Re-,

Please see inline.

Cheers,
Med (as editor)

De : Ketan Talaulikar <[email protected]>
Envoyé : mardi 3 juin 2025 13:34
À : BOUCADAIR Mohamed INNOV/NET <[email protected]>
Cc : The IESG <[email protected]>; [email protected]; 
[email protected]; [email protected]; [email protected]; 
[email protected]
Objet : Re: Ketan Talaulikar's Discuss on draft-ietf-netmod-rfc8407bis-25: 
(with DISCUSS and COMMENT)


Hi Med,

Thanks for your quick response and sharing the diff from your github edit 
buffer.

I understand that some of the points that I've raised are inherited from RFC 
8407. This is a pretty significant and important document. Therefore, there are 
some things that need discussion/update even if they were coming from RFC 8407 
to align with times.
[Med] I’m flexible but we need to be careful to not diverge much from the scope 
set for the bis this time: 
https://datatracker.ietf.org/meeting/117/materials/slides-117-netmod-7-guidelines-for-authors-and-reviewers-of-documents-containing-yang-data-models-00
 (slide 7) ;-) We need an iterative approach and clear scope for each round.

Please check inline below for clarifications. Ones without response can be 
considered as addressed/closed and can be trimmed out in further replies.


On Tue, Jun 3, 2025 at 2:27 PM 
<[email protected]<mailto:[email protected]>> wrote:
Hi Ketan,

Thanks for the review.

The changes made so far can be tracked at: 
https://author-tools.ietf.org/api/iddiff?url_1=https://netmod-wg.github.io/rfc8407bis/draft-ietf-netmod-rfc8407bis.txt&url_2=https://netmod-wg.github.io/rfc8407bis/KTcomments/draft-ietf-netmod-rfc8407bis.txt.

See more context inline.

Cheers,
Med (as editro)

> -----Message d'origine-----
> De : Ketan Talaulikar via Datatracker 
> <[email protected]<mailto:[email protected]>>
> Envoyé : mardi 3 juin 2025 08:04
> À : The IESG <[email protected]<mailto:[email protected]>>
> Cc : 
> [email protected]<mailto:[email protected]>;
>  [email protected]<mailto:[email protected]>;
> [email protected]<mailto:[email protected]>; 
> [email protected]<mailto:kent%[email protected]>; 
> [email protected]<mailto:[email protected]>;
> [email protected]<mailto:[email protected]>
> Objet : Ketan Talaulikar's Discuss on draft-ietf-netmod-rfc8407bis-
> 25: (with DISCUSS and COMMENT)
>
> Ketan Talaulikar has entered the following ballot position for
> draft-ietf-netmod-rfc8407bis-25: Discuss
>
> When responding, please keep the subject line intact and reply to
> all email addresses included in the To and CC lines. (Feel free to
> cut this introductory paragraph, however.)
>
>
> -------------------------------------------------------------------
> DISCUSS:
> -------------------------------------------------------------------
>
> Thanks to the authors and the WG for your work on this important
> document.
>
> I have one high-level point that I would like to discuss with the
> authors and
> the WG since is it not clear - this is regarding experimental and
> information
> track YANG module documents in IETF stream.
>
> At a high-level, I would like to discuss and understand whether
> YANG model
> documents can be experimental or informational. I think the answer
> is YES?

[Med] YANG modules defined in the IETF so far are defined in PS documents, even 
when covering protocols that are in Info/Exp. An example is RFC9105. You may 
refer to 
https://mailarchive.ietf.org/arch/msg/opsawg/7E9g02ZwMt5z9PT1vgSai4kW42U/ or 
the discussion at the time in the IESG.

KT> That email thread does not really cover my point. The email thread gives 
the position of an AD at that point of time. Four years down the line, we again 
have a couple of ADs discussing a somewhat similar point. What is needed is 
that the IETF consensus is captured in a BCP document. I believe this is the 
right BCP document to cover it.


The key point is that a YANG module (independent of the underlying objects it 
manipulates) is about making two peers (client/server) interoperable.

