don't render deformables for TinyRenderer/EGL if p.configureDebugVisualizer(p.COV_ENABLE_TINY_RENDERER,0)
Fix segmentation mask in tinyrenderer for deformables
1) allow to render deformables in 'getCameraImage', for TinyRenderer (tested OK) and EGL (untested)
2) allow to have textures for deformables. See deformable_ball.py, deformable_anchor.py and deformable_torus.py for examples
3) deformables: allow to request simulation mesh data (even if there is a render mesh) See deformable_anchor.py for an example usage
data = p.getMeshData(clothId, -1, flags=p.MESH_DATA_SIMULATION_MESH)
4) fix deletion of deformables, thanks to Fychuyan, https://github.com/bulletphysics/bullet3/pull/3048
5) allow to enable and disable double-sided rendering, p.changeVisualShape(objectUid, linkIndex, flags=p.VISUAL_SHAPE_DOUBLE_SIDED)
6) fix GripperGraspExample, model not found
7) Fix deformable anchor not attaching to multibody with object unique id of 0
8) Fix issue with assignment of unique ids in TinyRenderer/EGL renderer (always use broadphase uid)
9) Avoid crash/issue of simulation with pinned vertices (mass 0) in btDeformableBackwardEulerObjective::applyExplicitForce
10) Store uv/normal in btSoftBody::RenderNode to allow textured meshes
11) (uncomment in btSoftBodyHelpers.cpp): dump vertices and indices in obj wavefront format, when loading a VTK file, for quicker creation of a (textured) surface mesh
12) allow interpolateRenderMesh also for old position-based soft bodies (not only the shiny new FEM deformables)
13) fix a few premake targets
14) update build_visual_studio_vr_pybullet_double_cmake.bat so it suits c:\python37 and installs locally
for local install of Bullet, see also this example https://github.com/erwincoumans/hello_bullet_cmake
Add inverse dynamics / mass matrix code from DeepMimic, thanks to Xue Bin (Jason) Peng
Add example how to use stable PD control for humanoid with spherical joints (see humanoidMotionCapture.py)
Fix related to TinyRenderer object transforms not updating when using collision filtering
See createObstacleCourse.py for an example use. At the moment a limit of 1024 vertices and 1024 indices.
Will be lifted once we implement the streaming version (soon).
b3InitUpdateVisualShape doesn't take textureUniqueId by default. new API b3UpdateVisualShapeTexture to change texture (-1 will clear texture)
PyBullet/BulletRobotics: allow to reset the textureUniqueId to -1, to clear a texture
PyBullet/BulletRobotics: save all texture handles
enable follow cam in other Gym locomotion environments
add testing assets for multi-material obj files -> sdf conversion.
Also use ER_NO_SEGMENTATION_MASK flag for TinyRenderer/EGL plugin renderer
A fileIO plugin can override custom file IO operations. As a small test, load files from a zipfile in memory.
Default fileIO implementation is in examples/Utils/b3BulletDefaultFileIO.h
Affects URDF, SDF, MJCF, Wavefront OBJ, STL, DAE, images.
Apply clang-format-all.sh using the _clang-format file through all the cpp/.h files.
make sure not to apply it to certain serialization structures, since some parser expects the * as part of the name, instead of type.
This commit contains no other changes aside from adding and applying clang-format-all.sh
postpone the 'loadPlugin' for static eglPlugin, so that the init and exit happen in the same thread.
When you don't call unloadPlugin, the program may crash when exiting in SHARED_MEMORY_SERVER mode.
Problem description:
When you add a body, remove it and then add a new
body, the body id may be reused by Bullet.
Because the visual shape data was not being removed when a body was
removed, requesting the visual shape data for a 'recycled' body id
results in both the visual shape data of the new body as well as the old
one.
Removing the visual shape data when the body gets removed fixes this.
Changes the btAlignedObjectArray for visual shapes to a hashmap, so
that removing is faster. Additionally, functions like getNumVisualShape
don't perform a linear search anymore.
