Revision history [back]

I have a hunch that you need to specify the needed polynomials u, v better. Surely there's an infinity of solutions for them? Or am I missing something? I would be very surprised if a solve() algorithm could handle this infinity of solutions. In the case of polynomials with a single variable these coefficients are usually found with the help of Bezout's theorem. I guess the multivariate case is handled analogously.