KT> I am not going to debate that point (yet). However, I will note that there 
have been recommendations made over the years to include YANG modules in the 
same document as the protocol features. So, the question then becomes:

1) If the document is considered experimental (or informational) in view of the 
protocol specifications/extensions in it, then should it (a) not include a YANG 
and that should be in separate standards track document, or (b) no YANG work 
should be done for anything experimental or informational, or (c) the document 
needs to be standard track because it has a YANG (as in the proverbial "tail 
wagging the dog"? - to not be taken literally).

2) Independent of (1), the authors and WG seem to have not all considered even 
the possibility of YANG being documented in IETF stream as Experimental or 
Informational while artifacts from other older RFCs (but not the predecessors 
of this BCP) generally refer to IETF Stream documents and not standard tracks.  
I would like to discuss and understand why so. Further, whatever is the reason 
should be captured in this document as it is after all the BCP for 
writing/reviewing YANG documents.

[Med] My personal take on this is that there is no causality effect between the 
track of a base spec and its companion YANG module, especially when used for 
management purposes. I provided the example of 9105 on purpose to illustrate 
this (base=INFO, YANG=PS). I’m not aware of any YANG module that is published 
as Info or EXP to date. That’s said, there might be cases where publishing in 
the Info or Exp may make sense in the future but I don’t know those. It is 
difficult to create guidance for an hypothetical case. The preamble of Section 
4 is clear about the intent usage of the spec.


The following usage in Section 3 (inherited from 8407):

CURRENT:
   There are three usage scenarios for YANG that can appear in an I-D or
   RFC:

   o  normative module or submodule

   o  example module or submodule

   o  example YANG fragment not part of any module or submodule

KT> I do not see how this even relates to the 
standard/experimental/informational tracks. Can you please explain?
[Med] An example module is module which is provided for illustration purpose 
and is not registered within IANA, etc.

Also, perhaps I do not fully understand what an "example module" is used for 
from reading this document - is it something different from or beyond say 
https://www.rfc-editor.org/rfc/rfc9127.html#appendix-A.1 ?
[Med] That’s an example module.

I ask this question since you bring up "example module" when I bring up 
experimental/informational documents.


 But
> this is not clear. A follow-on question: what is the guidance for
> YANG models
> specified in standards track document being augmented by modules in
> experimental or informational track document? I think the answer is
> NO? But
> again, this is not clear.
>
> Please also see in the comments sections for some concerns that are
> related to
> this topic - those are provided inline for better context.
>
>
> -------------------------------------------------------------------
> COMMENT:
> -------------------------------------------------------------------
>
> Please find below some comments provided inline in the idnits
> output of v25 of
> this document.
>
> 14      Abstract
>
> 16         This memo provides guidelines for authors and reviewers
> of
>
> <nit> s/memo/document

[Med] OK, even if this was inherited from 8407 ;-)

>
> 232     1.1.  Changes Since RFC 8407
>
> <minor> Since this is an exhaustive list of changes, can it be
> moved to the
> appendix section? First time (new) readers don't need to be
> bothered by this in
> the body of the RFC?

[Med] This is a matter of taste, but more importantly follow the practice used 
in rfc8407 vs 6087.

>
> 688        Unless the modules comply with [RFC8791] or define YANG
> extensions
> 689        (e.g., [RFC7952]), the security section MUST be modeled
> after the
> 690        latest approved template (available at
> 691
> <major> I am not sure if a wiki pointer is appropriate here.

[Med] This practice is implemented and followed since many years now (including 
RFC editor, etc.). We need a stable reference to maintain the template out of 
the RFC itself. A reason why the template is maintained here and not completely 
offloaded is that we received a request from the IETF trust that we addressed 
(this text is used outside the IETF).

KT> First, I am happy that the template (as of the date of publishing this 
document) is going to be in the RFC itself (Appendix would have been better for 
my taste for all such templates - but I leave it to the authors/WG/AD). The 
problem is the normative MUST to reference something that is not going to be 
updated via the IETF Consensus process - my concern would be addressed by 
dropping that normative MUST. The document can even say that the template may 
be improved with community inputs and the latest version can be found at an URL 
that is placed as an informational reference.

