The Berkeley Software Distribution
==================================
UNIX is always litigious
by Bradford Morgan White
Feb 5, 2024
The first public presentation of UNIX was made at the Symposium on
Operating Systems Principles at the IBM Research Center in Yorktown
Heights in October of 1973. Dennis Ritchie is quoted as saying it was
beautiful day, and Ken Thompson layered his own memories with a thick
coating of modesty:
The audience was several hundred. I was pretty nervous. The
response was the normal, polite applause. I don't recall any
questions.
[The IBM Research Center in Yorktown Heights, image from IBM]
<
https://substackcdn.com/image/fetch/w_1456,c_limit,f_webp, q_auto:good,fl_progressive:steep/ https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2F images%2Fb9ace03e-5fa3-43df-875b-6989e389b156_1200x675.jpeg>
In contrast to Thompson stating that he didn't recall any questions...
the two were immediately asked for copies of the operating system,
and this wasn't a simple matter for AT&T. The American Telephone and
Telegraph company had been established as a legal, nation-wide
monopoly in the USA via the Kingsbury Commitment in late 1913. This
position was further cemented during World War I when the United
States' federal government nationalized the phone system. Following
the end of the war, the phone system went back into the hands of AT&T
and the company achieved some rather remarkable regulatory capture
with the Willis Graham Act of 1921 and the Communications Act of
1934. This complicated legal history presented a very serious
question to AT&T's legal department when people began asking for
UNIX: were computer operating systems part of the common carrier
services of the phone company and therefore required to be
distributed? If they were not, then the company needn't distribute
UNIX at all, but if they were indeed, then it was only a matter of
time before the FCC would force AT&T to distribute UNIX. In the end,
the decision was made to distribute UNIX to universities and research
centers at the cost of the media plus shipping. Somehow, quite
magically, this resulted in a nice round number of $150.00 (or around
$927 in 2024) for Katholieke Universiteit in Nijmegen, The
Netherlands in December of 1974. There are some rather important
points within the license that Katholieke was granted. Licensees were
granted source code for the operating system as computer systems of
the day weren't standardized in any meaningful way. The license then
granted free use and modification within the university, but
disallowed any spread outside. Specifically, the license mentioned
that employees and students had access.
[UNIX license from Western Electric to Katholieke]
<
https://www.abortretry.fail/api/v1/file/ ea99ff77-8f06-4c00-ae95-2a1f60410a44.pdf>
At this point in the computer industry, user groups were somewhat
common. IBM had SHARE, and that had inspired similar groups around
DEC, Burroughs, Rand, and so on. It was therefore somewhat natural
that a group would form around UNIX. Thus, Mel Ferentz and Lou Katz
organized a meeting of UNIX users in New York on the 15th of May in
1974. Around twenty people were in attendance, and by this time there
were just over thirty UNIX installations outside of AT&T and its
subsidiaries. This user group grew to become USENIX over time.
Following the user group's formation, a mailing list started. From
the first list on the 30th of July in 1975, we have the following
organizations listed as installation/user sites: AT&T, Brooklyn
College, Carleton College, Case Western Reserve University, The
Children's Museum, City University of New York, Columbia University,
Duke Medical Center, East Brunswick High School, Harvard University,
Hebrew University of Jerusalem, Heriot-Watt University, Johns Hopkins University, Knox College, Naval Postgraduate School, Oregon Museum of
Science, Polytechnic University of NY, Princeton University, Rand
Corporation, St. Olaf College, Stanford University, The Spence
School, University Catholique de Louvain, University of Alberta,
University of California (Berkeley), University of Manitoba,
University of North Carolina, University of Saskatchewan, University
of Texas (Dallas), University of Toronto, University of Utah,
University of Waterloo, and the University of Wisconsin. As we know
from the license granted to Katholieke, there were more UNIX user
sites than this, but they weren't members of the user group (or at
least not at first).
<
https://www.abortretry.fail/p/the-history-of-gm-naa-io-and-share>
[Dennis Ritchie (standing), Ken Thompson at the teletype, PDP-11
1972]
<
https://substackcdn.com/image/fetch/w_1456,c_limit,f_webp, q_auto:good,fl_progressive:steep/ https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2F images%2F3926f41f-c121-450b-8e10-e46ccc1c5448_652x522.jpeg>
In this early time period, UNIX only ran on the PDP-11, but that
changed at Princeton where UNIX was ported to the IBM 360 in 1976.
