>-----邮件原件-----
>发件人: maqiufang (A) 
>发送时间: 2021年12月13日 15:10
>收件人: Qin Wu <[email protected]>
>抄送: Michael Richardson <[email protected]>; [email protected]
>主题: RE: [Anima] Call for adoption: draft-richardson-anima-rfc8366bis, ends 
>December 19th, 2021

>Hi, Qin:
>Please see my reply inline.

>-----Original Message-----
>From: Anima [mailto:[email protected]] On Behalf Of Qin Wu
>Sent: Monday, December 13, 2021 11:19 AM
>To: Michael Richardson <[email protected]>; [email protected]
>Subject: Re: [Anima] Call for adoption: draft-richardson-anima-rfc8366bis, 
>ends December 19th, 2021

>Hi, Michael:
>>-----邮件原件-----
>>发件人: Michael Richardson [mailto:[email protected]]
>>发送时间: 2021年12月13日 4:36
>>收件人: Qin Wu <[email protected]>; [email protected]
>>主题: Re: [Anima] Call for adoption: draft-richardson-anima-rfc8366bis, 
>>ends December 19th, 2021


>>Qin Wu <[email protected]> wrote:
>>    > 1.       Since we still reuse rc:yang-data to model vouch-artifact, I
>>    > believe the model structure in section 5.1 of RFC8366 should keep as
>>    > is.

>>    > 2.       Secondly, vouch-artifact doesn’t need to be coupled with
>>    > RESTCONF protocol message, suggest to use YANG extension statement
>>    > “structure” defined in RFC8791 to replace yang-data.

>>Sorry, I'm confused here.  I think that #1 says that yang-data is okay.
>>I think that #2 says that yang-data is not okay, and we need sx:structure 
>>instead?

> [Qin Wu] :Good catch, I think sx:structure is preferable instead. The YANG 
> data structure in section 3 of RFC8791
>         can be followed.
>     structure <structure-name>:
>       +--<node>
>          +--<node>
>          |  +--<node>
>          +--<node>

>>Are you saying that we should update to RFC8791?
>>Do you think that this is a bug-fix?

> [Qin Wu] No, have clarified in the separate message, I think both yang-data 
> and sx:structure can be used, but sx:structure is not
>targeted to replace yang-data, but RESTCONFbis in the future may consider to 
>decouple yang-data from RFC8040.

>>I'm asking, because we have discussed that RFC8366bis might go forward 
>>as a
>> (step-two) Internet Standard, rather than cycling at Proposed Standard.

>>    > 3.       Do we expect iana-voucher-assertion-type to be changed in the
>>    > future, e.g., add a new enum value? Does it make sense to change
>>    > voucher-assertion typedef into identity data type.

>>Yes, we expect that new documents will Extend it.
>>I don't know what an identity data type is here.... I thought we did the 
>>right thing, so I do not quite understand what you are proposing.

> [Qin Wu] I have no strong opinion whether enum is the best option, the 
> difference between identity and enum is identity is a text name without 
>an associated value but enum is both a name and a value enum doesn't support 
>augment in YANG while 
>identity does. But the advantage of using Enum is Enum saves a lot of space 
>than identity.
>Here is what identity looks like:
>"
>   /* Identities */
>  identity voucher-assertion-type {
>   description
>   "Base identity for voucher assertion type. The voucher assertion 
>   is used to indicate what kind of ownership is being asserted by 
>   voucher.";
>  }

>  identity verified {
>   base voucher-assertion-type;
>   description
>   "Identity for verified voucher assertion, which indicates that 
>    the ownership has been positively verified by the MASA 
>       (e.g., through sales channel integration).";
>  }
>  identity logged {
>   base voucher-assertion-type;
>   description
>   "identity for logged voucher assertion. The logged voucher assertion 
>   indicates that the voucher has been issued after
>   minimal verification of ownership or control.  The
>   issuance has been logged for detection of
>   potential security issues (e.g., recipients of
>   vouchers might verify for themselves that unexpected
>   vouchers are not in the log).  This is similar to
>   unsecured trust-on-first-use principles but with the
>   logging providing a basis for detecting unexpected
>   events."
>  }
......
>"
> [Qiufang Ma] Given this is an IANA maintained YANG module, I think it's okay 
> to define it as enumeration type.
>If someone wants to add a new type, a request has to be made to IANA registry, 
>and IANA will update the YANG module and add a new "enum" statement 
>accordingly.
>An enumeration data type defined here makes it controlled by a single 
>authority.

[Qin Wu] Thanks for clarification, I believe your comment is relate to Lada's 
proposal
https://mailarchive.ietf.org/arch/msg/netmod/hxy7gKJ13yA0L_PYopb7usql870/
Yes, this proposal did work for me, which automate process not only for IANA 
registries update, but also for IANA maintained YANG module update, 
but I am not sure whether this proposal have been fully accepted and 
implemented.

>Best Regards,
>Qiufang Ma
>--
>Michael Richardson <[email protected]>   . o O ( IPv6 IøT consulting )
>           Sandelman Software Works Inc, Ottawa and Worldwide




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

Reply via email to