The right convention is also "obvious" to practitioners of combinatorics. The exponentiation x^y, for integer x and y, is the number of possible strings of leng

The right convention is also "obvious" to practitioners of combinatorics. The exponentiation x^y, for integer x and y, is the number of possible strings of length y from a set of letters of cardinality x. (Hence 28 possible bytes.) How many ways are there to make a string of length 0, regardless of the alphabet size? Just one... you don't do anything.