Subject: Re: Exploring the limits of free software: Cygnus, and GPL
From: Ian Lance Taylor <>
Date: 25 May 1999 13:11:18 -0400

   From: Jean Camp <>
   Date: Tue, 25 May 1999 12:04:10 -0400

   >Here's the kicker: if I figure out how to maximize social benefit, I
   >will probably be able to do the profit maximization (and any
   >combination in between) as a corollary.

   And I will guess they are mutually exclusive solutions to the equations
   that are the question.

I don't understand that comment.  Are you saying that profit is
incompatible with social benefit, or that they scale inversely?  That
seems unlikely to me.

   >_As the owner of that code_, I would be satisfied with my compromise,
   >because it maximizes social contribution.

   This does not follow without the following aside. Unless we are all working
   tirelessly only for the social good, for the people, waiting for the state
   to wither. I believe all experiments based on just this ideal have failed,
   excluding China, which is in slow failure mode.

I don't understand this comment either.  Are you saying that we are
only interested in maximizing social contribution if we are working
tirelessly only for the social good?  That defies observation.  Many
people prefer to maximize social contribution within the constraints
of their desired lifestyle.  Many people contribute time and money to
charity not to get a tax benefit but because they think it is the
right thing to do.

   Alternatively software can function quite well in the market, where value
   and packaging and price are strangely combined. Thus "lunchables", Anderson
   Consulting, insurance and other things where the bundling makes the
   otherwise useless valuable.  And this is the interesting question wrt
   Cygnus: how much value can they extract, how much do they need to put back
   in to be sustainable?  If Cygnus only keep FS sustainable out of the
   goodness of their hearts then how  is it possible to alter the balance,
   extract and concentrate the public good, make money and kill Linux doing
   it? (This is also why Liinux may be good for Microsoft -- they can extract
   value, cause ridiculous forking, add terrrible interfaces, bloatware,..
   crashware and put nothing back in. Be thankful the Justice Dept kept them
   occupied and they did not purchase,, etc.)

I'm having a hard time sorting out just what you are saying.  I don't
know if my comments will be to the point.

Cygnus gets a lot of money by modifying the software development tools
to order.  Accordingly, Cygnus needs to keep the software maintainable
in order to keep the costs to Cygnus of these modification services

Cygnus programmers are also motivated individually by issues other
than the love of lucre.  I believe that all good programmers naturally
prefer clean and elegant software (I can sustain that belief by
defining programmers who prefer ugly code as bad programmers).  Since
Cygnus programmers are good programmers, they naturally tend to
produce maintainable code as a byproduct of their profit maximizing

I have no idea where killing Linux comes in.  If software is a public
good, it is so only when it becomes infrastructure.  For example, this
has happened to TCP/IP.  It has almost happened to operating systems,
but not quite yet, and the issue remains unsettled.

Microsoft can do all the things you mention to Linux, but it will not
affect the existing Linux community.  Microsoft could purchase and, and it would not affect the existing Linux

   Is such a destruction of GNU/Linux inevitable when it becomes bundled or is
   bundling a necessary element of economic and political survival?  This
   question seems to me to be the core of Brian's concerns, as opposed to the
   trivial free==price//free==autonomy confusion.  Answering how could it be
   done means understanding how could it be prevented. Economics does not have
   the answer here, I fear.  Economics would only reward the person who acted
   to concentrate value without having the predictive power to understand how
   it would happen.

The economic and political survival of Linux is not at issue.  What is
at issue is whether it will grow an order of magnitude larger than it
already has.

Personally, I think Brian's concerns have much more to do with a
desire for ethical business behaviour, and a misnaming of that desire
with the term ``free software.''

Can Linux grow an order of magnitude larger while maintaining ethical
business behaviour in Brian's sense?  Probably not.  However, the
Linux community never maintained that ethical behaviour in the first
place, so abandoning what it never had will not destroy the community.

Can Linux grow an order of magnitude larger without being somehow
destroyed in some other sense?  Yes, of course.  The core of the
community, other than the kernel, is freedom as in autonomy.  Yes, it
will become increasingly disorganized and forked.  However, this
preserves the core values of the community; it does not destroy them.

   rms has a very important role. What if he is hit by the proverbial bus?
   What if he sells out? (Not likely to happen I hope in my heart but an
   interesting question given that the foundation will survive him.)  Has he
   constructed something which will prevent takeovers without the power of his
   passion and commitment behind it?

Yes, he has.  This may sound heretical, but from my position as a free
software developer, I would say that RMS has become increasingly
irrelevant to the continuing progress of free software.  His harping
on the Linux vs. GNU/Linux naming issue has helped to marginalize him,
as has to a lesser degree his insistence on the Xemacs fork and his
refusal to accept Tcl as an extension language.  The vast majority of
decisions within the free software community today are made without
reference to his wishes.

I don't think this is a good thing.  RMS has a great deal to continue
to offer the community, besides the fact that he basically created it
in its current form.  Unfortunately, some of the same personal
characteristics which permitted him to create it in the first place
tend to marginalize him as it gets larger.