My hobby project I am writing in my spare time. The idea is to perform all computations (bodies moving, collisions detection and resolving) on a GPU via compute shaders. For sure it will put some limitation on its possible functionality - e.g. we can't call collisions callbacks from compute shader immediately, but at least it should be (in theory) faster to compute everything on GPU instead of CPU.
TODO: provide more clear instruction on how to build (including environment, etc) If you want to, you can try to open it with Visual Studio 2022 as CMake project. Projects has some dependencies you may find in conanfile.txt. To install them, use 'conan install' with desired configuration (debug/release). You can do it in a directory out/build/x64-configuration/, VS should catch them up after that. Then you can open build and debug it within Visual Studio.
- Collision callbacks
- Event callbacks with client`s compute shader code (e.g. to manipulate on bodies somehow when they collide)
- Add support for Vulkan API (it is just easier to start with OpenGL)
- Linux/macOS support (though it is CMake+Conan project, it is currently tested only on Windows)