Punch card blues
The day I broke the university’s brand new Univac 1108
By Warwick Doll, PhD ’70
I came to Case Institute of Technology in 1965 as a master degree candidate in the Macromolecular Science Department under Dr. Jerome Lando. Professor Lando’s field of expertise was X-ray crystallography of polymers. Work in this area required a great deal of trial-and-error computations, still in the age of manual calculations, slide rules and mechanical calculators. There were no digital calculators or personal computers.
From my undergraduate studies in Missouri, I had a basic knowledge of writing computer programming and I taught myself how to write programs for the Univac 1107. The input for this computer was all on punched cards, and my program grew to a stack of about 850 cards, plus the input data cards. Each card contained one step of the program. It took about five minutes to compile and run my program on the Univac 1107.
About the start of my PhD program, CWRU bought a new Univac 1108, which had a much faster processing speed. The school moved the computing center to a new building and sold time on this computer to commercial businesses. CWRU was allotted a certain percentage of the computer time, and the rest was to be sold.
Even with the new computer, I still had to input all of my cards manually. While the computing time was cut by a factor of 4, it still took about a minute to compile the program before it would run. As I recall, the computer could only run one program at a time, no matter how big or small the program.
In my final year, the university purchased a state-of-the-art storage device. It looked like a long, rotating log but the computing center was quite proud. Since I had a large program, I was allowed to have a special user number and could store my entire program on the new device. Then all I had to do was use 8 to 10 cards to call the program off the computer storage unit and input my data.
I was still doing a lot of trial-and-error computations, and so I loaded two different sets of input data to be run one after another. Suddenly, the computer shut down and reported that all of the computing time allotted to the university for the entire year had been consumed. The computing staff expressed alarm. They were trying to figure out what had happened.
Since the disaster occurred while I was trying to run my two programs, one after the other, I was thought to be the culprit.
I was called in to see the head of the computer department, who was furious. I had to explain exactly what I had done. It turns out that by running the long program stored on the memory device, it still took some time to extract the program and run it with the data. The second set of data called for this same program and apparently the program was not available to run the new set of data while the old one was still running on the computer. This caused the computer to crash and consume all of CWRU’s computing time.
They were able to recover the computation time for the university, but I had to promise to never do that again. Today, this entire trial-and-error program would easily run on a lap-top.
Warwick Doll is retired and living outside of Spartanburg, South Carolina. You can reach him via firstname.lastname@example.org.
If you would like to share your own Case memory, email us at email@example.com.