Skip Montanaro wrote: > In the current 2.2 beta, int operations that would overflow silently return > longs: > > >>> 3**100 > 515377520732011331036461129765621272702107522001L Just curious: how bad is the performance impact from this (if any)? cheers, f