On Mon, Jul 19, 2021 at 1:38 PM Theodore Y. Ts'o <tytso@mit.edu> wrote:
However, I strongly contest the claim that Unix was "Open Source".
Unix was the UNPUBLISHED TRADE SECRET of AT&T, and students exposed to
Unix source code became contaminated with AT&T's "methods and
concepts" clause. 
But it was not.   And in fact, the whole thing about the AT&T court case was that most of thought it was about copyright, not if it was a trade secret (Which is why they lost) in the end.
Note Jack Haverty's words to me last night about UNIX in the mid-1970s when he was working the first PDP-11/40 TCP implementation show he was is thinking the same way as I am:  the fact that an incompetent novice can make changes to open source.

BTW: I asked Jack last night what type of license BBN had.  He responded that he did not think they really had one originally.   They were an ARPA contractor.  He (i.e. BBN) did not have to go into legal hoops until he wanted more documentation (a.k.a. the Lions text).

The concern was not that we were going to be sue for being mentally contaminated,  that whole idea did not come about until after the law suite.
So they couldn't even *reimplement* Unix without potentially getting sued by AT&T.
Which of course is what made the whole thing silly.   The first 'UNIX clone' was by an ex-BTL, Dave Plauger's Idris, an implementation of v6.  Many others would follow, from Coherent to Sol, Chorus, Minix, and Linux to name a few quickly.  In fact, AT&T did check on the Mark Williams code base.   Dennis once wrote about having to check out the sources for the lawyers.   As I understand it, the AT&T team concluded that while the MW team may not have directly taken the AT&T code, the MW folks clearly had seen it [which the Mark Williams folks I do not believe ever denied].  AT&T chose not to pursue them.  It was not until BSDi/UCB that they literally made of case of it.  Which again was why so many of us thought that case was about copyright, not trade secrets.   It had never been really discussed with us ->> on the outside<< of BTL (it would later learn from Dennis and few others that they had discussed TS with their lawyers at some point).


Your thinking would be reasonable iff AT&T had won the case, but the fact is the ideas (IP) and even the source to the basic UNIX was open and available.  The IP was published by but in places like CACM and from Prentiss-Hall.  Larry's point is a solid one, is that to get access was limited by having the means to afford the licenses but more importantly it was having the means to afford the hardware to run it.  So until a computer and 'mortal could own' on his own, AND that could support UNIX (i.e. a 386-based PC), the issue was moot.

I point out that the 1956 consent decree >>required<< AT&T to make UNIX (like the transistor before that) available to all 'interested parties' (see Pinheiro J. (1987). “AT&T Divestiture & the Telecommunications Market”, Berkeley Technical Law Journal, 303, September 1987, Volume 2, Issue 2, Article 5 if you don't believe me).   They had to make it available to research folks and we allowed to license its use to commercial people using 'fair and reasonable licensing terms regulated by the US Gov.   You can suggest that $20K was unreasonable for personal use, but again it was not unreasonable for a University ($150) or for the commercial sector for that matter and those were the people buying the computers in those times.

The practice of the day was to make the sources (which of course were written in assembler) to the customers of your hardware.  And by the way, IBM was not going to 'sue your pants off' as you mentioned.  The first 'clone market' was in fact the IBM 360 clones and IBM licensed their SW to non-IBM HW customers (like Amdahl, Nixdorf, and NEC to name 3) because IBM was afraid of being sued by the US Gov!!!   It was folks like DEC that sued Cal Data for 'cloning the PDP-11,' not IBM.

Yes, UNIX was licensed and yes the IP was owned by AT&T.  But that's really not much different than a GPL2 which is licensed and owned by someone else.  The rules of use are different, however.  But the source was just as open.  A difference was how it was distributed (you had to be part of the club in Larry's terms) but anyone that could pay the HW fees could join the club.

Larry has made an excellent point (which I agree with), is that in practice it was clubby.  But in my defense ... the was the same club as before.  You had to have the hardware and the need.  But if you had that you could get to it.  Hey Ted, you were part of that club too -- you had access to things at MIT Athena that most people did not see.  MIT had paid the club fees and gotten the HW.  Frankly, you personally had way more access to the sources as an MIT undergrad with a job at Athena, than say, Larry did at U Wis.

What changed was Moore's law and who afford (and thus get access) to the HW and economics associated with the desire to obtain but please don't try to say the behavior or intent was any different.   It just was not.   We once called this the 'hacker culture' -- we are all in it together and we shared what we had with each other.  

FWIW:  this is has been discussed in other books and areas too.   It was noted that the late 1960s 'hippie' sharing culture around the SFO area played into too.   Steve Levy's wonderful book 'Hackers' talks about it from the MIT Model RR club.  In fact, the last chapter of his book is dedicated to RMS and calls him 'the last hacker.'

Methinks the horse is dead  ... you can think it's new.  It just was not.  You can say, UNIX closed because you came upon it a timer when the versions that mattered (SunOS/Solaris/Ultrix/AIX/etc...)were becoming less available to you as a user.   But the fact is the core material of UNIX was open.  We all had access to it that's what made it great.  We did and could access and change it.  We could share it.  Sometimes we chose to clone it.  Sometimes we even improved on it (and sometimes like systemd, we can argue if we did).

It did not become more of a 'closed' until the HW economics changed the rules, which just happens to be when you and others came of age.  Which is fine, just please, please, please respect that the whole FOSS movement got its start because of foundations and ideas that came long before.   The cool part is the because of the new economics, you were able to do something with it and expand it.  I do celebrate and laud you for that.  But I do also ask that your respect the foundation which gave you that start.

Clem