bullet - Project Hosting on Google Code

Today 2 hours ago issue 323 (Inconsistent color values passed into btIDebugDraw::drawLine) Owner changed by erwin.coumans - Thanks for the patch, I'll look into applying it before 2.76 release. Owner: erwin.coumans Thanks for the patch, I'll look into applying it before 2.76 release. Owner: erwin.coumans 7 hours ago issue 324 (Adding parameter to addRigidBody to not set the gravity) commented on by erwin.coumans - We could use such optional argument (applyGravity), but you still have the problem that if you call world->setGravity, it will mess with the gravity of all rigid bodies. A flag for each body (ignoreWorldGravity) would solve this. We could use such optional argument (applyGravity), but you still have the problem that if you call world->setGravity, it will mess with the gravity of all rigid bodies. A flag for each body (ignoreWorldGravity) would solve this. Last 7 days Dec 30, 2009 issue 324 (Adding parameter to addRigidBody to not set the gravity) reported by rol...@rptd.ch - Based on the topic http://www.bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4496&start=0 . The method btDiscreteDynamicsWorld::addRigidBody sets the gravity while adding a rigid body. If the application in question though uses local gravity this gets overwritten by the gravity set in the world. This in turn requires tracking addRigidBody calls to reset the gravity. Better would be to not alter the gravity of the rigid body at all ( which in turn would make this call one without side effects which is desirable ). Proposed solution by S.Lundmark to add a parameter changing the method to look like this: void btDiscreteDynamicsWorld::addRigidBody(btRigidBody* body, bool applyGravity){ if( ! body->isStaticOrKinematicObject() && applyGravity ){ body->setGravity( m_gravity ); } (...) Based on the topic http://www.bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4496&start=0 . The method btDiscreteDynamicsWorld::addRigidBody sets the gravity while adding a rigid body. If the application in question though uses local gravity this gets overwritten by the gravity set in the world. This in turn requires tracking addRigidBody calls to reset the gravity. Better would be to not alter the gravity of the rigid body at all ( which in turn would make this call one without side effects which is desirable ). Proposed solution by S.Lundmark to add a parameter changing the method to look like this: void btDiscreteDynamicsWorld::addRigidBody(btRigidBody* body, bool applyGravity){ if( ! body->isStaticOrKinematicObject() && applyGravity ){ body->setGravity( m_gravity ); } (...) Dec 29, 2009 issue 323 (Inconsistent color values passed into btIDebugDraw::drawLine) reported by Alex.Silverman - I noticed that the color values used when drawing the object simplex are 0-1, but the color values used when drawing the actual geometry are 0-255. The patch attached simply makes the simplex values conform, as I thought there was less harm in assuming users had correctly rendering geometry and an incorrectly rendering simplex, rather than vice versa. - Alex I noticed that the color values used when drawing the object simplex are 0-1, but the color values used when drawing the actual geometry are 0-255. The patch attached simply makes the simplex values conform, as I thought there was less harm in assuming users had correctly rendering geometry and an incorrectly rendering simplex, rather than vice versa. - Alex Last 30 days Dec 24, 2009 r1872 (Various nail constraint improvements TODO: errors while sce...) committed by rponom - Various nail constraint improvements TODO: errors while scene loading / saving Various nail constraint improvements TODO: errors while scene loading / saving Dec 24, 2009 issue 90 (Glue / Fracture rigid bodies (based on btCompoundShape)) commented on by erwin.coumans - We need some way of calculating connectivity, so you don't have floating or disconnectd compounds. A method that precalculates contact points between all child shapes can do this automatically. We need some way of calculating connectivity, so you don't have floating or disconnectd compounds. A method that precalculates contact points between all child shapes can do this automatically. Dec 24, 2009 issue 90 (Glue / Fracture rigid bodies (based on btCompoundShape)) commented on by erwin.coumans - See http://bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=17&t=4498 And attached files. See http://bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=17&t=4498 And attached files. Dec 23, 2009 bullet-trunk.zip (Bullet Physics SDK Trunk SVN revision r1871) file uploaded by erwin.coumans - Labels: Type-Source OpSys-All Featured Labels: Type-Source OpSys-All Featured Dec 23, 2009 issue 311 (btCollisionWorld::rayTestSingle() does not support soft body) commented on by benoit.b...