[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: Possible nVidia driver bug? - (code sample attached)

Stephen J Baker wrote:


So it's very much worth your while to explore every other avenue and if
you truly cannot find anything wrong at your end, make life as easy as
possible for the guys at nVidia. eg: Get this to work in a GLUT application
instead of SDL - use a *simple* Makefile and link as few other libraries as
possible, make the code be as simple as it can be and yet still exhibit
the bug, give them a concise description of the problem, the hardware you
used, the driver revisions and any information about how it behaved with
other driver revisions - explain that the problem happens under Windows
as well as Linux, try to get someone to run it on an ATI card and hopefully
be able report that it works OK there.

That way, the next time you need their help, they'll be more likely to
give it because they'll get to know you as someone who doesn't cry
'Wolf' and they'll come to respect your abilities and actually value
your bug reports as genuine contributions.

I see your point. But aside from not using GLUT, I did all of the above. I did test on ATI cards, on various other nVidia cards and on another 6800 GT (around here they're kinda hard to get by :P). I tested in Win2000 and WinXP, and on Linux. I tested using a variety of drivers and as you must have seen, the sample program that replicates the problem is *very* simple: I get a GL context, initialize some extensions and that's it.
Thing is, I did explain all this to nVidia's support and still got no answer. It's surely no problem for me to port the sample app to GLUT, but then again, I fear having to hammer their mailbox in order to get some attention.

Best regards,

Miguel A. Osorio.
SomniumStudio - www.somniumstudio.com