Re: [lisp] Genart last call review of draft-ietf-lisp-rfc8113bis-01

Brian, 

Please see inline. 

Cheers,
Med

> -----Message d'origine-----
> De : Brian E Carpenter [mailto:brian.e.carpenter@gmail.com]
> Envoyé : jeudi 20 décembre 2018 20:56
> À : Dino Farinacci; BOUCADAIR Mohamed TGI/OLN
> Cc : Joel M. Halpern; gen-art@ietf.org; lisp@ietf.org; draft-ietf-lisp-
> rfc8113bis.all@ietf.org
> Objet : Re: [lisp] Genart last call review of draft-ietf-lisp-rfc8113bis-01
> 
> I may be missing something but I don't see how 8113bis can
> logically cite 8113, which it replaces.
> 

[Med] The change is for 6833bis NOT 8113bis. 6833bis already cites 8113, which describes the rules for assigning new types. 

> Frankly I think you've collectively created a plate of citation
> spaghetti by not moving the IANA considerations for the type field
> registry into 6830bis, which is where they naturally belong. If you
> don't want to do that, I think you have to leave them in 8113bis and
> simply lose the citation of 6833bis, which serves no purpose that
> I can see.
> 
> Regards
>    Brian
> 
> On 2018-12-21 06:32, Dino Farinacci wrote:
> > I’ll make that change if Brian thinks it fixes the issues he raised.
> >
> > Dino
> >
> >> On Dec 19, 2018, at 11:35 PM, <mohamed.boucadair@orange.com>
> <mohamed.boucadair@orange.com> wrote:
> >>
> >> Hi Dino,
> >>
> >> OLD:
> >>
> >>   Values in the "Not Assigned" range can be assigned according to
> >>   procedures in [RFC8126].
> >>
> >> NEW:
> >>
> >>   Values in the "Not Assigned" range can be assigned via Standards
> >>   Action [RFC8113].
> >>
> >> Cheers,
> >> Med
> >>
> >>> -----Message d'origine-----
> >>> De : Dino Farinacci [mailto:farinacci@gmail.com]
> >>> Envoyé : mercredi 19 décembre 2018 19:00
> >>> À : BOUCADAIR Mohamed TGI/OLN
> >>> Cc : Joel M. Halpern; Brian E Carpenter; gen-art@ietf.org; lisp@ietf.org;
> >>> draft-ietf-lisp-rfc8113bis.all@ietf.org
> >>> Objet : Re: [lisp] Genart last call review of draft-ietf-lisp-rfc8113bis-
> 01
> >>>
> >>> What does fixing in (1) mean?
> >>>
> >>> Dino
> >>>
> >>>> On Dec 19, 2018, at 3:51 AM, <mohamed.boucadair@orange.com>
> >>> <mohamed.boucadair@orange.com> wrote:
> >>>>
> >>>> Hi all,
> >>>>
> >>>> Brian, whether to maintain the document standalone was discussed by the
> WG.
> >>> You may refer, for example, to the message from Deborah which clarifies
> this
> >>> point: https://www.ietf.org/mail-archive/web/lisp/current/msg07886.html.
> One
> >>> of the outcomes of that discussion is to add an "updates" header to
> 8113bis.
> >>>>
> >>>> FWIW, one of the issues that led to that conclusion was whether to cite
> >>> rfc8113bis as normative in 6833bis (the approach I initially supported)
> and
> >>> agreed by Dino (https://www.ietf.org/mail-
> >>> archive/web/lisp/current/msg07882.html). Deborah convinced me that citing
> >>> 8113bis will lead to circular dependency. Which is a fair argument.
> >>>>
> >>>> The "updates" tag was justified as follows:
> >>>>
> >>>> (1)
> >>>>
> >>>> RFC6833bis includes the following:
> >>>>
> >>>>  Values in the "Not Assigned" range can be assigned according to
