Re: Deprecating IPv6 (Re: draft-bourbaki-6man-classless-ipv6-00)

On 08/06/2017 07:41, Fred Baker wrote:
> I'm struggling here. Thread is one adaptation of 6LowPAN etc; OCF (formerly OIC) describes and network "on IPv6" and treats 6LowPAN as an encoding of IPv6. For reasons that remain inscrutable to me, Thread acts as if the only link layer in the world were IEEE 802.15.4; OCF is extensible to (gasp) Ethernet, 802.11, ITU G.hn/G.9960 or IEEE 1901, and whatever else.
> 
> IPv6, and 6LowPAN, is the dog. The industry associations that use them are the tail. Why is the tail wagging the dog?

Well said. Also, the fact that rfc7084-bis might or might not require HNCP support doesn't prevent any vendor (or ISP procuring CE routers) requiring HNCP. In any case, it hasn't even reached WGLC yet so its content is all open for discussion.

   Brian

> https://docs.mbed.com/docs/arm-ipv66lowpan-stack/en/latest/thread_overview/
> https://workspace.openconnectivity.org/apps/org/workgroup/technology_sc_open/download.php/8950/Technology-Discussion-v3.pptx
> 
> On Jun 7, 2017, at 11:41 AM, james woodyatt <jhw@google.com> wrote:
>> On Jun 7, 2017, at 01:53, Lorenzo Colitti <lorenzo@google.com> wrote:
>>> On Wed, Jun 7, 2017 at 4:06 AM, james woodyatt <jhw@google.com> wrote:
>>> p1. Power conservative ND Proxy isn’t possible with Thread™ 1.1 (and earlier) networks. It may never be possible in future versions of Thread™.
>>>
>>> Can you explain to us why it doesn't work? One might naively think that if the BR is doing NAT for hosts behind it, it has to process a similar packets as it would have to process if it were doing ND.
>>
>> Thread™ 1.1 doesn’t even use RFC 4861 much less RFC 6775. A proxy for RFC 4861 at the Thread™ 1.1 border router would require DAD and NUD to be translated into prohibitively expensive multicast floods into the mesh. Use of IPv6/NAT allows the border router to make an entire Thread™ mesh reach the public Internet via the one stable IPv6 address that is reliably available on all residential networks with IPv6 providers.
>>
>> For years, we have been hoping that HOMENET would address the basic problem here, but now that it's clear the forthcoming update to RFC 7084 will not recommend adoption of the HOMENET protocol suite in IPv6 CPE residential gateways, Thread™ has no other option than to recommend IPv6/NAT to cope with operational reality.
>>
>>
>> --james woodyatt <jhw@google.com>
>>
>>
>>
>> --------------------------------------------------------------------
>> IETF IPv6 working group mailing list
>> ipv6@ietf.org
>> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
>> --------------------------------------------------------------------
> 
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> ipv6@ietf.org
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
> 