Quote:
Originally Posted by NeoGen
So... trial factoring would be the only possible (but absolutely unfeasible) method...

Why infeasible? If you can write the number in the form a^b+c with a and c reasonably small, trialfactoring is a very attractive approach.