> However, if you can use a permutation as a hash function, you might be fine with just using the identity function.
This certainly isn't true for 8-bit characters using ASCII.
The top-bit of all ASCII strings is always zero, its effectively a wasted bit. Permuting all ASCII characters to randomly use all 8-bits means a better distribution in virtually any hash-based data-structure. (ex: 64 slot hashtable will have fewer collisions after you permute the 7-bit ASCII into an 8-bit random permutation)
This certainly isn't true for 8-bit characters using ASCII.
The top-bit of all ASCII strings is always zero, its effectively a wasted bit. Permuting all ASCII characters to randomly use all 8-bits means a better distribution in virtually any hash-based data-structure. (ex: 64 slot hashtable will have fewer collisions after you permute the 7-bit ASCII into an 8-bit random permutation)