Open GoogleCodeExporter opened 8 years ago
I guess I can just use TriangleMesh for now as a workaround, unless I'm being
retarded and that's how I'm supposed to do it
Original comment by pyri...@gmail.com
on 23 Aug 2011 at 5:30
Yeah, StridingMeshInterface can't be subclassed, because I haven't found a good
way to make that work properly yet. Those abstract classes can be tricky to
wrap.
If you can use TriangleMesh instead, then that's great. I think the performance
will be better anyway since there won't be any calls to the managed code in the
subclass during simulation
Thanks for the feedback!
Original comment by andres.traks
on 23 Aug 2011 at 11:57
Yep, got TriangleMesh to work instead. Only problem now is that I have to have
it in a CollisionObject instead of a RigidBody (since the latter makes my
raycast vehicles freak out when they touch it), but the wheels of the vehicle
just go right through the mesh as if it wasn't there, with the chassis sitting
on top of it. Might be a problem with the raycaster, I don't know
Original comment by pyri...@gmail.com
on 24 Aug 2011 at 12:05
what about VehicleRaycaster? Same issue?
Original comment by pyri...@gmail.com
on 24 Aug 2011 at 8:31
Yeah, extending the wrapper classes isn't supported. I've tried to make those
classes extendable that are normally subclassed (like DebugDraw), but it's
always been a bit of a mess.
With BulletSharp it's more reasonable to avoid all kinds of virtual calls to
managed code such as overridden methods, because going in and out of managed
mode is expensive.
If you need the extended functionality, then you might want to try the native
.NET version of Bullet: http://code.google.com/p/bullet-xna/
Original comment by andres.traks
on 25 Aug 2011 at 11:04
I found workarounds for both of the things I asked you about so I guess it
doesn't really matter much to me any more. Thanks for your help though!
Original comment by pyri...@gmail.com
on 25 Aug 2011 at 11:16
Original issue reported on code.google.com by
pyri...@gmail.com
on 23 Aug 2011 at 4:51