These flexes use only 24 DOFs (3 per vertex of the bounding box), while colliding with the full high resolution mesh.
On an 8x8x8 cube, the performance using DOFs at all vertices is
```
Simulation time : 18.74 s
Steps per second : 533
Realtime factor : 0.53 x
Time per step : 1874.4 µs
Contacts per step : 114.88
Constraints per step : 3322.51
Degrees of freedom : 1536
```
With the new implementation, it is the following:
```
Simulation time : 1.82 s
Steps per second : 5507
Realtime factor : 5.51 x
Time per step : 181.6 µs
Contacts per step : 38.84
Constraints per step : 155.36
Degrees of freedom : 24
```
PiperOrigin-RevId: 721008829
Change-Id: I833df027527db578d86667cc4b24295bcf6f7d22
- `qLDiagSqrtInv` is only required for the dual solvers. It is now computed as-needed rather than unconditionally.
- `mj_solveM2` now requires a new input array `sqrtInvD` which contains the square root of the inverse diagonal D (formerly saved in `qLDiagSqrtInv`).
PiperOrigin-RevId: 710805133
Change-Id: I0622d6a8da3882916824e9c10bad9223c122c321
These functions names and argument ordering are more consistent with the rest of the API.
PiperOrigin-RevId: 643788290
Change-Id: I783eda8021b80b82098e23ed95669b102bb82508
Orthographic cameras are specified by setting the `orthographic` attribute of the `<camera>` element. The `fovy` attribute is still used to specify the field-of-view, but its semantic is different for orthographic cameras. For orthographic cameras, the field-of-view is expressed in units of length, rather than degrees.
Other related changes:
* Fix bug in the ordering of `cam_xxx` elements in `mjModel`.
* Make camera visualization translucent only when the frustum is visualized.
* Added a button to `simulate` to toggle between perspective and orthographic free cameras.
https://youtu.be/ZXBTEIDWHhs
PiperOrigin-RevId: 642293435
Change-Id: Id090a421ad88ab404b5b27ddbbd6bfc81ad49bc5
These bodies have `body_invweight0 == 0` (infinite mass) and no degrees of freedom (identically 0 Jacobians). This was breaking some assumptions of the perturbation code which assumed perturbation applied only to moving bodies and led to very large forces/torques being applied. While these huge forces had no effect on the dynamics, they were reported by sensors, leading to unreasonable sensor readings.
PiperOrigin-RevId: 614398354
Change-Id: If5efe1c9e57e2be28af33d109679db8b3129da7f
Also:
- Renamed recently introduced `mjvPerturb` member `reflocalpos` -> `refselpos` and corrected docstring.
- Added `mjvPerturb.localmass`: body spatial inertia at selection point.
- Removed `const` qualifier to `mjData` argument of `mjv_initPerturb()`, to facilitate spatial inertia computation.
- Improved `mj_Object{Velocity, Acceleration}` docstrings, clarifying that rotational velocities precede linear velocities.
- Removed redundant const qualifiers from non pointer arguments to `mju_dot` and `mju_writeNumBytes`.
PiperOrigin-RevId: 511002168
Change-Id: I27400f497947f330072d4abb7c9bf7e71e6c5d49
Since fdbbc8bbe0, mju_mulQuat allows in-place computations.
There were a few places where the result was unnecessarily assigned to a temporary variable.
PiperOrigin-RevId: 477746005
Change-Id: I7351ef4796e7d70a4a54b930037bfaadee781f12
Yuval Tassa