Here's the short account of the history of BSD UNIX I promised this list
I'd write.
Please inform me of any errors I have in this account, since for obvious
reasons I don't have personal experience with much of the UNIX versions
I will write about: when 1BSD came out, I was probably playing in a sandbox
in nursery school ;)

The morals of the following account are these:
1. The "open source" and "free software" sprits existed well before
GNU (1983) or Linux (1991).
2. Contrary to what Linus might think, "the pot-smoking hippies at Berkeley"
actually did contribute a lot to the design of modern UNIX.

Early history
-------------

The UNIX system was started in 1969 in AT&T's Bell Telephone Laboratories
(in New Jersey), by Ken Thompson and his colleagues, and originally written
for the PDP-7. Dennis Ritchie helped in 1970 to port the UNIX kernel to the
PDP-11. By 1972, UNIX started prospering in Bell Labs, and the second edition
of the UNIX manual included the following comment:

"The number of UNIX installations has grown to 10, with more expected."

By 1973, Ritchie had designed and wrote a compiler for the C language, and
then Thompson and Ritchie rewrote the UNIX kernel in C --- breaking away from
the tradition that system software should be written in assembly language.

Around 1974, Ritchie and Thompson report in an article (Communications of the
ACM, July 1974) of over 600 UNIX installations, and amusingly that "it can run
on hardware costing as little as $40,000". By that year, UNIX was licensed
to universities "for educational purposes" (for a very low charge) and a few
years later became available for commercial use.

Universities got for a tiny price (reportedly as little as $150) "source
licenses" to UNIX, meaning that their students and professors could actually
see UNIX's source code, study it (UNIX employed many of the OS-theory ideas
of the time), and even change it --- though the source code still belonged
to AT&T. A mere decade later, in 1984, Fortune runs an article saying that
out of 750 universities around the world, about 80% of those offering computer
science degrees, have UNIX licenses.

Thompson went in the fall of 1974 for one year of sabbatical at UC Berkeley,
where he left behind a tape with UNIX for the curious graduate students.
Led by Bill Joy, they started making changes to the UNIX kernel and utilities,
and distributing them (to other holders of the AT&T source license) as "BSD"
(Berkeley Software Distribution). BSD, the first version of which was released
in 1978, would eventually include some major improvements to AT&T's UNIX,
such as virtual memory, TCP/IP, job control, and even the VI editor (written
by Bill Joy). Many of these changes were later incorporated back into AT&T's
research and commercial UNIX versions.

Funded by a DARPA contract for incorporating TCP/IP into BSD, the CSRG
(Computer Systems Research Group) was formed in 1980 in Berkeley for working
on BSD UNIX, until the group was finally disbanded in 1995.

BSD releases included 1BSD (March 1978) and 2BSD (May 1979) based on AT&T
Sixth Edition Research UNIX, and later incorporated AT&T's Seventh Edition
and "32V" code into 3BSD (March 1980), 4.0BSD (October 1980),
4.1BSD (June 1981), 4.2BSD (September 1983), 4.3BSD (June 1986),
4.3BSD Tahoe and Net/1 (June 1988), 4.3BSD Reno and Net/2 (June 1990),
and finally 4.4BSD (June 1993).

Modern history
--------------

By 1992, most of AT&T's original code has already disappeared from BSD,
and a commercial company, BSDI (Berkeley Software Design, Inc.) was spun
off to sell "BSD/386", a commercial version of BSD. BSDI was immediately
sued by USL (UNIX Systems Laboratories, owned by AT&T, Novell, Sun, and a
few other companies, and holding the rights to the UNIX system).
After two years of legal battles, 4.4BSD-Lite is released, differing
from 4.4BSD in just 3 files (out of about 18,000). The last BSD from Berkeley,
4.4BSD-Lite Release 2, was released in 1995 and the CSRG was disbanded.

In the same year (1992), another 386 port spun off BSD: 386BSD, but this
one was meant to be freely distributable, rather than a commercial project.
William and Lynne Jolitz, writing a series of articles for the Dr. Dobb's
Journal, rewrote the pieces that were "tainted" by AT&T copyright and
finally released a working UNIX OS that "supposedly" didn't need an USL
source license. 386BSD version 0.1 (released on Bastille Day, 1992) became
the progenitor of the modern BSD operating systems (FreeBSD, NetBSD and
OpenBSD).

In the following year, Jordan Hubbard, Rod Grimes and Nate Williams worked
on 386BSD and released a set of patches known as the "Unofficial 386BSD
Patchkit". After Jolitz removed his approval of this project in early
1993, at the end of the year the patchkit got a new name: "FreeBSD 1.0" was
released. The BSDI/USL legal battle resulted in FreeBSD 2.0 being released
in November 1994, based on 4.4BSD Lite instead of Net/2.

NetBSD 0.8 was released about six months before FreeBSD 1.0. People in
Virgina Tech were frustrated by the pace of work on 386BSD, and took Net/2
and code from 386BSD and ported it to the Macintosh. The development effort
soon expanded to the Atari ST, Amiga, and PC. This project was named "NetBSD"
(as a tribute to the principal way of their collaboration: the Internet).
It soon became obvious that FreeBSD's niche would be the i386 and NetBSD
would provide BSD for other platforms. As with FreeBSD, before NetBSD 1.0
was released, the developers were forced to change it to be based on 4.4BSD
Lite, instead of the copyright-"encumbered" 4.3BSD.

In October 1995, OpenBSD was spun off NetBSD 1.1 by Theo de Raadt, who
had been responsible for the SPARC port of NetBSD as well as a few other
pieces of the system. OpenBSD quickly began focusing on security and strong
cryptography.


--
Nadav Har'El | Thursday, Jan 17 2002, 4 Shevat 5762
nyh@... |-----------------------------------------
Phone: +972-53-245868, ICQ 13349191 |An egotist is a person of low taste, more
http://nadav.harel.org.il |interested in himself than in me.