Hi,
I came up with a slightly different time step for the simulation:
dynamicsWorld->stepSimulation(btScalar(deltaT), 10, btScalar(0.015625));
That means 1/64. I think it could help reducing rounding errors as it fits better binary numbers (float expresses that value exactly as opposed to 1/60).
What do you think? Is my assumption correct? And does it do any difference at all?
Thanks.