Re: [Dri-devel] Question about MGA driver texturing patch

[Karl L. <kle...@ma...>]

Hi there, 

I'm really disappointed about this patch story, not because it hasn't
been applied yet (I've notive that yesterday too), but because I was
indirectly told why, exactly 7 months later. If I knew, it would have
been really easy to change my patch to make it conform to the DRI
patching rules (that I didn't know, of course). But now, I'm not sure I
have time to do it, so maybe somebody else?

Karl


On Thu, 2002-01-24 at 01:17, Keith Whitwell wrote:
> Lawrence Gold wrote:
> > 
> > Hi,
> > 
> > Back in June 2001, Karl Lessard posted the following patch to the mailing
> > list:
> > 
> >         http://www.geocrawler.com/lists/3/SourceForge/680/0/5955984/
> > 
> > Are there any plans to incorporate this?  I perused the latest CVS sources
> > and wasn't able to find this in there.
> 
> I'd prefer the patch if it didn't:
> 	- make whitespace changes
> 	- make changes to the names of constants
> 	
> and then
> 	- split into two patches, one for each of the problems he's addressing
> 	- didn't use magic numbers in mmesa->multitex
> 	
> but also, people have put some work into looking at GL_DECAL on mga - and
> concluded that it isn't possible to do correctly.  Maybe a 2-stage process can
> simulate single-texture blend, but in general it doesn't look possible.
> 
> Looking at the part of the patch that does address GL_DECAL, it seems bogus on
> a number of levels:
> 	- author goes to some effort to set multitex to magic number '2', indicating
> perhaps he intends to do use 2 texunits to get singletex GL_DECAL
> 	- but then when this value is tested later on, he just disables the second
> unit?
> 	- the patch seems to just enable some code that is commented out as not
> working, makes a few changes to that which themselves look bogus (mixing TD0
> and TD1 flags in TEXDUALSTAGE0 - this may just be syntactic but it's sloppy)
> 
> The code is disabled because it doesn't work.  If you think it is correct,
> test against glean and the texenv demo at least, and report a positive result
> and a minimal, readable patch, please.
> 
> If you just want a hack to allow fast, incorrect rendering for a particular
> game, maybe controlled by an env var, post that - I'm ok with that sort of
> thing, too.
> 
> Keith
> 
> _______________________________________________
> Dri-devel mailing list
> Dri...@li...
> https://lists.sourceforge.net/lists/listinfo/dri-devel
>