The Good Old Days

XKCD EditorsA recent XKCD on editors reminded me that I’ve been accumulating a number of articles on computer history I should clear out, because I’m a computing dinosaur.

  • With respect to the xkcd, there’s nothing new under the sun. I remember the days at UCLA when there was a pitched battle between the supporters of the Rand window editor (“e”, formally “ned”), and the vi editor (for those clueless, vim is a later reimplementation of vi, and of course, vi was the visual version of ex, which competed with the ed editor on Unix). Then there were the TECO stalwarts that came from the DEC world (I used TECO on RSTS/E), the editors such as TSO and URSA on the IBM 360/91 (later 370/3033), and the battles between emacs and vi stalwarts.
  • At the same time we were dealing with URSA and TSO, we were printing on a IBM 1403 Lineprinter. This wasn’t a dot matrix or a laser printer, kids: this used a chain of type and printed super fast. You could even play music, if you did your boldface right. IEEE Spectrum has a fascinating article on how the 1403 was able to print so fast, including the fact that it didn’t press the type against the paper — it pressed the paper (from behind) against the type.
  • Back in those days, we didn’t program in C++ or Java or even Ada. It was FORTRAN and COBOL and Algol and… Guess what? Folks are still using those languages. I had a CSSF submittal this year that was programmed in FORTRAN, and you can make a slew of money in banking if you can program in COBOL. All the old-time COBOL programmers are retiring (sometimes feet-first); and these newfangled kids don’t want to learn it. [As a PS: Dan Berry at one time had a cartoon that showed a 1950s housewife labeled COBOL, a 1950s engineer labeled FORTRAN, and a baby labeled PL/I…. and the milkman walking down the driveway labeled ALGOL. The caption: “Funny dear, he doesn’t look like me.” Does anyone have a scan of that cartoon?]
  • Jumping up to the 1980s: The news these days are filled with items on the death of support of Windows Vista and the first version of Windows 10. But there’s another milestone: Windows 3.1. Twenty years have passed, and we’re still living with many of the notions 3.1 introduced (it was the first stable and popular Windows version, cementing the fact that you should never trust even numbered Windows variants, remembering that Windows 10 is really Windows 9, but they screwed things up with 95 and 98)
  • Turning to the hardware: Chips used to be simple: instructions sets, memory mapping and such. Intel is starting to change all that, with multiple processor instruction sets on a single chip. One of Intel’s changes is a mix-and-match heterogeneous design where different types of cores can be put in a single chip package. Under the new design, it’ll be possible to mix different architectures on a single chip. Chip packages could also have cores made using different manufacturing processes. Now ask yourself: with hardware this complex, how do we know it is correctly implemented?