The next target was the Interdata 8/32 in 1977 which was undertaken
by Ritchie and Steve Johnson (author of yacc, lint, pcc). But,
porting efforts took off only after John Lions at the University of
New South Whales wrote a commentary on the UNIX sources and
distributed them as a book, Code and Commentary, for teaching
students about operating systems. Western Electric tried to stop
dissemination, but this was apparently impossible. Likewise,
modifications of UNIX began circulating following the release of
Lions' book, and a culture we would recognize today as “open source”
began to develop helped in no small part by AT&T's policies regarding
UNIX that essentially stated UNIX would have no advertising, no
support, no bug fixes, and payment in advance. To color the time
period more thoroughly, Bob Kahn and Vint Cerf published the first
description of TCP/IP in 1974 and by January of 1976 there were sixty
three hosts on ARPAnet, and UNIX while being used globally would run
only on hardware that cost over $9000 (around $48000 in 2024).
Professor Robert (Bob) Fabry was at the Symposium on Operating
Systems Principles where UNIX had first been announced and he was
very excited. Returning to UC Berkeley where he was then employed, he
assembled a group to purchase a PDP-11/45. As this was a large
purchase, he coordinated the departments of computer science, math,
and statistics. With the machine purchased, Fabry then ordered a tape
of UNIX from Thompson. The actual installation of UNIX was first
undertaken by Keith Standiford in January of 1974. In 1973/1974, it
was somewhat routine for Thompson himself to be involved in nearly
every UNIX installation for a licensee. The folks at Berkeley seemed
to be interested in doing everything themselves, but things didn't go
well. Eventually, Standiford reached out to Thompson, and Thompson
would connect to the University's 11/45 over a three hundred baud
acoustic coupler to remotely debug crash dumps from New Jersey. I
personally like to imagine that it was a Novation CAT 300, but I
haven't been able to find a model number of the modem, and I haven't
found any reference to what system Thompson was actually using. Plus,
the Novation CAT 300 wouldn't be released for several years.
Following the purchase of the PDP-11/45, the departments involved
began having issues with scheduling time on the machine. Berkeley
bought several more computers. One of these was a PDP-11/70, and its
arrival coincided with the arrival of Thompson as a visiting
professor. Thompson, Bob Kridle, and Jeff Schriebman then setup V7
UNIX on the 11/70. Shortly after the installation was completed, two
graduate students, Chuck Haley and William Nelson (Bill) Joy, arrived
on campus. They were intrigued by the computer system, and they began
hacking on Thompson's Pascal compiler. Within a few weeks (from what
I have been able to find), the teletypes attached to the 11/70 were
replaced with ADM-3 screen terminals.
[ADM-3, image by Chris Jacobs, CC BY-SA 3.0]
<
https://substackcdn.com/image/fetch/f_auto,q_auto:good,
fl_progressive:steep/ https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2F images%2F98f22ee6-2f34-4ffd-b8bd-b37fb059cbe5_1887x1860.jpeg>
For Bill Joy, using ed or em on a screen terminal wasn't really
sufficient. He took a detour from hacking on Pascal, and he created
the ex editor. Together with Pascal, the V7 UNIX at Berkeley was
notably better than other UNIX systems of the time. In early 1978,
Bill Joy began offering the Berkely Software Distribution. The first
copy we know to make it out of Berkeley was to Tom Ferrin at UCSF on
the 9th of March in 1978. The license was signed on the 13th, the
media was an 800 bpi tape, and on the tape was the “Unix Pascal
system” and the “Ex text editor.” Credits were made to W.N. Joy, S.L. Graham, C.B. Haley, K. Thompson for Pascal, and to W.N. Joy for Ex.
[Bill Joy, Silicon Valley Visionary, on the Future of Batteries ...]
