Unix and AT&T (flame)

[Geoff Kuenning]

In article <7294 at brl-tgr.ARPA> Barry Shein writes:

>RE: Message of Lauren Weinstein....
>
>Hear, hear. I agree. One of UNIX's strongest features
>is the wide availability of the sources to everything
>and it's design which makes it very source supportable
>by reasonably able people. I have wasted more of my
>life on systems that either wouldn't release the sources
>at a reasonable price (or any price) or, when you got
>the sources, were so impenetrable that they were nearly
>useless. It would be sad to see AT&T bitten badly because
>of their sanity. It would be worse yet to see them lose
>that sanity, I think some of the reactions in this column
>have been over fears of the latter (like all the unbundling
>now going on.)

Hang on there, Lauren and Barry!  Let's remember that this started out as a
question as to whether AT&T has the right to prohibit you from selling a
yacc-based program "because it includes /usr/lib/yaccpar".  The rest of the
discussion has been a legalistic one about whether they could make that
claim stand up in court.  It is not unnatural to extend such a discussion
to the more general question of whether AT&T could make ANY trade secret
claim about Unix stand up in court.  Most of these messages have been cross-
posted to net.legal, where this discussion belongs -- it's only in unix-
wizards because of the original note about yaccpar.

I cannot remember a single person stating or even implying that they approved
of stealing any part of Unix.  I certainly would not like to see AT&T lose
its ownership of Unix, because I think that would lead to impossible
fragmentation of versions.  But that does not prevent me from noticing that
AT&T has been rather sloppy in an area where the law does not permit
sloppiness.  Perhaps AT&T should be forgiven for being sloppy because they
were good guys;  to my mind that's a political question.

As to the original note about yaccpar, I am of the opinion that an attempt to
restrict the sale of yacc-derived parsers (other than perhaps collecting a
fair royalty on the yaccpar part) is restraint of trade.  You can fight that
with antitrust law, or you can fight it with the trade secret law.  The
principle that disallows patents, copyrights, and trade secrets on
knowledge available to the general public is designed to prevent such
restraint of trade.  In fact, (here I go again shooting my mouth off about
something I know nothing about) I wouldn't be surprised if you could beat
a royalty attempt on yaccpar by proving that it was substantially similar to
the one published by the Software Tools people.  That's an interesting part
of trade secret law that I don't know enough about.

In any case, I'm not suggesting stealing Unix (though I would fight an
attempt to keep me from selling yacc-processed programs just as hard as an
attempt to collect a royalty on cc-compiled programs).  I *am* suggesting
that AT&T continue its efforts to clean up its act (much as that causes
pain to the small 68000-type companies).  And I am also suggesting that
AT&T had better face up to the fact that part (not all, I hope) of Unix has
found its way into the public domain, and be prepared to deal with that
fact when somebody *does* try to do something unethical.  Personally, I'd
like AT&T Unix to include a directory full of "supported public domain
software".  Seems a pity for AT&T to not use its muscle to standardize
things like the strings packages and give them good support.
-- 

	Geoff Kuenning
	...!ihnp4!trwrb!desint!geoff