[11:02:31] --- m_ersue has joined
[11:02:38] --- m_ersue has left
[13:59:18] --- apetrescu has joined
[14:00:41] <apetrescu> Thomas Clausen is TC starts meeting
[14:00:45] <apetrescu> slide: MANEMO
[14:01:04] <apetrescu> slide: Practicalities
[14:01:37] <apetrescu> slide: Agenda
[14:02:17] --- dthaler has joined
[14:03:07] <apetrescu> slide: Introduction and Scope
[14:03:34] --- Andrew McGregor has joined
[14:04:27] <apetrescu> TC: this is not discussion on solution space
[14:04:40] <apetrescu> TC: this is discussion on problem and requirements
[14:05:02] <apetrescu> slide: a Nemo Mobile Router
[14:06:35] --- shubranshu has joined
[14:07:20] --- dku-tzi-org has joined
[14:10:06] <apetrescu> slide: Agenda
[14:10:12] <apetrescu> Teco Boot presenting
[14:10:46] <apetrescu> slide: MANEMO scenarios and requirements
[14:11:38] <apetrescu> (this is very funny, I realize people are sitting outside the room and looking at this room's projected slides :-)
[14:11:52] <apetrescu> slide: Military: Deployed and mobile networks
[14:11:58] --- igarashi has joined
[14:13:09] <dthaler> (is there a URL for the slides?)
[14:14:42] <apetrescu> slide: Public Safety: Metro6, U-2010, PSC-Europe
[14:15:34] <apetrescu> (slides don't seem to be on the IETF site, there isn't an IETF proceedings for IETF69)
[14:17:09] <apetrescu> slide: vehicle to vehicle: C2C-CC, CALM
[14:17:47] <apetrescu> Charles Perkins asked what is RSE?
[14:17:56] <apetrescu> TB: Roadside element, equipment
[14:18:16] <apetrescu> slide: Multiple use cases, similar requirements
[14:19:26] <apetrescu> slide: Communication between nearby nodes
[14:20:14] <apetrescu> slide: LFN to LFN communication
[14:20:52] <apetrescu> slide: Path to Home Agent may be broken
[14:21:31] <apetrescu> slide: LFN to LFN ncommnication
[14:21:50] <apetrescu> questioner: you have bulitlin assumption that somehting that is a mobile has a HA
[14:21:58] <apetrescu> questioner: but my mobile laptop doesnt
[14:22:19] <apetrescu> TC: this situation is related to the fact that _if_ you run NEMO and MIP then you have a HA. If not then different.
[14:22:26] <apetrescu> questioner: no MANET involved?
[14:22:42] <apetrescu> TC: connect to new place you get a new address.
[14:22:47] <apetrescu> (name of questioner?)
[14:23:08] <apetrescu> TB: connect to local nodes, wifi in adhoc mode. There are solutions forr this problem but I repesent here the reqs.
[14:23:14] <apetrescu> questioner: I miss some contact.
[14:23:28] <apetrescu> questioner: in a couple of senenceteces what are we trying to do, what's the purpose.
[14:23:30] <apetrescu> Thomas Narten
[14:23:39] <apetrescu> TN: think of MANEMO: where is the MANET part fit in?
[14:23:50] <apetrescu> TN: this is the generic Mobile IP problem where you're bound to HA
[14:23:58] <apetrescu> TC: in MANEMO there's no assumptions of MANET
[14:23:58] <shubranshu> Requirements presentation has just been uploaded, sorry for the delay.
[14:24:08] <apetrescu> (where? URL?)
[14:24:23] <apetrescu> TC: in the case of Mobile IP disconnect HA - same thing here
[14:24:31] <apetrescu> TN: in this particular case the problemm only occurs.
[14:24:43] <apetrescu> TN: the problme space is in the context of community that knows the MIP protocols
[14:24:43] --- kakima has joined
[14:24:48] <apetrescu> TN: what are the particular problems
[14:24:51] <apetrescu> Jari Arkko
[14:24:58] <apetrescu> TN: slide the problem we can tackle it.
