Erin Catto

Current summary: this is a feature request for convex decomposition

Yeah, the low level routines exist to support this. Completing this feature requires: - Implement swept AABB in dynamic tree - Create API in b2World

I looked at this more closely. The performance trade-off is between sorting the pair buffer and clearing the move flags. This trade-off is not clear to me since clearing flags...

Feature request: rounded polygonal shapes

Sounds reasonable. And yeah, I was also thinking about the number of separate resource loads. BTW, I really like the design of the variation and how they are visible within...

This page describes the process for adding natvis files manually in Visual Studio: https://msdn.microsoft.com/en-us/library/jj620914.aspx

Yes, that is what I have been doing. Visual Studio also supports adding files at the solution scope. The page I linked above shows the process for adding a natvis...

I found some other use cases that would make this nice to have: Doxyfile, .gitignore, premake5.lua, etc. This seems like a generally useful feature.

I think it is worth considering the cell based solver using k-means. It is possible to keep determinism by using a target cell count regardless of thread count. No parallelism...

Yeah, I think a local IdPool per cell is the way to go. Keep the IndexSet small. One thing appealing about this approach is that we can experiment with building...