Re: [math-fun] Gosper's "path invariance" in matrix product land