Preventing collisions about nphysics HOT 6 CLOSED

@Kerosene2000 No, it is not yet possible. There are still two features related to CollisionGroups to be implemented:

1. Allowing the user to pass its own collision group to the physics engine at rigid-body construction time. This should be implemented soon.
2. Allowing the user to change the collision group of a rigid body after it has been added to the scene. This one involves some intricate development on ncollide so it is much more complicated and won't be added in the very short term.

from nphysics.

Currently static bodies will report collisions but will trigger no force. A mechanism to filter collisions already exists on ncollide (cf. CollisionGroups) though they are not completely exploited by nphysics yet.

In the mean time, a work-around may be to test if the collisions that were reported to you involve only static rigid bodies (for which the rigid_body.can_move() method will return false).

from nphysics.

The problem is not that I want to detect if the colliding bodies are static, but that if I have lots of static bodies colliding it will slow down the simulation to crawl.

from nphysics.

@sebcrozet You probably just answer to my question in the message above, is it possible to use the CollisionGroups struct when adding a body to the world ?

It seems that .add_body() already fix collisionGroups, which gives a real world physic...

from nphysics.

I am actually working on a project that will need the collisionGroups at the rigid body creation.
What means soon for you ?

If you want, you can let me work on the code in progress, if you don't have time, I will be glad to work on this crate !

from nphysics.

@Kerosene2000 I'd say soon equals a week or so. I wanted to have a first implementation of #20 before tackling this. If you want to give it a try, I've provided some outlines on the issue #55 I just created for that.

from nphysics.