[Med] Not sure there is something broken in this process that is exercised 
since at least 2010. We need to trust the responsible ADs to do it right when 
major changes are to be made (consult with netmod, iesg, etc.).



 If it
> is a BCP
> level thing (beyond what is in 3.7.1 below) then please cover
> inline and
> remove the URL.
>
> 847     3.8.  IANA Considerations Section
>
> 849        In order to comply with IESG policy as set forth in
> 850
> <https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2F
> www.ietf.org<http://www.ietf.org/>%2Fid-
> info%2Fchecklist.html&data=05%7C02%7Cmohamed.boucadair%40orange.com<http://40orange.com/>
> %7C9b04b9803c344846ae4908dda2646796%7C90c7a20af34b40bfbc48b9253b6f5
> d20%7C0%7C0%7C638845274308491793%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU
> 1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsI
> ldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=qLf%2Fz4Eu%2BzB2Lm7ZIQrn3DcE7ldb
> eA5dZGnMvCj65RI%3D&reserved=0>, every I-D that is
> 851        submitted to the IESG for publication MUST contain an
> IANA
> 852        Considerations section.  The requirements for this
> section vary
>
> <major> The text gives an impression of referencing an IETF webpage
> (that is
> not updated via IETF consensus process) for a  normative (MUST)
> directive. More
> problematic is that it is not scoped to YANG documents alone. Given
> that the
> URL is already present in the template, would there be an issue if
> the text
> were to simply say "Every I-D specifying a YANG module that is
> submitted to the
> IESG for publication MUST contain an IANA Considerations section."
> ?

[Med] Even if this was in 8407/6087, removed that text.

KT> Thanks. But please note that I had not asked for its removal and instead 
provided an alternative text. Tomorrow, even if something similar comes in 
8126bis, it would have been helpful to retain the text mandating IANA 
considerations in this document as well since it is going to be referenced by 
YANG authors/reviewers.


>
> 884     3.8.2.  Documents That Extend an Existing Namespace
>
> 886        It is possible to extend an existing namespace using a
> YANG submodule
> 887        that belongs to an existing module already administered
> by IANA.  In
> 888        this case, the document containing the main module MUST
> be updated to
> 889        use the latest revision of the submodule.
>
> <major> What is exactly meant by "the document containing the main
> module MUST
> be updated"? That "document" has already been published as an RFC
> and that
> module is now being maintained by IANA on its website. Is this
> something that
> IANA needs to take care of by itself (I guess this is the case)? Or
> is it
> something that each new document doing such updates need to ask
> IANA to do
> this? Please clarify.

[Med] This section is under "3.8. IANA Considerations Section" which covers 
considerations in a document that extends the existing namespace.

KT> I am sorry, but it does not clarify. Could you please explain the intent 
here? Perhaps with an example?
[Med] I meant this section about the content of IANA cons section of a 
specification that extends an existing namespace.


>
> 954     3.10.  Validation Tools
>
> 956        All modules need to be validated before submission in an
> I-D.  The
> 957        'pyang' YANG compiler is freely available from GitHub:
>
> 959
> <https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2F
> github.com<http://github.com/>%2Fmbj4668%2Fpyang&data=05%7C02%7Cmohamed.boucadair%40ora
> nge.com<http://nge.com/>%7C9b04b9803c344846ae4908dda2646796%7C90c7a20af34b40bfbc48b9
> 253b6f5d20%7C0%7C0%7C638845274308504790%7CUnknown%7CTWFpbGZsb3d8eyJ
> FbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTW
> FpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=IM3BY3o8Q2%2BUVR0WGUo0HXt
> JhqWmnRYzrQQn1v0HCEs%3D&reserved=0>
>
> <minor> Should these GitHub links not be moved into the informative
> reference
> section?

[Med] Preserved the use from 8407.

