Dave Bessel
Recollections, 1960 - 1975
A short review of my Cornell computing recollections. I started as an undergraduate with Dave Waks when you [John Rudan] were Assistant Director of the Comp Center in 1960. I worked part time as an undergraduate until I graduated in 1962. Then I started Graduate School and I was a graduate assistant. Then I was married in about 1962 and my wife, Sue, started Graduate School so I went to work full time at CCC. You (Rudan) hired me as the first systems programmer. I remember being interviewed by the VP, Frank Long, because the salary was high for a CCC employee. I remember that because it was the time I walked in there and I had a shoe that was cut out because I dropped a big cabinet on my foot. It was an embarrassing start.
We grew a staff of part-timers. Jon Casher, Dick Stone, Walt Bilofsky and John Behrenberg were part of the earlier wave and then I recall that Eric Mintz and Alan Goodman came later. I was Assistant Director for Systems and then I was Acting Director for the closing of the Computing Center, which merged into OCS, with the arrival of the IBM 360/65.
I worked for Dick Cogger in systems at Langmuir when we did two or three generations of 360/65 spooling systems. Then I ran User Services where I worked with Joan Winters and Jim Manning among others. Thus ended my relationship with the computing center as I went back to Rand Hall and Computer Graphics.
I was there for some of the formative years of Cornell Computing. It was an exciting time period for Cornell computing and for me.
From Cornell I went to NCR, GE, and retired from Sony in San Diego.
I want to try and solve the mystery of where the CDC 1604 ended up after Cornell sold it. You recall that the 1604 was a white elephant and that Dick Conway was burning your (Rudan's) tail to get it out the door. He wanted the machine out so he could start with a clean slate. He brought in a 360/40 as a training machine in anticipation of delivery of the 360/67. He found a customer, Sergio Beltran, I'm not sure of the first name, who was supposed to be affiliated with the University of Mexico or some other university in Mexico. It turned out that his office was supposed to be over a feed store, although I'm not sure where I got this. In any case, he closed the deal and got an export license to ship the 1604 out of the country. That left me with the job of getting it shipped to him. I insisted that he send somebody here to be trained. He sent a guy, whose name I can't remember, but he had no practical knowledge of computing whatsoever. We had him in Rand Hall with Manning and Biata teaching how to run the thing, although I'm not sure what he learned as he spoke no English at all and we spoke no Spanish. We brought in CDC and crated it up and it disappeared. To my knowledge we never heard another word about what happened. Now there are various theories about what happened. One of them is that it ended up in Cuba and that's the most repeated story. However, in 2000, I sent you (Rudan) a copy of a newspaper clipping about a 1604 computer that had been retired from working in East Germany and sent to a museum. If I ever get to Berlin I would like to pop open the covers of the machine to check the serial number to see if it is Serial 13, which was the one of our machine. (Rudan noted that the Cornell lawyer, Ralph Barnard, was most pleased that he insisted that the payment be made before the machine was loaded on the truck. That did happen)
Things that I remember most are the humorous episodes. There was the great Petznick episode with the 220. George had decided that you could summary punch cards by connecting the 086 card punch to the 407 which was used as the B220 printer. He wheeled in the punch and he hooked it up. I remember him doing that as I was hanging around the 220 all hours of the day. I went home and came back the next day and the whole room was full with relays and parts of the Cardatron data buffer that connected the I/O equipment to the Burroughs 220. The whole room was filled with parts, as they had to rebuild that big box! I also remember that if the 220 didn't quite work right, you ran the diagnostics and tapped on the vacuum tubes with a little mallet. If the diagnostic hiccupped you replaced the tube.
The next most strange event was just before the 1604 was at the end of its life and I was the director. Bill Biata was the third shift operator. I get this call in the middle of the night from Biata, "Help, I'm bleeding to death!" I called the campus cops and threw some clothes on and drove hell bent to the Computing Center. There was Biata sitting on the step in the computing room with a cop on one side and a cop on the other and his head was covered with blood. He looked worried. About that time he was pretty bald as I recall and we found a missing patch of skin about the size of a fingernail. I think he had fallen through the window only lost that much skin!
You (Rudan) asked about the episode where I ran paper tape between the 220 and 1604 computers? Right, that was the first Cornell network - machine to machine communication. You just have to have a long leader on the tape. It is true, at Cornell; we replaced money with ingenuity. That was our hallmark. But then we also dug ourselves into corners too!
