Vignettes From the Early Days

I hope this set of pages containing little reminisces from the early days of computing will be amusing.  The Good Old Days weren't very good but they were fun and exciting. The equipment left something to be desired.  As a programmer, our programs often left something to be desired. Being an egotist (probably like most programmers), I will start the vignette section with some of my own.  Please send vignettes from the early day to roger_mills@acm.org.

Remembrances - Roger Mills - Started in the computing field in January 1951

In 1953, North American Aviation computing was divided between two divisions, the financial which operated the equipment and wrote utility subroutines and engineering which wrote application programs for the IBM 701.  Frank Wagner's Engineering Computing Group was the consultants for the engineers who programmed their own applications. In the Financial Division, Jack Strong's group was responsible for scheduling and operating the IBM 701 and all of the supporting punch card equipment. I was a member of the Engineering Computing Group.

Jack had a rule that only his people could maintain the card decks and no copies should be outside his area.  I was working with a programmer who probable had the largest applications program to be run on the 701.  It was a program to calculate loads on an airplane.  The program consisted of 21 binary programs, each calling the other to continue calculations. The decks of row binary cards almost filled a box which held 2000 IBM cards.  I could see a disaster in the making if that box was ever dropped, so I bootlegged a copy of the decks and kept them in my desk. I received a call from a computer operator asking if I could come to the computer room.  Since the box of cards had been marked along the top with magic marker, I recognized the loads program that was mostly on the floor. The operator asked if I had a copy. I told them to get Jack Strong out here.  When Jack came, I told him I has a copy of the deck.  He said that was great and they could make a copy of it.  I told him that I would give him the deck to make a copy if he rescinded the rule about programmers keeping a copy of their programs, otherwise, I would give him 21 memory printouts and he could reassemble the deck. Jack saw the light and programmers were allowed to legally keep copies of their decks.

Another bright idea Jack installed was the rule that no data cards could be run unless there was a verification punch in the card. For you non-punched card types, cards were initially punched by a keypunch operator and then given to a keypunch verifier to punch from the same input form.  If all of the punches matched, a verifying notch would be punched in the side of the card. I decided that what we needed was blank cards with a verifying punch on the sides.  Thus we could run programs with less turn around time than if we had to wait for keypunching. Jack confronted me and asked about the verified blank cards.  I offered him one from my pocket.  He said, "You can't do that!" I offered him another and said that I would make him a deal.  If they stopped checking data cards for the verifying punch, I would destroy all of the blanks I had and not generate any more.  That way, the verification punch would mean exactly what it was supposed to.  Jack saw the light and the programmers could punch their own cards.

My first Vignette from an old timer is from Jim Lynch about JOHNIAC.

***************************************

Jim Lynch, King George, Virginia

I was at McClellan AFB in Sacramento, California from 1965 to 1969.


During that time, a remote terminal was installed in our engineering offices. The terminal was connected to the JOHNNIAC computer in the bay area. I thought it (the computer) was at Berkeley, but after reading your note about the JOHNNIAC, I might be wrong about that. I assume the connection was over telephone lines.


The terminal was installed in a little cubical in our large office space in an old warehouse. I was an aerospace engineer there at the time and was shown how to use the terminal. The 'instruction' wasn't much more than how to turn it on. I found JOSS to be extremely easy to use. This was my first association with a computer. The officer in charge of the project didn't think that the ability to touch-type was important, but after observing us for awhile, he quickly realized that those who could touch type got far more out of the system.


You had to re-type in your "program" at every session. No one anticipated that programs would be more than a few lines long so this wasn't thought to be a problem. When I started to develop a center of gravity program for fuel tanks, it quickly grew to be quite a few lines and something of a chore to retype every time I had computer time. I remember telling the major that wouldn't it be great if the system could store your program and it wouldn't disappear when you logged off.


I theorized that it (the program) was obviously being saved somewhere while we were logged in and it shouldn't be that hard (I naively thought) to have it not go away when we logged out. Shortly thereafter the system was modified to include a storage capability for programs. You could "save" your program and then call it back up at your next session. I naturally viewed this as quite a boon!