Problems with Rolling friction in bullet 2.82

[mlazo]

Hey guys,

I've had problems when I try to use btDantzigSolver or btMLCPSolver solver in BPL 2.82-r2718

My scene consist with a set of spheres that collide with a trianglemesh. Spheres and the trianglemesh set the rolling friction parameter to 1.0.
When I use btSequentialImpulseConstraintSolver everything work good. But when I switch the solver to btDantzigSolver or btMLCPSolver the engine seems unaware of the change of rolling friction parameter. I've tried to use values greater or lower than 1.0 but nothing changes.

Any suggestions? Thanks very much!!

PD: Sorry for my elemental English ....

[Erwin Coumans]

The MLCP solvers and Featherstone don't have all features implemented, and rolling friction is missing.

Note that the Dantzig solver contains a hack to deal with friction, and this doesn't deal with rolling friction.
Thanks,
Erwin

[mlazo]

Hi Erwin, thanks for response.
Are you planning to implement this feature soon?

[Erwin Coumans]

I cannot promise it, I'm very busy and my focus should be on Bullet 3.x.

If you want that feature, your best bet is to file a request in the issue tracker at http://bullet.googlecode.com
Thanks,
Erwin