The Wayback Machine - https://web.archive.org/web/20140714222327/http://yarchive.net/comp/bsd.html
Index Home About Blog 
From: Dennis Ritchie <dmr@bell-labs.com>
Newsgroups: comp.arch,alt.folklore.computers
Subject: Re: Why is MS copying Sun???
Date: Thu, 26 Oct 2000 02:28:50 +0000

Mike Meredith at home wrote (I snip much):
> ...
> Source management could be one of the reasons why SystemIII/V and
> BSD diverged, but there's a number of other possible reasons too.
> For instance the management at AT&T (or whatever they were
> calling themselves at the time) may have been reluctant to use
> BSD code (although some code made it); they were also very
> interested in pushing the business uses of Unix.
>
No, source management had nothing to do with this divergence,
it was politics and technical goals.   AT&T's USG and then its computer
business were suspicious of incorporating things from college
students in a business product pushed mainly in the Bell System
prior to 1984, then fuller-out as a commercial thing after divestiture.
UCB's CSRG wanted to do their own things, some of which in the
end were more important, like TCP/IP, and earlier adoption of
virtual memory on the VAX.  They also really wanted to stay with
the Unix 32V license for their clients (universities, ARPA contractors,
then the workstation companies)--the cost of System III and V was going
up.  In the early 80's they would (at least officially) refuse even
to read System III manuals: they were on their own path.

> And of course source code management across a minimum of 2 groups
> on seperate coasts would have been ... interesting.

True, but that wasn't the issue.

> I wonder how much BSD code made it into Research Unix.

Research Unix 8th Edition started from (I think) BSD 4.1c, but
with enormous amounts scooped out and replaced by our own stuff.
This continued with 9th and 10th.  The ordinary user command-set
was, I guess, a bit more BSD-flavored than SysVish, but it
was pretty eclectic.

There's an enormously complicated story about what transpired between
then and now, but suffice it to say that lack of source code
control fails to explain the existence of the current *BSDs, Linux,
Solaris, IRIX, AIX, HP/UX, the SCO/Caldera offerings and on and on.
Unix in the 80s wasn't open software in the modern sense, but it
was an approximation, and its history shows some of the difficulties
attending thereto.

	Dennis
Index Home About Blog