Re: decryption algorithm

That was a cut-and-paste error on my part, not Michal's xor algorithm.
I have a library of 120 password/hash pairs and Michal's algorithm
correctly reproduced the encrypted hash for all passwords. The only
problem for me was figuring out missing elements in the "secret" array,
but as it turns out xor'ing any password/hash pair returns the key for
passwords of that length. And xor-ing the hash with the key returns the
unecrypted password. All amazingly symmetrical.