
delightful time -- the days when micros were sold on the
basis of their differences, not their similarities. And this
one's pretty different!
Like the early Ataris, the Sorcerer tried to bridge the gap
between a game machine and a general-purpose computer; but
with 32K RAM and a Z-80, it had power enough to be stiff
competition for an Apple II. An unusual (and nice) feature
is a full graphics set available from the keyboard with a
special shift key. Standard ROM-packs included a word
processor, a BASIC, and an assembler/debugger, making the
Sorcerer usable with nothing more than a monitor; third-
party software was generally supplied on audio cassettes.
"Serious" options included a six-slot S-100 expansion
chassis and a 10Mb hard disk, but even without all that, the
base computer could be a lot of fun for $895.
Corporate diffidence and minor design kinks combined,
unfortunately, to deprive this machine of its rightful
stature in history. All the more reason for your Association
to have one, especially this mint one, with a whole box of
add-on goodies and user-group newsletters. Thank you, Klaus
Krause.
HP 3000/42
-------------------------------------------------
Innovative Information Systems, Inc.
Civic-minded Kristen Helm of IIS called us (it's the price
The Analytical Engine, Volume 1, Number 4 April 1994 Page 35
of fame) and told us that a 3000 was headed for the
scrapyard. Well, we had to have a 3000, didn't we?
_Everybody_ had a 3000, didn't they? A lot of people
_still_.... Anyway, we weren't about to let history be
history. Roger and Michael slam-dunked it into the moving
van and we figured we'd sort things out later.
When that time comes we'll need some help. There's a 7933
disk drive, with a disk pack that may or may not have
survived, and a 7974 tape drive with plenty of tapes, but we
don't know which tapes. We have full MPE V docs but _no_
hardware manuals. And although some of the cables are
included, some are definitely missing.... Luckily, an
amazing number of CHAC members have HP 3000 experience.
We're not sure what we'll do with this after we boot it up,
but we'll think of something. Thanks to Kristen for alerting
us and to IIS for parting with this fine workhorse.
PACIFIC DATA SYSTEMS 1020
-------------------------------------------------
Duane Atkinson
"In contrast to other equipment," proudly states the
Engineer's Guide, "the products of PDS are designed
primarily for use by the man with the problem...the business
or technical man in any capacity." The clear implication is
that there are no whitecoats between you and the hardware;
you're welcome to sit down and hack.
Built by a "leading manufacturer of small, direct-access
computers" in Santa Ana, CA, the 1020 resembles nothing so
much as a robustly built desk with a couple of typewriters
on it. The one on the right is a numeric keypad and
instruction console with a paper tape reader/punch, and the
one on the left actually _is_ a typewriter; an IBM Selectric
I whose 15 ball-bats per second could barely keep up with
the comparative torrent of output from the computer. Memory
is installed in modules of 1024 signed words, each four
decimal digits; the standard complement was two modules, and
two more could be added. The "magnetostrictive delay lines"
(we'll have to look that one up) operate at a hair-raising
two-and-a-hair megabits per second. Scientific programming
was accomplished through the "PDS Engineering Interpreter,"
an English-like symbolic interpreter that proposed to take
the black magic out of programming.
This revisionist stance is typical of the whole computer --
and its docs. "[A] large computer installation," trumpets
the sales literature, "requires specialized facilities, and
most important, specialized personnel.... Many manufacturers
will lure you with specialized 'languages,' FORTRAN-type,
ALGOL-type, and other strange types. This is a siren song,
The Analytical Engine, Volume 1, Number 4 April 1994 Page 36
and you'll do well to beware of it...." The quotes around
"languages" are in the original. Ah! Do the old promises
ever lose their charm? Are they yet stale through
repetition? Here, says PDS, is a power-sipping 350-pound
mini that you can trundle into your own office (on fat
casters) and talk to like a faithful friend. No air
conditioning required.
Claims for ease of use may be exaggerated, but there's no
doubt that Duane Atkinson and his company pushed a fair
amount of work through this computer. It was quickly
discovered that sheet-fed output -- for the Selectric lacks
any tractor -- made essentially no sense; so Atkinson
constructed an unusual aluminum bracket that holds a roll of
Teletype paper behind the platen. A true engineer's touch on
this bracket is a lamp-and-photocell loop that automatically
stops output when the roll runs out.
We've never seen another PDS computer and we suspect that
they're scarce. According to the sales literature, the
company was founded in 1962; this 1020 dates from 1964, and
Mr. Atkinson recalls that PDS was purchased or otherwise
absorbed by Control Data Corporation in 1966. If you
remember more than that about this somewhat mysterious
manufacturer, send us detail by e-mail or drop a line to the
El Cerrito address.
HP 9100/9125
-------------------------------------------------
Duane Atkinson
HP's own product literature on the 9100 begins with a
straight-from-the-shoulder description: "a programmable,
electronic calculator which performs operations commonly
encountered in scientific and engineering problems...." Ah,
but then you get to the jaw-dropping list of what it'll do.
Arithmetic ops, log functions, trig functions, hyperbolic
functions, coordinate transforms....floating-point....40-
millisecond square root.... Gee. This poor thing never
realized it was a computer -- though the HP data sheet slyly
refers to its 2208 bits of core memory "[that] enables the
calculator to store instructions and constants...." Programs
can be read from and written to mag card, and an integral
CRT handles the display. Obviously, nothing was spared to
make it ready for day-in-day-out use. Results could be
tidily transmitted to the 9125 flat-bed pen plotter with a
10"x15" (25x38cm) plot area.
This is one of the all-time great desk calculators and, at
over US$7,000 for the calculator and plotter together, was
meant for the serious customer of the late sixties.
Naturally, though, it was part of a history with its own
The Analytical Engine, Volume 1, Number 4 April 1994 Page 37
inexorable logic. Bruce Flamm, in the newest issue of the
_International Calculator Collector_, quotes HP developer
Tom Osborn:
About fifteen minutes....after we finished the 9100,
Bill Hewlett said we should have one in a tenth the
volume, ten times as fast, and at a tenth the price.
Later he added that he wanted it to be a shirt-pocket
machine.
Hewlett's target, which seemed almost ludicrously distant in
1968, was comfortably met four years later by the HP-35 --
thirty thousand transistors of MOS/LSI in a nine-ounce box
for $395.
(When Doug Jones -- author of our January article on the IBM
701 -- heard that we had a 9100/9125, he mailed us a program
he'd written that does Spirograph-style decorative plots. If
we _ever_ have enough room around here for a decent test
bench, it's one of the first things we'll run.)
IBM 5100
-------------------------------------------------
Duane Atkinson
The IBM 5100, a luggable the size of a large briefcase, was
IBM's first foray into the market for authentically small
computers. As such it was very serious iron -- and very,
very expensive.