>
> 1053    4.1.  Module Naming Conventions
>
> 1055       Normative modules contained in Standards Track documents
> MUST be
> 1056       named according to the guidelines in the IANA
> Considerations section
> 1057       of [RFC6020].
>
> <major> The referenced RFC talks about "IETF stream documents" and
> not
> Standards Track documents alone. Could you please review all
> occurrences of
> reference to standards track document to review for correctness?
> They should
> also cover experimental and informational?

[Med] This part is untouched from 8407. The key part of the text you quoted is 
"normative modules". Other types (e.g., example modules) are not concerned to 
follow the full guidance.

KT> I think you are missing my point. The following is the text from RFC6020:


   The module name prefix 'ietf-' is reserved for IETF stream documents

   [RFC4844<https://www.rfc-editor.org/rfc/rfc4844>], while the module name 
prefix 'irtf-' is reserved for IRTF

   stream documents.  Modules published in other RFC streams MUST have a

   similar suitable prefix.

Why is this document then narrowing it down only to Standards Track documents? 
Note, you are focussing on "normative modules", but I am discussing the 
document track which is AFAIK independent.



>
> 1605    4.7.  YANG Definition Lifecycle Management
>
> 1607       The YANG status statement MUST be present within a
> definition if its
> 1608       value is "deprecated" or "obsolete".  The status SHOULD
> NOT be
> 1609       changed from "current" directly to "obsolete".  An
> object SHOULD be
> 1610       available for at least one year with a "deprecated"
> status before it
> 1611       is changed to "obsolete".
>
> <major> This at least one year limit - does it include time as an
> I-D or it is
> so that it can be changed by another RFC that gets published one
> year after
> the publication of the previous one. Can you please clarify?

[Med] Idem as previous one, this is inherited from 8407. As clarified in that 
same section, the reasoning is vs. published module, which is defined as 
follows:

   o  published: A stable release of a module or submodule.  For
      example, the "Request for Comments" described in Section 2.1 of
      [RFC2026] is considered a stable publication.

KT> That does not provide any clarity. The text should state clearly the 
procedure to deprecate and obsolete. This is something that perhaps gets 
exercise more often than what would be ideal, so it is important to cover it.
[Med] Added « publication date » per a suggestion from Éric.


>
> 1656       The "organization" statement MUST be present.  If the
> module is
> 1657       contained in a document intended for IETF Standards
> Track status,
> 1658       then the organization SHOULD be the IETF working group
> (WG) chartered
> 1659       to write the document.  For other standards
> organizations, a similar
> 1660       approach is also suggested.
>
> <major> This is another instance which talks only about standards
> track - what
> about the others? Also, not all standards track need to come via WG
> - some may
> be AD sponsored? Why not allow "IETF" alone for AD sponsored?

[Med] This is not excluded. This is covered as exception to the SHOULD. Added 
"Exceptions may be example modules, IANA-maintained modules, or modules 
contained in AD-sponsored documents."

KT> My original question is not answered. And the newly proposed text raises 
further questions. The IANA maintained modules are also originated via IETF 
standards track document - so why should they be an exception?
[Med] because IANA controls the registries! A YANG module is not more than 
another for to represent the content of a registry.

The AD-sponsored documents can also be IETF standards track - here the only 
exception is that they don't have IETF WGs. What still remains uncovered is 
experimental and informational ...
[Med] I don’t follow you here. The exception is in reference to ” SHOULD be the 
IETF working group”. There is no WG for AD-sponsored docs.


>
> 1824       Note that a different URN prefix string SHOULD be used
> for modules
> 1825       that are not Standards Track.  The string SHOULD be
> selected
> 1826       according to the guidelines in [RFC7950].
>
> <major> Which specific section of RFC7950 is being referenced here?
> And what
> is the guideline for experimental and informational track documents
> with YANG
> modules?

[Med] Updated to reference Section 5.3, where we have:

   XML namespaces for private modules are assigned by the organization
   owning the module without a central registry.  Namespace URIs MUST be
   chosen so they cannot collide with standard or other enterprise
   namespaces -- for example, by using the enterprise or organization
   name in the namespace.

KT> Thanks. But there is no guidance at that location for the namespace for 
experimental/information. There is a further redirection to 
https://www.rfc-editor.org/rfc/rfc6020#section-14 but that says "ietf-" is 
available for all of IETF stream and now this document is restricting it to 
standard track alone?


