> I prefer:
> 1. CBOR-compressed PKIX(5280) certificate.
> -and-
> 2. CBOR-native-encoded certificate.

(Assuming we at some point find a better word for “native”):

C509 is “CBOR native-encoded”.
What is different between C509 and the “CBOR-native” certificates is the 
derivation of the signing input; (1) tries to stay somewhat compatible with RFC 
5280 so your HSMs are still useful.
Note that there is no reason not to include both a RFC 5280-derived (1) and a 
COSE-derived (2) signature in a C509 item.

> (Yes, it uses RFC5280 semantics, but it's a new ecosystem)

Right, and step 3 (i.e., the interesting one) is actually addressing what we 
have learned about the data model that RFC 5290 (“X.509”) defines.
But that is not the subject of the C509 work.

> If we had to make up TLAs, I'd use something like CCPC (#1), and CNEC.

(*)
Thank god we don’t have to :-)

Grüße, Carsten


(*) And your proposals would be FLAs.

_______________________________________________
COSE mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to