[14:25:08] <apetrescu> JA: if use NEMO then in MIP they have some issues on HA
[14:25:17] <apetrescu> JA: for adhoc routing techniques then they have other issues.
[14:25:32] <apetrescu> JA: want to be independent of infra, but still want to have system survice and IP address stable.
[14:25:41] <apetrescu> TN: solutions are off the tables.
[14:25:51] <apetrescu> TB: good solutions could be used to what you have in mind
[14:26:08] <apetrescu> TC: TN asked the same question that I'm trying to ask: what is the problem we try to solve.
[14:26:26] <apetrescu> TC: clarify: MANEMO means Management of Network Mobility
[14:26:37] <apetrescu> TN: [resumably related to topic
[14:26:48] <apetrescu> JA: dont think it's desried to get...
[14:26:54] <apetrescu> JA: could be either MANET tech or NEMO
[14:27:04] <apetrescu> TN: speak to back of room to people on the hall.
[14:27:29] <apetrescu> JA: they have a business proboem they could use MIP or MANET... trying to find out whether the tech are specific or...
[14:27:43] <apetrescu> JA: dont know whether AUTOCONF or MANET is
[14:27:48] <apetrescu> questioner: who are they?
[14:27:55] <apetrescu> JA: TB is one
[14:28:00] <apetrescu> Jim Bound is JB
[14:28:27] <apetrescu> JB: what happens is issue you have is reactive or proactive
[14:28:50] <apetrescu> JB: if this network is split, the HA is split, a policeman, a doctor a soldier: the best connectivity is through home.
[14:28:59] <apetrescu> JB:explains slide.
[14:29:19] <apetrescu> JB: it's a problem space, people are going to solve it. Uniwireless (mesh networks) people ignore what we do.
[14:29:48] <apetrescu> JB: if you go with MANET there may be an answer, but there may not be. When an intermediary router blows up.
[14:29:59] <apetrescu> JB: the problem is there and MANEMO is trying to solve it with the French.
[14:30:06] <apetrescu> JB: in MANET they try to heal that.
[14:30:12] <apetrescu> TJ Kniveton
[14:30:18] <apetrescu> TJ: opening a wall
[14:30:27] <apetrescu> TJ: getting 30 people out in the hallway .
[14:30:43] <apetrescu> TB: me and others really interested in this kind of problem space.
[14:30:58] <apetrescu> slide: Satcom, UHF, Compoind, truck
[14:32:27] <apetrescu> slide: LFN to LFN communication
[14:33:49] <apetrescu> questioner: what do you mean by 'optimal'?
[14:34:03] <apetrescu> questioner's name sounds as 'lyshia'?
[14:34:41] <apetrescu> TN: isn't this a standard part of looking of an opitmal path? This scenario is common to all scenarios other than m obile too.
[14:34:52] <apetrescu> TB: specific to routing protocols and technologies?
[14:35:01] <apetrescu> TB: we have to analyze it
[14:35:13] <apetrescu> TN: in this picture, terminology right? HA1 and HA2 are home agents?
[14:35:15] <apetrescu> TB: yes
[14:35:24] <apetrescu> TN: is this the MIP Route Optimization problem?
[14:35:42] <apetrescu> TN: clear rules for about is this a MIP RO scenario or a different scenario?
[14:35:48] <apetrescu> TB:yes
[14:35:52] <apetrescu> CP Charles PErkins
[14:36:00] <apetrescu> CP: I thought you asked q in solution space?
[14:36:05] <apetrescu> CP: does RO solve this problem or not?
[14:36:12] <apetrescu> CP: until we figure the problem space we not...
[14:36:30] <apetrescu> CP: of course RO offers many possibilities but we're not quite there yet
[14:36:35] <apetrescu> slide: When using MANEMO ?
[14:37:31] <apetrescu> Fred Templin
[14:37:41] <apetrescu> FT: MANET specifically addresses heterogeneous link technologies.
[14:37:54] <apetrescu> TB: we also look for that kind of soluiontsons, but this is the problem area.
[14:38:03] <apetrescu> TC: anybody else questioning on Teco's presentation ?
