IETF-115 MANET WG meeting agenda

When: Thursday, November 10, 2022, Session III, 15:30 - 16:30 UTC

Where: Mezzanine 1-4 + online

Meetecho:
https://meetings.conf.meetecho.com/ietf115/?group=manet&short=&item=1
Chat: https://zulip.ietf.org/#narrow/stream/manet
Notepad: https://notes.ietf.org/notes-ietf-115-manet
WG link: https://datatracker.ietf.org/wg/manet/about/
Materials: https://datatracker.ietf.org/meeting/115/session/manet

Chairs: Don Fedyk, Ronald in 't Velt

1. Chairs' Introduction (5 min)

2. Internet-Draft status - chairs (10 min)

• credit-based flow control drafts

Ronald: working on shepherd write-ups, may need to get back to the
authors on some of the questions, notably implementation status.

Ronald: Does the recently introduced requirement for RTG Area early
review also apply to these I-Ds? Alvaro Retana (AD): Yes! WGs can ask
for these reviews at any time before asking for publication. Though not
applicable to the I-Ds anymore, the RTG ADs also asked for early
appointment of document shepherds.

Lou Berger: Suggest to ask for these reviews right now, remaining issues
from TSV ART review will not lead to significant changes.

Ronald: Revisiting the question of merging or not merging of drafts one
last time. Has been discussed at length. David Black (TSV ART reviewer)
strongly suggested to do so. WG decided against this at IETF 113. David
Black then asked for this decision to be recorded in the shepherd
write-ups, which is going to happen. We stick with 4 documents. However,
I am proposing to move the sub-data item specifications from the traffic
classification draft to -dlep-da-credit-extension and -dlep-ether-credit
extension, respectively. This would add a bit more content to the latter
drafts, as well as make -dlep-traffic-classification completely generic,
so it will not need to change if the WG ever specifies any additional
classification criteria, for example based on 5-tuple or 6-tuple.

Rick Taylor: +1, it's late in the process, but let's get it right.

Lou Berger: Agree on getting it right, but disagree on outcome. What if
we want to do traffic classification independent of credits and flow
control? Flow control and traffic classification are the two independent
building blocks. The other two drafts are intended for expression of
conformance by implementers.

Rick Taylor: That's a convincing argument; I retract my previous
statement.

Ronald: Not sure there will be proposals for other uses of traffic
classification.

Lou Berger: I have several documents waiting, that would use Traffic
Classification separately; I did not bring those to the WG yet, because
the current set of I-Ds are taking so long to progress.

Don: we will have to take this discussion to the mailing list

Lou Berger: Can we make a call now and stick with it? I'll live with
whatever you say.

Ronald: I wasn't aware there was more coming. In my opinion, we should
in that case stick with the current arrangement. However, it's for the
WG to decide.

Lou Berger: Will look for an answer on the ML. Absent that, will keep it
the way it is.

Ronald: Exactly.

• PHY-related individual I-Ds

Ronald: Will start another adoption call for each of these three drafts
on the ML. Hope to get some replies.

3. IETF-114 Joint Babel/ROLL/MANET session follow-up - chairs (10 min)

Alvaro Retana, AD: Action for MANET Chairs to contact the Roll And Babel
Chairs as follow-up. Last bullet on slide on connection to the outside
world, which you mention as apotential item for a new charter, is sort
of obvious to Babel and RIPL. Look for commonality.

4. DTNMA Updates – Sarah Heiner (15 min)

Lou Berger: Trying to understand relationship to RESTconf / COREconf. Is
this an alternate way to transport YANG models or something completely
different.
Sarah: Different than YANG.
Lou Berger: Sounds like you need a new way of transporting versus a
differnent way.
Sarah Agree. Looking for efficiency in encoding.
Lou: YANG is about modeling not about transport.
Ed Birrane: Part of the work is to determine is a useful autonomy model
on a resource-constrained device in a challenged environment. This is
probably a subset of what could be modelled in YANG.

Ronald: I think many MANETs can also be considered resource-constrained
and challenged.

5. Rechartering discussion - all (20 min)

Ronald: We have not addressed the request by Alvaro to discuss Babel and
MANET going forward.
Alvaro: DLEP Maintenance and Extensions. OSLR and DLEP. Babel there is
no where to take on the Babel protocol. MANET makes sense. DTNMA - not
just for MANET. IN DTN where they work with us on their framework.
Multicats may not get that for deployment. Try and focus the work and
get teh work moving.
Abdussalam: Support adding Mulitcast.
Juliusz: Closing Babel makes sense. But it work makes sense where thteis
a place to do Babel Maintenance.

Rick Taylor: What is happening with AOVDv2 ?
Alvaro: Charlie Perkins asked me to AD-sponsor it. He has not had time
to work on it. It is not moving. I am about to give up.
Donald: There is a bit of work finishing in Babel . Maintenance item
would make sense.
Alvaro: to avoid MANET gong the way of the dodo we need work and
contributions.