flame-engine
Creating performance benchmark suite for collision detection
What could be improved
It would be great to have a benchmark harness for the collision detection system.
Why should this be improved
So that it is easier to know if you're making improvements or regressions, and also a good test bed to try out new broadphase algorithms.
Any risks?
No change to the library code, so no risk.
More information
This could possibly be used: https://pub.dev/packages/benchmark_harness
What work needs to be done for this? Collision detection is a skillset of mine. I'm interested.
@TheMaverickProgrammer we'd like to have a benchmarking suite based on benchmark_harness, it should compare different collision detection scenarios and compare different broadphase algorithms.
If you're interested in collision detection, I'm sure there are some interesting new broadphase algorithms that could be implemented.
@TheMaverickProgrammer we'd like to have a benchmarking suite based on benchmark_harness, it should compare different collision detection scenarios and compare different broadphase algorithms.
If you're interested in collision detection, I'm sure there are some interesting new broadphase algorithms that could be implemented.
Yes there are a few. All of varying degrees of improvement depending on the environment and ratio of static-to-dynamic bodies. This is an area of research I've been nosing around in so it seems like the perfect place for me to contribute directly.