<
https://substackcdn.com/image/fetch/f_auto,q_auto:good,
fl_progressive:steep/ https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2F images%2Fd5b8a872-0970-4832-8b53-86c19f7c8543_1200x630.jpeg>
BSD (now referred to as 1BSD) shipped around thirty copies in the
first half of 1978. Somewhere around June, Pascal had been further
improved, the C shell had been written, vi had been written (by Joy),
and termcap had been written (by Joy). These new tools comprised the
bulk of the Second Berkely Software Distribution or 2BSD. Joy was the
man running the show, and he'd answer the phone, create the tapes,
incorporate feedback, package, and mail the software. Seventy five
copies of 2BSD were sent out.
The DEC VAX was first introduced in 1977. It was a 32 bit ISA with
virtual memory. In the first half of 1978, Professor Richard Fateman
at Berkeley was looking for a machine with a lot of memory for a
project of his. The VAX-11/780 met his requirements and his budget,
and he got some folks together to purchase it with a bit of help from
the National Science Foundation. The VAX did have one disadvantage in
the minds of those at UC Berkeley; the VAX ran VMS. Luckily, UNIX had
already been ported to VAX as UNIX/32V (V7 UNIX variant) by John
Reiser and Tom London at Bell Labs. This port, however, didn't take
advantage of the main VAX feature, virtual memory, which limited the
available memory to 1MB.
This had to be fixed. For Fateman, virtual memory was a requirement,
and UNIX was a requirement. He then contacted Professor Domenico
Ferrari about getting virtual memory support in UNIX. One of the
graduate students working with Ferrari, Özalp Babaoğlu, then set
about this task. Along the way, Babaoğlu reached out to Joy for
assistance. Joy helped integrate Babaoğlu's memory system into 32V
and helped him debug the resulting UNIX variant.
The new version of UNIX was good. Joy knew that the 32 bit VAX
running UNIX would render the 16 bit PDP-11 obsolete, and he started
porting 2BSD to 32V. Peter Kessler and Marshall Kirk McKusick worked
on porting Pascal, while Joy handled ex, vi, C shell, and many other
BSD utilities. The group had completed their work by the end of 1979,
and Joy shipped 3BSD in December.
At this point, it's important to delineate the family tree. 1BSD and
2BSD were improvements to UNIX version 6 while 3BSD was an
improvement to 32V which was itself of a port and modification of
UNIX version 7. During the creation of 3BSD, 2BSD had continued to
see additions, fixes, and releases. One important distinction to be
made is that the filename for the kernel in 3BSD became vmunix for
virtual memory UNIX.
Computer hardware and software were highly varied in 1979 and most
often nothing was compatible with anything else. For the Defense
Advanced Research Projects Agency, this was an issue. They felt that
the best they could hope for was unity at the operating system level
for the growing network of networks, ARPAnet, and to this end they
chose to standardize on UNIX due to its proven portability resulting
from having been written in C. This same system would also be used by
DARPA for work in the VLSI Project. In the autumn of 1979, Fabry
reached out to DARPA offering 3BSD as the solution to their problem.
Initially, this wasn't well received, but the success of 3BSD that
December changed opinions. Fabry secured an eighteen month contract
with DARPA that began in April of 1980. This contract stipulated that Berkeley's new Computer Systems Research Group would add the features
to 3BSD needed by DARPA which was left rather open. In a technical
report on the matter, the government doesn't actually list anything
too specific.
[ARPA Standard UNIX Report]
<
https://www.abortretry.fail/api/v1/file/ 2c12ca84-b921-4d68-97a1-c046f60372c4.pdf>
Fabry hired Laura Tong to be the project administrator, and Fabry
then set about finding a tech lead. On an evening in early March, Joy
rang Fabry at his home and expressed his desire to lead UNIX
development, and Fabry agreed. Unlike earlier BSD releases (and
on-going releases of 2BSD) a more robust system of distribution
needed to be in place to handle a higher number of orders. Tong setup
just such a system, and had Fabry coordinate with Bob Guffy at AT&T
as well as the university's lawyers to find license terms that would
satisfy all parties. The system that Joy and the software team would
create was 4BSD which was available in October of 1980. This release
brought job control to the C shell initially developed by Jim Kulp
and integrated by Joy, delivermail (predecessor of sendmail), job
control signals that worked more reliably (so if you sent SIGHUP the
job would actually hangup), the curses library, the control-z
suspend/resume functionality we know today, a filesystem that
supported block sizes up to 1K, and greater hardware support.
