On 26-May-99 Keith Whitwell wrote:
> Brian Paul wrote:
>>
>> Forwarding another reported bug in lighting...
>>
>> Joe Waters ([EMAIL PROTECTED]) wrote:
>>
>> > I just grabbed 3.1b2 and gave it a shot with BFRIS --
>> > unfortunately,
>> > somewhere betwixt a recent CVS snapshot I've got and 3.1b2, the
>> > lighting broke :-( BFRIS uses OpenGL lighting (both point and
>> > directional sources) and both worked with my most recent CVS build
>> > but
>> > with 3.1b2, it looks like all lighting is off. I'm still
>> > something of
>> > a CVS newbie, and I don't know how to tell when my last CVS update
>> > was... I pulled an ls -l under MesaCVS/CVS and it shows "Entries"
>> > with
>> > the date of May 14 17:19, which I think is about right for the
>> > last
>> > time I did a cvs update.
>> > ...
>> > Anyway, at least for BFRIS, the may 14 CVS build runs very well;
>> > there's a considerable improvement in geometry performance (I
>> > assume
>> > due to the vertex array and display list changes). Mesa 3.1b2
>> > runs
>> > speedily too, the only caveat being that everything is really
>> > dark...
>> >
>
> Joe - I've just committed a fix for lighting in the presence of
> display
> lists, gl_normalize and modelview matrices involving a scale
> component.
> If this is you, you may want to check if you problem is solved.
The latest CVS build did indeed fix the problems that I saw with the
release version of 3.1beta2. BFRIS uses (mostly) display lists and
occasional interleaved vertex arrays, sometimes with glScalef() and
either GL_NORMALIZE or GL_RESCALE_NORMALS for lighting. Both seem to
work properly with the latest CVS build.
Joe
---
Joe Waters ([EMAIL PROTECTED]) Aegis Simulation Technologies
Shoot stuff. Break things. Impress your friends.
We can help. Play BFRIS. http://www.aegistech.com
_______________________________________________
Mesa-dev maillist - [EMAIL PROTECTED]
http://lists.mesa3d.org/mailman/listinfo/mesa-dev
