That margin comment has been the teaser through the centuries. But it is doubtful that generations of fine mathers completely missed something that Fermat got correctly. The history of the problem shows much wreckage of battered hopes for easy solutions to it, so Fermat was probably wrong about having a valid proof. The problem now, following the solution by Wiles, Shimura, Taniyama and a bunch of other mathers, is to find a solution that is at least somewhat easier than this.