[ut2004] [OpenGLDrv.OpenGLRenderDevice]

[Dean Hamstead]

i resent that, given the huge amount of effort i have put in
trying to isolate for them when is going wrong

installing ia32 then back to amd64
installing fedora core, then going back to debian
adding a serial console (but board doesnt have legacy ports) to get
kernel dumps

up and down, up and down

bios updates etc etc

untill i stumble across what would seem to be the actual problem
and then everything suddenly goes quiet....

sounds dangerously like 'woops we dont actually know how to fix that'
and pointing the finger at 'bugs in kernel api' or 'changes in the 
kernel' is so much nicer of a reason for nvidia developers because that 
makes them look good and the evil open source kernel developers look bad.

hehe ;)

fact is, its probing agp details and its crashing. it knows its pcie,
so why cant it just ignore the calls or return some arbitrary value.
heck, could even return an equivalent value.

Dean

Aaron Gyes wrote:
>> I just noticed a post on the nvnews forums from someone mentioning the
>> same thing only to have his thread locked.
> 
> I've seen the nvnews whiners. They'll whine and whine and whine, bump
> their thread twenty times, mail developers (whom are not support
> people), etc. Linux support wasn't on the box your NVIDIA product came
> in.
> 
> Aaron Gyes
> 
>