Two mathematicians say they found a glaring hole in a proof that has convulsed the math community for years.