> >>>>  procedures in [RFC8126].
> >>>>
> >>>> That text is updated by RFC8113bis to be aligned with 8113:
> >>>>
> >>>>  Values can be assigned via Standards Action
> >>>>
> >>>> (2)
> >>>>
> >>>> RFC8113bis extends the type field to grab more bits/values when the
> >>> available types are exhausted. This is captured in 8113bis:
> >>>>
> >>>>  The values in the range 0-1023 are assigned via Standards Action.
> >>>>  This range is provisioned to anticipate, in particular, the
> >>>>  exhaustion of the LISP Packet types.
> >>>>
> >>>> Dino: If (1) is fixed directly in RFC6833bis, then I'm fine to remove
> the
> >>> "updates" header because (2) can be also seen as an extension.
> >>>>
> >>>> Cheers,
> >>>> Med
> >>>>
> >>>>> -----Message d'origine-----
> >>>>> De : Dino Farinacci [mailto:farinacci@gmail.com]
> >>>>> Envoyé : mercredi 19 décembre 2018 06:37
> >>>>> À : Joel M. Halpern
> >>>>> Cc : Brian E Carpenter; gen-art@ietf.org; lisp@ietf.org; draft-ietf-
> lisp-
> >>>>> rfc8113bis.all@ietf.org
> >>>>> Objet : Re: [lisp] Genart last call review of draft-ietf-lisp-
> rfc8113bis-
> >>> 01
> >>>>>
> >>>>> Mohmad to comment.
> >>>>>
> >>>>> Dino
> >>>>>
> >>>>>> On Dec 18, 2018, at 8:49 PM, Joel M. Halpern <jmh@joelhalpern.com>
> wrote:
> >>>>>>
> >>>>>> That is the other fix he offered.  Just remove the updates tag.
> >>>>>> I will leav eit to you and the the authors to determine which is
> correct.
> >>>>>> Yours,
> >>>>>> Joel
> >>>>>>
> >>>>>> On 12/18/18 11:43 PM, Dino Farinacci wrote:
> >>>>>>> 8113bis should say that is it *extending* the type field so we can
> have
> >>>>> more types. The word “update” I always had a problem with because it
> can
> >>> be
> >>>>> interpreted as “replacing". Replacing something to fix a problem.
> >>>>>>> 8113 is simply asking for one of the type value codepoint, so there
> can
> >>> be
> >>>>> another format to have more types.
> >>>>>>> Dino
> >>>>>>>> On Dec 18, 2018, at 9:24 PM, Joel M. Halpern <jmh@joelhalpern.com>
> >>> wrote:
> >>>>>>>>
> >>>>>>>> Authors: that sounds like a reasonable addition to me?
> >>>>>>>>
> >>>>>>>> Yours,
> >>>>>>>> Joel
> >>>>>>>>
> >>>>>>>> On 12/18/18 10:48 PM, Brian E Carpenter wrote:
> >>>>>>>>> On 2018-12-19 15:46, Joel M. Halpern wrote:
> >>>>>>>>>> This is part of the package to move the coherent set of base LISP
> >>> specs
> >>>>>>>>>> to PS.
> >>>>>>>>>>
> >>>>>>>>>> The reason we did this rather than folding it into 6830bis /
> 6833bis
> >>> is
> >>>>>>>>>> that we had originally simply cited 8113, and then realized that
> >>> needed
> >>>>>>>>>> to move to PS along with everything else.  It seemed (and is)
> simpler
> >>>>> to
> >>>>>>>>>> do it separately rather than to further modify 6830bis / 6933bis.
> >>>>>>>>>>
> >>>>>>>>>> As for why it updates 6833bis, that is because one of the cahnges
> in
> >>>>>>>>>> moving the set to PS was to improve the split as to which
> information
> >>>>>>>>>> belonged in which document.
> >>>>>>>>> OK, but I still don't find it logical The text doesn't explain
> which
> >>>>> part of
