Re: [homenet] Alvaro Retana's Discuss on draft-ietf-homenet-babel-profile-06: (with DISCUSS and COMMENT)

[Alvaro Retana <[e-mail]>]

Barbara:

Hi!

Thanks for the response!  This (along with the confirmation from Stephen
and Terry) was all the confirmation I wanted.  I agree with your points and
the rationale.

I’m clearing my DISCUSS.

Thanks!

Alvaro.

On May 9, 2018 at 6:19:56 PM, STARK, BARBARA H ([e-mail]) wrote:

Hi Alvaro,
Thanks for the comments. I agree we need to make sure the community is
comfortable with the decision. I've inserted my comments in-line.
Barbara

> I would like to DISCUSS about the Intended Status of this document --
with
> the Chairs and AD.
>
> I have to confess that I haven't been following the homenet WG as closely
as
> I
> probably should have. Hopefully that means that the topic below has been
> discussed and documented already, making this DISCUSS easy to resolve.
>
> Back in 2015, the then-Chairs and the AD posted a note titled "Routing
Design
> Team outcome and next steps"[1]; in it, they declared "rough consensus
that
> Babel[*] shall be the “mandatory to implement” routing protocol for
> Homenet
> routers, albeit only on an Experimental basis at this time...we solicit
> Experimental Internet Drafts to document Homenet-specific profiles of any
> applicable routing solution and to report results of any relevant
> experimentation and implementation. We expect that this decision will be
> revisited in a future Standards Track document based on specifications
and
> running code available at that time."
>
> My interpretation of the above text is that Babel is MTI,

<bhs> There is no draft or published RFC that proposes making Babel MTI.
The language of the quoted text is a bit odd (a MTI experimental
protocol?), which makes it open to interpretation. My interpretation
differs from yours. My interpretation was that people actively working to
create (experimental) code for homenet stacks needed to implement Babel if
they wanted to test out interoperability with other stacks. Those
experimental stacks were created. Interoperability was tested. We have
running code and a standards-track Babel draft (rfc6126bis) now. I don't
believe a WG discussion has the ability to make a protocol MTI in the
absence of a published RFC that states it is MTI in some context. No draft
exists or has ever been submitted to propose making Babel MTI.

> but that the work
> (documents) will be Experimental...and that this decision could change in
the
> future (most likely towards confirming and moving to the Standards
Track).
> This document was originally adopted as Experimental. I didn't find an
> explicit discussion on the list about changing that original overall
direction,
> nor another declaration by the Chairs/AD. I did find find a thread in
which
> one of the Chairs (Barbara) asked about the status for this document (and
> this document only)[2]; the initial question was framed around the
references
> being Standards Track documents (HNCP and rfc6126bis) -- just one answer
came
> back (from the author of this document)...
>
> I'm treating this point as a DISCUSS because I think that the WG
consensus
> may
> have not been determined to change the original declaration from the
> Chairs/AD
> (from 2015). In my interpretation of that original declaration, moving
Babel
> to the Standards Track means a recognition that it will be *the* protocol
> going
> forward (which changes that initial "only on an Experimental basis at
this
> time" phrase), is something that should be discussed explicitly, and not
just
> in light of this one document. That is the part that I haven't seen.

<bhs> I consider the lack of alternate proposals as implying there are no
technical objections to Babel being *the* protocol. It's hard not to be
*the* protocol when you're the only protocol. Whether or not the
homenet-babel-profile is Standards track, it is *the* protocol in the
absence of others. And after 2 years of no other proposal being submitted
and multiple Babel profile implementations, I suspect WG participants might
push back heavily against adding another protocol. Another protocol would
face a serious uphill battle. And this would be true independent of the
status attached to the Babel profile. Two years is plenty of time to have
submitted a draft.

But there is nothing in the language of the Babel profile that would
preclude submission of another protocol profile.

> I note that in the conclusion of the thread about the status of this
document
> [3] Barbara does include reasoning that may result in changing the
original
> declaration (as does the Shepherd writeup), for example: "there exist
> multiple,
> interoperable implementations" and "no drafts proposing other homenet
> routing
> protocol profiles have been submitted"...but those points don't seem to
> have
> been considered/discussed by the WG (they were not in the original
> message and
> I didn't find another thread -- I also looked at the minutes of the last
couple
> of IETF meetings).
>
> To be clear, I have no objection with Babel being used in homenet
> applications,
> or with it being the Standard protocol. My point here is that it is not
clear
> to me that the WG explicitly reached consensus to change the declaration
> from
> the Chairs/AD. I will be happy to clear this DISCUSS when the Chairs/AD
> point
> me to the discussion that I missed, or simply tell me that the
declaration from
> 2015 is no longer valid and that the WG knows, or that they believe that
the
> thread discussing this document is enough to call consensus...or
something
> to that effect.

