Add sighash versioning specification to ZIPs 246 and 230

[arya2]

No description provided.

[daira]

utACK with suggestion.

[nuttycom]

ACK with @daira @str4d @arya2 @SamHSmith modulo QEDit's confirmation that this will work for atomic swaps. cc/ @vivek-arte

[vivek-arte]

As mentioned in our discussion with @str4d last week, this approach should work for the Swaps use case as well.

However, as we discussed, there is a need to add sigHashInfo to the issuance authorization signatures as well -- to allow for the "update to the algorithm in case of a bug" setting (and similarly, to the key rotation bundles).
I believe the ZIP Editors will make a further pass through this in the near future, update for the above, and fill out the remaining parts. We will revisit this after those updates.

[ConstanceBeguier]

It is not specified how SighashInfo is incorporated into the digests.
I think we only need to update auth_digest and specifically:

• transparent_scripts_digest
• spend_auth_sigs (A.2b) in sapling_auth_digest
• binding_sig (A.2d) in sapling_auth_digest
• bindingSigOrchard (A.3b) in orchard_auth_digest
• spendAuthSigsOrchard (A.3a.ii) in orchard_auth_digest
• issueAuthSig (A.4a) in issuance_auth_digest

For each of these signatures, the signature should be replaced by (SighashInfo || signature) where SighashInfo = sighash_version || associated_data.

Do you agree?

[daira]

@ConstanceBeguier wrote:

For each of these signatures, the signature should be replaced by (SighashInfo || signature) where SighashInfo = sighash_version || associated_data.

Yes. I think I was confused in the meeting today between the sighash_version and the signature version byte. Your suggestion is completely fine as-is.

[nuttycom]

utACK afce126