From: Jason Daly (jdaly++at++ist.ucf.edu)
Date: 12/12/2000 14:00:11
On Tue, 12 Dec 2000, Allan Schaffer wrote:
> On Dec 12, 2:18pm, Guillermo E. Gutierrez wrote:
> > We think something might have changed in
> > either the way z-buffers or alpha blending is handled between the two
> > OS's (performer versions?).
> > Any light shed would be greatfully appreciated.
>
> [I'm not aware of any general problem introduced between 6.5.7 &
> 6.5.9 that would cause this problem.]
>
> This might not be entirely helpful, but --
>
> I've seen certain situations where the X Visual chosen by Performer
> is one without a Z-buffer, in the cases I saw it was due to the
> combination of the current display format, number of RM boards, and
> the requested framebuffer configuration. The details are hazy..
>
> Performer implements an iterative process for choosing a visual,
> falling back to less-demanding configurations if the better ones
> aren't available. So, pay particular attention to the visual that's
> being chosen, and make sure it's one that has a Z-buffer. Likewise,
> you might compare the number of RM boards and 'current' display
> format running on the systems being looked at; make sure that's not
> the operative difference between these two systems running 6.5.7 &
> 6.5.9.
Sorry, I deleted the original question before reading the details, so
this may or may not be relevant.
At our I/ITSEC booth we tried our software on an Onyx2 IR2 (Deskside,
so 2 RM's, right?) with Performer 2.2.9 installed, and the rendered scene
looked as if the Z-buffer wasn't being used. Originally, this software
ran on an Onyx RE2 (4 RM's, IRIX 6.5.6, Performer 2.2.7/IrisGL). When we
recompiled the code with Performer 2.2.9 and OpenGL, it worked fine. We
assumed it was an emulation issue (the Onyx2 emulating IrisGL), but maybe
not...
I don't know if this info helps anyone, but I thought I'd mention it :-)
--"J"
"I'm a castaway stranded in a desolate land,
I can see the footprints in the virtual sand."
--Neil Peart
This archive was generated by hypermail 2b29 : Tue Dec 12 2000 - 14:00:47 PST