I'm in the process to implement the collision system for our space-game project in school, and i have chosen Bullet for the collision tests. It seems like something we will have up and running quite quickly, and that is easy extendible later on. Note that this is the first contact I have with Bullet physics, and even third party physics library. We are planning to support different sizes of ships, ( big car vs big-ass cruiser containing the small big car ships. Maybe even big-ass stations containing the big-ass cruisers ). We are also planning to import a bounding hierarchy created in Maya, supporting different kind of basic collision shapes (boxes, spheres etc. ).
Now to my question:
What is the "correct" way to implement this support in Bullet? Do we gain on implementing a hierarchy object based on a collision object from bullet, with a specific collision callback? Is it even possible? Or is it just overhead, even unoptimized, vs the bullet physics own optimization? Could it be/is it more convenient to just load a bunch of collision objects from a file, and check them all within a hierarchy in btDbvtBroadphase? It is mainly meant for shooting and ship collisions, as physics for the ships already is implemented.
Any suggestions or thoughts are welcome. Will be in the process to test some things out, and implement Bullet in the process.
CAPE (Code And Play Engine(ere))
1 post • Page 1 of 1