1. 5
    1. 3

      I just wanted to point to the original paper describing this encoding technique:

      I found it interesting not so much for its initial purpose (encoding small binary tokens, multiple of 4 bytes like IP addresses), but as a technique of generating pronounceable passwords.