On Tue, May 09, 2000 at 07:17:59AM -0500, Stephen J Baker wrote:
|
| This seems even less acceptable than the original proposal since before, even
| when "new behavior" was the default ('C'), the implementation was free to export
| function prototypes for locally valid extensions.
That would make it impossible to check for naked references to
extension functions at compile time, with the highly unpleasant
side-effect that an app could fail at startup (because not all
libGL.so implementations contain entry points for the extension
functions that the app references).
I think it's worthwhile to design oglbase so that this problem can't
occur. That's one of the reasons why the current proposals are
structured as they are.
Allen
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h akin
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Stephen J Baker
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h/gle... akin
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h... Stephen J Baker
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h/gle... Thomas Roell
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h... Stephen J Baker
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Brett Johnson
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Andy Ross
- Re: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Stephen J Baker
- RE: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Michael I Gold
- RE: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Thomas Roell
- RE: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Michael I Gold
- RE: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Michael I Gold
- RE: [oglbase-discuss] CALL FOR VOTE: gl.h/glext.h Michael I Gold