With a built-in cartridge tape drive, integral keyboard,
numeric keypad, and APL or BASIC in firmware, the 5100 could
be used as a line-powered standalone -- although its 50
pounds precluded easy portability. On the other hand, when
set up with the optional 5103 dot-matrix printer, external
storage, and a CRT monitor, this could stand in capably for
a small IBM mini. Not a schematic descendant of IBM's bigger
iron, nor quite the technical ancestor of the 5150 PC, this
5100 still stands as the conceptual link between Big Blue's
large and small computers. Our example worked hard for a
living in California and will find a comfortable retirement
in your Association's museum.
ASR-33 TELETYPE
-------------------------------------------------
Duane Atkinson
Yes! We finally have a Teletype. Early micro hackers will
remember the days when an ASR-33 was _the_ I/O terminal of
choice, indispensable yet unattainable, miserably scarce on
the open market even if you could part with the small
fortune required. Now we have one reputed to be in working
The Analytical Engine, Volume 1, Number 4 April 1994 Page 38
order, ready to connect to (for example) our IMSAI
8080....and the Golden Era of micros will resound again.
Earplugs, anyone?
We thank Duane Atkinson for his comprehensive generosity and
hope that we can give his meticulously maintained hardware
another few decades of use. At a stroke, our collection was
expanded to calculators in one direction and minis in the
other. (DID I MENTION ANYTHING ABOUT NEEDING SPACE??!!)
HP 86
-------------------------------------------------
Alan Hawk
Thanks to Alan's public spirit, this computer came home to
us from Maryland....in two forty-pound boxes! Introduced in
late 1982, this proud member of "Series 80" (Models 85, 86
and 87 in their various flavors) is one machine that builds
the bridge between HP's programmable desktop calculators and
the Series 100 true micros.
This one includes a monochrome monitor, 128K RAM in two 64K
plug-in "drawer" modules, a 9130A floppy drive, and
carefully preserved manuals and software including HP
Word/80, HP File/80, and VisiCalc PLUS complete, and docs
(though no disks) for FORTRAN-77 and the UCSD p-System. This
computer worked for its living, whether heavily or lightly,
until 1993. In March of this year, with due care and
forethought, it was donated to our Association; we've said
this before and since, but people simply _don't_ throw out
Hewlett-Packard computers, and that's fine with us! Thank
you, Alan Hawk.
BYT-8 (OLSON 8080)
-------------------------------------------------
Frank McConnell
The BYT-8, which might be called the "definitive S-100 box,"
was the house-brand computer of Paul Terrell's Byte Shop --
the first retail computer store on the West Coast -- which
opened in Mountain View, CA, in December 1975. The one Frank
brought us is quite capable, being stuffed full of I/O and
boot hardware from Byte and Cromemco, and a Tanner memory
board that seems to hold 64K.
A nice extra for this machine is an Olson front panel. The
Olson 8080, which Haddock's _Collector's Guide_ describes as
"a variant of the Byte," was apparently identical to it
internally; but whereas the BYT-8 has a blank faceplate, the
Olson has a Real Front Panel(tm) with three rows of bat-handle
toggles and two rows of LED's to permit Altair-style access
to the works.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 39
Straightforward and significant, this is one of the machines
we'd love to hook up to our new ASR-33. And would anyone who
has full docs for it please contact us?
HP 2674A PRINTER
-------------------------------------------------
Frank McConnell
Frank heeded our call for an HP thermal printer and snagged
the next one that appeared to his practiced eye. As we
mentioned in the January issue, this clever device fits
under the top hatch of the HP 150 and spares the user from
lugging around a separate printer during field use -- or
taking up desk space with one in the office. We have no idea
what it originally used for paper but we're going to try fax
paper.... Actually, if anyone had ever written faxmodem
software for the 150, we'd make a fax machine out of it;
like most things HP ever built, it obviously prefers work to
leisure. Now we just need the hard disk.
Frank also brought us a box of miscellaneous publications
including two Symposia of the West Coast Computer Faires, a
CDC 6400 manual from UC Berkeley, some early issues of _Dr.
Dobb's Journal_ and the BAMDUA (Morrow) and BAKUP (Kaypro)
users' group newsletters, and a nearly flawless complete run
of the early-eighties CP/M publication _User's Guide_.
Gee....time for some Ziploc bags and another bookshelf.
Thanks, Frank!
APPLE MAGAZINES
-------------------------------------------------
Harold A. Layer
Two boxes hit the step the other day, and proved to contain
a donation from CHAC member and eminent collector Hal Layer
-- accumulations of several magazines related to early
Apples, including _Apple Orchard_, _Nibble_, the San
Francisco Apple users' group newsletter _Cider Press_, and a
long, long run of the justly renowned _Softalk_. Apple-
related publishing has been a roaring industry in
California, and we thank Hal for his help with the first
steps of a long hike toward comprehensive archiving.
APPLE PROJECT DOCS
-------------------------------------------------
David Craig
CHAC's newest member, David Craig, was kind enough to send a
bundle of technical reference documentation concerning the
Apple Lisa and the earliest days of the Mac -- material that
sheds light on a fascinating, and turbulent, period in the
The Analytical Engine, Volume 1, Number 4 April 1994 Page 40
history of a great company. At press time we've barely
looked at these documents (press time's like that) but we
know that they'll drastically improve our understanding of
both our Mac XL and our Mac 128. Thanks, David, and welcome
aboard.
-------------------------------------------------
LETTERS
-------------------------------------------------
MORE COMPUTER MUSIC
Most interested to read Tom Ellis' article in issue 2 about
making the tape drive sing. Reminds me that I wrote the
music program for my first computer, the Ferranti Perseus,
delivered in 1958 or -9. Ferranti computers in those days
always had a decent 9" loudspeaker and amplifier which could
be plugged into any useful waveform, so tone generation was
no problem. To get percussion, however, I found that reading
a zero length tape block from Tape Drive #0 made the pinch
roller make a satisfactory click, without stealing time from
tone generation. La Paloma, Flight of the Bumblebee and
Arrival of the Queen of Sheba were _pieces de resistance_,
with rhythmic backing to the foreground melody! Colleagues
sat up all night punching the notation on paper tapes.
-- from Chris P. Burton, Computer Conservation Society (UK,)
via Internet
MORE ON THE PLATO SYSTEM
Joe Cychosz sent me the short article by Doug Jones entitled
PLATO and SMALLTALK which appeared in your newsletter in
January. It was very interesting! Although I was a system
programmer at PLATO during the 70's, I wasn't aware of the
exchange between PLATO and Xerox PARC (or if I was, I forgot
about it.) I'd like to correct a couple of minor factual
errors in Doug Jones' article:
- He states that multi-user games were the single most-used
PLATO application. Actually, notesfiles were the most-used,
consuming about 34% of all hours of PLATO use. Games were
second, at about 20%.
