Re: [Roll] Dissenting technical arguments unwelcome (was: Re: trickle-mcast-04 - Clarify scope value of 3 - subnet-local)

[Don Sturek <[e-mail]>]

Hi Ulrich,

Let me say as an implementer of ROLL RPL (and Trickle Multicast) the topic
of multi-link subnets and the general topic of multicast address scope
continues to be a major concern.  For example, we needed to extend mDNS to
cover site specific addressing for this reason as well as need to define
another draft describing ULA prefix delegation rules and forwarding rules
for border routers (yet to be done).

While our current profile (ZigBee IP)  will continue to use multi-link
subnets, it would be fantastic to have a discussion on how multi-link
subnets might be avoided in the future (if that is even possible).  I have
to say after your last reference to the AUTOCONF RFC, I still have doubts
a /128 prefix solves the problems we are having (though interested in
seeing if I am wrong.....)

I for one am in favor of having a *technical* discussion on the topic on
the ROLL reflector.  I also think this topic applies to MANET as well.

 I did not see any discussion on this thread that was not focused on a
relevant technical topic.  Hopefully, the WG chairs/vice-chairs will
reconsider.

Don



On 7/25/13 10:52 AM, "Ulrich Herberg" <[e-mail]> wrote:

>Hi,
>
>Several people in the ROLL WG, amongst others Rob Caigie and Don
>Sturek, were interested in the topic of multi-link subnets and the
>related multicast scope for MPL / for IPv6 in general [1].
>
>This is an important technical topic and deserves due discussion on
>this mailing list and 6man.
>
>I think that both discussions (multi-link subnet and IPv6 multicast
>scope 0x03) are closely related; because using "subnet" for a scope
>that is meant to be network-wide would imply using (even mandating)
>multi-link subnets, which I think is a bad idea. So in order to
>explain why I think that the "subnet" scope is a bad idea, I needed to
>explain why multi-link subnets are a bad idea.
>
>I am, however, for the time being prohibited from continued
>participation on this technical matter. The ROLL WG chairs sent me a
>formal warning saying that:
>
> o my "comments about multi-hop subnets are not welcome on the ROLL list"
>
> o I should "refrain repeating them yet again, it is disruptive to the
>conversation, and confuses many people."
>
> o this was a formal warning "as per RFC2418/BCP25, and taking
>RFC3683/BPC83 and RFC3934/BCP94 into account, (particularly BCP93
>section 1, paragraph 3) and BCP94 section 2, paragraph 2."
>
>I am very sad to see that technical arguments are ejected from the
>discussion by the ROLL Working Group Chairs if these comments are not
>in line with their personal objectives.
>
>I have sent an appeal to the IESG regarding this formal warning,
>according to RFC2026, Section 6.5., "Process Failures", as I believe
>that an open exchange of technical arguments is key to the work in any
>IETF working group - and that issuing formal warnings simply to
>suppress dissenting technical arguments is not beneficial to the IETF,
>its participants, and its protocol designs.
>
>
>[1] http://www.ietf.org/mail-archive/web/roll/current/msg07951.html
>
>Best regards
>Ulrich
>
>On Wed, Jul 24, 2013 at 8:32 AM, Ulrich Herberg <[e-mail]>
>wrote:
>> I have pointed out two documents in my earlier email that explain in
>> long detail why I believe that multi-hop subnets are a really bad
>> idea. The AUTOCONF RFC 5889 presents an architecture for avoiding all
>> these problems by using /128 prefixes. In that regards, a subnet wide
>> flooding would be fairly uninteresting, as it would not go beyond a
>> single router.
>>
>> Regards
>> Ulrich
>>
>> On Wed, Jul 24, 2013 at 7:58 AM, Michael Richardson
>> <[e-mail]> wrote:
>>>
>>> I would still like an explanation of why "subnet" is the wrong term.
>>>
>>> When would scope-3 would be used such that it would not correspond to
>>>the set
>>> of links on which a "/64" (or other size) is used?
>>>
>>> --
>>> Michael Richardson <[e-mail]>, Sandelman Software Works
>>>
>>>
>>>
>>> _______________________________________________
>>> Roll mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/roll
>>>
>_______________________________________________
>Roll mailing list
>[e-mail]
>https://www.ietf.org/mailman/listinfo/roll