Subject: Re: Open letter to those who believe in a right to free software
From: Bernard Lang <Bernard.Lang@inria.fr>
Date: Sat, 30 Oct 1999 21:33:39 +0200

On Fri, Oct 29, 1999 at 09:37:25PM -0700, Karsten M. Self wrote:
> Bernard Lang wrote:
> > 
> > On Thu, Oct 28, 1999 at 03:15:39PM +0900, Stephen J. Turnbull wrote:
> 
> > >     Bernard> modularization is far from marginal 
> > > It _is_ marginal in the sense that the best-practice proprietary firms
> > > do it too.
> > 
> > Not so well apparently. I was yesterday on a panel with someone from
> > Computer Associates, and he told me that they are currently modifying
> > their production organisation to benefit from lessons learned from
> > free-software production.
> 
> I'd be inclined to agree with Bernard's comment that proprietary firms
> don't modularize well -- more from results than having seen processes. 
> As someone who's seen the insides of a lot of systems shops, the
> modularization and general understanding of "own code" is low.
> 
> Bernard:  details on what CA said?  I just gave an earful to an R&D
> manager for software I use heavily -- proprietary stuff.  Apparently all
> this free love and software's made me a bitter man....

We did not talk long ... not supposed to have private meetings on the
panel ... and I was in a rush at the end. Essentially that they are
beaking some development teams into geographically separated groups
that are allowed to communicate only about API.
 
> > And math develops slowly not because of its economic model, but very
> > simply because it is hard.
> 
> Does it and is it?

 I do not think it is that slow ... (was not my opinion ... but what
is the yardstick) ... but yes ... it is hard.

  Or put it another way. Has computer science progressed much faster
in the past ten years than in preceeding decades ... I believe not
despite patents and venture capital. The silicon valley bubbling is in
my opinion very superficial technically. On the other hand, it is used
to justify legal changes (notably about patents) that can introduce
much greater viscosity in the field, and possibly slow down the
progress by discouraging creators from entering a (mine) field where
independant work will be harder.

> Is math any harder, say, than chemistry or drug development?

I do not understand the question.

But one thing is certain, creation in math requires only bread and
water, and software creation require little more these days.
   Chemistry or drug development requires heavy investment.
   So, from an economic point of view, they can hardly be compared,
though the use of patents in biotech is often given as justification
for their use in infotech,..

> What about times like WWII and the Cold War era when certain areas of
> basic research (nuclear physics, crypto) were heavily funded for
> strategic reasons.

Funding can bring more people to bear on a field, and that is
especially effective for applications, for problems that are
semi-hard.
  But true innovation has its own incentives ... I just do not believe
it can be done for external gain. It eats the creator.

>  Is there an incentives problem in math?  Are we or are we not
> producing an economically efficient level of algorithms and proofs?

no amount of funding would have proved Fermat's theorem faster ... and
that is what really matters in the long run. Not the tehorem itself
... but the theories needed for the proof.


-- 
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