Re: [6lowpan] A suggestion ... With regards to re-chartering

[Jonathan Hui <[e-mail]>]

Hi Geoff,

On May 30, 2008, at 9:10 AM, Geoff Mulligan wrote:
> Jonathan,
>  We have the ID for routing requirements.  Could this, or a portion of
> it, be the starting point for the L2 meshing (mehs under) requirements
> document.

Just to be clear, I was referring to my earlier proposal of having a  
document something like: "requirements and suggestions for 6lowpan  
mesh-under subnetwork designs". Routing is only one piece of that, and  
we can certainly start by taking the existing ID for routing  
requirements. But there are a lot of other issues (e.g. whether or not  
multicast emulation is supported and if so what properties do we  
expect of it, what kind of latencies do we expect, what kind of  
traffic flows do we assume, if not arbitrary point-to-point, etc.).  
These are all useful items for other SDOs to refer to.

> I've been thinking about the Routing Requirements doc.  Should 6lowpan
> generate a ID that is a WG doc for use by ROLL as requirements for  
> Route
> Over or should we (members of 6lowpan) just make sure that the current
> drafts within roll address our needs and concerns and the the Routing
> Requirements draft should instead be focused on mesh under.

The primary value of having a doc come from 6LoWPAN is that we can  
state 6LoWPAN-specific requirements. The ROLL requirements have them  
in more general terms (e.g. low throughput, small MTU, etc.), while  
6LoWPAN can state some hard constraints (e.g. 250 kbps at 2.4GHz, 127  
byte MTU (smaller after headers), etc.). I'm not sure how important it  
is to state the specifics, but if we feel it is important then I'm  
fine with that.

--
Jonathan Hui

>
>
> 	geoff
>
> On Fri, 2008-05-30 at 08:14 -0700, Jonathan Hui wrote:
>> Hi Everyone,
>>
>> In support of most of this. For ND, we should add work on route-over
>> as well, which I've committed to at the last WG meeting. Architecture
>> is needed at this stage, and I'm willing to help drive it. RFC 4944
>> maintenance is needed as well.
>>
>> I still think we need some document to define requirements for a L2
>> meshing over 802.15.4 subnetwork. As I've said before, this could  
>> be a
>> separate document or a part of the architecture document.
>>
>> I hope we can close on this quickly...
>>
>> --
>> Jonathan Hui
>>
>> On May 30, 2008, at 5:48 AM, Pascal Thubert (pthubert) wrote:
>>
>>> Hi JP
>>>
>>> Works with me. I went through the published work and the recent
>>> thread "
>>> New charter for 6lowpan".
>>> My conclusion is as follows:
>>>
>>> From the reworked charter, we should keep Work Items 1, 3, and 5  
>>> which
>>> appear of foremost importance, drop 2 and 4 because pragmatically we
>>> are
>>> not advanced enough in these areas.
>>>
>>> For 1) we have draft-chakrabarti-6lowpan-ipv6-nd that we can couple
>>> with
>>> draft-thubert-6lowpan-backbone-router and we have most of the
>>> content we
>>> need to make a standard track doc.
>>>
>>> For 3) we have draft-culler-6lowpan-architecture. It needs  
>>> improvement
>>> in particular in explaining route over vs. mesh under in details  
>>> as we
>>> currently discuss in the ML. Also mobility, backbone... but we have
>>> people interested in the discussion (see current threads) so we  
>>> should
>>> work it out.
>>>
>>> For 5) we have draft-daniel-6lowpan-security-analysis. We need to  
>>> make
>>> sure we have people committed to the effort but the current draft
>>> looks
>>> good already.
>>>
>>> To those 3, I'd add:
>>>
>>> Explore requirements and usages.
>>> ---------------------------------
>>> We have a draft, draft-ekim-6lowpan-scenarios that we can leverage.
>>> I'd
>>> add to it some words on existing standards that need or use 6LowPAN.
>>> In
>>> particular, we need to place ISA100 requirements in there to be able
>>> to
>>> better serve them later.
>>>
>>> RFC 4944 maintenance and improvements
>>> -------------------------------------
>>> This should cover at least 6lowpan HC and fragment recovery efforts.
>>> We
>>> can discuss if we want ECN in that work item as well.
>>>
>>> What do you think?
>>>
>>> Pascal
>>> ________________________________________
>>> From: [e-mail] [mailto:[e-mail]] On
>>> Behalf Of Jean Philippe Vasseur (jvasseur)
>>> Sent: vendredi 30 mai 2008 04:25
>>> To: [e-mail]
>>> Subject: [6lowpan] A suggestion ... With regards to re-chartering
>>>
>>> Hi,
>>>
>>> Just a suggestion for the chairs and the WG.
>>>
>>> We have discussed many important items for 6lowpan.
>>>
>>> For several of them, all important, I think that there was a clear
>>> agreement: stateful header compression, security, Architecture ID,
>>> fragmentation, ..... For other ones such as the "Mesh-under" and
>>> "Route
>>> over" discussion, there are diverging point of views.
>>>
>>> So why not trying to quickly re-charter adding the items for which
>>> there
>>> is a consensus and continue the discussion on the open issues in the
>>> meantime until we have an agreement at which point the WG may re-
>>> charter
>>> ?
>>>
>>> As we all know, the WG has been fairly slow in term of progress  
>>> and it
>>> is I think now urgent to move on.
>>>
>>> Thoughts ?
>>>
>>> Thanks.
>>>
>>> JP.
>>> _______________________________________________
>>> 6lowpan mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/6lowpan
>>
>> _______________________________________________
>> 6lowpan mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/6lowpan
>

_______________________________________________
6lowpan mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/6lowpan