I agree that the at_hash definition is bizarre. I suggest adding a sentence when introducing the ath claim explaining that this is similar but different from at_hash. Thanks, -rohan
On Tue, Mar 29, 2022 at 6:14 AM Justin Richer <jric...@mit.edu> wrote: > Yes, it was considered, discussed, and rejected. The reason being > “at_hash” has a somewhat convoluted definition (left-bits of a hash of an > access token in the context of a JOSE object, etc), to fit some of the > design constraints of ID Tokens. DPoP proofs do not have those same > constraints. DPoP opted, correctly in my opinion, to simplify this by > declaring a single hashing algorithm and using its full output value. > Cryptographic agility would be achieved by defining a new claim with a new > hashing algorithm. > > — Justin > > On Mar 28, 2022, at 10:41 AM, Rohan Mahy <rohan=40wire....@dmarc.ietf.org> > wrote: > > Hi, > Did you consider using the (already IANA registered) at_hash claim defined > in: > https://openid.net/specs/openid-connect-core-1_0.html#CodeIDToken > instead of defining a new ath claim? > > It seems like if we don't use at_hash we should explain why ath is > better/different. > Thanks, > -rohan > _______________________________________________ > OAuth mailing list > OAuth@ietf.org > https://www.ietf.org/mailman/listinfo/oauth > > >
_______________________________________________ OAuth mailing list OAuth@ietf.org https://www.ietf.org/mailman/listinfo/oauth
