This PR replaces the matrix-vector math in the engine with GlmSharp as discussed.
These are only refactorings of existing content, and should exhibit the same behavior as before the commit (except with higher precision now). Using GlmSharp, we can use proper matrix-vector mathematics and also use different precision for vector types, such as double.
The tested code appears to work correctly within the visual test. The unit tests all work except for two infinity checks, which may or may not have been broken before already (@AdamNorberg also noticed this before this PR, so I'll ignore it for now and propose we look into it in a different PR).
This PR replaces the matrix-vector math in the engine with GlmSharp as discussed.
These are only refactorings of existing content, and should exhibit the same behavior as before the commit (except with higher precision now). Using GlmSharp, we can use proper matrix-vector mathematics and also use different precision for vector types, such as double.
The tested code appears to work correctly within the visual test. The unit tests all work except for two infinity checks, which may or may not have been broken before already (@AdamNorberg also noticed this before this PR, so I'll ignore it for now and propose we look into it in a different PR).