In your message of 9 September 1999 you write:
> On Thu, 9 Sep 1999, Jon Leech wrote:
>
> > > How does libGL map a (number of) unknown (at libGL compile
> > > time) extension function names to an equally unknown driver
> > > DLL name, and load that - on query? On glGetProcAddress?
> >
> > Query GL_EXTENSIONS string to identify supported extensions, then
> > query entry points corresponding to extensions used by the app.
>
> Hmm - I guess we frown upon people ignoring GL_EXTENSIONS and
> just checking for the NULL return from glGetFuncAddress to
> indicate the absence of an extension?
>
> If we do approve of that then we must say the glGetFuncAddress
> is REQUIRED to return NULL for unsupported extensions - or
> else I think there should be some words in the document to
> SAY that you are still required to check GL_EXTENSIONS.
>
> Painful though it is, I have to say that you should continue
> to check GL_EXTENSIONS. The existance of a function in the
> driver (which could be a function that exists but doesn't
> work on your rev of board or something) - doesn't guarantee
> it's a usable function.
Actually consider the GLX case, where the library has for example
support for a specific extension, while the X-Server doesn't.
- Thomas
--
Thomas Roell /\ An imperfect plan executed violently
Xi Graphics / \/\ _ is far superior to a perfect plan.
[EMAIL PROTECTED] / / \ \
/ Oelch! \ \ George Patton