- PLATO notesfiles were never moderated, at least not in
the sense of USENET moderated newsgroups. All postings
appeared immediately without any review. A notesfile
director had the power to delete a note after it was posted,
however. A 1991 rewrite of Notes on the NovaNET system added
the capability for moderated notesfiles for the first time.
With permission from you and Doug Jones, I'd like to repost
his article on NovaNET, where I think a lot of people would
The Analytical Engine, Volume 1, Number 4 April 1994 Page 41
be interested. (NovaNET is the new name for the PLATO system
at the University of Illinois.)
-- from David Woolley, via Internet
TIME-SHARING FOLKLORE (ET CETERA) WANTED
Greetings! I'm currently doing research on early time-
sharing/interfaces/programming, and I was wondering if your
organization might be able to provide me some leads towards
information on these topics. Any info would be greatly
appreciated. (I'm open to all sorts of bizarre topical
material in these fields.) Thanks.
-- from Blaine Jack, bjack@wvnvm.wvnet.edu, via Internet
LOOKING FOR DAVID AHL
Can anyone suggest how I can contact David H. Ahl? He was
the editor of `Creative Computing', a US computing magazine,
from the mid 70's until the mid 80's.
The last address I have for `Creative Computing' dates from
1985, and I doubt whether it's still valid (a letter sent
there has never been answered).
Thanks,
-- from Andrew Davison, ad@munta.cs.mu.oz.au, via Internet
TECH NOTE ON THE HP 5243
About a year ago, I was offered an HP 5243 counter for 1
pound. On the grounds that _anything_ HP is worth 1 pound,
of course I bought it. This thing has the most amazing
counter/latch/display boards you've ever seen - the only
active components for a 4-bit 1242 BCD counter, 4 bit latch,
and nixie-tube decoder driver are 8 transistors. Here's how
they did it:
The 8 transistors make 4 flip-flops, which are combined with
steering diodes to make the counter. Each flip-flop has 2
outputs, the collectors of the 2 transistors, which are
either at +19V or -10V (approx.) These outputs are fed
through 56k resistors to neon lamps, and from there to a
common 390k resistor to the -130V power supply. There are
diodes connected to the junction of the neon and the 56k
resistor, and the other ends of these diodes are connected
to the transfer control line. Normally, the transfer control
line is at +19V. Therefore, as the diodes are forward
biased, the voltage on that end of the neons can't drop
below 18.5V. Now, if one neon is lit, and dropping say 55V,
The Analytical Engine, Volume 1, Number 4 April 1994 Page 42
there will be insufficient voltage across the other one (a
maximum of 55.5V) to strike it. So, it stays in that state,
no matter what the flip-flop does.
Now, if the transfer control signal is pulled to -10V, then
the diodes are always reverse-biased. So, the neon connected
to the +19V output of the flip-flop now has sufficient
voltage across it, and lights. In so doing, it pulls the
junction with the 390K resistor up to 19-55 = -48V. Now, the
other neon has only -10-48V=38V across it, so it goes out.
The neons are now in the same state as the flip-flop.
Pulling the transfer control line up to +19V again returns
the system to the initial state, and latches the state of
the neons.
These neons are mounted in a plastic block lined with
reflecting metal (one reflector/neon) mounted on the counter
PCB. On top of this block is placed a ceramic thick-film CdS
cell array, which implements the decode tree needed to
convert 1242 code into decimal. This tree is in series with
the cathodes of the nixie tube display, and directly drives
it. So, a pattern of 1's and 0's on the neon matrix is
converted to a decimal display on the front of the
instrument.
Truly a wonderful hack, from the neon latches to the thick-
film decoder. They don't make them like that anymore....
-- from Tony Duell, via Internet
RING AND LOOP NETWORKS: REPLY TO KULIKOWSKI
Following up on the comments by Stan Kulikowski on the
history of ring and loop network technologies in ENGINE #3,
I looked in the bibliography of J. R. Pierce's "Network for
Block Switching of Data", Bell System Technical Journal,
Vol. 51, No. 6, July-Aug 1972. This is the paper in which
Pierce proposed the slotted loop network technology that was
the basis of the one of the first LAN systems at Bell Labs.
Pierce submitted his paper in late 1970, and among his
citations, perhaps the most relevant are a paper by Edgar H.
Steward, "A Loop Transmission System," in the 1970 IEEE
International Conference on Communications, and Farmer and
Newhall's paper, "An Experimental Distributed Switching
System to Handle High Speed Aperiodic Computer Traffic, in
the 1969 ACM Symposium on Problems on the Optimization of
Data Communications Systems.
It is hard to pin down the origin of an idea, but in the
case of J. R. Pierce's slotted ring network, we do know that
by 1972, the prototype network at Bell Labs included a
The Analytical Engine, Volume 1, Number 4 April 1994 Page 43
number of Honeywell (or 3C) 516 minicomputers. By 1973, when
I used the network, one minicomputer was configured as a
file server, while the other three were configured as
workstations, each with a CRT display, a mouse, and other
modern toys. I do not doubt that there were other early
local area networks, but few are likely to have been put to
uses that are so typical of the way we now use LAN
technology.
-- from Doug Jones, via Internet
TAKE A CRAY JUST BECAUSE IT'S THERE?
My university is in the process of trying to decide whether
to acquire a supercomputer. I understand it is some sort of
Cray, a fairly capable one. There is an apparently wonderful
deal by which a state agency will "give" us the computer.
All we have to do is operate it, support it, and (perhaps--
I'm not so sure about this) give them free use of it.
Part of the support will involve starting up an entire
empire of support and programming services, space to house
it, air conditioning, etc., etc.
I am not at all sure it's worth it. Some of the
alternatives, for example a network of high-end UNIX
workstations, maybe with a vector processor attached, sound
better to me. I've seen a couple of posts mentioning the
costs and other gotcha's involved in supporting these
beasts, including someone who mentioned that they had just
turned theirs off for good because of the expense of running
it. I would be very interested in hearing more about the
potential problems and expenses involved, and getting more
detail about sites that have acquired these beasts and lived
to regret it. Thanks in advance.
-- from Ross A. Alford, via Internet
HARDWARE RECYCLING: ASK, THEN DUCK
I'm looking for any and all information about uses people
have found for old and/or obsolete hardware; that is, apart
from junking it or turning it into aquaria. Some specific
things I have in mind are donating it to ex-Iron Curtain or
developing countries, local schools, etc.
If you know or know of someone who does this, any
information at all would be greatly appreciated. Thanks!
