pfFlux, fluxed geosets, fluxed geoset attributes

New Message Reply Date view Thread view Subject view Author view

From: christopher.g.dorosky++at++lmco.com
Date: 07/28/2000 09:54:14


pfHelp!!!

I have been looking at Angus' program for billboard trees. There was a
comment in there about how the geoset attributes could be fluxed. I decided
to try this, since I will have an application with a very large number of
trees.

I am assuming that the way to do this is with a pfFlux, not a fluxed geoset,
right? I worked through the fluxed_gset example in pguide, but still have no
understanding about when it is appropriate to use a fluxed geoset. Do you do
this when things about the geoset can change other than the attributes?

Here is what I tried.

I made the application run app_cull_draw. I rewrote the part that calls
UpdateBilbo, so that it was called from a post app callback (the idea being
that I could change it to post cull, etc). This worked fine.

Changing it to post cull made the scene not update. So I changed
billy->coords from a pfVec3 * to a pfFlux *. In the UpdateBilbo routine, I
changed the tmpcos (or whatever) pointer to point to
pfFluxGetWritableData(billy->coords). I allow the function to do the
coordinate changing like it did before, and do a fluxwritecomplete at the
end.

This works terribly. The trees jump all over the place and don't rotate.

Interestingly enough, if I change to APPCULLDRAW, it works fine.

What is going on here? I am very unclear about the following two things:
When is it appropriate to use a fluxed geoset?
When is it appropriate to use a pfFlux for a geoset attribute?

For the geoset attributes which are pfFluxes, does something automatically
do a pfFluxGetCurData each frame to make things update?

Thanks!

Chris

Eventually, I want to have the rotation done by the compute process.

Christopher Dorosky
Lead Electronic Systems Engineer - Real Time Simulation
Lockheed Martin Missiles and Fire Control - Dallas
christopher.g.dorosky++at++lmco.com
972-603-2349


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Fri Jul 28 2000 - 09:54:44 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.