Jeremy,

I have described much of this in previous messages. So to recap....

The first licensee was Columbia University (Lou Katz), 4th Edition and I >>believe<< Harvard was second, but it might have been someone else in NYC. I was always under the impression that Rand Corp was the first commercial license, when a couple of Harvard Students wanted to bring it west. CMU, MIT, UofI were all about 1-3 years later. I believe MIT and CMU started with 5th edition and UofI 6th. Chesson is no longer with us to verify, but I think Steve Holmgren was in on that, and I believe I have seem him on a couple of other mailing lists. CMU was definitely 5th edition to start, and quickly went to 6th. MIT had it early also since, like CMU, they had interns and OYOC students from the Labs, so code went freely both ways in those days.

The redistribution license is a little more hazy. It was either Peter Weiner at ISC or the folks at Wollongong. I had always been under the belief it was Peter, but Werner has some data that shows Wollongong was either at the same time or shortly thereafter. These were custom licenses for V6 and its not clear about certain details and Werner seems to think not exactly the same (i.e. Wollongong negotiated some special terms no one else had). Al Arms (AT&T legal) is likely to be the only person that really knows for sure, as he was the common thread on all it in the early days.

With Seventh Edition, Al wrote the first general commercial redistribution license with sliding fees etc. If you wanted a 7th edition license, any previous licenses we voided. There was great moaning about the fees. After about 6 months, Prof Dennis Allison of Stanford (who was consulting for just about all of us in those days), brokered a meeting at Ricki's Hyatt in Palo Alto, I want to say winter 1980. This was the beginning of the more global negotiation with what would become the System III license (all flavors). Again this license superseded all previous ones.

And thus the #1 president for the later OSF creation came into being, so called: 'Stable Licensing Terms.' When the System V license was released, AT&T changed things again. By the time of SVR3 the commercial folks had had enough and a war ensued.

On Sat, Nov 23, 2019 at 10:03 AM <reed@reedmedia.net> wrote:

Any more details about this?

The Dec. 1973 agreement with Univ. of California is "solely for
academic and educational purposes" and included "source program code".
(Their initial installs were 4th edition.)

The slightly revised Dec. 1974 agreement with Katholieke Universiteit
is also solely for academic and educational purposes" with nominal
service charge of $150. (This was signed in February 1975 a little
before the 6th edition came out.)

I read multiple times that the first "source" licensee may have been the
Univ. of Illinois. But I also read they were the first "source" licensee
for the "5th" edition so not any Unix source license in general.

Was Univ. of Illinois the first source licensee regardless of the
edition (so prior to Dec. 1973 / Jan. 1974)? Any docs/citations on
this?

https://web.archive.org/web/20160322042314/http://www.ece.ubc.ca/~gillies/mail/dbgillies_ken_thompson.txt
suggests the agreement with Univ. of Illinois happened a few months
before July 1975, so maybe it couldn't have been the first (since there
are copies of agreements prior years with other schools) even though
that says the first.

The (Univ. of Illinois) Network UNIX RFC 681 is dated March 18, 1975 and
NIC 32157 dated May 14, 1975. It references Fifth Edition and has:

BELL'S POLICY IN THE PAST HAS BEEN TO LISCENSE THE SYSTEM TO
UNIVERSITIES FOR A NOMINAL FEE, $150.00, AND UNFORTUNATELY FOR A COST
OF $20,000.00 TO "NONUNIVERSITY" INSTITUTIONS.

Since the NIC has later date that the RFC maybe this was updated later
than the RFC date so the commercial cost is really about the 6th
edition?

Does anyone have a copy of the software agreement for a non-university,
or without the "solely for academic" clause, or for $20,000 from early
1975 (or for 5th edition)?

This 1983 Byte magazine article
https://archive.org/stream/byte-magazine-1983-10/1983_10_BYTE_08-10_UNIX#page/n133/mode/2up
says the 6th edition was first Unix to be sold to commercial firms. A
company license was $20,000 and a educational license was $200.

Anyone have a copy of any 6th edition license agreement?
(The author of
https://wiki.tuhs.org/lib/exe/fetch.php?media=publications:theses:gmp_thesis.pdf
says has a copy signed May 12, 1977 but I couldn't get in contact with
yet.)

I understand the agreement for 7th Edition included a clause saying a
course curriculum couldn't discuss or describe the code. (John
Lions clause).

Does anyone have a 7th edition license? (Educational or otherwise?)
(Again the author of above thesis says has copy of a 7th edition
education license signed Feb. 20, 1981.)

Does anyone have Exhibit F -- the "32V Software Agreement" dated April
1, 1979 -- for the AT&T/USL vs. BSDI/Univ. of California Jun 1992
complaint (or Exhibit B to the DeFazio Affidavit)? (See
920724.complaint.txt and more details in 930107.amicus.txt and
930108.oppose.txt.) Or maybe the 1981 relicense?

What does it mean about the often mentioned first commercial
version from AT&T wasn't until System III (Unix Release 3.0) in 1982?
This is confusing since documented references to commercial versions in
early 1975. Is this about AT&T proper instead of Bell Labs? Or maybe
about official commercial support?

(RFC 681 which mentions the $20,000 non-university license fee also
mentions RAND, Lincoln Laboratories, and Inco had Network UNIX source
code, but unsure if that means that had the commercial license too.
There are other commercial licenses long prior to System III, but maybe
it is about professional support.)