Real-Time Physics Simulation Forum

Check Erwin's reply in this thread http://bulletphysics.org/Bullet/phpBB3/ ... f=9&t=4737

While it's not exactly what Erwin says I think it means both shapes in a collision query can't be btBvhTriangleMeshShape.

btQuaternion::setEuler() expects radians as far as I know. Also if your program run at a high framerate adding 2 degrees each time might still be very fast, I suggest you multiply with the deltatime to get the rotation to be frame rate independent. Lastly is that a dynamic rigid body? It's not a goo...

You seem to set all child shapes at the same position:

T.setOrigin(btVector3(0,0,Radius));

So as far as I can see it's normal that you get the same origin for all child shapes.

You can only use TriangleMesh shapes with static bodies. It's not supported for dynamic bodies.

Yes but I'm not sure. I know that triangle meshes aren't supported for dynamic bodies. Therefore there's no case where bullet need to do triangle mesh vs. triangle mesh collision which makes me think it's not supported/not implemented.

I don't think mesh vs. mesh collision is supported if that's what you're doing.

I think you could just do this to get what you want. btTransform _btTransform; _btTransform.setIdentity(); _btTransform.setFromOpenGLMatrix( _SIO2object->_SIO2transform->mat ); btVector3 lightPosition( lightposx, lightposy, lightposz); btVector3 worldLightPosition = _btTransform.getRotation() * ligh...

From btBroadphasePair you can get the two btCollisionObject in the pair. btCollisionObject's can have a user pointer ( setUserPointer/getUserPointer) which can either contain straight up flags for if something custom should be done for the collision or simply point to your own class which can handle...

Check btCollisionDispatcher::setNearCallback()

Maybe you could try use the character controller and let it solve the collision for you, it probably a fairly good match to what you try to do. Another approach could be to use convexSweepTest and move the sphere to where it start collide instead of first moving and then compensate which it sounds l...

I think it could be margins, just try with a 1.0 and 5.0 unit box combo instead.

So my question is if there is any info about how hard it would be to port the current math aspects of bullet (which I guess are 90% of the source code) to XDK. Is there any person here with experience There should be nothing to port for the maths, just compile the bullet math lib and it should work...