On Tue, Mar 14, 2023 at 12:30:56PM -0700, Laurence Lundblade wrote:
> -04 is a good improvement. Thank you.

It even seems implementable. :-)


> The kid parameter is always optional in COSE. It shouldn’t be
> mandatory in 3.1.1 and 3.1.2. I think this is a mandatory change for
> this draft. Section 5.2 of 9052 even discourages use of kid in and
> Encrypt0:
> 
> The COSE_Encrypt0 encrypted structure does not have the ability to
> specify recipients of the message.  The structure assumes that the
> recipient of the object will already know the identity of the key
> to be used in order to decrypt the message.  If a key needs to be
> identified to the recipient, the enveloped structure ought to be
> used.

I think the COSE_Encrypt0 case here is actually analogous to Direct
Key Agreement, because COSE-HPKE smashes two layers together (as HPKE
makes that natural to do). And DKA usually uses kid.

In my implementation, there is shortcoming[1] that causes things to blow
up if one tries to decrypt COSE_Encrypt0 messages without kid using
multiple decryption keys (it does kid matching, but can only try to
decrypt once, COSE_Encrypt does exhaustive trial decryption after
kid matching).


> I’d like for “HPKE” to be in the name of the structure since it is
> HPKE specific — perhaps “HPKE_sender_info”.

Yeah, I think that is a better name.


> Section 4.3 is clear now and well-aligned with what I was thinking
> it should be. 

Yeah, I also tought that techically that section is good, but that it
could be phrased better (relying more on RFC 9052).



[1] At least it does not make the whole system with 8GB RAM to go
Chernobyl if one encrypts/decrypts 20GB+ message...





-Ilari

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

Reply via email to