Jason Posted June 11, 2010 Author Share Posted June 11, 2010 Personally I believe that the best way to make a SOP Solver version would be wrap a DOPnet into a SOP OTL, but also develop a DOP which would analyze the geometry for other things, like constraint hints and stuff. In that way you could keep the solver in DOPs, and have it masquerade as a SOP. Best of both worlds, right? Quote Link to comment Share on other sites More sharing options...
rony Posted June 11, 2010 Share Posted June 11, 2010 Personally I believe that the best way to make a SOP Solver version would be wrap a DOPnet into a SOP OTL, but also develop a DOP which would analyze the geometry for other things, like constraint hints and stuff. In that way you could keep the solver in DOPs, and have it masquerade as a SOP. Best of both worlds, right? I totally agree. I think the biggest challenge is to recreate the DOP network work flow into a simplified system if you want to do it in SOPs. The problem is the linear evaluation of the SOP network. The best thing in DOPs is data access and data update. So yeah it would be a challenge to make it a scalable solver that behaves as a state machine exclusively in SOPs. To give an example, we'd have to build a library for constraints, so that a constraint "knows" what to do when it needs to get updated or evaluated, unless we derive from the SIM constraint. But that may be reinventing the wheel. And that's not the only thing that would need to be implemented. Relationships, forces, fields etc. It would be great to design such a system in SOPs if we have time but that will also require the users some knowledge of the system. So yeah a DOP network in a SOP network would be the best of both worlds I think. I still haven't had a chance to update the latest plug-in that handles hinge and pin constraints and I probably won't have time for a while but I'll get to it as soon as I can. Cheers, Ron Quote Link to comment Share on other sites More sharing options...
Ratman Posted June 12, 2010 Share Posted June 12, 2010 Just to contribute something quick, what Van Aarde meant was an actual SOP level bullet solver, not a Bullet SOP Solver involving DOPs. Just felt like correcting that :-D Quote Link to comment Share on other sites More sharing options...
com48 Posted June 12, 2010 Share Posted June 12, 2010 Just to contribute something quick, what Van Aarde meant was an actual SOP level bullet solver, not a Bullet SOP Solver involving DOPs. Just felt like correcting that :-D Quote Link to comment Share on other sites More sharing options...
goldleaf Posted June 12, 2010 Share Posted June 12, 2010 (edited) Yeah, the SOP version sounds way cool. Can't wait to see what happens with that; it would be cool to contribute to it. BTW, I put more of seth_ro18's changes into the branch at http://gitorious.org/~crydalch/bullet-physics-solver/rydalch-bullet-physics-solver. Biggest additions are deformable static objects (called kinematic objects in bullet, see this video), impacts data, and good multisphere support. I'm updating a gitorious wiki page here with my notes, to-dos, etc if anyone's interested; I'll get the Bullet page on the OdWiki updated as well if anyone thinks that would help. Seth also made lots of notes and comments in the code, so it should be explanatory to anyone wanting to work on it, or to anyone new to all of this trying to figure out how it all works. Enjoy! Edited June 12, 2010 by goldleaf Quote Link to comment Share on other sites More sharing options...
com48 Posted June 13, 2010 Share Posted June 13, 2010 thats awesome goldleaf! going to be checking that out tomorrow! Quote Link to comment Share on other sites More sharing options...
com48 Posted June 14, 2010 Share Posted June 14, 2010 Hey goldleaf. Just been looking at the version seth_ro18 and you have been working on and it works great. just found that if I changed getDeepestImpactOnly to false it will crash houdini on first impact. Quote Link to comment Share on other sites More sharing options...
goldleaf Posted June 14, 2010 Share Posted June 14, 2010 Thanks for the heads up! Quote Link to comment Share on other sites More sharing options...
goldleaf Posted June 24, 2010 Share Posted June 24, 2010 (edited) New update to the solver. Thanks lars for the heads up on that bug; the crashing when changing getDeepestImpactOnly is now gone. Also, added Cone primitive; bullet has two others, which are X and Z pointing; right now, only the y-direction primitive is used. The size is dictated by the Prim Sphere Radius, and Prim Capsule Length parameters. You can get it here. Edited June 24, 2010 by goldleaf Quote Link to comment Share on other sites More sharing options...