@online.be - Here is the patch. Thanks for looking into it. Here is the patch. Thanks for looking into it. Dec 23, 2009 issue 311 (btCollisionWorld::rayTestSingle() does not support soft body) commented on by erwin.coumans - I was planning to look into a solution like that (virtual btCollisionWorld::rayTest), but without code duplication. If you have the patch, please attach it. I'll spend some time avoiding the code duplication. Thanks a lot for your help and feedback! Erwin I was planning to look into a solution like that (virtual btCollisionWorld::rayTest), but without code duplication. If you have the patch, please attach it. I'll spend some time avoiding the code duplication. Thanks a lot for your help and feedback! Erwin Dec 23, 2009 issue 311 (btCollisionWorld::rayTestSingle() does not support soft body) commented on by benoit.b...@online.be - hmm, I had also prepared another patch where I made btCollisionWorld::rayTest() virtual with a different implementation in btSoftRigidDynamicWorld but it wasn't so nice: lots of code duplication (the callback is nearly the same) I can provide you with this patch if you want. hmm, I had also prepared another patch where I made btCollisionWorld::rayTest() virtual with a different implementation in btSoftRigidDynamicWorld but it wasn't so nice: lots of code duplication (the callback is nearly the same) I can provide you with this patch if you want. Dec 23, 2009 issue 311 (btCollisionWorld::rayTestSingle() does not support soft body) Labels changed by erwin.coumans - Thanks a lot for all the effort, and I'm sorry this is not so easy: BulletCollision/CollisionShapes has no dependency on BulletCollision/CollisionDispatch, so we can't pass btCollisionObject around. I will fix this issue before 2.76 release, preparing for this in the upcoming weeks. Labels: Milestone-2.76 Thanks a lot for all the effort, and I'm sorry this is not so easy: BulletCollision/CollisionShapes has no dependency on BulletCollision/CollisionDispatch, so we can't pass btCollisionObject around. I will fix this issue before 2.76 release, preparing for this in the upcoming weeks. Labels: Milestone-2.76 Dec 23, 2009 issue 311 (btCollisionWorld::rayTestSingle() does not support soft body) commented on by benoit.b...@online.be - Ok, I've reworked the patch by adding a virtual btCollisionShape::rayTest() and dispatching the code that was originally in btCollisionWorld::rayTestSingle() to the various shapes. Let me know that works for you. Ok, I've reworked the patch by adding a virtual btCollisionShape::rayTest() and dispatching the code that was originally in btCollisionWorld::rayTestSingle() to the various shapes. Let me know that works for you. Dec 23, 2009 issue 311 (btCollisionWorld::rayTestSingle() does not support soft body) commented on by benoit.b...@online.be - Ok, I've reworked the patch by adding a virtual btCollisionShape::rayTest() and distributing the code that was originally in btCollisionWorld::rayTestSingle() to the various type of collision shape. Let me know if that works for you. Ok, I've reworked the patch by adding a virtual btCollisionShape::rayTest() and distributing the code that was originally in btCollisionWorld::rayTestSingle() to the various type of collision shape. Let me know if that works for you. Dec 22, 2009 r1871 (added new msvc 9 projectfiles, generated using cmake (experi...) committed by erwin.coumans - added new msvc 9 projectfiles, generated using cmake (experimental) added new msvc 9 projectfiles, generated using cmake (experimental) Dec 22, 2009 r1870 (test Win32DemoApplication instead of GlutDemoApplication) committed by erwin.coumans - test Win32DemoApplication instead of GlutDemoApplication test Win32DemoApplication instead of GlutDemoApplication Dec 22, 2009 r1869 (add link to src/OpenGL) committed by erwin.coumans - add link to src/OpenGL add link to src/OpenGL Dec 22, 2009 r1868 (fixes to make VC6 compiler (and its users) happy ) committed by erwin.coumans - fixes to make VC6 compiler (and its users) happy fixes to make VC6 compiler (and its users) happy Dec 22, 2009 issue 322 (internal issue with convex