Hello all, I think I'm missing something with respect to how to position kinematic rigid bodies. Right now I have a custom MotionState that gets the global transform of each body from the model (which assembles it from sensor readings and forward kinematics) and returns it. Breaking there reveals that plenty of the transforms being returned are non-identity, but yet everything displays centered at (what I presume to be) 0,0,0!
Any idea what I might be doing wrong?