Subject: Re: Cygnus and proprietary software
From: Bernard Lang <Bernard.Lang@inria.fr>
Date: Tue, 23 Dec 1997 19:18:11 +0100 (MET)

-----------
I'm still unclear on one thing, Brian.  Why do you want to discuss a new
free software model with less business potential than the GPL on the free
software business list?  Put on your business hat and explain what you want
to achieve as a business model, rather than your user hat under which you
offer concepts that you personally would like to have in software that you
use.  I'm sure you are thinking of something deeper than you may be
communicating.  I'm not clear, though, on whether it is relevant to a
pro-business discussion.
-----------

   I have been wondering too.

 2 or 3 points seem important to me.

  1- for various sociological and historical reasons, the software
business is largely dominated by commercial software. I personally
think that this is bad, not for moral reasons (though I sympathise
with RMS views), but rather because software development is much like
the development of mathematical theories (BTW this is not just an
analogy, profs and program have been shown formally to have the same
mathematical nature). The production of good software (i.e. with the
proper functionnal interfaces, the right choice of data
representation, and elegant and efficient bugless implementations,
...) can seldom be achieved at first try, or by a single individual.
It is a long social process of comparison, confrontation with
competing solutions, usage assessment in various contexts, etc... that
make the right choices and the best implementation emerge. The
maturation unix environmment (GNU) is an exemple of that social
process.
  But commercial development is done in secrecy, the code (proofs) is
not available, and sometimes not even proper specification of
functionalities (definitions and theorems). Hence the social process
cannot take place, and this slows down technological progress.

 2- however, as I said software is dominated by the commercial world.
That is currently a hard fact. Hence, the next best thing to freeing
it from commercial constraints is to create free software in a form
that is attractive to business, by whatever means, so that business
credibility will help it to spread, while leaving doors open for the
kind of social process I have been describing above. Besides, short of
being financed otherwise (for example as mathematicians are), free
software developpers do have to make a living. I do like them happy so
that they make more :-)

 3- conceivably in the future, the economic value of freeness will be
recognized, and other economic models can develop for the production
of free software, not necessarily business models.
     But even then, there will be room for commercial software in
specific application areas, exactly as they are now mathematicians who
do applicative work for hire, and keep their results for the companies
that pay them. But this is seldom fundamental mathematics, on which
fundamental research would be based.
   In other words, while freedom is desirable for most of fundamental
software packages or system (thoses that are widely used, especially
as component or support of others), there is no reason freedom should
be expected on all software (for example a military command system).
The same company could release as free some "fundamental" stuff (like
they do already for mathematical results), while keeping specific
applications for its business. There is no contradiction between
releasing some software as free (whatever the licence), and keeping
other software proprietary or secret to make money from it.

  In an economy where most thing have to be paid for (you may dispute
that it is a good system, but it is the one that is), someone,
somewhere, has to pay for the production of free software. I would
hardly blame a company for doing it on its own, without subsidies, by
financing it with proprietary software.
  And free software must be made attractive to business, if it is to
have any success, which I do hope.

Cordialement

Bernard

 Bernard.Lang@inria.fr             ,_  /\o    \o/    Tel  +33 (1) 3963 5644
 http://pauillac.inria.fr/~lang/  ^^^^^^^^^^^^^^^^^  Fax  +33 (1) 3963 5469
             INRIA / B.P. 105 / 78153 Le Chesnay CEDEX / France
          Je n'exprime que mon opinion - I express only my opinion
                  CAGED BEHIND WINDOWS or FREE WITH LINUX