I was working in a lab at Penn in 1954 (while a student) and there was an analog computer there that, a year later, I became the guy who operated it and kept it in repair (mostly replacing the vacuum tubes, but they were subminiature and had to be soldered into the packages that were used throughout the machine. Someone claimed it was the largest electronic analog computer ever built, although I have no particular reason to believe that.
Around 1955 or 56 Remington-Rand donated to Penn a Univac I. By that time, it was essentially obsolete. It occupied a large room, plus several side rooms. It had 1000 memory locations, each of which held one 72 bit word (actually, 12 6-bit bytes). At any given instant, 900 of these words were in a mercury delay line and 100 were in the electronic memory. I wasn’t involved, but several people were engaged in writing a program that would make my analog computer obsolete, at least if it ever got written and debugged. The eventual program was 80,000 words long, half of which were instructions to bring the next procedure off the tape and into the memory. The instructions were written for fixed addresses directly into the memory. Later, a very primitive assembler came along.
My next experience was when my department bought a Wang mini in 1975. It cost $20,000 and the annual maintenance contract cost $2000. The only thing I did on was take my 8 and 9 year old kids down on Saturdays to play games on it. BASIC was the OS and editing a line just meant writing out a line using the same line number. There was no actual editor. They lost interest. Then in 1980, a friend visited bringing with he Apple 2. We had agreed to write book together and were trying to figure out the logistics. Mail between Montreal and Cleveland took 2 weeks minimum in those days (even special delivery was one week at best). I spoke to our computer centre and they gave me an account on their mainframe with instructions how my friend could sign on to it, using Tymnet in the US to connect to Datapac in Canada and then to my account. I could get his stuff printed out in the computer centre and type out my stuff in my office, connecting to the CC from the Wang using something called a Gandalf box. Then a colleague going on leave lent me a teletype machine and I bought a 300 baud acoustic modem. You dialed a number on a phone and then plugged the handpiece into the modem. Suddenly, I could work from home. Good thing too because the department stopped paying the maintenance on the Wang. Then in 1982 I brought home an IBM-PC with its IBM-DOS operating system, unnumbered although retronumbered 1.0 when 1.05 came out. It had one disk drive that could store 160 KB on one-sided disk drives. Version 1.05 allowed 2-sided drives and also upped the storage on each side from 160 to 180 by changing from 8 sectors/track to 9. At first I used the acoustic modem but then bought a 1200 baud modem card. The book was finished this way and published in 1984. I wrote in Basic a primitive editor. Following an example in the BASIC manual that came with the computer (in those days you got manuals not help(less) screens), I wrote a modem control program that the CC distributed around campus.
In later 1984, we got email. Changed our collaboration and we started writing another book. In 1987, I got a computer with a hard drive (20 MB IIRC). And modems kept getting faster. Also in 1984 I got a really good editor, one that I use to this day. I have never used Word and never want to use it.
I went through the usual upgrades to DOS, played with Win up to 3.1, but never liked it. Finally Win-95 came and I thought that was usable. Then NT (which my who used to play computer games in 1976 was an NT programmer), which I liked. Then XT, which I liked better. And then later the OSs get worse. At some point, I got DSL from the phone company, but their incompetence drove me to the cable company.