[14:38:12] <apetrescu> Ryuji Wakikawa going to present.
[14:38:32] <apetrescu> slide: Architectural Considerations and Relationship within the IETF
[14:39:10] <apetrescu> slide: MANEMO Problems
[14:40:29] <apetrescu> slide: Issue 1: Sub-optimal path...
[14:40:36] <apetrescu> slide: Optimal Path for internal communication
[14:40:58] <apetrescu> slide: MANET/AUTOCONF Applicability
[14:42:02] <apetrescu> slide: Considerations
[14:44:28] <apetrescu> FT: specific use case scenarios and specific link types are what?
[14:44:49] <apetrescu> FT: wifi? wimax? 3g? wll-deba?
[14:45:10] <apetrescu> RW: depends on what is the interface MR has for mobile network, likely 802.11b
[14:45:20] <apetrescu> RW: dont think we can have wimax or 3g for this purpose
[14:45:28] <apetrescu> FT: you think for multihp networks?
[14:45:31] <apetrescu> RW: depends on scenarios
[14:45:55] <apetrescu> RW: some operators may use wifi, some operator may use different wireless devices like for vehicles they have DSRC... other wireless medii,
[14:46:04] <apetrescu> FT: useful to get down to the specific link technologies.
[14:46:09] <apetrescu> FT: is like Ethernet?
[14:46:24] <apetrescu> RW: yes... in a NEMO it doesn't say we should use a specific link tech.
[14:46:32] <apetrescu> FT: it would be useful to know that.
[14:46:48] <apetrescu> FT: on slide. you talk about prefix obtained from u pper MR as oopposed from AR.
[14:46:59] <apetrescu> FT: mobile network moves... re-number.
[14:47:05] <apetrescu> FT: mobile network travel as a group.
[14:47:11] <apetrescu> FT: you can do both I think.
[14:47:15] <apetrescu> RW: more slides later.
[14:47:35] <apetrescu> TC: to answer FT's comment, egress and ingress means whether or not a device is running one or another part of NDP.
[14:47:41] <apetrescu> FT: that ...
[14:47:48] <apetrescu> Andrew McGregor: on link types.
[14:48:05] <apetrescu> AMG: for types: pointotpoint, ether, star-like (wimax) and mesh.
[14:48:20] <apetrescu> slide: MANET/AUTOCONF applicability
[14:49:05] <apetrescu> Joe MAcker
[14:49:13] <apetrescu> JM: clarification
[14:49:28] <apetrescu> JM: NHDP does: sharing information on 2-hop? no, on 1hop but has to be...
[14:49:41] <apetrescu> JM: you gain 2-hop information but only on 1-hop.
[14:49:49] <apetrescu> RW: with a single hop neighbour.
[14:50:03] <apetrescu> Christian Dearlove: actually you could (2-hop).
[14:50:12] <apetrescu> slidE: NHDP vs NDP
[14:52:16] <apetrescu> FT: you're actually about NDP and NHDP on _same_ interface, issues with, right?
[14:52:19] <apetrescu> RW: yes
[14:52:33] <apetrescu> FT: you could run NHDP on one interface and NDP on another, right?
[14:52:34] <apetrescu> RW: yes\
[14:52:46] <apetrescu> slide: Issue 2: sub-optimal ....
[14:53:20] <apetrescu> CP: I dont believe conflicts on reachablity... about NDHP and NDP.
[14:53:33] <apetrescu> CP: if you don't have a subnet. I dont see the conflict, it's just the concept.
[14:53:40] <apetrescu> RW: similar functionalities.
[14:53:53] <apetrescu> Samita Chakrabarti
[14:54:03] <apetrescu> SC: do you need to run same protocols on same interface?
[14:54:15] <apetrescu> RW: if we decide to use MANET on MR then we have to run ND and also the NHDP.
[14:54:23] <apetrescu> SC: you say you need nd for initial address prefixes?
[14:54:45] <apetrescu> RW: if CP agrees we should use NHDP for all MIP protocols then fine, but if MIP needs ND then.
[14:54:56] <apetrescu> RW: if we want to run MANET then we must support NHDP too.
