fastily / ftools Goto Github PK

I was linked to text2hash from this Wikipedia page. I then wanted to make sure I could replicate the results and when I tried, I found that other hash tools resulted in a different hash.

I believe this is because the function that is being called from CryptoJS is actually using the Keccak-512 algorithm. If you look at the Google Code Archive for CryptoJS, and scroll to the NOTE they mention that the SHA3 function is misnamed.

If you use the input hello world as an example, when using text2hash in the sha3-512 mode, its result matches the output from the Keccak-512 hash algorithm.

text2hash (sha3-512):
3ee2b40047b8060f68c67242175660f4174d0af5c01d47168ec20ed619b0b7c42181f40aa1046f39e2ef9efc6910782a998e0013d172458957957fac9405b67d

emn178's tool (sha3-512):
840006653e9ac9e95117a15c915caab81662918e925de9e004f774ff82d7079a40d4d27b1b372657c61d46d470304c88c788b3a4527ad074d1dccbee5dbaa99a

emn178's tool (Keccak-512):
3ee2b40047b8060f68c67242175660f4174d0af5c01d47168ec20ed619b0b7c42181f40aa1046f39e2ef9efc6910782a998e0013d172458957957fac9405b67d

I also tested this locally with the program HashTab which verified my conclusion.

To fix this, I think you should relabel the dropdown option of "sha3-512 (strong)" to "Keccak-512 (strong)". People on Wikipedia have been using this tool to create their Confirmed Identity Hashes, and if someone checks the hash without using this specific tool, they will get the "wrong" result for sha3-512. I suggested adding a note to the website explaining this change.