Notably, 4BSD supported the VAX-11/750. Like prior BSD releases, 4BSD
included the Pascal compiler and that saw still more improvements,
and it included the Franz Lisp system by Richard Fateman. 4BSD saw
nine months as the system de jour, and about one hundred fifty copies
were shipped. Licenses were per institution and not per machine, and
estimates are that the distribution was running on about five hundred computers.
4BSD drew criticism from David Kashtan at Stanford Research
Institute. He'd run some benchmarks on VMS and BSD and he claimed
that VMS was the clear winner. This didn't sit well with Joy. He went
and made a series of performance improvements to BSD (mostly in
vmunix), and a few weeks later released a paper rebutting Kashtan's
claims and showing that BSD was every bit the match to VMS. The
improved kernel was coupled with Kevin Robert Elz's
auto-configuration and a system featuring them was released as 4.1BSD
in June of 1981, and 4.1BSD for VAX became 2.8BSD for the PDP-11..
This version was current for two years and shipped four hundred
copies. This version was good enough to win CSRG another two years of
funding from DARPA, and the CS department at Berkeley also saw
funding.
Interestingly, the reason for a point release was political. The CSRG
had wanted to call this release 5BSD, but AT&T blocked that name.
AT&T were releasing System V UNIX at the time, and they felt that a
Berkeley release also named “five” would confuse customers. Point
releases of 4.?BSD were then the agreed upon solution.
The new funding round came with some more concrete goals.
Specifically, DARPA wanted a better filesystem with higher
throughput, support for multi-gigabyte adress spaces, better IPC, and integrated networking. Marshall Kirk McKusick outlines how these
decisions were made:
To assist in defining the new system, Duane Adams, Berkeley's
contract monitor at DARPA, formed a group known as the "steering
committee" to help guide the design work and ensure that the
research community's needs were addressed. This committee met twice
a year between April 1981 and June 1983. It included Bob Fabry,
Bill Joy, and Sam Leffler of the University of California at
Berkeley; Alan Nemeth and Rob Gurwitz of Bolt, Beranek, and Newman;
Dennis Ritchie of Bell Laboratories; Keith Lantz of Stanford
University; Rick Rashid of Carnegie-Mellon University; Bert
Halstead of the Massachusetts Institute of Technology; Dan Lynch of
The Information Sciences Institute; Duane Adams and Bob Baker of
DARPA; and Jerry Popek of the University of California at Los
Angeles. Beginning in 1984, these meetings were supplanted by
workshops that were expanded to include many more people.
What was developed largely by McKusick, Joy, Sam Leffler, and Rob
Gurwitz is astounding. These developments were iterated in releases
4.1a, 4.1b, and 4.1c but ultimately culminated in 4.2 which was
released in August of 1983. The two largest innovations were the
Berkeley Sockets API and the Berkeley Fast File System. The sockets
interface allowed multiple different network protocols to be used at
any time and exposed them as files. The Berkeley FFS allowed
blocksizes from 128 bytes to larger than 4096 bytes if needed. Thus,
with Berkeley's FFS, a system operator could optimize either for disk
use or for disk performance. 4.2BSD included a full TCP/IP stack and
NFS support as well. Starting in the early iterated releases, small
tools like rcp, rsh, rlogin, and rwho appeared to demonstrate
capabilities. These were intended to be short lived, but those
familiar with UNIX will certainly recognize them. 4.2BSD also brought
disk quota facilities, a better install process, better
documentation, and some new filesystem related system calls.
Events less visible to users also occurred between 1982 and 1983. Joy
left the project in late spring of 1982 for Sun Microsystems, but
still spent a little time that summer working on IPC and reorganizing
the UNIX kernel sources to isolate machine dependent code. Once at
Sun, contributions to BSD continued with Sun sending their
modifications for running BSD on the Motorola 68000 back to Berkeley.