> >>>>>>>>> 6833bis is impacted, and normally these days we require such an
> >>>>> explanation.
> >>>>>>>>> And if there is an impact, you're missing the opportunity of fixing
> >>> the
> >>>>> error
> >>>>>>>>> or gap in 6833bis, so the reader of 6833bis will be none the wiser
> >>>>> unless
> >>>>>>>>> you insert a reference to 8113bis.
> >>>>>>>>> On the other hand, if there is no error or gap, you don't need
> >>>>> "Updates:"
> >>>>>>>>> at all. (Unfortunately, we don't have an "Extends:" header.)
> >>>>>>>>>  Brian
> >>>>>>>>>>
> >>>>>>>>>> Yours,
> >>>>>>>>>> Joel
> >>>>>>>>>>
> >>>>>>>>>> On 12/18/18 9:25 PM, Brian Carpenter wrote:
> >>>>>>>>>>> Reviewer: Brian Carpenter
> >>>>>>>>>>> Review result: Ready with Issues
> >>>>>>>>>>>
> >>>>>>>>>>> Gen-ART Last Call review of draft-ietf-lisp-rfc8113bis-01
> >>>>>>>>>>>
> >>>>>>>>>>> I am the assigned Gen-ART reviewer for this draft. The General
> Area
> >>>>>>>>>>> Review Team (Gen-ART) reviews all IETF documents being processed
> >>>>>>>>>>> by the IESG for the IETF Chair.  Please treat these comments just
> >>>>>>>>>>> like any other last call comments.
> >>>>>>>>>>>
> >>>>>>>>>>> For more information, please see the FAQ at
> >>>>>>>>>>> <http://wiki.tools.ietf.org/area/gen/trac/wiki/GenArtfaq>.
> >>>>>>>>>>>
> >>>>>>>>>>> Document: draft-ietf-lisp-rfc8113bis-01.txt
> >>>>>>>>>>> Reviewer: Brian Carpenter
> >>>>>>>>>>> Review Date: 2018-12-19
> >>>>>>>>>>> IETF LC End Date: 2018-12-27
> >>>>>>>>>>> IESG Telechat date:
> >>>>>>>>>>>
> >>>>>>>>>>> Summary: Ready with issues
> >>>>>>>>>>> --------
> >>>>>>>>>>>
> >>>>>>>>>>> Comments:
> >>>>>>>>>>> ---------
> >>>>>>>>>>>
> >>>>>>>>>>> I note that this is being raised from Experimental to the
> standards
> >>>>> track.
> >>>>>>>>>>> Presumably that depends on the base LISP spec becoming PS.
> >>>>>>>>>>>
> >>>>>>>>>>> Minor issues:
> >>>>>>>>>>> -------------
> >>>>>>>>>>>
> >>>>>>>>>>> "This document updates I-D.ietf-lisp-rfc6833bis." The text
> doesn't
> >>>>>>>>>>> explain which text is updated. This is in contrast to RFC8113,
> which
> >>>>>>>>>>> explains clearly how it updates RFC6830 (*not* RFC6833). Why
> doesn't
> >>>>>>>>>>> this draft claim to update rfc6830bis? I'm going to assume that
> >>>>>>>>>>> is an error.
> >>>>>>>>>>>
> >>>>>>>>>>> In fact, why wasn't the definition of the LISP Packet Types
> registry
> >>>>>>>>>>> moved into the base spec (rfc6830bis)? That is where it belongs.
> >>>>>>>>>>>
> >>>>>>>>>>> Since rfc6830bis (and rfc6833bis) are still under IESG review,
> >>>>> anything
> >>>>>>>>>>> in them that needs updating should be updated! The fact is that
> >>>>> rfc8113bis
> >>>>>>>>>>> extends rfc6830bis, which is not the same thing as "updates".
> >>>>>>>>>>> If the WG thinks that implementers of 6830bis need to read
> 8113bis,
> >>>>>>>>>>> there should be a normative reference in 6830bis to 8113bis.
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>
> >>>>>>>> _______________________________________________
> >>>>>>>> lisp mailing list
> >>>>>>>> lisp@ietf.org
> >>>>>>>> https://www.ietf.org/mailman/listinfo/lisp
> >>>>
> >>
> >
> >