I started my first computing job in October of 1955. You may be surprised, but things have really changed since I was 20!
In those days, a computer was a rare and awesome thing. The first machine I worked with was called Illiac, at the University of Illinois in Urbana. The computer took up two floors of the building, used a lot of electricity, and had a major air-conditioning requirement (vacuum tubes make a lot of heat).
It was by today's standards a pitiful gadget. It cost over $1 million to make -- hand-built, it was one of a kind -- and was also expensive to operate. Today you can buy more computing power than it had for about $20, hold it easily in the palm of one hand, and run it simply by daylight.
To leave the history of computing machinery behind quickly, there followed mainframes, batch processing, punched cards, compilers, timesharing, desktop computing, graphical user interfaces, networking, the Internet, and more unfolding every day.
I became a programmer in a series of small steps. At first I did rather student-type jobs for the high-energy physics research group that hired me: preparing data, graphing results, and so forth. It was important work nevertheless, because the physics people in our group were literally on the leading edge of designing new kinds of particle accelerators. Every day, major discoveries were being made through the group's computing efforts.
The project's goal was to design and build a new kind of proton synchrotron, an immense instrument physicists use to delve into the basic nature of matter. Not only would it be large, it would also far exceed the funding resources of any one institution. So 15 universities (the Big Ten and five others) formed a consortium called the Midwestern Universities Research Association, or MURA, and jointly obtained grants from the Atomic Energy Commission (precursor of the Department of Energy in Washington, DC) to fund this basic research.
As the theoretical work was proceeding, planning was also underway to select a suitable site for the national laboratory that would grow up around the device. As you might imagine, this would mean a substantial influx of cash into someone's local economy, but the requirements for geological stability, adequate water supply, and affordable electric power made site selection a complex technical matter as well as a sizable practical and political one.
Eventually, a site near Madison was chosen as the location of an interim laboratory, the research effort having outgrown the facilities at Urbana. Among other things, the project would need its own computer. In 1956, the cat's meow of computers was the IBM 704, which we leased and set up temporarily in a refurbished two-story garage -- a former auto dealership -- near the UW campus.
Though I had written code for the Illiac while still in Urbana, I didn't really become a programmer in any serious sense until we got the 704. But I worked full time, and and it was not long until I had done a lot of programming. I also resumed my schooling in 1958, going to classes by day and working at night. My academic life was something of a hodge-podge, I guess, for I was working a lot, taking academic courses, and (from 1958-62) studying the piano, all at the same time.
At work, what we were doing, basically, was making mathematical models of particle accelerators, using large systems of simultaneous differential and partial differential equations to manipulate magnetic fields and particle orbits in the virtual machine, testing various design parameters. This was an absolutely novel approach at the time. In due course it was decided to build a working model to demonstrate the validity of the method. These are very complicated machines we're talking about, and the model brought us all great joy by working perfectly the first time it was turned on!
Everyone was now ready for the next phase, which was to move again, to buildings erected for the purpose on the rural site, 17 miles south of Madison, where we hoped the full-sized accelerator would be built. And it would have been, except for a national disaster that took place on November 22, 1963. President Kennedy had reportedly personally chosen the Madison site as the final location for the machine and was to announce his decision on the Monday after that infamous Friday. But following the assassination, President Johnson found he had to have the support of Illinois Senator Everett Dirksen (a Republican) to get his legislative program enacted. The new national laboratory would be worth an estimated $700 million in jobs and new infrastructure, and Dirksen wanted that plum for Illinois. He got it, and that's how Fermilab, as it's now known, came to be in northern Illinois and not in southern Wisconsin.
As the core research group made ready for the move to the Illinois site, things began to wind down in Madison. In 1965, the University bought the facility and took on the staff who remained behind, including me. I had received a generous offer from the Illinois group, but I did not want to move because I was still in school and intended to finish here. Furthermore, I was living with a lover in a stable home environment, situations that were important to me. The University planned to use the facility, which had a good-sized machine shop, computers, and a large, very specialized knowledge base in its staff, for special research projects.
The new entity was called the Physical Sciences Laboratory, PSL. It still exists. I worked there until December, 1969, when yet another budget crunch (I've now been through many) cut the computing program to nothing and moved the computing staff to campus. It would also prove to be the end of my scientific programming career, though I didn't realize that at the time.