changes are backwards compatible (btBulletWorldImporter can load old .bullet files)
but not forwards compatible (constraints in new .bullet files are ignored/unrecognized by old Bullet SDK)
This commit is for Issue 734. Some more work needs to be done for btGImpactMeshShapeDoubleData and thus btStridingMeshInterfaceDoubleData and btMeshPartDoubleData
fix vehicle demo (heightfield is broken, when using very large scaling factor)
add btConeShape serialization, see Issue 725
add btGearConstraint get/set, see Issue 685 and Issue 671
rayResult.m_flags |= btTriangleRaycastCallback::kF_UseGjkConvexRaytest;
See Demos/RaytestDemo/RaytestDemo.cpp for usage
This fixes issue 146, longest outstanding from 2008.
with hash pair caching and tandem tree traversal. It should be much faster, when using many child shapes for both compounds.
Fix iOS compilation, added header.
Fixes Issue 710
error C2084: function 'XMVECTOR operator +(FXMVECTOR,FXMVECTOR)' already has a body
error C2084: function 'XMVECTOR operator -(FXMVECTOR,FXMVECTOR)' already has a body
error C2084: function 'XMVECTOR operator *(FXMVECTOR,FXMVECTOR)' already has a body
Enable btGImpact raycast optimization, thanks to C0DEFACE, see Issue 664
Cull triangle/AABB for concave/heightfield shapes, thanks to Danny Chapman
fix btGetCpuCapabilities, thanks to Ian Ollman! See https://code.google.com/p/bullet/issues/detail?id=738
(converted raw data to .cpp so it doesn't need to be loaded from disk)
Some optimizations for btHeightfieldShape, thanks to Danny Chapman
Quick check on AABB (was not done) and reduce calls to 'getVertex'
fix some crash in btSoftBody, related to running out-of-memory. You can configure the default maximum memory allocation for the signed distance field using worldInfo->m_sparsesdf.Initialize(hash, maxAllocation);
fix 'exploding' btSoftBody, related to very small masses. You can configure maximum displacement per frame using worldInfo->m_maxDisplacement
avoid some crash in the world importer
In some degenerate cases the contact normal is pointing the wrong direction
so fix it now (until we can deal with all degenerate cases in GJK and EPA)
contact normals need to point from B to A in all cases, so we can simply check if the contact normal really points from B to A
We like to use a dot product of the normal against the difference of the centroids,
once the centroid is available in the API
until then we use the center of the aabb to approximate the centroid
The idea is to create tetrahedra from each surface triangle, by moving its centroid inwards along the collision normal using depth.
If the centroid hits an internal triangle, it will stop there.
The commit also includes part of the gimpact raycast performance improvement, except for the actual btCollisionWorld part
(that file was out-of-date and needs manual patching)
after 'loadFileFromMemory' you can save the existing memory back to disk converted to native architecture endianness.
Fixed a bug that prevented big endian machines to load cloth data (due to wrong endian swap of pointer array)
Also improved the 'preswap' so that you can swap endianness on big endian machines (to little endian)
