LisaGreen wrote:Ahhhhh.... the old days!! I still have a soft spot for er... let's call it unisys... THough there was a joke going around at the time... prob buried in the past now that it stood for univac is still your supplier!! hehehe.
I don't know how things are at the moment but I just loved the idea that they used a simple text editor coupled with a stream generator to create ALL of their code for EVERYTHING!!
Am I correct in remembering that Burroughs didn't use an assembler?
Lisa
When the name was announced after the merger, we said it sounded like "Aunt Eunice's Computer Company".
Our coding was done in what was essentially a flat-file database, where every line was stored as a record with metadata about its sequence, date of creation, and patch number. If necessary, any section of code could be backed out through some or all previous patches, all the way back to ground zero. We could also pinpoint which patch introduced a particular bug. Every time I described this to non-Burroughs programmers, I could see them turning green with envy.
Some machines in the B5000/B5500 large systems series were optimized to run Algol programs (
https://en.wikipedia.org/wiki/Burroughs_large_systems). If I'm correct, the instruction set was a one-for-one match to the Algol constructs, so the compiler was essentially an assembler. Fortunately, I never had to worry about the machine-code details.