Philipp Reist

Hi Niklas, The equations you have for backward Euler look correct to me, so you should be able to predict the PGS results using these equations. You could test with...

I am not sure I understand your question. A soft limit is continuous, so it does not make sense to formulate it with an LCP I think. I.e. there is...

If there are no collisions or other constraints, the solver will just do the backward euler step I mentioned above. The differences between TGS and PGS should be due to...

Hi - you should be able to just scale the local pose position of the joint and it should work. I.e. have a look at `setLocalPose` E.g. local pose position...

Thank you for the report @shaneasd! We'll have a look. Internal tracking: PX-4664

Hi @npestov9 - thank you for the report. From the stack alone this is not straightforward to debug. Any chance you can repro the issue by adapting a snippet in...

Hi @npestov9 - could you isolate the random mesh generator into a snippet? Given the same seed, you should be able to repro the crash reliably.

Internal Tracking: PX-4530

Unfortunately not, our testing suite is internal only.

Internal tracking: PX-4475