IME it's easy to get an error in bigint libraries if you do something like 2 ** (2 ** n) for some n greater than a few dozen. Because then you will have

IME it's easy to get an error in bigint libraries if you do something like 2 ** (2 ** n) for some n greater than a few dozen. Because then you will have 2**n digits, and the internal value holding the number of digits is limited to 2**32 or 2**64