Taking your last point on what charters say, I would interpret that as:
- A solution just intended for LLNs, or which that is the dominant case for, belongs to ROLL (unless they don't want it).
- A wide ranging solution (which could include LLNs as a special case) belongs in MANET.
- A specialist solution that is not an LLN specialist solution is permissible in MANET, though a specialised WG could be created for it.

But all subject to the overriding condition there has to be a reason to do it, not just a protocol with no user base looking for a home. And of course I'm just talking about ad hoc routing protocols (by which I mean, roughly speaking, decentralised multi-hop routing in dynamic or otherwise unknown characteristic networks) or closely related things here.

-- 
Christopher Dearlove
Senior Principal Engineer, Communications Group
Communications, Networks and Image Analysis Capability
BAE Systems Advanced Technology Centre
West Hanningfield Road, Great Baddow, Chelmsford, CM2 8HN, UK
Tel: +44 1245 242194 |  Fax: +44 1245 242124
chris.dearlove@baesystems.com | http://www.baesystems.com

BAE Systems (Operations) Limited
Registered Office: Warwick House, PO Box 87, Farnborough Aerospace Centre, Farnborough, Hants, GU14 6YU, UK
Registered in England & Wales No: 1996687


-----Original Message-----
From: Stan Ratliff (sratliff) [mailto:sratliff@cisco.com] 
Sent: 02 August 2012 16:54
To: Dearlove, Christopher (UK)
Cc: Abdussalam Baryun; C Chauvenet; manet
Subject: Re: [manet] Some LLNs are NOT MANETs

----------------------! WARNING ! ----------------------
This message originates from outside our organisation,
either from an external partner or from the internet.
Keep this in mind if you answer this message.
Follow the 'Report Suspicious Emails' link on IT matters
for instructions on reporting suspicious email messages.
--------------------------------------------------------

Chris, 

On Aug 2, 2012, at 5:12 AM, Dearlove, Christopher (UK) wrote:

> I would subtly differently emphasise your point. (I think we are basically in agreement.) I would view ad hoc networks (I'm going to avoid the MANET term as it's overloaded here) as a broad area. LLNs are a specialist sub-area within that area. The MANET WG in principle covers the whole of the area, and most of its work has created protocols that can cover most of that area. However where there are specialised, and in particular more limited requirement, sub-areas, specialised protocols may make a better trade-off (typically be simpler). And you have emphasised one of the key differences, the differing traffic pattern, and in particular the dominant role of an egress (or ingress, but egress is often more important)  gateway rather than peer to peer communications (which in an LLN are typically either absent, or rare enough that the otherwise inefficient go via gateway, possibly optimised to clip off reused links, is acceptable).
> 
> (This bit may or may not be in agreement with Stan.) Note that I'm not saying that you can't use e.g. OLSRv2 in an LLN, but that if you have more limited/specialised requirements you may be able to do better. Nor am I saying RPL or LOADng is (or is not) that better solution, I've studied neither in enough detail to have a view.
> 

No, we agree. As a mentor of mine once said, "You can do anything you're big enough to try". ;-) ;-) The definitions we're using (MANET and LLN) are somewhat vague, and somewhat overlapping. And any specific network's environment could well lead to multiple choices for a routing protocol. 

On another note, and more to the list at-large (instead of just replying to Chris), we can argue this for days, or even weeks. But the truth of the matter is that our charter references MANET networks, while the charter for ROLL references LLN's. It's pretty much just that simple. 

