On Fri, Jul 08, 2022 at 10:34:11PM +0200, Anders Rundgren wrote:
> I can only iterate my claim that overloading an existing definition is an 
> unusual solution.
> 
> However, since even the author of 
> https://datatracker.ietf.org/doc/html/rfc7517#section-4.1
> have abandoned [*] the (IMO) most logical interpretation, I guess
> this is it:
> 
> All new keys having a "crv" SHOULD have "kty" = "OKP".

Well, if the new key is some prime-field weierstass elliptic curve,
then it should be EC (EC2 in COSE). EC/EC2 was designed for those
things.

BLS is not prime-field, so it does not fit EC/EC2.



-Ilari

_______________________________________________
COSE mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/cose

Reply via email to