-- from Gal Kaplan, gal@das.harvard.edu, via Internet
The Analytical Engine, Volume 1, Number 4 April 1994 Page 44
LOGO'S TURTLE: REPLY TO JONES
In _The Children's Machine_, Seymour Papert writes: "The
turtle came from thinking about how on earth a child could
capture in computational form something physical like
drawing or walking. The answer was a yellow robot shaped
rather like R2D2 and, like him, mounted on wheels.... In
those days, the turtle was a large object, almost as big as
the children who were using it, connected by wires and
telephone links to a faraway computer that filled a room."
The turtle was developed some time after Logo debuted, but
he doesn't give a date for either nor any measure of the
time lapse.
GUI HISTORY: REPLY TO KULIKOWSKI
_The NeWS Book_ (Gosling, Rosenthal, and Arden, Springer-
Verlag, 1989) has a chapter on the genealogy of GUIs. It
starts with the Alto at Xerox PARC "in the early 1970's,"
running Smalltalk and the associated windowing system, which
had all elements of the system running in one address space,
communicating via procedure calls. DLisp is mentioned next,
also at Xerox PARC, developed in 1977. The Altos didn't have
enough computing power, and their PDP-10 didn't have
graphics, so they developed a Lisp-based windowing system
which ran on the PDP-10 but displayed on Altos connected via
Ethernet. "After DLisp, Xerox PARC developed a number of
window systems supporting multiple processes in a single
address space. These were all implemented in the Mesa
programming language." Tajo, one of these systems,
introduced icons and was the first notification-based
system. Star, apparently another Mesa-based system, was
Xerox's attempt to go commercial.
"In 1981 at MIT, the advent of the Motorola 68000 led to an
attempt to build a workstation and its software environment
called NU." This was a UNIX-based system, and all the GUI
code was integrated into the UNIX kernel.
The W Window System was developed in 1982 for the
experimental V operating system running on Sun hardware.
SunWindows, introduced in 1983, is next on the list, billed
as the first widely-used UNIX window system. (It must be
mentioned here that this book was in fact published by Sun
and may be biased.) Some of the GUI code is in the kernel
and some is linked into each application. Another system,
Andrew, was developed in the same year by Gosling and
Rosenthal, designed for a system which did not exist through
most of the development, the PC/RT. The GUI code is in a
user-level server process. "Andrew was the first practical
The Analytical Engine, Volume 1, Number 4 April 1994 Page 45
UNIX networked window server."
The Apple Macintosh GUI, introduced in 1984, is in many ways
a throwback to the very early days; once again, all elements
of the system are in one address space and communicating via
procedure calls.
X, based on W, was developed in 1984-85. X10 was the first
widely-available version, but turned out to be relatively
unportable. The following version, X11, was much more
successful. X was similar to previous systems in that the
GUI code was a user-level server process, but different in
that the window manager was yet another user-level server
process.
In 1986, NeWS was developed, and there the history ends.
Microsoft Windows is not mentioned, presumably because at
the time of publication it was still a market nonentity,
since 3.0 had not been introduced yet.
-- from James W. Birdsall, jwbirdsa@picarefy.com, via
Internet
NEW OLD-IRON CLUB AT CORNELL
First thing next semester (August 20th or so) we'll begin
getting the paperwork together to get the Cornell University
Classic Computer Club registered as a genuine University
Club, with a faculty advisor and all that other wonderful
stuff. We have been quite active in our formation process.
It all started late last September when an avid DEC
enthusiast/Cornell Senior was running out of room in his
apartment, and posted on a local newsgroup that he had a
MINC-11 free for the taking. I had been interested in the
history of computers for quite some time, but I'd never
actually *seen* anything older than the TRS-80's used at my
elementary school, so I snapped it up in a fit of nostalgia.
Needless to say I was hooked. My friends and I spent long
hours fiddling with it, getting the power supply to work,
and generally admiring the RX-02 floppies. We named it
'Sparky' ....and decided that we *really* had to get it
working, and *really* had to see what else we could find.
Thus was the CUCCC born.
Our one goal is to preserve and protect classic computers,
and make sure people never forget the humble (and not so
humble) beginnings of that multimedia whiz-bang box they
bought at ComputerLand and don't properly appreciate. ("Only
80 WinMarks? What kind of computer is that!??")
As of today, we've got much more equipment than we've got
The Analytical Engine, Volume 1, Number 4 April 1994 Page 46
space to put it in. Several PDP-11's of various flavors,
IMSAI's, DEC Rainbows, terminals, bits n' pieces, a
Honeywell DPS6, etc., so we need to get organized and find
an on-campus home for these. Once we're official, people
will be able to donate things to us and get a lovely tax
deduction, because donated computers will be property of the
club, ergo of Cornell. Nice how that works. Eventually we
want to get a proper display of classic computers going, a
hands-on look at computer history (granted, a very small
slice of computer history) somewhere on campus.
Starting in late August, we'll have a newsletter (name
currently being discussed) and other general club info
available via WWW, gopher, and ftp at
motherbrain.ithaca.ny.us. (Unfortunately, we haven't got a
good place to send snail-mail to yet.) We'll have an FAQ
made up RSN, er, we hope....
-- from Seth J. Morabito (sjm1@cornell.edu), Cornell
University Classic Computer Club, Cornell University,
Ithaca, NY
[Seth,
Welcome to the great Club of Clubs! You have some
interesting times in front of you, as we know all too well,
and we wish you the very best of luck. Please keep the
ENGINE informed of your progress and, especially, send along
your newsletter when it appears.
-- KC]
-------------------------------------------------
QUERIES
-------------------------------------------------
AEGIS/APOLLO WORKSTATIONS: ANYTHING WELCOME....
In addition to collecting old DEC 12 & 18 bit machines I've
started collecting old Apollo computers. So far I've got
several DN300, DN400, DN500 series machines. I've developed
my own CAD system for the Apollo and eventually hope to
store the schematics for the DEC machines online. The idea
is to use old workstation technology from the early 80's to
assist in maintaining old machines from the 60's.
I've set up an archive on "nickel.ucs.indiana.edu" for
storing old PDP8 source code, binaries, and documentation.
I'd like to start archiving old Apollo stuff before it
starts to disappear.
General Questions:
The Analytical Engine, Volume 1, Number 4 April 1994 Page 47
1) Has anyone written a book about the history of Apollo and
Poduska (the founder) yet? If not, is anyone familiar with
the company's history interested in writing down a few pages
of text for the archive?
2) What is the status of Aegis 9.7 source code? Is there any
chance HP/Apollo will release it?
3) What about schematics for the old, long obsolete
machines?
4) What journal articles/conference papers related to Apollo
still exist in ftp'able format? So far the only one I've
found is "An Extensible I/O System" that was presented at
the 1986 Summer USENIX conference.
5) What is the status of the Apollo Users Group? Do they
have an archive anywhere? Does anyone have proceedings of
old user group conferences?
