Closed TimoGlastra closed 1 year ago
I think that a verifier would/should always create/use their own nonce in a presentation request. The use of the proof-req format in the proposal
is just out of convenience, and does not/should not imply that the verifier would just blindly use an unchanged version of the proof request data structure in their proof. Further, by removing the nonce from the proposal, the two data structures become different, which means more specifying.
Perhaps a clarification that the nonce from the proposal should NOT be used in the request should be added?
Appears resolved.
Present Proof v2 proposal and request both use
hlindy/proof-req@v2.0
format. This is different from v1 where it used the protocol native presentation proposal.One thing I've noticed is that the proof request specifies a nonce, however in the case of a holder making a proposal this doesn't make a lot of sense. The nonce should be provided by the verifier to prevent replay attacks.
By allowing the holder to propose a nonce we lose this. In AFJ we now just always generate a new nonce, but it's still quite odd that it's in the RFC.
Should we only allow the
nonce
for requests and not when used as a proposal?