Going back to ingenuity, one of the things that I thought was extremely significant, which has probably got lost by now, was Cortran! Cortran - in core (memory to you new comers) Fortran, was my project and design. I got sick with mononucleosis and I was really sick. To keep from going crazy I designed it in about two weeks in bed. I gave it to Dick Stone and Tatjana Grenewitz and they and some others coded it in assembler; they did a great job. Cortran was what kept our ass out of the fire that last year. It was an exact copy, almost, of the Control Data COOP production Fortran. It was so fast that we made money two ways. It would actually compile and run programs faster than the loader on the production Fortran would load programs, and we made money by the rounding of clock time. I think I have a copy of Dick Stone's master thesis or 5th year project, which was the implementation of Cortran..
Do you remember the guy who sent a tape recorder up in a balloon? This was in the 1604 days. He had an incremental IBM compatible tape drive that he put in a balloon and the instrumentation package was supposed to be parachuted down. Except, the parachute didn't open and the thing came back with the tape drive folded over itself. At any rate, he extracted the short pieces of tape and he had the CDC technician, Cleve Yoder, wire the error mechanism shut on the tape drive and he literally spliced the tape together to retrieve whatever information he could.
One of the major inflection points that I remember was the selection process for equipment prior to acquiring the 360/65. That was at the time that IBM and other people were just beginning build address translation boxes better known as paging and segmentation. The three main candidates were the IBM 360/67, the GE 645 and CDC 6600. The CDC 3600 had been around for a while as a successor to the 1604 and the 6600 replaced it. The Burroughs 5500 or 8500 were not in the running. Conway and I went to a meeting in Buffalo where this was discussed in great detail by universities. MIT was supporting the GE 645. Michigan was driving the IBM 360/67. Most every one else was watching as things developed. It was a critical time for the manufacturers. IBM had a great story and a good deal and won many important universities. GE sold a few 645 s to MIT and then called it quits. CDC filled government basements with 6600s. Shortly before the delivery date for the 360/67 IBM announced that the operating system, TSO, was a total failure. They converted most of the 67 orders to 360/65s. Thus, IBM managed to dominate the university large computer market.
Aside from the early things the next thing I remember is the Cornell version of HASP spooling system. In retrospect we did a really good job. It was done under Cogger's direction. That was another one of the same maneuvers - ingenuity substituting for money!
One of the first things I did when the 360 was installed was to build the first instrument for hardware measurement of the resources being used on the 360/65. It was used on the 360/65 and measured idle CPU time by wait state integration. It was a little socket, which plugged into the wait state lamp on the 360 console, and basically it integrated the amount of time the light was on over some time constant over a few seconds or so.
Do you remember the Herb Weiner episode with disk drives? Herb made up a
disk pack out of LP records. It was the wrong color and half the weight of a
real IBM disk pack. He ran a job that called for it to be mounted. I think
Biata was the operator and he hung it on a disk drive. I came in the next day
and it was almost a reliving of the Petznick story. The place was floating in
hydraulic oil and the IBM technicians had bits and pieces of the disk drive all
over the floor. I am sure Herb was punished. However I forgot what the
punishment was.
(Rudan recalled his most notable episode with Milt Hoenig who was a physics
grad student who was running the 220 during the midnight shift. The 407
printer ribbon broke and there was no replacement. Being ingenuous he loaded
5 or 7-part carbon paper which was stored near the printer and then discarded all
but the first carbon copy which he kept. However, he was so inept that he left all
the other copies piled up behind the 407 in this big mound which Dave Pulleyn
discovered the next morning and was quite upset. I can't recall the punishment
for this episode.)
Let me end this with a partial list of some of the notable system software created during this period. My apologies for omissions and mistakes.
CAP - 220 Assembly Program - David Waks 1959
CORC - Educational language - Conway, Maxwell et al
CUPL - Educational Compiler - Conway, Maxwell et al
CORTRAN - High speed Fortran - Bessel, Stone et al
160A - 1604 Spooling System - Casher, Bessel, et al
CLASP - 360 Spooling System - Cogger et al
HASP - Modified HASP Spooling System - Cogger et al
David H. Bessel 8/29/04
Prepared by John W. Rudan September 25, 2001