Regards,
Stan


> -- 
> Christopher Dearlove
> Senior Principal Engineer, Communications Group
> Communications, Networks and Image Analysis Capability
> BAE Systems Advanced Technology Centre
> West Hanningfield Road, Great Baddow, Chelmsford, CM2 8HN, UK
> Tel: +44 1245 242194 |  Fax: +44 1245 242124
> chris.dearlove@baesystems.com | http://www.baesystems.com
> 
> BAE Systems (Operations) Limited
> Registered Office: Warwick House, PO Box 87, Farnborough Aerospace Centre, Farnborough, Hants, GU14 6YU, UK
> Registered in England & Wales No: 1996687
> 
> 
> -----Original Message-----
> From: Stan Ratliff (sratliff) [mailto:sratliff@cisco.com] 
> Sent: 01 August 2012 17:23
> To: Dearlove, Christopher (UK)
> Cc: Abdussalam Baryun; C Chauvenet; manet
> Subject: Re: [manet] Some LLNs are NOT MANETs
> 
> ----------------------! WARNING ! ----------------------
> This message originates from outside our organisation,
> either from an external partner or from the internet.
> Keep this in mind if you answer this message.
> Follow the 'Report Suspicious Emails' link on IT matters
> for instructions on reporting suspicious email messages.
> --------------------------------------------------------
> 
> Precisely, Chris. 
> 
> I've come to view MANETs more and more in the terms of an "opportunistic network". By that, I mean a network with a fairly large amount of dynamism, and also a network that contains heterogeneity of links (e.g., different radio types, and/or some wired portions). The challenge of these networks is to opportunistically (and maximally) use the resources available to it at any point in time. 
> 
> And at least for me (and I know this will draw some fire in return), another distinction in MANET as opposed to LLN is a greater requirement for peer-to-peer (node-to-node) communication, From what I've seen of LLN deployments (and I'm no expert), they tend to have more of a "multiple source/single sink" model for the data flow.
> 
> Again, just one opinion. 
> 
> Regards,
> Stan
> 
> On Aug 1, 2012, at 11:26 AM, Dearlove, Christopher (UK) wrote:
> 
>> The statement that MANETs refers only to networks with physical mobility is simply false. While the expansion of the acronym might suggest that, in fact MANET has become a term of art, and that art includes many networks with some or all nodes that are not physically moving.
>> 
>> -- 
>> Christopher Dearlove
>> Senior Principal Engineer, Communications Group
>> Communications, Networks and Image Analysis Capability
>> BAE Systems Advanced Technology Centre
>> West Hanningfield Road, Great Baddow, Chelmsford, CM2 8HN, UK
>> Tel: +44 1245 242194 |  Fax: +44 1245 242124
>> chris.dearlove@baesystems.com | http://www.baesystems.com
>> 
>> BAE Systems (Operations) Limited
>> Registered Office: Warwick House, PO Box 87, Farnborough Aerospace Centre, Farnborough, Hants, GU14 6YU, UK
>> Registered in England & Wales No: 1996687
>> 
>> 
>> -----Original Message-----
>> From: manet-bounces@ietf.org [mailto:manet-bounces@ietf.org] On Behalf Of Abdussalam Baryun
>> Sent: 01 August 2012 14:21
>> To: C Chauvenet
>> Cc: manet
>> Subject: Re: [manet] Some LLNs are NOT MANETs
>> 
>> ----------------------! WARNING ! ----------------------
>> This message originates from outside our organisation,
>> either from an external partner or from the internet.
>> Keep this in mind if you answer this message.
>> Follow the 'Report Suspicious Emails' link on IT matters
>> for instructions on reporting suspicious email messages.
>> --------------------------------------------------------
>> 
>> Hi Cédric,
>> 
>> Ok we can discuss efficiently when we read the documents (production
>> of such WG)as you agreed.
>> 
>>> When you say "IMHO, as long we have a ROLL WG in IETF we need to forward
>>> work to it which has no mobility behavior", do you mean that the difference
>>> between ROLL and MANET is limited to mobility considerations ?
>> 
>> *Mobility* It is the mostly difference clearly spoted between the
>> documents produced by the both WGs, also the name MANET starts with
>> Mobile, the word mobile is not equal to change/dynamic. So MANET
>> refers to mobile node or mobile network, which means location change
>> *NOT* topology change/dynamic. In the old days there was no ROLL WG so
>> MANET WG was doing the job.
>> 
>>> 
>>> Based on the docs you mentioned, I see that RFC5548, RFC5826, and RFC5867
>>> all include the usual IoT constraints : Memory, Processing, Power, Battery,
>>> Cost and Size of device.
>>> RFC5673 does not mention explicitly processing and size constraints, but
>>> mention all the others.
>>> 
>> 
>> Ok,
>> 
>>> RFC 2501 speak about energy and power, as you previously mentioned, but does