cast versus btQuanzedBVH, related...) commented on by erwin.coumans - See http://code.google.com/p/bullet/source/detail?r=1867 See http://code.google.com/p/bullet/source/detail?r=1867 Dec 22, 2009 issue 322 (internal issue with convex cast versus btQuanzedBVH, related...) Status changed by erwin.coumans - Has been fixed in latest trunk. Reproduction case is attached. Status: Fixed Has been fixed in latest trunk. Reproduction case is attached. Status: Fixed Dec 22, 2009 r1867 (bugfix in aabb expansion, causing the convexSweepTest to fai...) committed by erwin.coumans - bugfix in aabb expansion, causing the convexSweepTest to fail in certain cases (when the AABB is not symmetric around the origin) Thanks to fishboy82 for the report and fix: See http://bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4484 http://code.google.com/p/bullet/issues/detail?id=322 bugfix in aabb expansion, causing the convexSweepTest to fail in certain cases (when the AABB is not symmetric around the origin) Thanks to fishboy82 for the report and fix: See http://bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4484 http://code.google.com/p/bullet/issues/detail?id=322 Dec 22, 2009 issue 322 (internal issue with convex cast versus btQuanzedBVH, related...) reported by erwin.coumans - It is the minkowski difference A-B, so the bounds need to be subtracted: bounds[0] = rootNode->m_aabbMinOrg; bounds[1] = rootNode->m_aabbMaxOrg; /* Add box cast extents / bounds[0] -= aabbMax; bounds[1] -= aabbMin; See http://bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4484 It is the minkowski difference A-B, so the bounds need to be subtracted: bounds[0] = rootNode->m_aabbMinOrg; bounds[1] = rootNode->m_aabbMaxOrg; / Add box cast extents / bounds[0] -= aabbMax; bounds[1] -= aabbMin; See http://bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4484 Dec 21, 2009 issue 306 (COLLADA loader crash) commented on by erwin.coumans - What tools are you using? If you are using Blender, there is already a parser that can extract all information from a .blend file, including Bullet physics data. See http://gamekit.googlecode.com If you are using Maya, Max etc. you could use FBX and upcoming .bullet format, and find the matching graphics geometry <-> physics geometry by node name. What tools are you using? If you are using Blender, there is already a parser that can extract all information from a .blend file, including Bullet physics data. See http://gamekit.googlecode.com If you are using Maya, Max etc. you could use FBX and upcoming .bullet format, and find the matching graphics geometry <-> physics geometry by node name. Dec 21, 2009 r1866 (removed obsolete tree) committed by erwin.coumans - removed obsolete tree removed obsolete tree Dec 21, 2009 issue 320 (Unicode bug in Win32ThreadSupport::startThreads()) Status changed by erwin.coumans - Ok, it has been fixed in latest trunk: http://code.google.com/p/bullet/source/detail?r=1865 Thanks for the report+fix! Erwin Status: Fixed Ok, it has been fixed in latest trunk: http://code.google.com/p/bullet/source/detail?r=1865 Thanks for the report+fix! Erwin Status: Fixed Dec 21, 2009 r1865 (Use CreateEventA instead of CreateEvent to avoid breaking un...) committed by erwin.coumans - Use CreateEventA instead of CreateEvent to avoid breaking unicode builds. Thanks tom@sickheadgames.com for the report/fix, see Issue 320 Use CreateEventA instead of CreateEvent to avoid breaking unicode builds. Thanks tom@sickheadgames.com for the report/fix, see Issue 320 Dec 21, 2009 issue 268 ([PATCH] Fix M_PI redefinition warning in ConstraintDemo.cpp) Status changed by erwin.coumans - It has been fixed in latest trunk: http://code.google.com/p/bullet/source/detail?r=1864 Thanks for the report! Status: Fixed It has been fixed in latest trunk: http://code.google.com/p/bullet/source/detail?r=1864 Thanks for the report! Status: Fixed Dec 21, 2009 r1864 (Don't re-define M_PI but use SIMD_PI instead Thanks philiplo...) committed by erwin.coumans - Don't re-define M_PI but use SIMD_PI instead Thanks philiplowman for the report, See Issue 268 Don't re-define M_PI but use SIMD_PI instead Thanks philiplowman for the report, See Issue 268 Dec 21, 2009 issue 255 (btConvexConvexAlgorithm optimization.) Status changed by erwin.coumans - it has been fixed in this commit: http://code.google.com/p/bullet/source/detail?spec=svn1863&r=1839 Thanks for the report! Erwin Status: Fixed it has been fixed in this commit: