iamcsharper / bullet-xna

Automatically exported from code.google.com/p/bullet-xna
0 stars 0 forks source link

Missing GImpactMeshShape and GImpactCollisionAlgorithm #2

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago
The description says that all collision shapes are implemented, however 
GImpactMeshShape and GImpactCollisionAlgorithm are not.   

Currently, the only concave triangle mesh that bullet supports in Dynamics is 
the GImpactMeshShape.  The bvhTriangleMesh works for stationary objects.

Might want to put this in the 'todo' section.

Original issue reported on code.google.com by dcoliva...@gmail.com on 7 Apr 2011 at 5:29

GoogleCodeExporter commented 9 years ago
Thanks, must be suffering from code blindness. Will add to do list and look at 
porting.

Original comment by xexu...@gmail.com on 11 Apr 2011 at 7:55

GoogleCodeExporter commented 9 years ago
No worries again. :)   GImpact is probably another 'if you're feeling brave' 
thing.   There's a strange combination of macros, static methods, and a new set 
of types to merge.  

If you think it would help, there's a jbullet implementation of the gimpact 
types that may give you another look at the same thing.

Original comment by dcoliva...@gmail.com on 13 Apr 2011 at 12:56

GoogleCodeExporter commented 9 years ago
first pass of GImpact code added

Original comment by xexu...@gmail.com on 13 May 2011 at 3:46

GoogleCodeExporter commented 9 years ago
[deleted comment]
GoogleCodeExporter commented 9 years ago
I got the latest trunk, merged in andres.traks MonoXNA type updates, and tried 
it. There were a few compile errors where Count() was used instead of Count in 
the GImpactCollisionAlgorithm.     

Aside from that, It compiled. I do not get any collision response.    I see 
some of the methods being called but i'll have to do some more testing to see 
where it's exiting the loop prematurely.   

I did call 
GImpactMeshShape bmesh = new GImpactMeshShape(btshapeArray);
bmesh.UpdateBound();   To ensure that the AABB is updated to trigger NearPhase

And

GImpactCollisionAlgorithm.RegisterAlgorithm(m_dispatcher); to ensure that 
NearPhase uses the GImpact Algorithm for the NearPhase.

Original comment by dcoliva...@gmail.com on 29 May 2011 at 5:18

GoogleCodeExporter commented 9 years ago
Thanks. I've put off looking properly at the gimpact stuff as the initial demo 
was such a perforamance hog. Now that i've got most of the 2.78 updates in I'll 
spend some more time on it.

Original comment by xexu...@gmail.com on 29 May 2011 at 3:09

GoogleCodeExporter commented 9 years ago
I know that you're still working on it..   and I wanted to give it a try and 
report my experience.   

I tried it with the latest r55 and I can walk through GImpact objects.   Every 
so often, I'll notice a collision and the world stall for a moment.

Original comment by dcoliva...@gmail.com on 3 Jun 2011 at 12:43

GoogleCodeExporter commented 9 years ago
[deleted comment]
GoogleCodeExporter commented 9 years ago
Looks like I'm getting stuck collisions also..   where capsuleZ collides with 
GImpact and restitution forces are applied indefinately.

Original comment by dcoliva...@gmail.com on 3 Jun 2011 at 1:26

GoogleCodeExporter commented 9 years ago
It doesn't look like they collide with the HeightFieldShape correctly.    Fall 
through the ground.   Doesn't fall through a cube though.

Original comment by dcoliva...@gmail.com on 3 Jun 2011 at 6:56

GoogleCodeExporter commented 9 years ago
Thanks. I've just been updating the c++ and c# gimpact code with lots more 
exciting debug stuff <yawn> to get an idea on errors. There was a very obvious 
dumb one where the vertices and indices in the mesh weren't being read but 
thats just the first of many...

Original comment by xexu...@gmail.com on 3 Jun 2011 at 9:48