[14:55:13] <apetrescu> TC: not too much time on this issue. Note just that there is a potential conflict between ND and NHDP.
[14:55:30] <apetrescu> TC: need to be aware there is a potential problem. Not discussion beneficial for progress. Proceed.
[14:56:05] <apetrescu> slide: MANEMO/AUTOCONF Applicability
[14:57:49] <apetrescu> slide: Considerations
[14:59:19] <apetrescu> FT: back to point: it's not requireed in AUTOCONF to get a prefix from AR, could get it from MR, maybe topologically incorrect, then maybe go to HA... that's also a possiblity in AUTOCONF.
[14:59:35] <apetrescu> JM: trying to understand, what do you mean by routing scope?
[14:59:46] <apetrescu> JM: is this discovery scope? A discovery gateway?
[14:59:51] <apetrescu> JM: I knew that.
[15:00:08] <apetrescu> JM: you don't need routing scope, but discovery scope. SMF doesn't require NHDP you could n-scope.
[15:00:19] <apetrescu> NHDO comma you could n-scope.
[15:00:23] <apetrescu> JM: anycast
[15:00:33] <apetrescu> JM: discovery first and then routing infrastructure to next step?
[15:00:43] <apetrescu> RW: point a is globally correct topologically address.
[15:00:50] <apetrescu> JM: true if you have a correct prefix.
[15:00:55] <apetrescu> JM: looks as a discovery problem.
[15:01:12] <apetrescu> RW: yes, I may have a q to MANET community: relationship between AR discovery and ...
[15:01:21] <apetrescu> RW: how to set up a rout between a... and a.
[15:01:28] <apetrescu> RW: what we try to be in routing scope is this.
[15:01:45] <apetrescu> JM: good q to follow up on, sometimes the discovery mechanism isn't the same as the routing approach.
[15:02:03] <apetrescu> slide: MANET/AUTOCONF applicability
[15:02:36] <apetrescu> TC: rather like the first one to be phraseD: a req that could be discussed with the AUTOCONF WG.
[15:02:56] <apetrescu> TC: it is _a_ req by what has been... MANEMO could present to AUTOCONF: hey do you have, or can you make something like this.
[15:03:06] <apetrescu> slidE: AR/IGW involvement, deployment consideration
[15:04:27] <apetrescu> questioner Hejse?
[15:04:33] <apetrescu> q: is there a single...
[15:04:36] <apetrescu> RW: to MR3.
[15:06:38] <apetrescu> slide: Possible Issue: divide a Nested NEMO
[15:08:23] <apetrescu> JA: why is that really a problem? We have bigger areas of protocols, update routers in between.
[15:08:30] <apetrescu> JA: you see usecases where that is common?
[15:08:48] <apetrescu> RW: this FR doesn't support any extension for MANEMO, then it will divide the operation.
[15:09:15] <apetrescu> JA: if you run these things in a particular deployment then why is that to update
[15:09:18] <apetrescu> RW: no problem
[15:09:24] <apetrescu> JA: you can't get out of this
[15:09:55] <apetrescu> TC: explains
[15:10:07] <apetrescu> TC: transitorial or fallbakck of evolutionary or interemdiary
[15:10:21] <apetrescu> JA: you're saying this req is that you can't break anything.
[15:10:32] <apetrescu> JA: you don't require this automatic optimization . right
[15:11:09] <apetrescu> JB: as a customer, this is portraying a problem space, we can't for years, there are MANET protocols out there.
[15:11:49] <apetrescu> JB: it's happeninng on DoD when off and build a MANET and fixed networks, that doesn't always work. My feeling we build this flat network. There's a whole bunch of other places.
[15:11:58] <apetrescu> JB: a new problem presented here, and that's fair.
[15:12:07] <apetrescu> JB: a new set of problems
[15:12:26] <apetrescu> FT: there is a lot that is going on in the military. a lot we can't talk about.
[15:12:35] <apetrescu> FT: they don't all ... est in that contest.
[15:12:53] <apetrescu> TC: unless you folks military you present scenario, you can't expect IETF or any other org to provide solutions.
