Hey all, I've got a grid of objects 100 seperate objects for a 2d game on iOS. I'm using a raycast when I tap to select the objects (each object has its own box collider)
My question is if it's expensive to have a box collider on the objects even though they are idle 99% of the time and do not use any type of physics, they do, however move via transform.position = newpos and end up intersecting eachother at points.
Is there any kind of performance penalty associated with doing this? Again, I'm not using any sort of physics, just the collider for raycasting to hit when a player taps the screen. Thanks!
asked Jan 31 '11 at 08:53 PM
Moving colliders without rigidbodies is actually a very bad idea, performance-wise, because it forces PhysX to recalculate whatever is necessary, for static colliders. Put kinematic rigidbodies on them and you should experience as good of performance as possible.
answered Jan 31 '11 at 10:18 PM