I appreciate your feedback too and agree that security related code is a sensitive thing. I did not loose any thought on the actual hashing and as written in the readme this is just a prototype. The key aspect of Passpod and on what I'd love to get some feedback is the idea to save the passwords in a giant set of hashes instead of the commonly used map of clear usernames and hashed passwords. I wrongfully supposed that it would be obvious that the actual hashing is just meant to be a placeholder. So I changed sha512 to md5 and documented that the actual hashing is broken accordingly.