Leffler then took over Joy's previous duties. Joy's work, however,
would make BSD the single most easily ported operating system for
about the next twenty years. Pauline Schwartz was hired to take over distribution duties in April of 1983. In June of 1983, Fabry went on
sabbatical and Ferrari and Susan Graham took over running the CSRG.
In 1984, Leffler went to work for Lucasfilm and Mike Karels took over
as the UNIX dev lead. He'd previously worked on the 2.?BSD series for
the PDP-11. Later that year McKusick joined CSRG full-time.
4.2BSD gained over a thousand site licenses in a year and a half, and
most UNIX system vendors actually shipped 4.2BSD (or a system based
on it) rather than AT&T's System V UNIX. Likewise, many varieties of
corporate UNIX would be based upon 4.1c or 4.2BSD such as SunOS and
DEC Ultrix. An important point at this time is that BSD's codebase
was largely divergent from that of AT&T while still mostly
maintaining compatibility. Additionally, AT&T had started shipping
UNIX without source, and those customers who bought AT&T's commercial
UNIX would often then separately obtain BSD sources if they wanted to
modify some aspect of the system. For those using microcomputers,
Microsoft's (and SCO's) XENIX was the standard.
Over on the PDP-11, a mixture of 4.1a and 4.1c formed 2.9BSD, but
4.2BSD never showed up.
In 1976, comic artist Phil Foglio drew the first versions of Beastie
as T-shirt art for Mike O'Brien as payment for unlocking a safe.
[T-shirt art of BSD by Phil Foglio]
<
https://substackcdn.com/image/fetch/w_1456,c_limit,f_webp, q_auto:good,fl_progressive:steep/ https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2F images%2F1b00a913-e45d-44e7-b206-763ebac84cd4_688x833.jpeg>
Much of this is largely a pun on services being called daemons, and
UNIX making heavy use of pipes, with /dev/null being a bit bucket.
The daemon, Beastie, gained lasting association with BSD via a
drawing by John Lasseterf Lucasfilm being used on the cover of the
Unix System Manger's Manual that was published in 1984 by USENIX for
4.2BSD.
Cover of the manual, image from jacobelder.com <
https://substackcdn.com/image/fetch/w_1456,c_limit,f_webp, q_auto:good,fl_progressive:steep/ https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2F images%2F0cfe4090-7d76-4d07-acfd-0eadab7936cd_640x905.png>
While 4.2BSD was successful and mostly well received, it did get some complaints. The majority of the complaints were centered on
performance. The team then spent two years improving performance,
refining the networking stack, and they felt that they were ready to
announce an impending release at USENIX in June of 1985. This didn't
go well. The fine folks of Bolt, Beranek, and Newman (part of the
steering commitee) noted that 4.2BSD had shipped without the final
version of the their networking code, and was instead using a heavily
modified version of their initial prototype. After some bickering
back and forth, DARPA provided both network stacks to Mike Muuss
(author of ping) of the Ballistics Research Laboratory for
testing.erkeley's code was better. 4.3BSD was released in June of
1986.
Keith Bostic joined CSRG in 1986 with the condition that he be
allowed to continue and to complete a prior project. He was working
to port 4.3BSD to the PDP-11. This meant that a 250K minimal system
on VAX would be made to fit in the 64K address space of the PDP-11...
no one thought this would work. No one. The man was clearly mad, but
despite what madness may have lurked within him, Bostic made it work
using a set of overlays and auxiliary processor states, and this
formed the 2.10BSD release. He also began attending USENIX and he'd
announce the progress of the removal of all AT&T code from BSD which
started in 1986 at thirty five percent AT&T license free, and the
announcement was met with widespread cheers and applause. This became
important due to price increases from AT&T. As of the 24th of
February in 1984, the price for a commercial license of UNIX System V
Release 2 with source for a single CPU stood at $43000 (about $126000
in 2024), and each additional CPU was a further $16000 (nearly $47000
in 2024). For educational institutions, the price was lower at $800
(or $2300 in 2024) and an additional $400 for each CPU.
