Newsgroups: comp.misc
Path: utzoo!utgpu!watserv1!watdragon!rose!ccplumb
From: ccplumb@rose.uwaterloo.ca (Colin Plumb)
Subject: Re: rms says...
Message-ID: <1991Feb16.234642.23311@watdragon.waterloo.edu>
Sender: daemon@watdragon.waterloo.edu (Owner of Many System Processes)
Organization: University of Waterloo
References: <21327@yunexus.YorkU.CA> <4607@lib.tmc.edu> <1991Jan29.201935.840@watdragon.waterloo.edu> <4687@lib.tmc.edu>
Date: Sat, 16 Feb 1991 23:46:42 GMT
Lines: 61

jmaynard@thesis1.hsch.utexas.edu (Jay Maynard) wrote:
> If he's not interested in maximizing the reuse of software, then he
> should remove the claim [from the GPL] that people should place their
> code under it so that reuse will be maximized and mankind will
> therefore benefit.

My apologies; I wasn't addressing the issue properly.  To see if the
GPL encourages software reuse, compare it to the public domain:  the
GPL scares off some people (less code reuse), but encourages others to
distribute their code when they might not bother if it was public
domain (which promotes reuse).  NeXT's objective-C patches and Intel's
i960 port of GCC come to mind; there are certainly other cases.  To be
fair, I know of examples for the other case.  Some Microsoft code in
development I know of uses Henry Spencer's regexp code.

I don't know enough to say which side of this equation wins, but I
don't think it's obvious.  Even if it's a net loss now, as more GNUware
is used successfully (becuase it's shipped by OS vendors, for example),
fewer people will be inhibited, and the balance may shift.  (I don't
claim that this decline goes to zero, however.)

> Except that I still have to install them, and hack them up to get them
> going on my SysVr2-based system. Therefore, I have to have seen their
> source code. Microsoft, Lotus, et al can afford to have separate people
> for installation and maintenance of GPV-infected code. There's only one
> of me (I can hear several "Thankfully!"s around the net now... :-).

This is not an issue.  First, the FSF is not going to use its limited
funds to sue you for something as weak as that.  Second, there was one
part of the Intel/NEC microcode suit that's relavent (in the U.S.).
The judge ruled that, even though the author of the NEC microcode had
made a detailed analysis of the Intel microcode (this is a lot more
than looking at it; he wrote a paper on the subject), the knowledge he
gained was "background" and its use in his writing of the functionally
equivalent NEC microcode (which had some similar parts) did not infringe
Intel's copyright.

In other words, reading someone's code does not contaminate you.  Even
if you have carnal knowledge (i.e. very intimate) of someone else's
code, and write an exact clone with some obvious similarities, you don't
infringe their copyright.  See the "Micro Law" column in past issues of
IEEE Micro for an analysis of the decision.  Take heart, compile
GNUware, and be happy.

For example, I've had discussions with Microsoft's compiler authors
regarding the innards of GCC vs. MSC.  At Microsoft, there are no
clean-room procedures surrounding GNUware.

> All that'd have to happen to make me happy is to disinfect the virus. I
> realize that's not going to happen, though...since RMS wishes to coerce
> others into his utopia. Oh well.

I don't know whether RMS wishes he could coerce others into his utopia or
not, but he can't.  He can try to make it easy to enter and hard to leave,
but coerecion reminds me of a quote from Gerald Durrell's siblings over
breakfast (_My_Family_and_Other_Animals_, probably):
Margo: [My weight] "is all your fault, forcing me to eat all that rich food."
Larry: "It's just being put on the table - nobody's forcing you to eat it."
Margo: "Well, you know I can't say no, so it's forcing."
-- 
	-Colin
