> it would fail for instance if some solution involved a value greater than 10^100 And that's just what happens apparently for (at least) 77869, so I'll work on redoing the table with a larger cutoff.