http://code.google.com/p/bullet/source/detail?spec=svn1863&r=1839 Thanks for the report! Erwin Status: Fixed Dec 21, 2009 issue 100 (Add btTransform::operator(btQuaternion) method) Status changed by erwin.coumans - Thanks for the report and heads up, let's close it then! Status: Fixed Thanks for the report and heads up, let's close it then! Status: Fixed Dec 21, 2009 issue 237 (btQuaternion::angle returns nan in some (rare) cases (maybe ...) Status changed by erwin.coumans - Status: Fixed Status: Fixed Dec 21, 2009 issue 140 (btQuaternion angle and slerp do not deal with redundancy in ...) Status changed by erwin.coumans - ok, with clamping in place we removed the assert. http://code.google.com/p/bullet/source/detail?r=1863 Thanks for the report. Erwin Status: Fixed ok, with clamping in place we removed the assert. http://code.google.com/p/bullet/source/detail?r=1863 Thanks for the report. Erwin Status: Fixed Dec 21, 2009 r1863 (removed normalize (btAcos has clamping) and removed assert i...) committed by erwin.coumans - removed normalize (btAcos has clamping) and removed assert in btAcos. removed normalize (btAcos has clamping) and removed assert in btAcos. Dec 21, 2009 issue 266 ([PATCH] CharacterDemo.cpp causes gcc warnings re. constness ...) Status changed by erwin.coumans - We don't introduce std::string or other STL in Bullet. We look into removing warning before next release. Thanks for the report. Status: Fixed We don't introduce std::string or other STL in Bullet. We look into removing warning before next release. Thanks for the report. Status: Fixed Dec 21, 2009 issue 283 (Seperate different products into seperate releases e.g. Bull...) commented on by erwin.coumans - Once we have binary serialization, we probably move BulletColladaConverter to a separate project/tree. Once we have binary serialization, we probably move BulletColladaConverter to a separate project/tree. Dec 21, 2009 issue 254 (Dynamica Maya Plugin fails to load on Linux x86_64) Status changed by erwin.coumans - Thanks for the heads up! Status: Fixed Thanks for the heads up! Status: Fixed Dec 21, 2009 issue 203 (maya linux plugin) commented on by erwin.coumans - cmake generates the libraries in <relative_library_folder>/Debug and <relative_library_folder>/release src/LinearMath/Debug/libLinearMath.a src/BulletCollision/Debug/libBulletCollision.a src/BulletDynamics/Debug/libBulletDynamics.a Note that is it possible to build out-of-source tree using cmake, so above path are relative to where you run cmake. cmake generates the libraries in <relative_library_folder>/Debug and <relative_library_folder>/release src/LinearMath/Debug/libLinearMath.a src/BulletCollision/Debug/libBulletCollision.a src/BulletDynamics/Debug/libBulletDynamics.a Note that is it possible to build out-of-source tree using cmake, so above path are relative to where you run cmake. Dec 21, 2009 issue 306 (COLLADA loader crash) commented on by jan.ciger - Thanks! The native format is good news. The main issue for me was the availability of exporters - COLLADA allows exporting both the geometry to be loaded by the graphic engine and also the physics at the same time. This is essential for complex models, where there is separate geometry for the visualization and simplified set of shapes for the physics simulation and they need to be kept consistent. COLLADA exporters and tools leave things to be desired, though. Thanks! The native format is good news. The main issue for me was the availability of exporters - COLLADA allows exporting both the geometry to be loaded by the graphic engine and also the physics at the same time. This is essential for complex models, where there is separate geometry for the visualization and simplified set of shapes for the physics simulation and they need to be kept consistent. COLLADA exporters and tools leave things to be desired, though. Dec 21, 2009 issue 272 (Feature Request: Generate geometric decals) Status changed by erwin.coumans - Status: WontFix Status: WontFix Dec 21, 2009 issue 237 (btQuaternion::angle returns nan in some (rare) cases (maybe ...) commented on by erwin.coumans - Commit 1806 should fix this issue, it clamps the range to [-1,1] inside btAcos http://code.google.com/p/bullet/source/detail?spec=svn1862&r=1806# Can you double-check latest trunk? Commit 1806 should fix this issue, it clamps the range to [-1,1] inside btAcos http://code.google.com/p/bullet/source/detail?spec=svn1862&r=1806# Can you double-check latest trunk? Dec 21, 2009 