Previously, it was assumed that attaching was the last operation performed on the mjSpec, so keyframes resulted in the incorrect size. In order to prevent information loss, we now only allow the resize function to expand the keyframe array, since attach should never remove a degree of freedom.
PiperOrigin-RevId: 804917828
Change-Id: Icfacd207ced1c6aac600d514f8d9065ceff2d347
This extends both the spec and the XML definition to allow specification of a material name on a mesh. This material acts as a fallback in the case that the referencing geom element does not specify a material of it's own.
PiperOrigin-RevId: 798232378
Change-Id: Iaf300c727732744ea1b613e64334ee505ac6e209
Number of octree nodes for cow example:
Before change 45217
After change 28921
PiperOrigin-RevId: 783748606
Change-Id: I278ac7bc969dc1463d6dd0f7cdc8b267b7ce6c92
Number of octree nodes for the cow example:
Before change 57849
After change 45217
PiperOrigin-RevId: 783338447
Change-Id: I0e5178a4aa5bbe066f2a0ac1eb965b84e541dfd4
Temporary changes to the octree:
- Changed frame from mesh to geom
Before change (tolerance 1e-3):
```
Simulation time : 2.34 s
Steps per second : 4275
Realtime factor : 8.55 x
Time per step : 233.9 µs
Newton iters / step : 2.26
Contacts / step : 3.69
Constraints / step : 14.76
Degrees of freedom : 12
Dynamic memory usage : 0.7% of 14M
```
After change (6 octree levels):
```
Simulation time : 2.24 s
Steps per second : 4472
Realtime factor : 8.94 x
Time per step : 223.6 µs
Newton iters / step : 2.47
Contacts / step : 3.37
Constraints / step : 13.49
Degrees of freedom : 12
Dynamic memory usage : 0.2% of 14M
```
PiperOrigin-RevId: 781019754
Change-Id: Ib15581244dfe9e571e1c6c2cac4a101cf3d8ba3d
Saved XMLs are filled with
```
<joint type="free" limited="false" actuatorforcelimited="false"/>
```
but since these are hardcoded to "false" at compile time, there is no point in saving them to XML.
PiperOrigin-RevId: 776606516
Change-Id: Iac61c5126231f776077a8c265be732f9215fbfbb
For the MJCF -> USD plugin, and I suspect other usecases for user values, it's necessary to give ownership of the object to Mujoco. However since they get type erased, we can't clean up the data automatically for the user. Instead this allows c++ clients to provide a cleanup function with their data.
PiperOrigin-RevId: 756832010
Change-Id: I80b8e7822e1a0e399a0d19dcaa57ee69b3ccc16a
The signature now contains the necessary information to safely perform `bind`. The private UIDs are now removed.
Fixes an issue of changing signature when compiling a copy of an mjSpec.
PiperOrigin-RevId: 744670626
Change-Id: Id3c66419cf2afbe78e91bc4b37d2299f5ec00ab1
Also raise error if the frame is not found in the mjCBody copy constructor.
Fixes#2543.
PiperOrigin-RevId: 742624013
Change-Id: I4999b3165c97f8d079412214d027ca7a8dae8cb3
Previously, we stored the source `mjSpec` during a copy as a hack for having access to the compiler options, but this is not robust since we cannot guarantee that 1) the source `mjSpec` is not destroyed before we need to look up the compiler options nor 2) that the `mjSpec` was appended without a copy.
While 2) could be solved by simply handling an additional case in `mjCModel::FindSpec`, using a map also solves 1) and it is easier to understand.
PiperOrigin-RevId: 741504264
Change-Id: Iab1bfd9e61299a94fa8caf3a244c067b09d54384
This simplifies the compiler logic since an updated tree list was necessary at many stages (e.g. attach and keyframes resizing) and it will be anyway required when computing the mjSpec signature.
PiperOrigin-RevId: 740356447
Change-Id: I7f2ec25b27b8d4ca4364801c9a401c40c6d84569
Alessio Quaglino