Re: [GiNaC-list] gcd: wrong sign of cofactor (heur_gcd is broken?)