I finished work on a recursive determinant, although as stated above, it’s redundantly recursive. Initially I thought I’d build up a TMap mapping Matrices to determinant solutions, and only work out determinants I hadn’t worked out prior. But that could lead to an huge memory hog. In the future since I know I’ll be generating cofactor matrices for calculating inverse matrices, I’ll probably just make that a member of the struct, so I can use it as a scratch pad for cofactor heavy calculations. Only problem is I have to use dirty bits or something to indicate when a previously calculated cofactor is invalidated. I’ll work on that later, for now this works fine.