Closed aknik closed 8 years ago
Brainwallet.io uses the scrypt standard. I just double-checked myself, and got matching results.
I used the link you provided above, entered "hello" as the passphrase, "world" as the salt, N=262144, r=8, p=1, and # bytes = 32. The result was: 1752aa0001d505168a308bac976ee2c0e94e7ca8235c6de404e6428eba946796
I did the same thing in brainwallet.io, and the result was identical. If you want to see for yourself, download the HTML, then insert the following directly below line 1414:
alert(result);
You should see the same scrypt result generated by the github link above.
You are right. The 'problem' is that brainwallet.io uses the result as passphrase (sha256)for generating the private key not as the private key itself (or secret exponent). My fault. I'm sorry.
Thanks for your fast reply. Best regards.
No worries!
There's actually a very good reason why I am also feeding it through a round of SHA256. If brainwallet.io were to ever disappear, and you don't have a copy of the website, you could still recover your keys using the following method:
You will arrive at the same keys.
I just check and the result from brainwallet.io is different from http://kclnn.github.io/js-scrypt-async/test_scrypt_browser.html I check it in python and nodejs with a standard scrypt implementation and the result is different from brainwallet , too
The process is as follows (pseudocode): key = scrypt(passphrase, salt, N=218, r=8, p=1, dkLen=32) keypair = generate_bitcoin_keypair(sha256(key))
So, is brainwallet.io not scrypt hash standard ?