[TUHS] System Economics

[TUHS] System Economics

[Doug McIlroy]

"Open" was certainly not a work heard in the Unix lab,
where our lawyers made sure we knew it was a "trade secret".
John Lions was brought into the lab both because we admired
his work and because the lawyers wanted to reel that work
back in-house.

Out in the field, the trade secret was treated in many
different ways. Perhaps the most extreme was MIT, whose
lawyers believed it could not be adequately protected in
academia and forbade its use there. I don't know what                           eventually broke the logjam.

Doug

[TUHS] System Economics

[Clem Cole]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1446 bytes --]

On Thu, Mar 16, 2017 at 3:33 PM, Doug McIlroy <doug at cs.dartmouth.edu> wrote:

> Perhaps the most extreme was MIT, whose
> lawyers believed it could not be adequately protected in
> academia and forbade its use there.
>

​Doug - that is interesting.  Do you know the time frame of the
banishment?  Noel any memories of what allowed it be used?  Clearly, once
the restriction was removed, it spread.

At CMU, for the OS course, we had a sign a document with the university
stating something that we understood it was AT&Ts IP and we were using it
as a teaching tool.  I remember thinking that whole thing was weird, the
students could not get accounts on the OS development machines for the
course to get get source access -- but I already had one because I was
working for the EE dept before I took the official OS course and had been
kernel hacking and already helped write fsck with Ted that summer before.
So it felt like the CMU lawyers were trying close the barn doors after they
were already open with a couple of us (their were probably 5 or so that
were in that same spot).

Maybe if we can find Bill Wulf, whom I think was the one that shepherded
that document through the CMU lawyers.  His wife, Anita was teaching the
course at the time; IIRC.  He might remember.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170316/1058ef3c/attachment.html>

[TUHS] System Economics

[Paul Winalski]

On 3/16/17, Doug McIlroy <doug at cs.dartmouth.edu> wrote:
> "Open" was certainly not a work heard in the Unix lab,
> where our lawyers made sure we knew it was a "trade secret".
> John Lions was brought into the lab both because we admired
> his work and because the lawyers wanted to reel that work
> back in-house.

That matches my recollection:  AT&T treated the UNIX sources as a
trade secret.  When I worked on DEC's port of the VAX/VMS linker to
Ultrix, our team was very careful to work from the a.out specification
only, and to avoid any contact with the sources to ld.  We wanted to
avoid any chance of AT&T claiming that our VMS linker port in any way
used their proprietary technology.

AT&T made the sources available pretty widely in academia, for use as
a teaching tool, and some of the universities involved seemed to play
pretty fast and loose with the NDA.  A lot of CS students I talked to
were under the impression that the UNIX sources were freely open and
hackable at their college.  Because of this I always wondered whether,
if push came to shove, AT&T would be able to legally enforce its trade
secret claims.  I don't think the issue was ever actually litigated.

-Paul W.

[TUHS] System Economics

[Wesley Parish]

In relation to which, a google search on "site:groklaw.net unix methods" yields some interesting 
observations on this very topic of "trade secrets" wrt Unix.

Wesley Parish

Quoting Paul Winalski <paul.winalski at gmail.com>:

> On 3/16/17, Doug McIlroy <doug at cs.dartmouth.edu> wrote:
> > "Open" was certainly not a work heard in the Unix lab,
> > where our lawyers made sure we knew it was a "trade secret".
> > John Lions was brought into the lab both because we admired
> > his work and because the lawyers wanted to reel that work
> > back in-house.
> 
> That matches my recollection: AT&T treated the UNIX sources as a
> trade secret. When I worked on DEC's port of the VAX/VMS linker to
> Ultrix, our team was very careful to work from the a.out specification
> only, and to avoid any contact with the sources to ld. We wanted to
> avoid any chance of AT&T claiming that our VMS linker port in any way
> used their proprietary technology.
> 
> AT&T made the sources available pretty widely in academia, for use as
> a teaching tool, and some of the universities involved seemed to play
> pretty fast and loose with the NDA. A lot of CS students I talked to
> were under the impression that the UNIX sources were freely open and
> hackable at their college. Because of this I always wondered whether,
> if push came to shove, AT&T would be able to legally enforce its trade
> secret claims. I don't think the issue was ever actually litigated.
> 
> -Paul W.
>  



"I have supposed that he who buys a Method means to learn it." - Ferdinand Sor,
Method for Guitar

"A verbal contract isn't worth the paper it's written on." -- Samuel Goldwyn

[TUHS] System Economics

[Steve Johnson]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1850 bytes --]

Yes, Doug is spot on.  Being told by the govt. that we had to patent
everything we could and license it fairly got rather strange when
software was involved -- there was a lot of question whether software
could be patented at all, and the Labs had to patent software if it
could be patented (e.g., the setuid bit).

The biggest issue, which still hasn't gone away, is that software
moves so much faster than the law.  The lawyers seemed to take the
standpoint that if something was questionable, just wait for five
years until there are better legal precedents. 

At one point I made a major push to get the PCC grammar for C released
in the public domain.  I still think this would have brought
standardization about much sooner, and my managers were in favor of
it.  But the lawyers delayed and delayed.  The next thing you know,
we had "far pointers" and all sorts of other gook in the language that
took the standards committee additional years to wring out.  Sigh...

Steve

----- Original Message -----
From: "Doug McIlroy" <doug@cs.dartmouth.edu>
To:<tuhs at tuhs.org>
Cc:
Sent:Thu, 16 Mar 2017 15:33:10 -0400
Subject:Re: [TUHS] System Economics

 "Open" was certainly not a work heard in the Unix lab,
 where our lawyers made sure we knew it was a "trade secret".
 John Lions was brought into the lab both because we admired
 his work and because the lawyers wanted to reel that work
 back in-house.

 Out in the field, the trade secret was treated in many
 different ways. Perhaps the most extreme was MIT, whose
 lawyers believed it could not be adequately protected in
 academia and forbade its use there. I don't know what eventually
broke the logjam.

 Doug

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170316/e174f549/attachment-0001.html>

[TUHS] System Economics

[Noel Chiappa]

    > From: Clem Cole

    > Do you know the time frame of the banishment? Noel any memories of what
    > allowed it be used?

Sorry, this is something I know nothing of; it must have happened while I was
still an early undergrad.

The first Unix I knew of at MIT was the one in the DSSR/RTS group in LCS,
which arrived (I think) roughly around the start of my sophmore year (so
early '76 or so) - I have a memory of one of my friends (who was an undergrad
working in that group) telling me about it, and showing it to me. (I remember
being totally blown away with the way you could write a command 'foo',
compile it, and jut say 'foo' to run it...)

Actually, it may have shown up well before that - perhaps they had it well
before I first saw it.

Certainly by the time I showed up at LCS (fall of '77) it had already spread
to CSR; they had an 11/40 with Unix on it, cloned from the DSSR system.
Again, I don't know if there was any paperwork that had to happen, or if that
system was already covered under whatever license the DSSR machine was under.

Of course, this was all DARPA-funded work, and there may have been something
there that allowed it. We certainly passed Unix source around with other
DARPA projects (e.g. at BBN) without, AFAICR, worrying much about paperwork.

    > we had a sign a document with the university stating something that we
    > understood it was AT&Ts IP

I don't recall anything like that at MIT; maybe in the very early days, there
was something, but certainly not by '77.


If it's important to know what happened, I can ask (e.g. Prof. Ward, head of
DSSR).

	Noel