It's Saturday night and I'm reading a 1989 interview with Ken Thompson, one of the creators of Unix. I'll tweet the highlights in this thread. Source: princeton.edu/~hos/mike/tran…
Straight into it! Ken's talking about the MULTICS influences on Unix (e.g., it's where they stole the idea of a shell from). MULTICS users talked about pipes but they weren't implemented. "A discussion on whether we should go to six or eight bytes", LOL. 
C programmers will appreciate these nuggets about BCPL and B, which are C's forebears. BCPL was the high-level language they used on CTSS and GECOS (operating systems), but UNIX was first made for machines with 4K of RAM hence B, aka interpreted BCPL.
The genesis of all those buffer overflows ...
Bootstrapping language compilers always seemed like magic to me, even when I knew how it worked. "Von Neumann machines in the real sense. There’s a lot of power in executing data --generating data and executing data."
"I imagine people programmed in Fortran for the same reason they took three-legged races." *snort*
We almost didn't get Unix because Bell Labs wanted their computer scientists to get out of writing useful things. Like, programming isn't computer science. "Why did you persist?" "It's what I do." Staunch.
"The idea that time-sharing would never survive because, you’re spending all of your time on this big mainframe, you know, all floating point hardware and all of this stuff, you know, fielding these ratty little interims from people typing on flex-o-writers."
"We always wanted to expand it and turn it back into communal things. [...] You can get your own work done, but you can really work faster if there’s a community of ideas, a community of help." The Unix creators' philosophy on sharing.
The Unix filesystem was first simulated on the GE-635 before it was implemented "in a day or two". It sounds like 1960s human-powered voice transcription was less awesome than modern ML-powered tech too :)
Ken Thompson wrote a space travel game, and in porting it to the PDP-7 he built the reusable OS and library components that became Unix. en.wikipedia.org/wiki/Space_Tra… I'm a Bryan Adams fan, so I'm delighted this happened in the Summer of '69.
Wife and the kids went away, so Ken Thompson wrote the operating system, shell, editor, and assembler ... one per component. "During the month she was gone, which was in the summer of ‘69, it was totally rewritten in a form that looked like an operating system."
As originally conceived, Unix pipes were multidimensional craziness talked about with hand-waves until finally Ken figured out the simplest thing that could be done.
The CS researchers' struggle with the Bell Labs management just to get a PDP-11 -- request after request for funds: denied. The psychology department down the hall eventually bought the machine for the CS department!
"the early reason UNIX thrived is because it had no vendor competition at all, none. It was the only software around for the DEC machine [PDP-11]". Never underestimate the power of a slow competitor! (for whatever value of "competition" there was)
Early versions of the operating system spread around different Bell Labs groups like versions of Word documents around a 2000s office. If only there were some kind of version control! (SCCS, the first source code control system, would be developed three years later, at Bell Labs)
The strategy of a genius. (Indistinguishable from the strategy of a bastard, so you can't tell the genius from their strategy)
Huh, Ken fertilised and watered Unix at UCB. I did not know that. "Most of the names you know were in fact students of mine."
And that's where the interview ends. Wow, what a legend. I still love the idea of him vanishing to Berkeley when managing Unix got too much, and by the time he came back he wasn't needed for Unix to progress. That's a position that every creator wants to get to!
