Re: [Lake] COSE IANA registrations in EDHOC (Was: New Version Notification for draft-ietf-lake-edhoc-09.txt)

[John Mattsson <[e-mail]>]

The ID_CRED_x fields are just COSE header maps. The only thing EDHOC adds is how to format CRED_x. I think a separate COSE draft defining ‘cwt’ and ‘ucss’ header parameters should at least be considered. There are supposed to be useful also in other uses of COSE outside of EHDOC which might have different security consideration. At a minimum COSE should be involved to make sure that ‘cwt’ and ‘ucss’ are useful and secure also in non-EDHOC uses of COSE. If there are not a lot of additional things that need to be said regarding non-EHOC use of ‘cwt’ and ‘ucss’, I agree that keeping them in the EHDOC draft is the easier and fastest option.

John

From: Lake <[e-mail]> on behalf of Carsten Bormann <[e-mail]>
Date: Tuesday, 24 August 2021 at 10:05
To: Göran Selander <goran.selander=[e-mail]>
Cc: [e-mail] <[e-mail]>, Michael Richardson <[e-mail]>, [e-mail] <[e-mail]>
Subject: Re: [Lake] COSE IANA registrations in EDHOC (Was: New Version Notification for draft-ietf-lake-edhoc-09.txt)
I see.

So, you are saying, this will be a “using EDHOC in COSE” specification, still normative, but referenced from EDHOC as informative as EDHOC works without COSE.

Yes, it is always hard to position a “using X in Y” draft between the X and Y working groups — after all, the two ends of this draft need to fit X and Y, respectively.  If the EDHOC specification truly doesn’t need the contents of this specification, then I can see moving them into a COSE document.  But I think it is as expedient to keep them together in one document.  The only strong reason to split the document would be to avoid a long wait while COSE is deciding on some controversial content of the extracted spec.  Do we foresee such a delay?

Grüße, Carsten


> On 2021-08-24, at 09:35, Göran Selander <goran.selander=[e-mail]> wrote:
>
> Combining the responses from Carsten and Michael, and including COSE.
>
>> On 2021-08-23, 19:17, "Michael Richardson" <[e-mail]> wrote:
>>
>>   Göran Selander wrote:
>>> * The key identifier ‘kid’ is extended to also support CBOR ints,
>>> making ‘kid2’ introduced in -08 redundant. This change was based on
>>> feedback from the COSE WG [1]. One potential next step is to move all
>>> COSE-related IANA registrations from this draft to a separate COSE
>>> draft and make an informative reference.
>>
>>> [1] https://mailarchive.ietf.org/arch/msg/cose/qGngdte4s3SEZEKM-xBEoXYUgKc/
>>
>>   I understanding splitting the document so that it is easier to update,
>>   but I think that the reference should be normative.
>>
>>   I think we want to publish the documents together.
>
>
> > On 2021-08-23, 21:42, "Carsten Bormann" <[e-mail]> wrote:
>>
>>> One potential next step is to move all COSE-related IANA registrations from this draft
>> to a separate COSE draft and make an informative reference.
>>
>>   Why?
>>
>
>
> The registrations in question are in section 8.5 -  8.7 of draft-ietf-lake-edhoc-09: The extension of 'kid' to int (both as a reference and in the referenced object) and the registration of 'cwt' to signify that the value is a CWT or UCCS.
>
> A few reasons have been mentioned for moving this from EDHOC to a COSE draft, I don't know what is most relevant, if anything:
>
> * In case of 'kid', these registrations would make EDHOC an update of draft-ietf-cose-rfc8152bis-struct (RFC-to-be 9052). I don't know if LAKE or COSE wants that.
>
> * These registrations are independent of the base EDHOC protocol, but enables the use of CWT and UCCS as credentials, and more compact identification of credentials. Therefore they could instead be referenced from EDHOC. I don't see why the reference needs to be normative.
>
> * These registrations belong to the COSE domain and may gain better awareness and reviews if put into a COSE draft.
>
>
> Göran
>
>
>
>
>
>
>
>
> --
> Lake mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/lake

--
Lake mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/lake