ilay Posted July 29, 2010 Share Posted July 29, 2010 New update to the solver. Thanks lars for the heads up on that bug; the crashing when changing getDeepestImpactOnly is now gone. Also, added Cone primitive; bullet has two others, which are X and Z pointing; right now, only the y-direction primitive is used. The size is dictated by the Prim Sphere Radius, and Prim Capsule Length parameters. You can get it here. Hello After raw compile under v11.0(win32/64), gets always 3772: Fatal error: Segmentation Fault Quote Link to comment Share on other sites More sharing options...
goldleaf Posted August 12, 2010 Share Posted August 12, 2010 (edited) I haven't done any Windows compiling, nor have I tried with 11, sorry. Anyone else out there working in Windows having success? Edited August 12, 2010 by goldleaf Quote Link to comment Share on other sites More sharing options...
infinity_spiral Posted August 29, 2010 Share Posted August 29, 2010 Can someone make a step by step tutorial how to compile this with Visual Studio and win7x64.I am trying with Visual Studio 2010 with no success.Thank you... Quote Link to comment Share on other sites More sharing options...
luoqiulin Posted September 13, 2010 Share Posted September 13, 2010 Can someone make a step by step tutorial how to compile this with Visual Studio and win7x64.I am trying with Visual Studio 2010 with no success.Thank you... up , i need this help too, I am trying with Visual Studio 2005,no success!! Quote Link to comment Share on other sites More sharing options...
almatea Posted October 31, 2010 Share Posted October 31, 2010 (edited) I make a modification of bullet solver. Added a fixed time step parameter of a solver - now it is possible to have, for example, stable tower of plenty bricks. Value of fixed time step parameter 240 is normal. Set it to zero to get a mode of previous versions. It tested on bullet 2.77, houdini 11. bullet-physics-solver-0.12-almatea-modification.tar.gz Edited October 31, 2010 by almatea Quote Link to comment Share on other sites More sharing options...
almatea Posted November 1, 2010 Share Posted November 1, 2010 (edited) added a git fork http://gitorious.org/~almatea/bullet-physics-solver/almateas-bullet-physics-solver Edited November 1, 2010 by almatea Quote Link to comment Share on other sites More sharing options...
Netvudu Posted December 2, 2010 Share Posted December 2, 2010 almatea, this last version you uploaded is it for Windows, or just Linux? Quote Link to comment Share on other sites More sharing options...
Denis Savin Posted March 6, 2011 Share Posted March 6, 2011 Biggest additions are deformable static objects (called kinematic objects in bullet, see this video) Goldleaf, please attach an example scene to setup deformable static objects. Checked ‘Deformable Mesh’ on the RBD Object DOP and ‘Kinematic’ on the Bullet Data DOP does not work. Thank you... Quote Link to comment Share on other sites More sharing options...
zzangxx Posted April 26, 2011 Share Posted April 26, 2011 (edited) up , i need this help too, I am trying with Visual Studio 2005,no success!! Did you solve this problem? I also got trouble to compile with Visual Studio 2005. I try to compile using Visual Studio 2005. Here are my error message. Please give me advices. 1>LINK : warning LNK4098: defaultlib 'MSVCRT' conflicts with use of other libs; use /NODEFAULTLIB:library 1>SIM_SolverBullet.obj : error LNK2019: unresolved external symbol "__declspec(dllimport) public: double __cdecl SIM_BulletData::getCollisionMargin(void)const " (__imp_?getCollisionMargin@SIM_BulletData@@QEBANXZ) referenced in function "protected: virtual class std::_Tree<class std::_Tmap_traits<int,struct bulletBodystr,struct std::less<int>,class std::allocator<struct std::pair<int const ,struct bulletBodystr> >,0> >::iterator __cdecl SIM_SolverBullet::addBulletBody(class SIM_Object *)" (?addBulletBody@SIM_SolverBullet@@MEAA?AViterator@?$_Tree@V?$_Tmap_traits@HUbulletBodystr@@U?$less@H@std@@V?$allocator@U?$pair@$$CBHUbulletBodystr@@@std@@@3@$0A@@std@@@std@@PEAVSIM_Object@@@Z) Edited April 26, 2011 by zzangxx Quote Link to comment Share on other sites More sharing options...
edward Posted March 15, 2014 Share Posted March 15, 2014 Should we un-sticky this thread? Quote Link to comment Share on other sites More sharing options...
Marc Posted March 15, 2014 Share Posted March 15, 2014 uh yeah, I think so . Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.