On Thu, Mar 10, 2022 at 12:41:16PM -0500, Mike Prorock wrote: > +1 Mike - that is helpful. > > Any thoughts of applying that in this case? Seems like it would > line up well with key types for the two main things in play here: > 1) post quantum, hash based algorithms > 2) post quantum, lattice based algorithms > > This would also help on the implementation side as well per Ander's > note
All of Falcon, Dilithium and SPHINCS+ use the same key format (this was due to NISTPQC algorithm requirements). Which is byte string public/private key, which would map to OKP in both COSE and JOSE. As for mixing algorithms and keys, I think all wrong combinations are guaranteed to blow up as key and signature sizes just do not match up. -Ilari _______________________________________________ COSE mailing list [email protected] https://www.ietf.org/mailman/listinfo/cose