[15:13:40] <apetrescu> JB: RFID full traceabaility, add configure vs planders and tons of healthcare applications. This is not just military.
[15:13:55] <apetrescu> Doug Street
[15:13:57] <apetrescu> (?)
[15:14:24] <apetrescu> DS: different kind of problem. from the DoD perspective: we try to solve mobility in the general case. It's not mobility at the edge, but how far can we push in the general case.
[15:14:51] <apetrescu> DS: is that a Routing RG or AUTOCONF? Lots of ways to tackle. My perspective in IETF is sort of take small part of problem we think we can work and wor out from there.
[15:15:28] <apetrescu> DS: most of MANET work tries to seem to solve som elimitation. Somehow we have to simplify the problem: is that that requires revisiting problems at architectural level or mechanisms that override the existing architecture?
[15:15:37] <apetrescu> JM: 96 I've been on MANET
[15:16:08] <apetrescu> JM: trying to solve the flat problem or not. MANET rescoped back, hierarchcal gets complicated. Routing problem, doesn't scale. Engineers work on archis and hierarchis.
[15:16:17] <apetrescu> JM: point of manet was to solve some specific problem
[15:16:44] <apetrescu> JM: really the problem here is to scale the nested NEMO. PEople say the problem is MANET but not, maybe we should use routing instead of nest and nest and nest.
[15:17:19] <apetrescu> JM: people though about dividing scope, two -tier potentially n=tier models. Research groups didn't work out very well. Let's look back in the pas where we came from.
[15:17:28] <apetrescu> JM: if you try to solve NEMO then it's about NEMO.
[15:17:39] <apetrescu> JM: if we talk building archis then in another group.
[15:17:44] <apetrescu> RW: we have real reqs from industry,.
[15:18:00] <apetrescu> RW: remember MANET worked on similar topic. I've been working on IGW.
[15:18:18] <apetrescu> RW: but now that we have NEMO protocol and a real problem, I think this is a good time to try to solve this problem.
[15:18:30] <apetrescu> AMG: copuld of other things to be aware of,
[15:18:53] <apetrescu> AMG: there are out there in the env, one of those is the id-locator stuff and hip, very relevant. What is the nature of referral and find each othe.r
[15:18:59] <apetrescu> AMG: that was one.
[15:19:01] <apetrescu> AMG: lost the other one.
[15:19:30] <apetrescu> JB: the point is the end result as someone like you deploying this stuff. I worked on NEtwork centric operations... vendor.
[15:20:06] <apetrescu> JB: really hard core waveform people, when the MANET name is inappropriate. You can have an ad-hoc network but doesn't move, so MIP is gone. MANET is misnomer.
[15:20:18] <apetrescu> JB: and we need these,.
[15:20:32] <apetrescu> JB: MANEMO is taking it to another level, we have to have failover in the infra on routing protocols.
[15:20:51] <apetrescu> JB: products, reliability. Gets exacerbated by wireless, and transport won't probably work.
[15:21:00] <apetrescu> JB: in a market of ??...
[15:21:08] <apetrescu> JB: PMIP look what 's happening.
[15:21:23] <apetrescu> JB: time for WG or someone or JA to make a decision what you want to do with this work.
[15:21:40] <apetrescu> JB: when I hear you don't hear ther reqs, I don't understand what do you want? A business plkan/
[15:21:48] <apetrescu> JM: it doesn't have to be mobile.
[15:22:05] <apetrescu> JM: if you feel like you need a routing protocol then we may need to work together if not then that's fine too.
[15:22:19] <apetrescu> JM: some protocols with convergence issue and dynamic links, exploratory discussion.
[15:22:25] <apetrescu> TN: uplevel
[15:22:37] <apetrescu> TN: get the feeling there is an intent to boil the ocean.
[15:22:42] <apetrescu> TN: it's focused on nested NEMO.
[15:22:48] <apetrescu> TN: if we restrict the problem
[15:22:58] <apetrescu> TN: you have some real constraints when you have addresses fixed.
[15:23:05] <apetrescu> TN: not leak route into the routing infra.
