Re: what DOES the GPL really say?

[ig25@fg70.rz.uni-karlsruhe.de (Thomas Koenig)]

In gnu.misc.discuss, dewar@merv.cs.nyu.edu (Robert Dewar) wrote:
><<This is now false.  See the egcs project at Cygnus at:
>
>        http://www.cygnus.com/egcs>>
>
>We will have to see how this develops. At best it will be an interesting
>experimental environment for trying out new ideas in a situation where
>stability is not the first order of business.

I think you underestimate the people on that project.  At best, this will
become a quite stable release which will provide the free user community
with a stable, working gcc which can do RISC scheduling and which has
useful C++ templates.  The FSF, to whom all changes are assigned, can
then integrate them into 2.8 in whatever way they please.

>At worst, it will just be
>a further divergence in gcc versions.

It is already resulting in reducing these versions.  Quite a number of
groups which have brought out diverging gcc versions in the past (pgcc,
the Linux people, ...) are now working together under a single umbrella.
This is definitely a Good Thing.

>Actually it would be very nice if there was more experimental activity
>with GNAT in terms of interesting new experimental ideas,

Why doesn't somebody from ACT take part in the egcs project? So far,
egcs doesn't have anybody working on the Ada frontend to gcc.
-- 
Thomas Koenig, Thomas.Koenig@ciw.uni-karlsruhe.de, ig25@dkauni2.bitnet.
The joy of engineering is to find a straight line on a double
logarithmic diagram.