>>> not say something on the following constraints :  Memory, Processing, Cost
>>> and Size.
>> 
>> Yes your right, but be careful RFC2501 does not exclude thoes
>> constraints either. However, if the MANET network has some nodes with
>> constraints but clearly mobility is the main constraint.
>> 
>>> 
>>> Do you think that these constraints are in the MANET scope ?
>>> I guess that they may be (obviously for Cost), but not in the same order of
>>> magnitude as the type of devices described in the requirements RFC of ROLL.
>> 
>> Really don't mind either way for MANET WG, and will think that the WG
>> community wants it in between In-Scope and Out-Scope, that may be
>> strange but seems that is what is happening as I read documents. Out
>> of IETF, MANET is mostly understood as mobile nodes connecting, not
>> LLNs.
>> 
>> Overall, IMHO, in the end it is the decision of the IETF community to
>> decide of any new idea or any I-D acceptance/adoption, but for the WG
>> it MAY take over a WORK as an IETF I-D, but it may not be successful
>> to convince the IETF community to continue the work or make it become
>> a RFC, we always have to see through to make the work efforts flow
>> successfully. We do not forget that many participants work in both
>> WGs.
>> 
>> Regards
>> Abdussalam
>> ==========================================================
>> On 8/1/12, C Chauvenet <c.chauvenet@watteco.com> wrote:
>>> Hi,
>>> 
>>> Thank you for your answer,
>>> 
>>> See inline.
>>> 
>>> Le 1 août 2012 à 13:41, Abdussalam Baryun a écrit :
>>> 
>>>> Hi Cédric,
>>>> 
>>>> I think the RFC2501 is more important than the WG charter, because the
>>>> charter MAY change but RFCs never changes (only can be updated,
>>>> obsoleted or replaced).
>>> 
>>> Good Point, 100% agree.
>>> 
>>>> So far now all MANET-protocols are refering to
>>>> RFC2501 which is good and SHOULD continue, and I like this referencing
>>>> to documents (even if they are old or expired) not referencing to
>>>> charters, because for example, if in a journal paper we reference to
>>>> the charter of MANET WG or ROLL WG the information is not solid like
>>>> if you reference a published document (publisher date), but still we
>>>> can reference the charter as web reference sited date.
>>>> 
>>>> I read some paper that reference sited web documents but they may be
>>>> not valid and cannot be read for some reason. Therefore, IMHO, we need
>>>> to stick to the following document to make a right decision in
>>>> definitions:
>>>> 
>>>> 1) [RFC2501] for MANETs. (published in 1999)
>>>> 2) [RFC5548], [RFC5673], [RFCRFC5826], and [RFC5867] for LLNs.
>>>> (published in 2009)
>>> 
>>> I think they are good references to build some arguments into this
>>> discussion.
>>> 
>>>> 
>>>> Low power was indicated in RFC2501 in section 5.1 as <possibly power
>>>> constraints> and also the word *sink*, in page 7 you read <sleep mode
>>>> for energy conservation>. IMHO, as long we have a ROLL WG in IETF we
>>>> need to forward work to it which has no mobility behavior. Delegation
>>>> will help make work flowing and more focused.
>>> 
>>> When you say "IMHO, as long we have a ROLL WG in IETF we need to forward
>>> work to it which has no mobility behavior", do you mean that the difference
>>> between ROLL and MANET is limited to mobility considerations ?
>>> 
>>> Based on the docs you mentioned, I see that RFC5548, RFC5826, and RFC5867
>>> all include the usual IoT constraints : Memory, Processing, Power, Battery,
>>> Cost and Size of device.
>>> RFC5673 does not mention explicitly processing and size constraints, but
>>> mention all the others.
>>> 
>>> RFC 2501 speak about energy and power, as you previously mentioned, but does
>>> not say something on the following constraints :  Memory, Processing, Cost
>>> and Size.
>>> 
>>> Do you think that these constraints are in the MANET scope ?
>>> I guess that they may be (obviously for Cost), but not in the same order of
>>> magnitude as the type of devices described in the requirements RFC of ROLL.
>>> 
>>> Regards,
>>> 
>>> Cédric.
>>> 
>>>> 
>>>> Regards
>>>> AB
>>>> =======
>>>> 
>>>> On 8/1/12, C Chauvenet <c.chauvenet@watteco.com> wrote:
>>>>> Hi,
>>>>> 
>>>>> This is an interesting discussion.
>>>>> 
>>>>> My understanding is that both MANET and ROLL considers lossy/dynamic
>>>>> links
>>>>> in the way described in the MANET charter :  "static and dynamic
>>>>> topologies
>>>>> with increased dynamics due to node motion or other factors." I also
>>>>> agree
>>>>> that dynamicity of links is not hard wired to node mobility.
>>>>> 
>>>>> So, in the LLN Vs MANET debate, I think they share the "N" for Networks,