Wanted:
1) I'm interested in buying old Apollo workstations/disks.
If it's junk to you I'm probably interested in it.
DN3000/4000 machines are still too modern for me (unless the
price is right).
2) If you've got some old Apollo sales literature buried in
your filing cabinet I'd be interested in it.
3) Aegis 9.7 on 8" floppies and tape cartridges.
4) Old Apollo training materials. Did Apollo offer any
training classes to customers?
Aegis Questions:
1) What is the format of object files produced by Aegis
compilers?
2) What is the format of the executables in the sau
directories?
3) Are there any manuals on Aegis 9.7 internals?
Thanks in advance.
-- from Jeff Russ, via Internet
ALTOS 2086: RELUCTANT DISK
I have an Altos 2086 that I'd really like to get running
The Analytical Engine, Volume 1, Number 4 April 1994 Page 48
again, but unfortunately it lost its hard drive some time
ago. I have a 'new' drive I can use, but no way of low-level
formatting it.
If anyone can supply a program to perform a low-level format
I'd be most grateful. I have the SDX disk, but it seems to
only be able to format floppies; I've tried low-leveling the
drive on a XENIX peecee with an MFM controller, but the
Altos wants its drive formatted with 16 sectors. Thanks in
advance...
-- from Geoff McCaughan, via Internet
AT&T IN/ix: DOCS AND CONTEXT NEEDED
I have a port of UNIX System V Release 2, known as IN/ix, to
the AT&T 6300 (a PC/XT-class machine) by Interactive. As I
understand it, this system was never officially marketed; I
obtained my copy when a warehouse-full of AT&T 3B1 machines
and associated items was discovered and sold off to the 3B1
community. It seems to have been distributed by a company
called Media Software and Systems, Inc., of Aurora,
Illinois. That company has either gone out of business or
moved, and Interactive has not answered my queries about
this product. I am particularly looking for manuals, since I
have only the introduction/install manual, and there are no
online man pages. Between my manuals for its predecessor,
PC/IX, and my System V manuals for my 3B1, I can operate it,
but there are tantalizing hints of features, such as
overlayed executables (to escape the 64K+64K I+D program-
size restriction), which I have no idea how to access. Any
help is welcome.
-- from James W. Birdsall, jwbirdsa@picarefy.com, via
Internet
ATARI x00: STAR RAIDERS
Does anyone else remember the greatest video game of all
time, _Star Raiders_, on the Atari 400/800 computers? An
identical version was later released for the 5200 system. I
would like to find out exactly who designed the game.
-- from Matt McCullar, via Internet
ATARI 800: FASTER THAN THE OTHER WAY AROUND....
I heard there is an Atari 800 emulator for IBM PC/AT
machines floating around, does anyone know where I can get
this? Any help is appreciated.
-- from David Fox, via Internet
The Analytical Engine, Volume 1, Number 4 April 1994 Page 49
ATARI 2600: REPLY TO PATRICK FLEMING'S QUERY
There was talk of a project to make an Atari 2600 emulator a
few years ago but nobody ever followed through. If you want
to attempt to code the 2600 there is hope though.
You can buy the new 2nd version of the Atari 2600/7800
assembler cart. Basically it allows you to assemble 4k of
code in 2600 mode and 8k of code in 7800 mode (it works on
both systems). I'm not sure if you can actually save
anything to disk or tape with any kind of cable...but I'm
under the impression you can't.
It comes with a 250 page official Atari system document for
the 2600 and 7800 along with sample code from some games to
get you started. And as you already know there are several
good guides to coding the 2600 on the net. Pick up a good
book on 6502 and you're all set...
If you want the address I can go dig it out for you...it's
about $60. (Basically it's a hacked Hat Trick cartridge for
the 7800, so you NEED a 7800 to run it, even if you just
want to use the 2600 side.)
-- from Ralph Barbagallo, via Internet
C PREPROCESSOR: EARLIEST MENTION
I'd like to know the first mention of the C preprocessor in
literature. Was it in K&R, or even earlier? Also, since when
does it exist?
Many thanks in advance!
-- from Lars Duening, via Internet
CARDS: DIG DEEP FOR THESE DOCS!
I have two cards that I am trying to ID. One is a full-
length card with a full-length daughterboard marked:
JATEK Design Corp.
A/DFLGIFB
It appears to be some A/D board. There are 2 small 1/8"
phone plugs on the mounting bracket along with a 2x12 square
male connector where I assume the analog connections are
made. I have searched my CDROM telephone book, but only turn
up a construction company is Illinois. Phone number or
operating instructions would be appreciated.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 50
The other card has no manufacturer listed, but contains a Z-
8 processor, 8K of static ram and a 27C128 EPROM. There is a
red dipswitch which sets the base address of the card. It
appears to have 4 ports which supply some type of valid
data; they are in pairs 220-221 and 226-227, for instance,
when SW1-5 is in the off position. The value returned from
the first pair of ports is usually 79$, which seems fairly
close to a 8-bit signed 0 (80$), although adjusting the pot
on the mounting bracket did not seem to change this offset.
There is an RCA jack on the mounting bracket just below the
trim pot. The card has "p222" printed towards the front of
the board on the component side. The EPROM does not appear
to be in the PC's address space, so it must just drive the
conversion and scaling. Help, in the form of operating
instructions, would be appreciated.
-- from Guy Cox, via Internet
COCO: GETTING COMPREHENSIVE
I'm working on a bibliography of references (books,
articles, etc) for or about the TRS-80/Tandy Color Computer.
If you know of any references, please send me the
bibliographic information (author, title, date of
publication, pertinent pages, etc), or at least enough
information so that others can look up the reference. Thank
you.
-- from aaron.banerjee@his.com, via Internet
COMPAQ PORTABLE PLUS: OPAQUE DIPSWITCHES
I need some information about the Compaq Portable Plus. This
machine was manufactured in 1982 by Compaq Computers. It has
2 banks of 8 DIP switches on the motherboard. I need to
reconfigure the machine, but I don't have the switch
names/functions. I would appreciate it if anyone who has a
manual would send me a listing of the switches and their
functions. I have asked a few people locally, but no one has
any information on this machine. (Some people don't even
know what DIP switches are: "Just hold down DEL and wait for
the CMOS menu to pop up!")
-- from Bryan M. Armstrong, via Internet
COMPUCOLOR
I've been curious about the CompuColor. From what (little) I
can remember, it was an all-in-one system (like the TRS-80)
driven by a Z80 processor. It sold for about $1000 and
The Analytical Engine, Volume 1, Number 4 April 1994 Page 51
included one floppy drive (back in about 1981). It had low-
resolution (128x64?) 16-color graphics.
The company was based in Minneapolis, MN, and I think the
system could run CP/M. Does anyone know anything about this
interesting computer artifact?
