Eelco Hoogendoorn

Is the angular damping supposed to take negative values? Not sure, just shooting from the hip. Ive had similar issues, when dealing with situations where multiple contacts can be generated...

Ive ran into something similar recently; without tracing the latest master... im guessing this is because the dispatch of OneWayCollider assumes that its always the second body participating in the...

It seems like almost everything in the physics submodule is indeed specific to the physics simulator provided out of the box with brax currently; except for the math.py which provides...

Another interesting exercise in generality along the same lines would be to add support for a 2d backend as well. Having that also tends to be nice for debugability and...

Hi Daniel; Thanks for a cool project, and thanks for taking the time to reply! > Re XPBD: We had not thought about XPBD at all, though we have vague...

> Hi Eelco - thanks so much for sharing these thoughts. Super cool. Thanks for making an awesome package! > +1 to Daniel's suggestion of sharing his colab. We have...

Was planning on spending a solid block of time working on this earlier in the year, but life decided otherwise. However, just wanted to let you know this is still...

Ah awesome; ill go check it out, must be nice to get rid of stability worries. Is there a net efficiency gain, on typical benchmarks? Guess youd want to compare...

Some quick feedback; looking good! It appears in one of my test cases that the effective static friction has gone down? As far as I can tell from looking at...

Cant say ive fully digested it yet; but one oddity I noticed is that the OneWayCollider is dealing with the friction coefficient in its _position_contact method, whereas the TwoWayCollider isnt....