issue 145 (add isZero method to btVector3) Status changed by erwin.coumans - Thanks for the suggestion, it has been added now: http://code.google.com/p/bullet/source/detail?r=1862 Status: Fixed Thanks for the suggestion, it has been added now: http://code.google.com/p/bullet/source/detail?r=1862 Status: Fixed Dec 21, 2009 r1862 (added isZero and fuzzyZero to btVector3 Thanks jazztickets f...) committed by erwin.coumans - added isZero and fuzzyZero to btVector3 Thanks jazztickets for the report, see Issue 145 added isZero and fuzzyZero to btVector3 Thanks jazztickets for the report, see Issue 145 Dec 21, 2009 issue 61 (btCollisionWorld::convexSweepTest fails on convex meshes whe...) Status changed by erwin.coumans - Thanks for the report Status: Fixed Thanks for the report Status: Fixed Dec 21, 2009 issue 281 (use acceleration structures for convexSweepCompound and rayT...) changed by erwin.coumans - Status: Duplicate Status: Duplicate Dec 21, 2009 issue 25 (improve performance of raycast versus compound shapes) commented on by erwin.coumans - Issue 281 has been merged into this issue. Issue 281 has been merged into this issue. Dec 21, 2009 issue 297 (Add SPH fluid iso-surface generation) Labels changed by erwin.coumans - Labels: Milestone-3.x Labels: Milestone-3.x Dec 21, 2009 issue 296 (Add SPH fluid interaction with rigid body / cloth simulation) Labels changed by erwin.coumans - Labels: Milestone-3.x Labels: Milestone-3.x Dec 21, 2009 issue 290 (Vehicle maxSuspension force.) Status changed by erwin.coumans - Thanks for reporting this, it has been applied now: http://code.google.com/p/bullet/source/detail?r=1861 Status: Fixed Thanks for reporting this, it has been applied now: http://code.google.com/p/bullet/source/detail?r=1861 Status: Fixed Dec 21, 2009 r1861 ( Expose vehicle maxSuspension force (instead of hard-coded 6...) committed by erwin.coumans - Expose vehicle maxSuspension force (instead of hard-coded 6000 value). Thanks blakflow for the patch, see Issue 290 Expose vehicle maxSuspension force (instead of hard-coded 6000 value). Thanks blakflow for the patch, see Issue 290 Dec 21, 2009 issue 306 (COLLADA loader crash) Status changed by erwin.coumans - The patch has been applied, thank a lot. http://code.google.com/p/bullet/source/detail?r=1860 Note that we are going to provide a new binary physics format for Bullet (.bullet) in an upcoming version. It will be natively supported as part of the core SDK. See Issue 308 Status: Fixed The patch has been applied, thank a lot. http://code.google.com/p/bullet/source/detail?r=1860 Note that we are going to provide a new binary physics format for Bullet (.bullet) in an upcoming version. It will be natively supported as part of the core SDK. See Issue 308 Status: Fixed

issue 324 (Adding parameter to addRigidBody to not set the gravity) reported by rol...@rptd.ch - Based on the topic http://www.bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4496&start=0 . The method btDiscreteDynamicsWorld::addRigidBody sets the gravity while adding a rigid body. If the application in question though uses local gravity this gets overwritten by the gravity set in the world. This in turn requires tracking addRigidBody calls to reset the gravity. Better would be to not alter the gravity of the rigid body at all ( which in turn would make this call one without side effects which is desirable ). Proposed solution by S.Lundmark to add a parameter changing the method to look like this: void btDiscreteDynamicsWorld::addRigidBody(btRigidBody* body, bool applyGravity){ if( ! body->isStaticOrKinematicObject() && applyGravity ){ body->setGravity( m_gravity ); } (...)

Based on the topic http://www.bulletphysics.org/Bullet/phpBB3/viewtopic.php?f=9&t=4496&start=0 . The method btDiscreteDynamicsWorld::addRigidBody sets the gravity while adding a rigid body. If the application in question though uses local gravity this gets overwritten by the gravity set in the world. This in turn requires tracking addRigidBody calls to reset the gravity. Better would be to not alter the gravity of the rigid body at all ( which in turn would make this call one without side effects which is desirable ). Proposed solution by S.Lundmark to add a parameter changing the method to look like this: void btDiscreteDynamicsWorld::addRigidBody(btRigidBody* body, bool applyGravity){ if( ! body->isStaticOrKinematicObject() && applyGravity ){ body->setGravity( m_gravity ); } (...)

Today

Last 7 days

Last 30 days