btaba

Hi @breakds , as a first pass, we'll probably implement something pretty similar to what is used in MuJoCo, at least for sphere-hfield, and we can vmap over the hfield...

Are you thinking to use boxes for now? Go for it! If you have a lot of boxes, consider adding these fields to your model https://github.com/google-deepmind/mujoco/blob/a8db22f0d077aee86f771808fd24ea60a148c93f/mjx/mujoco/mjx/test_data/shadow_hand/right_hand.xml#L8-L11 `max_geom_pairs` will do some...

Hi @r-aristov , the issue referenced for Humanoid was closed, Can you please provide a complete example that we can reproduce? I don't really understand what the expected behavior is...

Hi @Servo97 , have you considered doing an `mjx.get_data` and using the same pipeline you have for MuJoCo? You can also take a look at the [MJX colab tutorial](https://colab.sandbox.google.com/github/google-deepmind/mujoco/blob/main/mjx/tutorial.ipynb) which...

Hi @bheijden thanks for the bug report! In principal, MJX gets 4 contact points for any convex-primitive or convex-convex collision. In practice, it looks like box-plane picks 1 degenerate point,...

The contact point part of the fix https://github.com/google-deepmind/mujoco/commit/805d862d5feffe5a4449afc0605b990baf7aab07

Ok I played with a version of the box example at HEAD, and it looks good to me. @bheijden let us know if you still see weird behavior. XML ```XML...

Any updates @bheijden ? Is this still an issue?

Hi @cdagher , thanks for the bug report! This is a known issue, see a previous bug for more context https://github.com/google-deepmind/mujoco/issues/1182 If you can use the Newton solver and `iterations=1`...

The recommended way to do this now is: do a `mjx.get_data` or `mjx.put_data`, and use the get/set State API from MuJoCo. Does that not fit the use-case you have? Adding...