Re: [Rats] Call for adoption (after draft rename) for Yang module draft

[Henk Birkholz <[e-mail]>]

For convenience of some readers, please find the reply at the bottom.

On 13.11.19 12:28, Michael Richardson wrote:
> 
> 
> On 2019-11-12 6:10 p.m., Henk Birkholz wrote:
>>
>>
>> On 12.11.19 10:44, Michael Richardson wrote:
>>>>>> on one hand, we have to address the overlap between YANG and EAT
>>>>>> information elements (statements & Claims) and how to deal with them
>>>>>> (one obvious issue, for example, would be potential redundant
>>>>>> information model content in two different drafts).
>>>>>
>>>>
>>>> Examples of the same information elements that are used in different
>>>> data models via Claims or statements:
>>>>
>>>> * EAT Time stamp / YANG clock info
>>>> * EAT Origination / YANG attestation key cert, or
>>>> * EAT Uptime / YANG tpm ticks
>>>
>>> Those aren't EATs, those are attributes of claims that an EAT could
>>> include.
>>> If you are saying that we could write a YANG module describing those
>>> claims, and that they could be used to describe both claims of an EAT,
>>> but also results from a TPM 1.2/2.0 call, then I understand finally.
>>
>> Yes, that is basically a way to phrase the question, I think.
>>
>> Specialized questions are: do we need to do that? Do we want to do
>> that? Is it technically okay to instead convey opaque EAT blobs via a
>> YANG statement?
> 
> Sorry to be more pendantic.
> Do you mean: Is it okay to convey attributes of claims that an EAT could
> include as blobs in a YANG model?
> Or you mean: Is it okay to convey a binary serialization (say to CWT) of
> an EAT in a binary blog in a YANG model?
> 
> (The answer to both is yes.  I don't know if it's useful)
> 

Isn't that supposedly my job?

To the point. I meant: Is it okay to convey a binary serialization (say 
to CWT) of an EAT in a binary blog in a YANG model?

Why am I asking that? (well, actually now you did)

YANG modules thrive on well defined semantics for each statement. I just 
want to make sure that it is okay to convey an unknown composition of 
Claims (opaque semantics) via the web token format in a YANG module 
statement.

Laurence brought up a good example, I think. If "Identity Document" is 
good enough. Probably "Attestation Evidence" is good enough, too.

To comment on Michael's quote: Do "we need this in the same document 
(the same model?) as the TPM stuff?". I think "need" would be a very 
strong word here and therefore I would not recommend to make it 
mandatory and impose more code on YANG datastores/rpcs.

Viele Grüße,

Henk