On 9. Dec 2022, at 12:17, Hannes Tschofenig <[email protected]> wrote: > > I am OK with registering “HPKEv1-BASE” as the algorithm id since I don’t see > much difference to the currently proposed “HPKE”.
While HPKE doesn’t actually seem to have versions, I would be OK with this, as it makes it quite clear that we do not need to jam any revision of HPKE into the same algorithm-ID (*). (Note that the “version-label” that is used in used in LabeledExtract() and LabeledExpand() but not otherwise defined in RFC 9180 is "HPKE-v1”, so we probably should keep that second dash in there, even if it actually is just human-readable information in the registry.) Grüße, Carsten (*) well, unless we want to make a forward-compatible (**) revision work with old systems. But we can cross that bridge if we ever get to it. (**) * Backward compatibility = old data, new system. * Forward compatibility = new data, old system. _______________________________________________ COSE mailing list [email protected] https://www.ietf.org/mailman/listinfo/cose
