>>13662673i finally worked out how to think about tensors over finite dimensional vectorspaces in simple terms from first principles. i get the coordinate free formulation too. while my understanding doesn't extend to infinite dimensions, it's still a good starting point for finding more general rules that are equivalent in the finite case.
also i have started to fuck with cuda in earnest. i'm needing to reverse engineer some PTX code and i'm well on my way at this point
also you can adjust launch fuel levels in KSP so you don't have to build ridiculous shit to meet awkward contract requirements
