> Brian and oglbase,
>
> Even including <windows.h> with WIN32_LEAN_AND_MEAN defined
> creates a large amount of name space pollution.
That is true. I usually remove windows.h..,, that si why lean_and_mean is
there;-)
Suhaib
>
> It is a tad involved, but GLUT goes through careful contortions
> to be able to include the <GL/gl.h> without including
> <windows.h>.
>
> Should a similar scheme be used for <GL/glext.h>?
>
> Ultimately, what'd I'd like is a <GL/gl.h> and <GL/glu.h>
> for Win32 that doesn't require including <windows.h>.
>It is
> not all that hard, though it is unlikely that Microsoft would
> distribute such a file. It might be nice though to have
> a <GL/gl.h> and <GL/glu.h> suitable for Win32 compilation
> that doesn't require <windows.h> namespace pollution
>
> - Mark
>
> -----Original Message-----
> From: Brian Paul [mailto:[EMAIL PROTECTED]]
> Sent: Monday, May 22, 2000 12:18 PM
> To: [EMAIL PROTECTED]
> Subject: Re: [oglbase-discuss] glext.h
>
>
> Suhaib Siddiqi wrote:
> >
> > I noticed in the new OpenGL glext.h,
> >
> > #if defined(_WIN32) && !defined(APIENTRY) && !defined(__CYGWIN32)
> > #define WIN32_LEAN_AND_MEAN 1
> > #include <windows.h>
> > #endif
> >
> > This should be
> >
> > #if defined(_WIN32) && !defined(APIENTRY) && !defined(__CYGWIN__)
> > #define WIN32_LEAN_AND_MEAN 1
> > #include <windows.h>
> > #endif
>
> Should the same thing be used in Mesa 3.3's glext.h?
>
> There's a few issues to be addressed before I switch to SGI's glext.h
>
> -Brian