-- from Jon Dunn, via Internet
CONQUEST (GAME) ORIGINS
I've begun hacking away at the source of the venerable game
'Conquest'. Its source is, to put it lightly, a mess.... If
anyone can help me to find out who, where and when it was
originally made, I would be very grateful.
-- from Lars Clausen, via Internet
CONVERGENT TECHNOLOGY MINIFRAME: HELP WANTED
I have recently inherited 3 CT MiniFrames (caught them
before they hit the landfill) and have some general
questions.
3B1/7300 Binary Compatibility: As I understand it, a
MiniFrame is somewhat binary compatible with the AT&T
3B1/7300 (which is based on it). What are the limits of this
compatibility? (Obviously I can't run anything which uses
7300-specific HW, but what other limits exist?) Can I run
some of the 7300 packages available on the Internet?
Specifically, gcc (any version) or BNU?
Disks: It appears, from looking at it, that the MF can
support disks w/ >7 heads, but is limited to 1024 cylinders
by the WD1010 controller. Can I up the cylinder limit by
replacing the 1010 w/ a 2010? Would I need other patches?
Availability of Hardware: Only one of my systems has more
than the base 512K RAM on the motherboard. Obviously, I'd
love to find more somewhere. Is MF hardware easy to find?
How expensive, say, for a memory upgrade? What about an
Ethernet board and TCP/IP? (A brochure I have says that they
did exist at one time.) What about disk mounting brackets? I
*really* need one for a primary disk and could use a couple
for secondary disks. (No second disks in any of my systems.)
Discussion of MiniFrames: What is the correct USENET
newsgroup?
Thanks in advance...
-- from John Ruschmeyer, via Internet
The Analytical Engine, Volume 1, Number 4 April 1994 Page 52
CYRNET NETWORKING: LACK OF DOCS
Does anyone know anything about a networking package called
Cyrnet?
It's currently in use at the City Of Richmond Radio Repair
Department, and they wanted to expand from the current setup
of 4 computers to [more]; unfortunately, the manuals have
all been lost. They called the company that made it, and
were told: "We haven't sold software for more than 5 years.
We don't even have any of the people from that division
still working here. Sorry, we can't help you."
If anyone knows anything, It'll probably be a help, A spare
manual would be a godsend.
-- from William W. Arnold, via Internet
DATA GENERAL ECLIPSE MV/1400 DC: NOT MUCH TELLING
I have obtained an old Data General Eclipse MV/1400 DC
minicomputer running AOS/VS (installed on 38 Meg(?) HD)
Problem is: I don't know the system password, I don't have
the system tape, and I don't know AOS/VS.
In short: any info on this beast including the OS (or, if
anyone knows, where I can obtain DG/UX) is *very* welcome!
-- from Johannes Elg, via Internet
DIGITIZING PUCKS NEEDED
I need help with two old digitizing tablets:
Summagraphics MM961: I need *any* style of the cursor (puck)
or stylus. This model was normally shipped without one,
expecting the user to purchase it separately.
Talos 4020 "Wedge": I need a manual for this unit. I would
also like to obtain a cursor/puck if one is available (I
already have the stylus).
-- from Alan Frisbie, via Internet
FRONTIER TERMINALS: DOCS NEEDED
Recently I came into an Altos 486 computer running Xenix.
The machine was equipped with two Frontier terminals. These
seem to be standard RS232 interfaced terminals with pretty
nice graphics capabilities. Unfortunately, no docs came with
the terminals. I would like to use them in a project where a
The Analytical Engine, Volume 1, Number 4 April 1994 Page 53
graphics display would be handy. If you have any info on the
escape sequences which generate the graphics commands,
please email me. Thanks!
-- From dcongdon@delphi.com, via Internet
HP: MASQUERADING AS AT&T
I have this Hewlett-Packard portable terminal/computer that
I'm trying to find out about. It says AT&T Information
Systems on the front but, is a HP unit underneath. The main
unit is a calculator type qwerty keyboard with a very small
LCD display. It also has a Mag strip reader. It is encased
in another unit that has NiCad batteries and phone jacks.The
back says:
2450A07925 64K RAM
It has one 2-pin plug (power?), a 3-pin jack (DIN type?),
and two other 3-pin jacks that are some kind of in and out
as they are reversed (M/F) from each other. Questions: What
is it?, What voltages does the power plug take?, and what
can it do? Help. I have nowhere else to turn.
-- from Tom Reese, via Internet
HONEYWELL DELTA 2000: OBSCURE DATA STORE
I'm looking for people who have experience with Honeywell's
Delta 2000 computer system. In the Delta 2000's printer
controller, the data for the formatting of the printer
output is, as one might imagine, stored in a ROM. My
question relates to the buffers in which the data to be
output was stored -- the data storage apparatus -- and its
relationship to the data processor in the printer
controller. Did the data processor in the printer controller
use the data storage apparatus as a variable data store
during processing? Also, I wonder what sort of hardware was
used? The information that I have found on this does not
indicate whether it is RAM, core, or something worse.
If someone could point me at someone whom might know the
answer to these questions or could suggest some publication,
library, etc. that might be of help, I would be very much
obliged.
Andrew Robertson
Department of the History of Science
Harvard University
The Analytical Engine, Volume 1, Number 4 April 1994 Page 54
HUSKY HUNTER: WHAT WAS IT?
....I amuse myself by messing around with old and strange
hardware, and recently I've managed to come across a
portable computer with the rather dubious name of Husky
Hunter.
Not only does the name remind me of a washed-out B-actress,
but it appears that this British marvel of engineering is
also supposed to be connected to some military hardware or
similar....
Does anyone know anything else about this greengrayish piece
of obsolete hardware? It's quite cute in a brutal and
slightly fascist kind of way.
-- from Jan Besehanic, via Internet
LISP MACHINE MONITORS: THIS HAS TO BE SOMEWHERE
Does anyone know any technical specs of the large monochrome
monitors that were used on LMI Lisp Machines? We have four
of them (three in working order) that we would like to put
into service in some other way. We can do some electronic
modifications as needed if the basic information can be
found.
Also, anyone who wants to deal with the State of Georgia
surplus property system could purchase various parts of
defunct Lisp machines from us. I'm the person to contact.
The red tape may be non-trivial, but we'd like the equipment
put to good use somewhere.
-- from Michael A. Covington, University of Georgia,
mcovingt@ai.uga.edu
MICROLOG CONTROLLER: WHERE'D IT GO?
I picked up a pair of surplus 8" disk drives with a Microlog
controller card that fits in a standard 8 bit ISA bus slot.
I'd like to find someone who knows what formats this
hardware was able to read, and ideally find the software for
this combo. I've got some old CP/M disks I'd like to read,
as well as some more oddball formats such as NBI word
processor disks.
