Yair Kurzion (yair++at++polygon.engr.sgi.com)
Wed, 28 Apr 1999 14:01:05 -0700 (PDT)
Your problem sounds a lot like the problem Kenneth Russell reported (see other
reply). The latter was a bug in the way pfGetSets draw fluxed attributes.
The fix for this bug went into pf 2.2.5 which should start shipping in
two weeks.
If you wish to verify that the fix is good for your problem too, please
send me your sample program and I will test it against the 2.2.5 code.
-yair
> I recently decided to give pfFluxes a try and thought I understood it
> until I ran my code. It turns out when I assign a pfFlux pointer to a
> geoset for the coordinates, the geometry is not rendered as intended in
> that it appears that certain elements are missing, or some of the
> vertices have moved. I can tell that it is close to what I am supposed
> to have, but it isn't correct. My first guess was that I had a bad
> flux pointer (not pointing at the beginning of the pfVec3 array) or that
> the indexing was incorrect and the coordinates were not being
> "connected" correctly.
>
> In trying to debug this problem, I went back to experiment with the
> Performer sample "morph_engine.c" to make sure that I could follow how
> it worked. Since I wasn't trying to set the colors in my code, I
> decided to remove the use of the color_flux pointer for the geoset by
> commenting out the following line in the sample code:
>
> pfGSetAttr(gset, PFGS_COLOR4, PFGS_PER_VERTEX, (void*)color_flux, NULL);
>
> I thought that it would be ok to do that since even though the code was
> still working with the color_flux buffer, it wasn't assigned to the
> geoset so there wouldn't be any problems other than no color animation.
> When running the sample code with the only change being that line
> commented out, there is erratic flashing that looks nothing like the
> intended geometry.
>
> I then went one step further in the sample program and commented out ALL
> of the code that deals with colors so that the only flux activity left
> was for the coordinates and normals. I then discovered that I see a
> result similar to what I saw in my code. The geometry displayed looks
> like it is missing a tristrip or two and there appears to be an errant
> vertex or two. So why does the fact that I don't update the colors
> affect the geometry?
>
> I can send my modified "morph_engine.c" code but I figured that I'm
> probably missing something pretty basic that someone might be able to
> point out.
--
\_________ \_____ \__ \__ \_____ Yair Kurzion
\_________ \_____ \__ \__ \_____ yair++at++sgi.com
\__ \__ \____\__ \__ http://reality.sgi.com/yair
\__ \__ \__ Work: (650) 933-6502
\__ \__ \__ Home: (408) 226-9771
\__ \__ \__
This archive was generated by hypermail 2.0b2 on Wed Apr 28 1999 - 14:01:25 PDT