[AT&T UNIX prices 1984]
<
https://www.abortretry.fail/api/v1/file/ caa106e2-deee-4f25-b938-54f15bcd1f64.pdf>
Getting well into the 1980s, it became rather obvious to CSRG members
that VAX wouldn't go on forever. Computer Consoles Inc had opened a
development center in Irvine and there they developed a proprietary minicomputer called the Power 6/32 code named Tahoe. This new
platform was aimed to compete directly with VAX. In need of an
operating system, CCI turned to CSRG. CCI provided the team several
machines and the team set about porting 4.3BSD to the 6/32. This
became 4.3BSD-Tahoe in June of 1988. Importantly, they were able to
push forward Joy's work of separating machine dependent code from the
rest of the system, and 4.3BSD-Tahoe introduced an OSI network
protocol stack, improved the kernel's virtual memory system, and
introduced more efficient TCP/IP support. Unfortunately, 4.3BSD-Tahoe
was short lived as CCI changed their company's focus, and Sperry and
Burroughs released rebranded minicomputers based on the Power 6/32
platform. Where this effort did live on, however, was in a merger of 4.3BSD-Tahoe and 2.10BSD into 2.11BSD for the PDP-11. The Tahoe
effort also meant that much of the BSD codebase had been rewritten
furthering the aim of removing AT&T licensed code.
BSD was released in source format only. Any prospective user would be
required to compile his or her system from source entirely. Given
this, any user would first need to acquire an AT&T source license
before he/she would be able to make use of BSD. As noted previously,
UNIX license fees were ridiculous. Yet, the TCP/IP stack in 4.3BSD
was unique to Berkeley. Several software developers requested that
BSD's networking code be offered separately from UNIX, and this was
done in June of 1989 as Networking Release 1. Pricing for a tape from
Berkeley was $400 (around $983 in 2024), but the license terms
allowed for free modification, free redistribution, and free
application to any use case provided that the copyright notices on
Berkeley's code remain in place, and that products incorporating the
code mention in documentation that code from the University of
California and its contributors was included. It was also available
via anonymous FTP shortly following the initial release. This
newsgroup post was made on the 7th of December in 1988, but other
documentation states public availability was November:
Path:
utzoo!utgpu!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!
ames!pasteur!ucbvax!OKEEFFE.BERKELEY.EDU!bostic
From: bos...@OKEEFFE.BERKELEY.EDU (Keith Bostic)
Newsgroups: comp.bugs.4bsd.ucb-fixes
Subject: V1.73 (BSD Networking Software, Release #1)
Message-ID: <8812070154.AA18358@okeeffe.Berkeley.EDU>
Date: 7 Dec 88 01:54:54 GMT
Sender: dae...@ucbvax.BERKELEY.EDU
Organization: University of California at Berkeley
Lines: 374
Approved: ucb-fi...@okeeffe.berkeley.edu
We are happy to announce the availability of the first release of
the BSD networking software. It consists of the standard user level applications, (along with their manual pages and some related
documentation) and some kernel and C library support. It should be
noted that this software has only been tested for compilation and
operation on 4.3BSD and 4.3BSD-tahoe. A complete list of files is
attached to this message.
The TCP and IP code is approximately the same as that recently made
available via the ARPANET and Usenet. Several new algorithms are
used in TCP, in particular Van Jacobson's slow start and dynamic
window size selection algorithms and Phil Karn's modification to
the roundtrip timing algorithm. These changes increase throughput
and reduce congestion and retransmission. Several fixes were made
in the handling of IP options and other gateway support.
This software suite is copyright The Regents of the University of
California and may be freely redistributed. No previous license,
either AT&T or Berkeley is required. The release costs $400.00 US.
To request an order form, please contact our distribution office by
phone at 415-642-7780, or by email at bsd-d...@ucbarpa.berkeley.edu
or uunet!ucbarpa!bsd-dist, or by U.S. Mail at:
CSRG, Computer Science Division
University of California
Berkeley, CA 94720
Mike Karels
Kirk McKusick
Rather than including the file list, I will simply note that this
release was very complete. It included arp, ftp clients and servers
(yes, plural on both), route, telnet, dns tools, ifconfig, inetd,
pieces of the BSD libc, sendmail, syslog, ping, and uucp.
While the networking release was being made, 4.3BSD development
continued. A new virtual memory system was implemented from MACH at Carnegie-Melon with the porting and merging work done by Mike Hibler.
