In your message of 27 March 2000 you write:
> Of the members of this group, I may be the only one who doesn't develop
> OpenGL implementations or work for a big games house or something.
>
> It looks like I'm the one who speaks for the the bulk of non-commercial
> Linux programmers out there.
>
> Putting just the extensions I personally need into just the one
> libGL.so that I happen to use would solve MY needs - but that's
> hardly fair to the people I'm speaking for.
I think you are not alone. Despite the fact that we do our own OpenGL
implementation, we are at the same time on the other side of the
fence, as we have to support people who port their applications from
SGI/SUN/IBM and yes, from Windows to LINUX. The fewer things we do to
break source code, the better it willbe. The ideal scenario is that
SGI/SUN/IBM code is will compile under Linux right out of the box
(well, apart from the usual include file changes). But forcing them to
have significant source changes in the logic of their code would mean
a big hurdle for those people to port things.
And what a lot of folks on this list don't seem to understand is the
writers of applications don't really care so much about OpenGL per se,
what they care about is how to make their application work. OpenGL is
just a means to do that. If we decide here on a standard that makes
OpenGL on LINUX different from OpenGL on SGI, or OpenGL on Windows,
then OpenGL on LINUX is considered broken. And THIS is a big problem.
I don't really care to much, whether gl.h includes glext.h or not, or
whether a version symbol is defined for this api, and which format it
has. I really do care about changing the default behaviour of
including gl.h and the established standard all other OpenGL
implementations of how extensions are being handled right now.
- Thomas
--
Thomas Roell /\ An imperfect plan executed violently
Xi Graphics / \/\ _ is far superior to a perfect plan.
[EMAIL PROTECTED] / / \ \
/ Oelch! \ \ George Patton