[15:23:12] <apetrescu> TN: you have tunnels and suboptimal routing.
[15:23:21] <apetrescu> TN: routing scaling problem and you can't have all things you want.
[15:23:36] <apetrescu> TN: if really nice, ... really nice to have optimal routing and go peertopeer right away.
[15:23:50] <apetrescu> TN: there are problems. you can restrict the mobility.
[15:23:55] <apetrescu> TN: these are choices.
[15:24:02] <apetrescu> TN: one of these or solution unoptimal.
[15:24:08] <apetrescu> FT: Ethernet again.
[15:24:12] <apetrescu> FT: id-locator split.
[15:24:26] <apetrescu> FT: there _are_ other ways of doing network mobility other thna NEMO.
[15:24:48] <apetrescu> CP: I want to follo w up with TN< there shouldnt be framed black or white.
[15:24:54] <apetrescu> CP: reduce tunnels, then more signalling.
[15:25:08] <apetrescu> CP: optimization is more important then maybe more signalling is necessary.
[15:25:16] <apetrescu> CP: applications that are purely local.
[15:25:33] <apetrescu> CP: these are not household familiar sort of operations, that's why it's important to take a look what are the...
[15:25:45] <apetrescu> CP: no bw distinction at this time because we won't achieve much.
[15:25:55] <apetrescu> Timothy Shepard is TS
[15:26:04] <apetrescu> TS: a lot of problems here connected together.
[15:26:21] <apetrescu> TS: havent seen something like multiominh, like shim6,. I had some hopes for a while.
[15:26:51] <apetrescu> TS: connected simultaneously to ... then maybe that secnario earlier where you're disconnected from a larger network but connected locally...
[15:26:58] <apetrescu> TS: multiple networks connected.
[15:27:08] <apetrescu> TS: cope with that maybe that might lead into thinking.
[15:27:20] <apetrescu> TS: then mobility may be easier to solve.
[15:27:24] <apetrescu> TS: id-locator maybe.
[15:27:42] <apetrescu> TS: maybe not that id-locator split, but maybe id-locator is just a solution but there may be others.
[15:28:08] <apetrescu> TS: referrals, discover services, cope with that scenario but a little bit of less adhoc'ish.
[15:28:22] <apetrescu> TS: embrace the problem of being multiple connected then maybe more hope.
[15:28:38] <apetrescu> slide: Impact of MR movement.
[15:30:16] <apetrescu> slide: The NEMO Addressing: Movement Transparency
[15:30:59] <apetrescu> FT: periodically I mentioned: that is not what AUTOCONF is stipulating.
[15:31:10] <apetrescu> FT: AUTOCONF gets prefix on from MR
[15:31:12] <apetrescu> RW: document
[15:31:18] <apetrescu> FT: yes, manet autoconf....
[15:31:23] <apetrescu> RW: yes
[15:31:36] <apetrescu> slide: Multihoming Capability
[15:32:26] <apetrescu> slide: Monami6 Applicability??
[15:33:06] <apetrescu> JA: why doesn't it solve it? Why doesnt monami6 help you?
[15:33:14] <apetrescu> RW: because MR registers both addresses to HA
[15:33:26] <apetrescu> RW: what we want path address from AR so we can bypass the HA
[15:33:44] <apetrescu> JA: ok
[15:33:50] <apetrescu> RW: Some Overall Observations
[15:36:28] --- kakima has left
[15:37:19] --- igarashi has left
[15:37:58] <apetrescu> JA: there's a lot of issues on the table,
[15:38:14] <apetrescu> JA: everyone involved on NEMO and MANET is aware of limitations, tunnelling and scalability.
[15:38:23] <apetrescu> JA: not sure we can solve every problem we have on MANETMO.
[15:38:38] <apetrescu> JA: we have to select what's appropriate to business situation, appropriate.
[15:38:44] <apetrescu> JA: make a decision, nail down something.
[15:38:59] <apetrescu> JA: one is that you want to use MIP amke it better in terms of RO, that can be done.
[15:39:20] <apetrescu> JA: to customers they tell me you can't really rely on HA at all. That's not compatible with using MIP or NEMO. You can optimize but.
