[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: GL extensions
On 12-Jan-2000 Keith Lucas wrote:
>>I was talking to some guys a while back about a game they were working on,
>>one of them said that they won't do a linux port because linux has bad opengl
>>support. After some questions, I found out that he meant it didn't have
>>extensions that he likes (mostly concerning multi-texturing, I think). Is
>>stopping gamehouse developers from working on linux ports? How is linux with
>>extensions? I haven't even looked at them... is that something we'll have to
>>wait for drivers to mature before it gets tackled?
> That would be a "erm, we need an excuse" reason.
> Windows, of course, has fantastic OpenGL support, each card having several
> iterations of drivers for it. Almost half the drivers have no obvious bugs
> some of them are even quick.
> I'd rather have the BASE openGL working at some consistent speed before
> start panicking about multi-texturing extensions.
> There is a reason Quake doesn't go near that stuff and basically does it all
> hand: because on one card it'll take orders of magnitude more time to do
> something than on another. The behaviour is unpredictable. Quake uses a very
> small set of OpenGL activities and they documented exactly what so people
> optimise the drivers. Of course, ID could be /wrong/... in fact they must be,
> because they support Linux.
> I'd rather think that people have gone: "oh.. our code isn't portable...
> was it supposed to be? We can't say it isn't.. erm... Linux doesn't work.
> that's it."
> I've seen that happen in commercial environments, I don't see why it wouldn't
> happen in games development.
that was my initial interpretation, but I felt compelled to mention it incase
it was legitimate :) sorry about wasting everyones time.
I'd rather have a good solid correct stable opengl implementation than one
that's super fast. I see windows drivers popping up 'artifacts' all the time
but ppl don't mind cuz they get an extra 2 fps. I'd rather run a bit slower and
have it RIGHT (well, I'd rather have it working faster and flawless, but...)
I've seen mesa used to demonstrate flaws in microsofts "reference" drivers
(which are supposed to be 110% correct, but not terribly fast). If drivers or
glx servers pop up with the same correctness (and open source/gpl'd), I don't
see how they'd fail to outdo the best of the best windows drivers for those
cards in terms of correctness, stability, and (mebbe) speed.
unfortunantly, the extension stuff is an excuse they can throw at us. :) If
someone somewhere had written up the extensions (or a fake interface), then
that'd be one less excuse for them...
-Erik <email@example.com> [http://math.smsu.edu/~br0ke]
The opinions expressed by me are not necessarily opinions. In all
probability, they are random rambling, and to be ignored. Failure to ignore
may result in severe boredom or confusion. Shake well before opening. Keep