[manet] Re: MANET WG Charter and MANET Internetworking

Christopher, turning back now to your initial post:

> -----Original Message-----
> From: Christopher Dearlove <[email protected]>
> Sent: Thursday, June 26, 2025 11:46 AM
> To: Donald Eastlake <[email protected]>
> Cc: [email protected] List <[email protected]>; [email protected]
> Subject: Re: MANET WG Charter and MANET Internetworking
> 
> With regard to the charter, there are two protocols the WG created that aren’t mentioned. The first is NHDP. That originated as part of OLSR,
> and you can consider it to be part of OLSRv2 and thus not needing specific reference. But it was separated out because some people thought
> they had uses for it outside OLSRv2. Should it thus be explicitly mentioned alongside OLSRv2 as a supported protocol? (I could go either way
> on that, but I think it worth considering.)

The OMNI approach uses adaptation layer encapsulation of IPv6 ND messages sent over a
virtual link spanning the MANET. So, functions like Address Resolution, Neighbor Coordination
and Neighbor Unreachability Detection occur in the same fashion as they would on an
ordinary IPv6 link (e.g., "virtual Ethernet"). I'm not sure whether that means that the IPv6
ND over virtual link approach would be used in addition to, or instead of, NHDP?

> The second is SMF. Do we consider this an experiment that has run its course?

I think SMF should remain on the table as a potentially useful too going forward.

> (There’s also the format in RFC 5444, but not changing that I think is a good thing. But note RFC 8245.)
> 
> With regard to Fred’s proposed text, I think it’s too specific, it’s presupposing part of the solution.
> 
> But considering the broader question should there be work on handling MANETs as anything other than fixed attachment stub networks? Or
> even just ensuring those work seamlessly? The obvious answer is yes, that’s something the WG should cover. But it’s worth asking why, in that
> case, why was the last attempt in this area separated out as a separate WG (Autoconf)? I think the key issue was that the IETF was quite
> happy with a MANET WG that proposed solutions in its area - but as long as it didn’t mean something had to be done in the rest of the
> Internet. A new address type use only in a MANET was OK (we’ve already got one), but Autoconf’s main issue - and really all it achieved - was
> to keep people happy that MANETs didn’t break the “all the world is ethernet” view of subnetworks, I.e. that they are all informed. It was
> quite an uphill struggle to convince people that radio isn’t like that. The solution of course was that MANET nodes are routers, and subnets can
> be within platforms, but not across platforms,s - except that possibly the whole MANET can look like a subnet to the outside world. But how?
> And here was where Autoconf ran into the ground and Fred is right, there is work here. Which might go down Fred’s route, but that’s not the
> starting point. What existing art is there?

It might be useful to begin by looking at the archived Autoconf WG charter captured at the time that group closed:

https://datatracker.ietf.org/wg/autoconf/about/

They clearly saw a path forward that included DHCPv6 operation over MANET, and that is the same path that
the AERO/OMNI approach is taking. But, the WG decided to declare success with the publication of RFC5889
and shut down before any new work was taken on. Many people involved then stopped work on Autoconf
and went off to do other things. I did not stop, and AERO/OMNI is the result.

> All of course, at least in my opinion and recollection.
> 
> As ever of course, the killer question for the WG is that it’s easy to say what MANET could do - I just did - but the issue is who will do the work.
> And for how long. Work out how long you think the work should take and it will be longer, and not just slightly longer. We all know that, but
> think this time it will be better. On which subject there’s an anecdote that just was so familiar in Kahnemann’s Thinking Fast and Slow. (On a
> syllabus, not a protocol, but the point is the same.)

I think it is wise to make conservative estimates of how long work should take. But, this
would be a "build on substantial existing work" rather than "start from scratch" scenario.
And, by existing work I mean to also include all of the good work done in MANET over the years.

> As for me, I can advise (and even write one draft) on OLSRv2 (including NHDP and other related matters) but unless a long shot happens - and
> it hasn’t for nearly a year - that’s about my limit.

It would be good to have whatever energy you might be able to contribute, yes.

Thank you - Fred

> 
> Christopher
> 
> 
> > On 26 Jun 2025, at 18:38, Donald Eastlake <[email protected]> wrote:
> >
> > Hi,
> >
> > We have been working on a draft new Charter for the MANET WG for some
> > time. One question that needs to be resolved is whether, as a WG, we
> > want the scope to include the questions Fred Templin has presented,
> > for example here:
> > https://datatracker.ietf.org/meeting/122/materials/slides-122-manet-manet-internetworking-01
> > https://datatracker.ietf.org/meeting/121/materials/slides-121-manet-aeroomni-autoconfiguration-services-for-manet-internetworking-00
> >
> > Please respond with your thoughts on this. Note the question is not
> > whether to include a particular solution but whether or not the
> > problem should be covered by the Charter.
> >
> > After we agree on the new Charter, it will still have to go through
> > multiple steps of approval. The most recent draft is here:
> > https://datatracker.ietf.org/meeting/122/materials/slides-122-manet-draft-charter-text-ietf-122-01
> > Comments on any aspect of the draft are welcome.
> >
> > Thanks,
> > Donald
> > ===============================
> > Donald E. Eastlake 3rd   +1-508-333-2270 (cell)
> > 2386 Panoramic Circle, Apopka, FL 32703 USA
> > [email protected]
> >
> > _______________________________________________
> > manet mailing list -- [email protected]
> > To unsubscribe send an email to [email protected]
> 
> _______________________________________________
> manet mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
_______________________________________________
manet mailing list -- [email protected]
To unsubscribe send an email to [email protected]