[15:39:32] <apetrescu> JA: that's one approach: go for NEMO and make it better,
[15:39:51] <apetrescu> JA: other approach is to comboine NEMO and use some adhoc techniques.
[15:39:58] <apetrescu> JA: not convicended NHDP vs ND
[15:40:01] <apetrescu> JA: maybe.
[15:40:25] <apetrescu> JA: or as fourth, as id-locator split, that is an architectural direction that will provide a clearn solution but legacy.
[15:40:40] <apetrescu> JA: before making out something of this then we have to decide something.
[15:41:05] <apetrescu> TC: thanks TJ for room
[15:42:02] <apetrescu> CP: this isn't exactly anwering JA
[15:42:04] <apetrescu> CP: but
[15:42:20] <apetrescu> CP: for using MIP you have connectivity to Internet, without it then MANET.
[15:42:28] <apetrescu> CP: how to have a mechanism that....
[15:42:36] <apetrescu> CP: MIP isn't really solving the problem in that case.
[15:42:46] <apetrescu> CP: in MANET scenario, not always IP subnet structure.
[15:43:18] <apetrescu> CP: in MANET good but doesn't give you anything to connect to Internet. Put those two together then you still want to haev a smmoothly OS ressources available
[15:43:23] <apetrescu> CP: this hasn't been dnne
[15:43:29] <apetrescu> CP: these solutions arent stdized.
[15:43:48] <apetrescu> CP: identify how to make sense of ND messages for 2-hop backbone in MANET, it's possible
[15:44:20] <apetrescu> CP: seems to me from experience we had doing some of these things, solutions _are_ available, but that's a lot different statement than where are those products conforming to these stds.
[15:44:29] <apetrescu> TC: so what are the issues.
[15:45:03] <apetrescu> TC: there are issues and what they are with nested NEMO?
[15:45:25] <apetrescu> TC: is there something we need to somehow address and think about in any way?
[15:45:29] <apetrescu> Gabriel Monteneggro
[15:45:40] <apetrescu> GM: way forward to prioritize and we heard different types of reqs
[15:45:45] <apetrescu> GM: optimizations?
[15:45:58] <apetrescu> GM: but JB talked about fundamental problems, different priorities.
[15:46:14] <apetrescu> GM: makes sense go address fundamental problems first or at least recognize they have higher priority.
[15:46:37] <apetrescu> GM: one part of MANET through other part through Internet, you could use Teredo, not necessariliy MIP.
[15:46:54] <apetrescu> GM: for partitioned manet, rediscover themselves.
[15:47:06] <apetrescu> GM: they are just partitioned MANETs and that should be part of MANET protocols.
[15:47:12] <apetrescu> TC: lots of stuff has to do with NEMO
[15:47:21] <apetrescu> TC: there is another WG looking at NEMO RO
[15:47:53] <apetrescu> TJK: I have a hard time answering the q, there aren't that many nested nemos out there. But to th e point of conversation of MANEMO it's clear there are real engineering problems that
[15:48:02] <apetrescu> TJK: need to be solved for a type of products and.
[15:48:22] <apetrescu> TJK: focusing on the use cases and archis that people are trying to biuidl, topdown.
[15:48:50] <apetrescu> TJK: guidance from JA about how to progress here.
[15:49:05] <apetrescu> TJK: bground and assumptions in engineering these protocols are different than what we see in products.
[15:49:17] <apetrescu> TJK: at first sight it may be different types of protocols...
[15:49:30] <apetrescu> TJK: soldier in battlefield, shooting at me, I just want it to work.
[15:49:48] <apetrescu> TJK: reuse the code, what's been done. but situation is different.
[15:50:13] <apetrescu> Steve Pratt
[15:50:19] <apetrescu> SP: military scenario.
[15:50:24] <apetrescu> SP: similar types of situations
[15:50:40] <apetrescu> SP: we have a great start we show the tcenario but not the issues with the scenarios.
[15:50:57] <apetrescu> SP: drive from scenario issues see where people have problems.