>>>>> and
>>>>> the 2nd "L" for Lossy.
>>>>> So the remaining point is the first L, meaning Low Power.
>>>>> 
>>>>> Low Power requirements is explicit in the ROLL charter and not mentioned
>>>>> at
>>>>> all in the MANET charter.
>>>>> Is the power efficiency consideration the big difference between ROLL
>>>>> and
>>>>> MANET ?
>>>>> 
>>>>> Cédric.
>>>>> 
>>>>> -----Message d'origine-----
>>>>> De : roll-bounces@ietf.org [mailto:roll-bounces@ietf.org] De la part de
>>>>> Abdussalam Baryun
>>>>> Envoyé : mercredi 1 août 2012 09:22
>>>>> À : Henning Rogge
>>>>> Cc : roll; manet
>>>>> Objet : [Roll] Some LLNs are NOT MANETs
>>>>> 
>>>>> Hi Henning,
>>>>> 
>>>>> I am about to say many LLNs are NOT MANETs but it seems like the market
>>>>> or
>>>>> community will decide the outcome, but surly that some LLNs are NOT
>>>>> MANETs.
>>>>> 
>>>>>> Could you state an example what would be considered a LLN but not a
>>>>>> MANET. I normally consider LLNs a subset of what we call MANETs.
>>>>> 
>>>>> I not totally agree with that, because we need to be considering both
>>>>> NETs
>>>>> use case and applicability in the vision of the different WGs (MANET and
>>>>> ROLL). There are many examples we can find them in the [RFC2501] for
>>>>> MANETs
>>>>> characteristics and applicability, and for LLNs characteristics in
>>>>> [RFC5548], [RFC5673], [RFCRFC5826], and [RFC5867] including LLNs
>>>>> requirements. That is why I suggested before that
>>>>> OLSRv2 and AODVv2 should mention their applicability to LLN if they do.
>>>>> They
>>>>> just refer to RFC2501, but RFC2501 is OLD and does not mention LLN but
>>>>> describes the meaning. The authors of RFC2501 still not responded to my
>>>>> update suggestions.
>>>>> 
>>>>> I understood from one discussion in MANET WG that few don't have time to
>>>>> read many pages of documents, so that is why I suggested to have
>>>>> terminology
>>>>> I-D [1] as we have ROLL terminology [ROLL] . I also taken initiative to
>>>>> make
>>>>> new draft of  MANET subnet technologies which include only related LLNs
>>>>> [AB2].
>>>>> 
>>>>> Therefore, I will add the definition for MANET and LLN into my
>>>>> manet-terminology draft [AB1] (propose that authors of [ROLL] define LLN
>>>>> more details) to assist discussions as it is proved now in the list that
>>>>> there still is problems in definitions in MANET WG or in some I-D
>>>>> editorial
>>>>> content.
>>>>> 
>>>>> [AB1] http://www.ietf.org/id/draft-baryun-manet-terminology-00.txt
>>>>> [AB2] http://www.ietf.org/id/draft-baryun-manet-technology-00.txt
>>>>> [ROLL] http://tools.ietf.org/id/draft-ietf-roll-terminology-06.txt
>>>>> 
>>>>> Best Wishes
>>>>> 
>>>>> Abdussalam Baryun
>>>>> University of Glamorgan, UK
>>>>> 
>>>>> ====================================================
>>>>> On 7/31/12, Henning Rogge <hrogge@googlemail.com> wrote:
>>>>>> On Tue, Jul 31, 2012 at 6:45 AM, Abdussalam Baryun
>>>>>> subject: Re: [manet] Discussing LOADng suggestions
>>>>>> <abdussalambaryun@gmail.com> wrote:
>>>>>>> IMHO this protocol was intended as for ROLL WG not for MANET WG, but
>>>>>>> then changed its direction to MANET [3]. However, please note that
>>>>>>> *ONLY* some LLNs are MANETs, and *ONLY* some MANETs are LLNs. That
>>>>>>> said, LOADng SHOULD specfy where is its limits. Then we can discuss
>>>>>>> adoption.
>>>>>> 
>>>>>> Could you state an example what would be considered a LLN but not a
>>>>>> MANET. I normally consider LLNs a subset of what we call MANETs.
>>>>>> 
>>>>>> Henning Rogge
>>>>>> 
>>>>>> --
>>>>>> Steven Hawkings about cosmic inflation: "An increase of billions of
>>>>>> billions of percent in a tiny fraction of a second. Of course, that
>>>>>> was before the present government."
>>>>>> 
>>>>> _______________________________________________
>>>>> Roll mailing list
>>>>> Roll@ietf.org
>>>>> https://www.ietf.org/mailman/listinfo/roll
>>>>> 
>>>>> 
>>>>> 
>>>> 
>>> 
>>> 
>>> 
>> _______________________________________________
>> manet mailing list
>> manet@ietf.org
>> https://www.ietf.org/mailman/listinfo/manet
>> 
>> 
>> ********************************************************************
>> This email and any attachments are confidential to the intended
>> recipient and may also be privileged. If you are not the intended
>> recipient please delete it from your system and notify the sender.
>> You should not copy it or use it for any purpose nor disclose or
>> distribute its contents to any other person.
>> ********************************************************************
>> 
>> _______________________________________________
>> manet mailing list
>> manet@ietf.org
>> https://www.ietf.org/mailman/listinfo/manet
> 
>