I did consider the emails you pointed to as having adequately advised the
WG of the proposed change and invited discussion and/or objection. I did
also get a private email response pointing to the 2015 decisions you noted
but not objecting to the proposed status change. I discussed with Stephen
(co-chair) and Terry (AD) who both indicated they were satisfied with the
lack of objection to my emails. It is my belief that members of the WG are
very, very aware (many still bear deep scars) of the 2015 discussion. They
are also aware that the WG has been open to receiving other routing
protocol profile proposals for 2 years now, and that none has been
received. Several people told me privately they were glad the status was
being changed so we could get this over with (and that homenet was already
in danger of being obsolete and losing everyone if it couldn't move on).
They didn't want to say *anything* on the email thread, for fear that it
might actually trigger re-opening the discussion in all it's glory. There
is a strong desire not to have that discussion again. I do believe that,
given the hyper-awareness of the 2015 discussion, the WG knew what it was
agreeing to (or not objecting to). I believe the WG is knowingly saying it
wants to move on. Naming and security need solutions, and we need to be
working on those.

But if you still think I need to explore this some more with the WG, I
will.
Barbara

> [1] https://urldefense.proofpoint.com/v2/url?u=https-
> 3A__mailarchive.ietf.org_arch_msg_homenet_kiI7pIYfpgT2Qrfx1VBAwng7-
> 5FQY&d=DwIDaQ&c=LFYZ-o9_HUMeMTSQicvjIg&r=LoGzhC-
> 8sc8SY8Tq4vrfog&m=LS5wCRK83QbSm0A7c-uQ50-sORylXeseG-
> zQI_D8ABU&s=3czWjkOM36nRroRWTFHYjcWEZF_WsTqCHLjLrGBRxmk&e=
> [2] https://urldefense.proofpoint.com/v2/url?u=https-
> 3A__mailarchive.ietf.org_arch_msg_homenet_5L5WYN14gDCamP7qlknJmW
> keU5M&d=DwIDaQ&c=LFYZ-o9_HUMeMTSQicvjIg&r=LoGzhC-
> 8sc8SY8Tq4vrfog&m=LS5wCRK83QbSm0A7c-uQ50-sORylXeseG-
> zQI_D8ABU&s=TntrJlZrFtZU8GHp86i_BfjU5BPtSn5lqHwDVSEHfv0&e=
> [3] https://urldefense.proofpoint.com/v2/url?u=https-
> 3A__mailarchive.ietf.org_arch_msg_homenet_35EU8oBr8hunvvSRYUStypZI
> PVU&d=DwIDaQ&c=LFYZ-o9_HUMeMTSQicvjIg&r=LoGzhC-
> 8sc8SY8Tq4vrfog&m=LS5wCRK83QbSm0A7c-uQ50-sORylXeseG-
> zQI_D8ABU&s=8kr7ZyZhA8zZEKfNDbVouzU-BmWT-t_WOgi7WcZKYkM&e=
>
>
> ----------------------------------------------------------------------
> COMMENT:
> ----------------------------------------------------------------------
>
> (1) I think that this document walks a fine line when Normatively
referring to
> Appendix A in rfc6126bis given that it is an informative appendix. In
general,
> it does a good job at it -- I do, however, have one nit: "The algorithm
> described in Appendix A of RFC 6126bis MAY be used." I think that
changing
> that to something non-normative would be good, perhaps something like
> "Appendix
> A provides an example of an algorithm..." or simply s/MAY/may
>
> (2) This reminds me; please use rfc8174 template (for Normative
language).
>
> (3) The Non-requirements sections (2.2/3.2) are confusing to me. Maybe
it's
> the "negative logic"...
>
> (3.1) What do these non-requirements represent? Are they requirements
> that
> were considered at some point, but discarded? Using rfc2119 language adds
> to
> the confusion -- consider describing these non-requirements not using it.
>
> NR2, for example, is worded as a requirement that was considered, and the
> rationale explains why not: an "implementation of Babel MAY include
> support for
> other extensions"...this is not a requirement because "with the exception
of
> source-specific routing, no extensions are required". Ok.
>
> (3.2) Are implementers to interpret that the converse is true/expected?
In
> the
> case of NR2, is a true interpretation that implementations SHOULD NOT
> include
> support for other extensions? IOW, while the option of other extensions
is
> not
> a requirement, is not having them one?
>
> (3.3) The non-requirements in §3.2 seem a lot more confusing to me:
>
> (3.3.1) NR3 -- The text says that the requirement not considered
> (non-requirement) is such that "an HNCP node that receives a DHCPv6
prefix
> delegation MAY announce a non-specific IPv6 default route", but the
> rationale
> says that "announcing an additional non-specific route is allowed". I'm
> confused. Is announcing the additional route ok, or not? Is it ok to
assume
> that optionally advertising the additional route is ok? If it's allowed,
then
> why is this a non-requirement?
>
> (3.3.2) For NR4, is the non-requirement, i.e. one that was not
considered,
> that
> the source-specific route SHOULD NOT be announced? This piece is also
> confusing to me because the rationale says (at least the way I read it)
that it
> may be ok to advertise. It seems to me that the text is saying that in
fact
> the route SHOULD NOT (or even MUST NOT be announced)...which brings
> me to the
> question: what is the requirement that was not considered? What am I
> missing?
>