[15:51:04] <apetrescu> SP: work with emergency dispatch system.
[15:51:18] <apetrescu> SP: policy, hosptiatlas, all connected together.
[15:51:19] <apetrescu> TC: I agree
[15:51:32] <apetrescu> TC: scenarios where you have direct experience with this?
[15:51:38] <apetrescu> SP: yes, absolutely these are.
[15:52:18] <apetrescu> JM: backin up having experience is. if you fragemnet the network is fragmenting the internet - a very different set of issues that comes out, very important.
[15:52:34] <apetrescu> JM: IP addressing layer, or other ways of doing it.
[15:52:47] <apetrescu> JM: solve both of them can lead to very different solutions.
[15:53:11] <apetrescu> TC: is this something a topic that is worth solving in IETF and ready to be solved in IETF
[15:53:32] <apetrescu> TN: the difficulty, what is the constraints?
[15:53:43] <apetrescu> TN: what does the solution should look like?
[15:54:02] <apetrescu> TC: requirements?
[15:54:53] <apetrescu> AMG: seems most real applications, to do with oceanic air traffic control
[15:55:10] <apetrescu> AMG: all situations have something in common: mobile, likely to be disconnected, links disruptive
[15:55:25] <apetrescu> AMG: apps keep working when Internet go away.
[15:55:48] <apetrescu> CP: question, it's all about getting packets in the right places, in the routing group.
[15:55:57] <apetrescu> CP: why here? whynot in the routing area?
[15:56:02] <apetrescu> TN: tell them
[15:56:09] <apetrescu> CP: Brian McLaughlin
[15:56:15] <apetrescu> BML: I work with U-2010
[15:56:37] <apetrescu> BML: problem when we went to emergency, we tell 'IP' they say 'IP?' they care about bodies on the ground.
[15:56:44] <apetrescu> BML: I want more batteries more radios.
[15:57:14] <apetrescu> BML: very good accurate up to date information, successful manner. If the sattelite link disappears then I dont use the police connection I use the ambulance
[15:57:21] <apetrescu> BML: we need to look a little bit further.
[15:57:47] <apetrescu> DS: use cases and scenarios are beneficial to get people thinking in the right dir.
[15:58:11] <apetrescu> DS: but risk of getting use case A and design something forr case A but in field you find in the field case A-prime.
[15:58:24] <apetrescu> DS: general set of problems is beneficial but there are harder problems to solve.
[15:58:31] <apetrescu> Justin Dean
[15:58:56] <apetrescu> JD: a lot of these issues just come from the problem and issues from the different kind of boxed soltions. MANET just solves routing, but other protocol might need.
[15:59:39] <apetrescu> JS: finding out what doesnt work and why doesn't work that well, specifically in AUTOCONF. It doesn't work when interacting; they work by themselves, but not together. That's the cruz (by me) of what we try to solve here.
[16:00:01] <apetrescu> TC: is going to publish the slides
[16:00:09] <apetrescu> TC: http://www.mobileip.jp/MANEMO
[16:00:16] --- Andrew McGregor has left
[16:00:18] <apetrescu> TC: WGs.
[16:00:29] <apetrescu> TC: grab RW
[16:00:43] <apetrescu> TC: suitable next steps?
[16:00:59] <apetrescu> TC: thanks
[16:01:06] --- shubranshu has left
[16:01:11] <apetrescu> TC: adjourns.
[16:01:13] --- apetrescu has left
[16:05:26] --- apetrescu has joined
[16:05:39] --- apetrescu has left
[16:23:02] --- dthaler has left
[16:30:48] --- dku-tzi-org has left
[16:38:53] --- apetrescu has joined
[16:39:14] <apetrescu> slides were at http://www3.ietf.org/proceedings/07jul/slides/autoconf-0.ppt and http://www3.ietf.org/proceedings/07jul/slides/autoconf-1.ppt
[16:39:59] --- apetrescu has left
[16:43:15] --- dku-tzi-org has joined
[16:50:05] --- alexpetrescu has joined
[16:50:13] --- alexpetrescu has left
[18:56:13] --- dku-tzi-org has left