The interface for that VM system was, however, a purely Berkeley
design adhering to the architecture descriptions found in 4.2BSD. The
NFS system was upgraded to be Sun-compatible via the use of Rick
Macklem's work at the University of Geulph. This became 4.3BSD-Reno.
The original BSD license included in parts of Tahoe and Reno, and in
all of NET/1 and NET/2 read:
Copyright (c) <year> <copyright holder>. All rights reserved.
Redistribution and use in source and binary forms are permitted
provided that the above copyright notice and this paragraph are
duplicated in all such forms and that any documentation,
advertising materials, and other materials related to such
distribution and use acknowledge that the software was developed by
the <copyright holder>. The name of the <copyright holder> may not
be used to endorse or promote products derived from this software
without specific prior written permission. THIS SOFTWARE IS
PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES,
INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
Keith Bostic was still as avid as ever to rid BSD of AT&T code and
free it from the heavy costs that came with that code. At a meeting
of CSRG, he mentioned the popularity of the networking release and he
proposed an expanded release that would include more than networking. Discussion went on for a while, but eventually McKusick and Karels
took up the kernel work with Bostic handling the utilities and C
library. Obviously, this would be a seriously large undertaking for a
single individual, and Bostic figured that others would be willing to
help. Bostic then encouraged people to coordinate over the budding
Internet with people across the globe submitting contributions. He
also encouraged people to contribute at USENIX. In a little over a
year, most of the utilities and libraries had been rewritten with
major contributions coming from Bill Jolitz, Donn Seeley, Trent Hein,
Vadim Antonov, Mike Karels, Igor Belchinsky, Pace Willisson, Jeff
Polk, and Tony Sanders. Karels and McKusick hadn't actually expected
Bostic to succeed, but with the work completed, Bostic walked into
their office with his head held high and inquired as to their
progress on the kernel. The two then went off to work on it file by
file, removing everything originally included in the 32V release, but
they were short by six files. Rather than getting a new license and
name created by the university's lawyers, the group reused what they
had and this became Network Release 2 announced on the 3rd of July in
1991.
Path: gmdzi!unido!fauern!ira.uka.de!sol.ctr.columbia.edu!spool.mu.edu! caen!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!ucbvax! OKEEFFE.BERKELEY.EDU!bostic
From: bos...@OKEEFFE.BERKELEY.EDU (Keith Bostic)
Newsgroups: comp.bugs.4bsd.ucb-fixes
Subject: V1.95 (BSD Networking Software, Release #2)
Message-ID: <9107032314.AA06592@okeeffe.Berkeley.EDU>
Date: 3 Jul 91 23:14:59 GMT
Sender: dae...@ucbvax.BERKELEY.EDU
Organization: University of California at Berkeley
Lines: 36
Approved: ucb-fi...@okeeffe.berkeley.edu
We are happy to announce the availability of the second release of the
BSD networking software. The distribution includes approximately 75%
of
the utilities distributed as part of 4.3BSD-Reno and the C library
(along with manual pages and some related documentation), and much
of the kernel. We wish to *strongly* emphasize, however, that
significant portions of the kernel are missing and that no binary
support is supplied for any architecture. Please note also that
this software has only been tested for compilation and operation on 4.3BSD-Reno.
This release is intended for system developers and others who wish to
preview or experiment with the most recent Berkeley system. It may
also
be useful as an update to earlier BSD or BSD-derived systems, although substantial work will be required to integrate portions of this
release
into older systems. This distribution is *not* intended to be used on production systems, nor is it intended for sites without the expertise
to find and fix problems that are encountered.
This software suite is Copyright (C) 1991 The Regents of the
University of California and may be freely redistributed without
further charge. No previous license, either from AT&T or Berkeley
is required. The release costs $850.00 US on 6250 BPI 9-track
magnetic tape or 8mm Exabyte cassette or $950.00 US on 1600 BPI
9-track magnetic tape. The distribution is approximately 90Mb in
size. To request an order form, please contact our distribution
office by phone at 415-642-7780, or by sending email to bsd-d...@ucbarpa.berkeley.edu or uunet!ucbarpa!bsd-dist, or by U.S.
Mail at:
[continued in next message]
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)