Failing that, does anyone know of a current
hardware/software solution to read disks from old systems?
-- from Dave Lacey, via Internet
The Analytical Engine, Volume 1, Number 4 April 1994 Page 55
NORTHSTAR ADVANTAGE: GETTING A GRIP
I recently purchased an old run-down NorthStar* Advantage
computer. Unfortunately, the person I bought it from didn't
have any type of technical references or documentation on
many of the programs included. I don't even know how to use
the ED.COM program (doesn't work the same as DOS's EDLIN).
Any ex-Advantage gurus out there? Have an old Advantage
manual laying around? I'd be interested in any helpful
information you could provide.
Some of the software packages I'm having trouble with are:
F80.COM, etc. (FORTRAN compiler)
ED.COM (commands, etc.)
ASM.COM (instructions, etc.)
using the BASIC compiler.
The OS for the machine appears to be:
NorthStar* Graphics CP/M (R) Release 2.2
1.1.0 AQH ADVANTAGE(tm) Version
Thanks for any help you can provide! Let me know if you need
any further info in order to help....
-- from Bob Galles, via Internet
SIEMENS PC-MX2: NOISY BUT NEAT
Well, I just bought....a Siemens PC-MX2 UNIX box with 3
CPU's: a main NS32016, on the serial card an 8085, and on
(what I think is) the network adapter, an 80188. It is a
neat little computer although it makes a hell of a noise (3
fans). If anyone knows anything about the PC-MX2, I am
interested in information about this adapter; I have some
TCP/IP software for the computer but I don't know anything
about the hardware. The "net-card" has 2 DB-style
connectors, one 25- and one 15-pin.
-- from Michael Christensen, via Internet
SGI 1200: ANYTHING AT ALL?
Hi. The subject says it... I have seen an old SGI 1200
computer in a flea market near here and want to know
something about it. Does it do graphics? How old is it? etc
etc... Thanks for info! EMail is better of course!
-- From bmcbrine@hakatac.almanac.bc.ca, via Internet
The Analytical Engine, Volume 1, Number 4 April 1994 Page 56
SPERRY/VARIAN V77 MINI WANTED
....Years ago I worked for a long time on Varian Data
Machines V77 series mini computers. Later they sold out to
Sperry. I remember that Sperry at some point sold the rights
to a company called Second Source Computers Inc in Tustin
CA.
The machines I'm specifically interested in are the V77-600
& V77-200, although any models will do. If you have any info
or know where I can get some I would like to hear from you.
Especially if you have one in use or want to get rid of a
working one! or if you have any info on Second Source.
-- from jonathan, via Internet
TECHNICAL DESIGN LABS: WHERE ARE THEY NOW?
Whatever became of Roger Amidon and the fantastic Technical
Design Labs of Princeton NJ??
They produced some of the finest ahead-of-the-pack S-100
products going. A quality operation. Their software was far
superior to the lame stuff that got us to where we are
today. Their hardware was feature-rich and had the ability
to grow beyond the state-of-the-art at the time - their CPU
board had the ability to run at 4Mhz at a time when Zilog
only produced processors rated at 2.5Mhz.
I still have (on PPT) their ZAPPLE monitor, Z80 Macro
Assembler, Text Editor and output processor, and both the 8k
and 12k BASICs. This was just prior to CP/M dominating the
market. They mention in a flyer that they were about to
release their FDOS, in December 1976. So where did they go
and what became of their Pro-from-Dover Roger Amidon?
-- from David K. Bryant, via Internet
UNIX BOOKS: FERVENTLY WANTED
In November last year (!), I ordered a copy of volumes I and
II of "UNIX System Software Readings" from Prentice-Hall
(two books which I believe are reprints of the UNIX special
issues of the AT&T Bell Systems Technical Journal (BSTJ)).
Last week (!!), volume II arrived. Great!
TODAY, I had a call from the bookstore, to say that volume I
would *never* arrive, because Prentice-Hall have stopped
printing it! Why do all of the *good* books, the *classics*,
go out of print.... [smoking snarl abridged]
The Analytical Engine, Volume 1, Number 4 April 1994 Page 57
If there is anyone in the U.S. (or anywhere else) that knows
of somewhere that has a copy of Volume I still in stock, I
would appreciate hearing about it so I can snap it up! (The
same goes for "The Multics System" [Organick], which is also
out of print; I don't know how long ago).
-- from Adrian Booth, via Internet
WANG CALCULATOR: TECH REF NEEDED
Perchance does anyone have tech data on the Wang LOCI-2, a
mid 60's desktop programmable calculator? I have the
manuals, but no tech info (e.g. schematics).
-- from Michael Dunn, via Internet
WICAT 156: BACKGROUND, DOCS, AND UNIX WANTED
I have been offered a Wicat machine, and I am looking for
any information anyone might have on it or the company. It
is billed as model 156, a 68000-based machine with a 12-
megabyte hard drive, from the early 1980's. It originally
ran a UNIX variant, but that did not work very well so the
owner had it replaced with an operating system he describes
as "VMS-like." In particular, if anyone has or knows where
to get the UNIX variant for this machine, I would like to
hear from them.
-- from James W. Birdsall, jwbirdsa@picarefy.com, via
Internet
ZENITH MINISPORT ZL-1: ELBOW-DEEP IN THE HARDWARE
A friend....bought an old Zenith Minisport laptop from a
swap meet and was wondering about it. The laptop is model
ZL-1, has a 2" floppy drive, came with a power AC adapter,
and has 1Mb of RAM.
1) What can it do? What are the stats on the machine? Is DOS
in ROM?
2) Is there a source for 2" floppy disks, that can be used
by the Zenith drive?
3) The unit comes with password protection. We can short two
pins on the EEPROM and reset the machine so it can bypass
the password protection. Anyone know which chips to jumper?
Zenith describes using a jumper between two pins. I already
took out the backup batteries and shorted the CMOS using 2
quarters (fortunately the backup batteries are the size of
quarters....) Any info would be helpful and is there a
The Analytical Engine, Volume 1, Number 4 April 1994 Page 58
source for technical manuals?
Please help me. Replies can be sent to:
lewj@nextnet.ccs.csus.edu. Thanks in advance.
-- from James Lew, via Internet
ZILOG Z80: C CROSS-COMPILER SOUGHT
For reasons too ridiculous to explain I am looking for a C
cross compiler (ANSI C would be nice, but is not necessary)
to executables for the Z80 processor. The compiler need not
be freeware or shareware, and can run on a UNIX, Mac, or DOS
platform, although DOS is the preferred environment. Any
leads or other information would be greatly appreciated.
