Re: [trill] [Lime] Adoption call for draft-deepak-trill-yang-pm (12/15/2014 to 12/30/2014)

[Donald Eastlake <[e-mail]>]

Hi Carlos,

On Wed, Dec 17, 2014 at 12:46 PM, Carlos Pignataro (cpignata)
<[e-mail]> wrote:
> Hi Donald,
>
> Thanks for the follow-up. Just for clarity, I did not scream “process!”.

Sorry if I misinterpreted you message. I guess it was Qin Wu who said
"it is process issue...".

> I had two main points in my quick review — 1. that key references that
> should be Normative are Informative, and 2. that there is a large number of
> idnits (3 errors, 19 warnings), which severely impact readability and
> understanding of links to other pieces of work.
>
> At the time of WG call for adoption, I do think it’s important to understand
> dependencies and normative ongoing work, to make an adoption determination,
> and therefore I believe it to be important to fix these before adoption. And
> FWIW I disagree this is a “process issue” and the other comment about
> sequentiality of adoption.

I guess I still disagree with you, although somewhat weakly. A draft
can evolve quite a bit from the version adopted to a later version
that is approved so, for example, dependencies can change and, in any
case, at this stage I would depend more on my judgement based on the
draft text than on how the reference was classified by the authors.
(From my point of view, it was not that long ago in the IETF that
references in drafts ad RFCs were just a single list without the
current Normative/Informational classification.) Perhaps it just
depends on the particular draft and situation and reasonable people
could come to different conclusions.

In any case, it seems like these defects in the draft will be fixed
pretty quickly.

Thanks,
Donald
=============================
 Donald E. Eastlake 3rd   [phone] (cell)
 155 Beaver Street, Milford, MA 01757 USA
 [e-mail]

> Hope these help,
>
> Carlos.
>
> On Dec 17, 2014, at 12:04 PM, Donald Eastlake <[e-mail]> wrote:
>
> Hi Carlos and Qin,
>
> Thanks, Carlos, for reviewing and commenting on this draft. I'm sure your
> comments will help improve the draft.
>
> However, I do not agree on the process points. It is important to
> distinguish between adopting a draft, that is taking a personal draft under
> the control of its authors and making it a Working Group draft under the
> control of the Working Group, and the final Working Group consensus
> determination (usually involving a WG Last Call) to request RFC publication
> of a draft.
>
> More polished drafts are obviously better than drafts with lots of nits,
> references incorrectly categorized are Normative or Informative, and the
> like, That sort of thing needs to be fixed before RFC publication is
> requested, but I do not think it has to be fixed before adopting the draft.
> A draft can be adopted by a Working Group even if there are much worse
> problems in the draft, such as serious technical flaws or inconsistencies,
> as long as the Working Group thinks it is a plausible starting point to work
> from. Just adopting the draft is not a guarantee that there will later be
> Working Group consensus to request RFC publication. Of course, you would
> expect a draft to be improved after it has been adopted and before RFC
> publication is requested.
>
> The normative references process point is a bit less clear cut. It is very
> common for drafts to be fully approved all the way through being in the RFC
> Editor's queue despite normative dependencies on other drafts that are at a
> much earlier stage in the process. If you go to
> http://www.rfc-editor.org/queue2.html
> and search for "MISSREF" you will find many fully approved drafts whose
> publication is being held up by normative references to other drafts that
> are not yet approved (or, in some cases, held up by a normative reference to
> an approved draft that in turn normatively references anther draft that is
> not yet approved or even longer normative reference chains with the last
> draft in the chain not approved). The TRILL WG has frequently approved
> drafts with one or more normative dependencies that had not yet been
> approved. On the other hand, if the Working Group feels it is appropriate to
> delay, it certainly can hold up on requesting publication of a draft until
> it is satisfied with the status of normative references from that draft.
>
> Thanks,
> Donald
> =============================
>  Donald E. Eastlake 3rd   [phone] (cell)
>  155 Beaver Street, Milford, MA 01757 USA
>  [e-mail]
>
> On Wed, Dec 17, 2014 at 3:36 AM, Qin Wu <[e-mail]> wrote:
>>
>> Agree with Carlos, it is process issue. Usually a I-D will not be adopted
>> until all the normative referenced drafts get proceeded.
>>
>>
>>
>> Regards!
>>
>> -Qin
>>
>> 发件人: Lime [mailto:[e-mail]] 代表 Carlos Pignataro (cpignata)
>> 发送时间: 2014年12月16日 23:03
>> 收件人: Susan Hares
>> 抄送: Donald Eastlake; [e-mail]; [e-mail]; [e-mail]
>> 主题: Re: [Lime] Adoption call for draft-deepak-trill-yang-pm (12/15/2014 to
>> 12/30/2014)
>>
>>
>>
>> Sue, Jon,
>>
>>
>>
>> I have a few concerns with this document:
>>
>> The Abstract says “It extends Yang Data Model for TRILL OAM
>> [TRILLOAMYANG]”, however, [TRILLOAMYANG] is listed as an Informative
>> reference — I believe it should be Normative (all the definitions are coming
>> from it, and this doc extends that model)
>>
>> It seems the same issue exists with draft-ietf-trill-yang-oam and
>> [GENYANGOAM], which should be Normative there.
>> By the way, a nit as well (the Abstract includes citations).
>>
>> Ditto for [GENYANGOAM], as the text says “In this document we extend the
>> YANG model defined in [GENYANGOAM]”
>> There’s a non-insignificant set of Nits which frankly should be taken care
>> of before this call for adoption (see
>> http://www.ietf.org/tools/idnits?url=http://www.ietf.org/archive/id/draft-deepak-trill-yang-pm-00.txt)
>>
>> This severely detracts from readability.
>>
>>
>>
>> I believe these are blocking for adoption.
>>
>>
>>
>> Thanks,
>>
>>
>>
>> Carlos.
>>
>>
>>
>> On Dec 15, 2014, at 11:05 PM, Susan Hares <[e-mail]> wrote:
>>
>>
>>
>> This begins a 2 week WG adoption call for draft-deepak-trill-yang-pm-00
>> which can be found at:
>>
>>
>>
>> http://datatracker.ietf.org/doc/draft-deepak-trill-yang-pm/
>>
>>
>>
>> This draft will be reviewed along with other yang modules at the
>> 12/17/2014 interim of trill held from 9-10pm ET (6-7pm PT, 10-11am Beijing
>> on 12/18/2014).  In your review, please indicate the following:  a) if this
>> yang module provides information necessary to manage or monitor TRILL, and
>> b) if you support adoption this work into the TRILL WG.
>>
>>
>>
>> Sue Hares and Jon Hudson
>>
>> Co-chairs
>>
>>
>>
>>
>>
>> _______________________________________________
>> Lime mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/lime
>>
>>
>
>