highperfocused/bips
1
0
Fork 0
mirror of https://github.com/bitcoin/bips.git synced 2025-04-10 05:10:57 +02:00
Code Issues Packages Projects Releases Wiki Activity

minor comment about "byte"

Greg Tonoski 2024-10-25 10:29:16 +02:00
parent b5528fa921
commit ed7f4d10e9
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Comments:BIP-0340.md

@ -3,4 +3,6 @@ Removed a wall of [misinformation](https://github.com/bitcoin/bips/wiki/Comments
----
The comments above appear to be misinformed. Nearly all ECDSA implementations today, including all the ones used in Bitcoin software that I know about, already use derandomized RFC6979 nonce generation for secp256k1. BIP340 too specifies such a nonce generation algorithm - one that is inspired by Ed25519's in fact. It permits adding randomness as research has shown this improves resistance to certain fault & side-channel attacks, but the randomness is not critical for security (it is purely additive). Lastly, the entire nonce generation concern is an implementation aspect that's orthogonal for the signature scheme - it can be done well, or badly, either with ECDSA or Schnorr/BIP340. BIP340 chooses to specify it, in the hope that implementations adopt it as a best practice, but circumstances may call for alternative nonce generation algorithms too. -- Pieter Wuille, 2021 Mar 07.
The comments above appear to be misinformed. Nearly all ECDSA implementations today, including all the ones used in Bitcoin software that I know about, already use derandomized RFC6979 nonce generation for secp256k1. BIP340 too specifies such a nonce generation algorithm - one that is inspired by Ed25519's in fact. It permits adding randomness as research has shown this improves resistance to certain fault & side-channel attacks, but the randomness is not critical for security (it is purely additive). Lastly, the entire nonce generation concern is an implementation aspect that's orthogonal for the signature scheme - it can be done well, or badly, either with ECDSA or Schnorr/BIP340. BIP340 chooses to specify it, in the hope that implementations adopt it as a best practice, but circumstances may call for alternative nonce generation algorithms too. -- Pieter Wuille, 2021 Mar 07.
2024-Oct-25 minor comment: "byte" is used in the specification without definition and the one from C does not suit well in this context. I would suggest either defining bytes as consecutive groups of 8 bits or not using it at all (e.g. 512 bits instead of "64-byte[s]") for clarity.