Hello!
I was looking for a data-structure that is meant to handle sorting of 3d-objects, where every object, within trees or child trees, will potentially move/change their position, hence no static objects.
I am mainly curious if there is anything similar to an octree with a cheap update-complexity, because updating one element takes quite some iterations, e.g. recursively entering parent octrees, checking whether they fit in them or not...
I considered to roll out a 3d spatial hashing instead, as octrees' precision seems quite unimportant for my project as well. It would seem to be easier to update compared than an octree.
Why do I worry about performance? I really just would like to use the semantically most fitting structure. I often hear, quad- and octrees shine with static objects, because they allow a high precision via their branching of smaller trees within each tree. On the other hand, I totally do not need this, there is no complex collider either, just cubes.
I might be totally wrong on my performance-observations, though!
Summing it up, what would you recommend/use for a 3d-world with simple cube-colliders where every object is dynamic?
Thanks for your time!
Edit: Fixed terribly screwed sentences that were not even complete, sorry!
↧