signature format, or what am i doing wrong? about curvy

[libitx]

Apologies - only just noticed this issue. Will take a look this week and get back to you.

[libitx]

@faried ah I see...

Curvy by default always hashes the message with sha256 - in your case it looks like the message is already hashed. We don't want to hash it again so can pass hash: false to recover_key:

Curvy.recover_key(sig, message, recovery_id: 1, hash: false)

That will give you: 0316b9dc66ec8b319b3d0a0ed0ec2ed433ab080f72c43fac6d51f48a0b8696e832

[faried]

Thanks! That solves my problem.

If it's not too much of a bother, can you describe how you found the solution?

[libitx]

Sure. I could see that the message you were providing in your example above was a 32 byte hexadecimal string, so there was a reasonable chance it was already a hash. I know that curvy hashes messages automatically so in all likelihood the reason for your discrepancy was because Curvy was hashing what was already a hash and the other software you were comparing against was not doing that.