   The module name prefix 'ietf-' is reserved for IETF stream documents

   [RFC4844<https://www.rfc-editor.org/rfc/rfc4844>], while the module name 
prefix 'irtf-' is reserved for IRTF

   stream documents.  Modules published in other RFC streams MUST have a

   similar suitable prefix.



>
> 3154          The initial version of this YANG module is part of
> RFC IIII; see
> 3155          the RFC itself for full legal notices.
>
> <minor> This RFC IIII is used here and other places in the context
> of the IANA
> related aspects. The use is described in the following paragraph
> but can you
> also please cover in Section 2?

[Med] Happily. Done.

>
> 3196    4.30.3.  Guidance for Writing the IANA Considerations for
> RFCs Defining
> 3197             IANA-Maintained Modules
>
> <major> What is missing is guidance for future documents (i.e. not
> RFC IIII)

[Med] There won't be any ;-) This is the beauty of IANA-maintained modules.

KT> Exactly! But you should not be surprised that this may not be as commonly 
known. Therefore, adding a sentence in this document saying "there isn't 
anything to do" would help at least some folks like to point to during reviews.


> that allocate code points from a registry that is associated with
> an
> IANA-maintained YANG module. I guess the instruction for such a
> document is to
> not give any specific instruction related to such a module (e.g.,
> don't try to
> repeat the updated module in appendix or such?) - all of this
> should be taken
> care of by IANA automatically based on instructions provided in RFC
> IIII ?
>
> 3257       *  A note that unassigned or reserved values must not be
> present in
> 3258          the IANA-maintained module.
>
> <major> Is that a MUST NOT? There is a mix of lower and upper case
> normative
> language. Some consistency would be good.

[Med] The note will be included in IANA consideration section. The current 
wording is consistent with the guidance fro that section at: 
https://datatracker.ietf.org/doc/statement-iesg-statement-on-clarifying-the-use-of-bcp-14-key-words/#inappropriate-uses-of-key-words

KT> You are correct. I misread that part.

Thanks,
Ketan



>
> 3623    7.1.  Normative References
>
> 3625       [ID-Guidelines]
> 3626                  IETF, "Guidelines to Authors of Internet-
> Drafts", n.d.,
> 3627
> <https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2F
> authors.ietf.org<http://authors.ietf.org/>%2Fen%2Fcontent-guidelines-
> overview&data=05%7C02%7Cmohamed.boucadair%40orange.com<http://40orange.com/>%7C9b04b9803c
> 344846ae4908dda2646796%7C90c7a20af34b40bfbc48b9253b6f5d20%7C0%7C0%7
> C638845274308518244%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWU
> sIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%
> 3D%7C0%7C%7C%7C&sdata=FQf7Fjalpe9gC57J1hzJvUkDGELGP9Mk2g4tioa%2BkPM
> %3D&reserved=0>.
>
> <major> Is it normal for this to be a normative reference to an
> IETF webpage
> that is not updated via IETF consensus? Why not move it to
> informative
> references?

[Med] This is inherited from 8407. I think that we can move it around as we 
softened now the generic I-D guidance.

>
> <EoRv25>
>
>

____________________________________________________________________________________________________________
Ce message et ses pieces jointes peuvent contenir des informations 
confidentielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu ce 
message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages 
electroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deforme ou 
falsifie. Merci.

This message and its attachments may contain confidential or privileged 
information that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and delete 
this message and its attachments.
As emails may be altered, Orange is not liable for messages that have been 
modified, changed or falsified.
Thank you.
____________________________________________________________________________________________________________
Ce message et ses pieces jointes peuvent contenir des informations 
confidentielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu ce 
message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages 
electroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deforme ou 
falsifie. Merci.

This message and its attachments may contain confidential or privileged 
information that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and delete 
this message and its attachments.
As emails may be altered, Orange is not liable for messages that have been 
modified, changed or falsified.
Thank you.
_______________________________________________
netmod mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to