>> I think it's easy to see the disagreement just comparing the responses
>> from Ole and James.
>> 
>> Ole says: 
>> > I don't see the point of doing address resolution on links without
>> > addresses.
>
> If you've actually got links with no addresses at all, then I agree.
> That seems odd, though, given the usual automatic assignment of at
> least link-locals.

sorry, I meant links without link-layer addressing.

>> > If one implementation sends a NS and
>> > expects to see a NA before sending the first message to the neighbors
>> > address but the other implementation doesn't use NS/NA messages on the
>> > PPP link, there is a problem.
>> 
>> We have now observed the problem Markus predicted, with real
>> implementations.  That is, I know of two non-interoperable
>> implementations of RFC 2472.
>
> Sigh.
>
>> I think James' argument does have real substance.  The quote from
>> 2461 section 3.2 is (especially the last sentence):
>> 
>> >     point-to-point - Neighbor Discovery handles such links just like
>> >                      multicast links.  (Multicast can be trivially
>> >                      provided on point to point links, and interfaces
>> >                      can be assigned link-local addresses.)  Neighbor
>> >                      Discovery should be implemented as described in
>> >                      this document.
>> 
>> If this is the behavior we agree on, then I think that answers my
>> question.
>
> Changing that 'should' to a 'MUST' may fix the problem with the
> document.
>
> As for implementations, I'd expect the implementor to have a really
> good explanation of why he ignored fairly clear (if not capitalized)
> 'should' advice like that, if it was in fact done on purpose.  If it's
> just a bug, then it's a bug, and might not need to be immortalized.
>
> My reading of that text is that if the peers have prior agreement that
> they're not going to do ND, and if the designer or user of the system
> has a Very Good Reason to believe that omitting ND is desirable, then
> it's possible to do.  Otherwise, as with any recommendation, you
> 'should' do what it says, and use ND as usual.
>
> I don't know of any cases where omitting ND for IPv6 addresses makes
> much sense.

OK, so at least we've clarified that we're in disagreement. I don't see
support in the specs for doing address resolution on links without
link-layer addresses. what would the purpose be in doing address
resolution on a link without link-layer addresses?

other ND functions, of course they should be supported.

cheers,
Ole

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to