Re: [Roll] [6lowpan] draft-kelsey-intarea-mesh-link-establishment-03.txt

Hi Thomas,

Sorry it was late when I wrote that response :-)

We debated a long time on MLE.  We knew we needed a one hop information
exchange between nodes to help with 6LoWPAN ND (eg exchange long address
information), link quality (eg. for creation and maintenance of symmetric
links), etc.   We looked at a lot of transport options and UDP seemed to
make the most sense.  One nice thing about having MLE use UDP:  If someone
wishes to extend MLE they don't need to have work done in ICMP or ROLL to
make it happen. 

I think we arrived at the Internet area for the draft since part of MLE
helps with 6LoWPAN ND, other functions help with ROLL RPL parent
selection.  I think we made these generic so others with similar needs
could make use of MLE without necessarily using either 6LoWPAN ND or ROLL
RPL.  That said, if ROLL is interested in this draft, Richard Kelsey (the
drafts author) has plenty of experience there :-)

Don

On 6/15/12 3:32 AM, "Thomas Heide Clausen" <ietf@thomasclausen.org> wrote:

>Not sure how fantastic (or not) it is - it is not immediately clear to me
>how tied MLE should be to RPL - if it truly aims at being for _MESH_ link
>establishment, then it would appear to be a much larger scope, and should
>not be tied narrowly to a special-purpose protocol's type-space (&
>conventions etc., that do not apply universally).
>
>I guess I'm expressing some sort of support for Don's position...
>
>-- 
>Thomas Heide Clausen
>http://www.thomasclausen.org/
>
>"Any simple problem can be made insoluble if enough meetings are held to
> discuss it."
>   -- Mitchell's Law of Committees
>
>
>On 15 Jun 2012, at 01:43, Don Sturek <d.sturek@att.net> wrote:
>
>> All sounds fantastic but we don't have time for all these changes so
>>will
>> opt to use MLE as written using UDP ( at least for our application....)
>> 
>> Don
>> 
>> 
>> 
>> On 6/14/12 1:13 PM, "Michael Richardson" <mcr+ietf@sandelman.ca> wrote:
>> 
>>> 
>>> In draft-kelsey-intarea-mesh-link-establishment-03.txt it says that
>>>  MLE messages are sent using UDP.
>>> 
>>> I want to suggest that MLE messages should be sent at least as IPv6
>>> ICMP messages. 
>>> 
>>> Further, I think that it could use the RPL type 155, and RPL would
>>> allocate a new "code" codepoint.
>>> 
>>> There would be some preference to have the general structure of MLE
>>> match that of RPL, but I think that this isn't particularly a strong
>>> requirement.   Allocating a new ICMP type code for MLE alone might be
>>> seen as excessive, but I can not speak for the folks in 6man.
>>> 
>>> -- 
>>> Michael Richardson <mcr+IETF@sandelman.ca>, Sandelman Software Works
>>> IETF ROLL WG co-chair.    http://datatracker.ietf.org/wg/roll/charter/
>>> 
>>> _______________________________________________
>>> Roll mailing list
>>> Roll@ietf.org
>>> https://www.ietf.org/mailman/listinfo/roll
>> 
>> 
>> _______________________________________________
>> 6lowpan mailing list
>> 6lowpan@ietf.org
>> https://www.ietf.org/mailman/listinfo/6lowpan