Daryll Strauss wrote:
>
> On Thu, May 31, 2001 at 03:39:08AM -0400, Mike A. Harris wrote:
> > Glide is dead as a developmental library for 3D and is really
> > only useful as an in between for DRI.
> >
> > Would it make sense to make Glide3 a native part of XFree86? I
> > think it would simplify bug reporting and bug fixing, etc.
> >
> > Are there any licensing problems, or other reasons that preclude
> > this from occuring? We include it in our XFree86 package, but
> > I'd just as soon see it natively added to XFree86 and made part
> > of the Imake build system. It is unlikely to be used or
> > developed much outside of XFree86 anyway IMHO.
> >
> > Just curious if others are feeling this also, and to find out if
> > it is something that might happen or be feasible for the trunk
> > code or DRI project.
>
> It's a big ugly hairy mess that no one is supporting. How does moving it
> into XFree help?
>
> What needs to happen is that Glide and X need to be separated not
> integrated. There's an easy way to do it, and a better way. The easy way
> is to make the driver dynamically load Glide and make the X tree include
> only those header files it needs to compile the 3D driver. That's work
> we're already doing.
I've got the dlopen() stuff working but haven't put the glide headers
into the XFree86/DRI tree yet. Any opinions on where they should go?
I think either xc/lib/GL/mesa/src/drv/tdfx/ or xc/extras/glide/include/
are the best candidates. Comments?
Putting the Glide headers into the XFree86 tree should finally
eliminate tdfx compilation problems.
-Brian
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/dri-devel
