bullet3
bullet3 copied to clipboard
Fix NaN due to division by zero
I encountered NaN when simulating with a kinematic rigid body, one dynamic rigid body, and one rope anchored between them.
It was solved by avoiding division-by-zero when computing solverConstraint.m_jacDiagABInv
.
Hello @erwincoumans ! Would you kindly review this PR? If there are any concerns or comments, I would greatly appreciate your input and am happy to address them. Thank you very much for your time and consideration!
I would also appreciate clarification on this, as I also ran into this NaN when trying to use a fixed point type for btScalar, and I'm not certain what the proper behavior is supposed to be.