11 Nov
2025
11 Nov
'25
1:17 p.m.
On Tue, 2025-11-11 at 15:13 -0500, Paul Koning via cctalk wrote:
Via one of the articles about that Unix V4 tape I
found this
https://eylenburg.github.io/os_familytree.htm which is quite an
impressive chart of operating systems old and new.
A few bits are clearly incorrect or missing; for example, P/OS and
TRAX are both not there. I was going to mention them to the author
of that chart but I don't actually know the correct placement. (Is
P/OS an RSX-11/M or M+ derivative? I should know but don't.)
Neat to see RSX-15 mentioned as the ancestor of RSX-11/D. Did the
earlier RSX-11 versions ever ship? I know I have seen traces of /A
and/or /C...
The DOS-11 and RSTS entries aren't quite right either, I can send
feedback for those.
Also, while VMS is reasonably listed as RSX-11/M derivative, I always
thought that VAXElan was unrelated. Along the same lines, where
would MicroPower/Pascal fit?
paul
Erstwhile Bell Labs denizen Stu Feldman, the author of "make" and the
first (and slowest) FORTRAN 77 ciompiler, told me decades ago that when
he was at MIT he and several others decided Multics was too big and too
complicated, so UNICS was (originally) designed in reaction as a
simpler single-user system — hence UNI instead of MULTI. Dessis Richie
was a grad student at Harvard at the time, but he had a part time job
in Project MAC. Ken Thompson also worked on Project MAC even though he
had studied at Berkeley. I think it was Dennis who said about BSD
"never, ever, give your source code to a grad student." He had given
UNIX source code to Bill Joy, who used BSD to found Sun Microsystems.