Re: Owsley and BSD

[Richard Welty <rwelty@domain.elided>]

this way off the officially accepted topics for ihc, but i will provide the
necessary historical background, nonetheless. as list admin, i'm allowed to
do that...

back in the 1960s, there was a big honking governement project to build a
super multitasking multiuser time sharing system; it was named multics.
originally, GE and AT&T were involved. GE sold their computer division to
Honeywell and AT&T dropped out; honeywell plugged along for a while,
building  a facinating, secure and very expensive system. AT&T followed a
different path.

there were two ex AT&T multics types in bell labs -- Thompson and Ritchie.
they had a derelict DEC PDP-7 in the corner of their lab -- an old, 18bit
mini (it wasn't that old at the time, but it was not in use, certainly.)
they wanted to build a decent, cheap system. as multics refugees, they had
some relatively sophisticated ideas for the time, but they had no budget
and no resources. they improvised.

they developed a computer language, named B, based on the existing langauge
BCPL, and wrote a primative time sharing operating system. they developed
the langauge further, and called it C. they outgrew the PDP-7, and
convinced their management to buy them a newfangled PDP-11. since bell
wasn't allowd in the computer business, based on an agreement with the
justice department, they basically licensed the source code for this system
(named Unix, a weak pun on Multics) to universities and educational
institutions for nothing. unix started to become a force in the early 70s.
in the late 70s, the first bits and pieces of what we now know as Usenet
came into being in the Raleigh/Durham/Chapel Hill neighborhoods of NC --
based on Unix systems.

at the same time, DEC was introducing their first true 32 bit computer --
the vax 11 (which i remember from my co-op term at dec large systems in
marlboro massachusetts in 1978 -- as we used to heap abuse on those silly
vaxen, which would never pose a threat to our wonderful decsystem 10s and
20s.)

in this same period, the old ARPANet, precursor to the modern internet, was
going through some profound changes, as the Department of Defense was
looking to get out of the internet business. the University of California,
Berkeley, was comissioned to take the 32V distribution of AT&T Unix (the
first Vax version) and make it into a "standard" version of Unix for
internet sites running vaxen and large pdp-11 models. in 1982, i personally
installed my first copy of BSD (Berkeley Standard Distribution) Unix 4.1 on
a vax 11/780 in the computer science lab at RPI here in Upstate NY.

the folks at Berkeley had a somewhat broader vision. they looked at the
Motorola 68000 cpu (just coming out) and at some other hardware like bitmap
screens and cheap mouses, and concluded that workstations were a realistic
prospect -- bigger than PCs, but smaller than minis. the early "sun board"
workstations were failures, licenses to companies that didn't "get it", so
the berkeley and stanford academics founded Sun Microsystems in order to
get it right.

in parallel with the Berkeley activity, Bell continued developing unix,
creating a parallel and distinctly different flavor. Sys V unix from
AT&T/Bell is the dominant strain today, because everyone else (sun
included) has moved towards this version. FreeBSD is an inexpensive (nearly
free) incaration which is Sys V compliant but derived from the BSD tree.
Linux is a different creature; it is its own version, but is pretty much
Sys V compliant from the viewpoint that well written Unix programs will run
on Linux without major work.

am i a big unix fan? well, yes and no. i'm really a lisp programmer at
heart, and i miss my symbolics 3675. unix is a poor substitute for a good
computer, but it's the best i've got.

cheers,
  richard