Many thanks in advance,
-- from John Todd West, via Internet
-------------------------------------------------
PUBLICATIONS RECEIVED
-------------------------------------------------
_ International Calculator Collector_, Volume 2 Number 1,
Spring 1994. HP 35; Rockwell; Early advertising; Photo
Album; Pricing trends; Novelty Calculators; more. US$8 per
year with membership (goes to $10 June 30). From Guy Ball.
_Historically Brewed_, newsletter of the Historical Computer
Society.
Issue #4, Mar/Apr 1994. Apple II part 2; Kaypro Korner;
Calculating Computers; My First Computer; more. 16 pp.
Issue #5, May/Jun 1994. Apple II part 3; Kaypro Korner;
Atanasoff's Computer; Reviews of Ranade and Nash's _Best of
BYTE_, Levy's _Insanely Great_, Cringely's _Accidental
Empires_; Computer history bibliography; Computer museums;
more. 16 pp.
US$15.00 per year; Can$20.00; International, US$24.00. From
David Greelish.
_The Z-Letter_, newsletter of the CP/M and Z-System
community.
Number 28, November/December 1993. Bondwell 2 laptop;
Evolution of ZDB Z-System database; correspondence,
resources and technical discussion. 20 pp.
Number 29, January/February 1994. HP 125 and 120; HELLO
source listing; correspondence, resources and technical
discussion. 20 pp.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 59
US$18 for 12 issues (2 years); Canada/Mexico, US$22;
International, US$36. From David A. J. McGlone.
-------------------------------------------------
ADDRESSES OF CORRESPONDING ORGANIZATIONS
-------------------------------------------------
International Association of Calculator Collectors, 14561
Livingston Street, Tustin CA 92680-2618. Guy Ball, Bruce L.
Flamm, directors.
Historical Computer Society, 10928 Ted Williams Place, El
Paso TX 79934. CompuServe 100116,217. David A. Greelish,
director and editor.
Lambda Software Publishing, 149 West Hilliard Lane, Eugene
OR 97404. David A. J. McGlone, editor and publisher.
The Perham Foundation, 101 First Street #394, Los Altos CA
94022. Donald F. Koijane, president; Mike Adams, editor-in-
chief.
-------------------------------------------------
THANKS TO....
-------------------------------------------------
Aaron Alpar for copious contributions of space, time and
money.
Barrie Grennell for lots of wise advice on fundraising.
Bob Kushner and National Productions for Pomona booth space.
David and Tamara Greelish for generously sharing their booth
at the LA Computer Fair -- and having a good time doing it.
Erwin Tomash for a working lunch that really worked.
Hilary Crosby for flawless pathfinding through bureaucracy,
surgically exact advice on logistics, and that picnic!
Jean at Atkinson Dynamics for coordinating the rescue of the
PDS 1020.
Joan Piker for ideas on how to pack 'em in at Pomona.
Joe Schopplein for fabulous photography.
Jodi Redmon for meticulous transcription.
Kevin Hogan and Rich Karlgaard for the coverage in _Forbes
ASAP_; and Ray Healey for making sure we knew about it.
Lana Taber for Pomona booth banners.
Max Elbaum for taking part in the latest round of
interviews.
Melissa Leventon for much discussion of accession and
registration.
Michael Oliver for driving, and loading, and unloading that
truck.
Michael Tague et al. for slogging through Kentucky's worst
snow in a century, and sitting with the Witchcraft server to
keep it up, instead of staying home and keeping warm.
Mike Malone for sharp writing in the NY Times article, en_ we
needed it.
US Printing for photocopying done quickly, inexpensively and on time.
-------------------------------------------------
NEXT ISSUE
-------------------------------------------------
Interview: Salad Days at PARC, part one. Techstuff. Letters.
Queries. And more and more and more....
-------------------------------------------------
GUIDELINES FOR DISTRIBUTION
-------------------------------------------------
The ANALYTICAL ENGINE is intellectual shareware.
Distribution of _complete, verbatim_ copies through online
The Analytical Engine, Volume 1, Number 4 April 1994 Page 61
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 600 to 1500
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 CHAC. Article deadlines are the first
of each month prior to publication: June 1 for the July
issue, September 1 for the October issue, December 1 for the
January issue, and March 1 for the April issue.
Each author may publish a maximum of one signed article per
year. This restriction does not apply to letters, queries or
interviews. Thank you for cooperating to protect diversity
of voices and topics. We regret that we cannot republish
previously published material, except in the case of clearly
attributed quotations or citations.
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
DOS or Mac 3.5" diskettes. Alternatively, please provide an
ASCII file attached to Internet mail. Please avoid
submitting on paper unless absolutely necessary.
The Analytical Engine, Volume 1, Number 4 April 1994 Page 62
-------------------------------------------------
NINES-CARD
-------------------------------------------------
704x/709x TRIVIA CONTEST:
Two real stumpers from Joe Morris, MITRE
1) The IBM 7040 and 7044 had the same basic architecture as
did the 704, 709, 7090, and 7094. The 7040/44, however, had
two memory reference instructions which involved *37* bits,
S,P,1-35. (They are *not* zero-the-accumulator instructions;
that's not a memory reference.) For extra credit, what would
these two instructions do if executed on a 7090? HINT: the
original 7040 WATFOR used these instructions to provide a
critical feature in a *very* slick way.
2) The 709x had two fullword integer divide instructions
which performed exactly the same operations on the specified
data. Another two instructions performed exactly the same
floating-point division, and a third pair of opcodes
performed integer division on a user-specified divisor bit
count. The difference between the two instructions in each
pair is the same. What is this difference? (Incidentally, I
had completely forgotten the pairing until I noticed it just
now.) Hint: one of the instructions in each pair was utterly
useless for all but a *very* small number of programmers.
[First correct answer or answers we receive by our September
deadline will earn the respondent an extra issue on his or
her ENGINE subscription. -- Editors ]
The Analytical Engine, Volume 1, Number 4 April 1994 Page 63
-------------------------------------------------
ADD MONEY, MAIL....
-------------------------------------------------
____ Yes! Please enroll me in the Computer History
Association of California for the next year. I enclose
Individual membership: $25 on-line / $35 paper (circle)
Corporate/institutional membership: $75 on-line / $85 paper
(circle)
Low-income/student membership: $15 on-line / $25 paper
(circle)
and send four issues of the ANALYTICAL ENGINE to
____ Internet address: ______________________________
____ CompuServe ID#: ________________________________
____ Other gateway: _________________________________
My mailing address is:
Name: _______________________________________________
Company: ___________________________________________
Suite, Box, Mail stop: ______________________________
Street: _____________________________________________
City: _____________________ Zip/postcode: ____________
Country: _______________ Phone: _____________________
____ I will submit an article to the ANALYTICAL ENGINE.
(Please refer to the guidelines for submission, above.)
____ I will help produce the ANALYTICAL ENGINE or do other
work for the Association. Please contact me at:
_____________________________________________________