Re: [DNA] Confirming today's face-to-face meeting decisions

[Erik Nordmark <erik.nordmark@sun.com>]

Vijay Devarapalli wrote:

> the framework seems to recommend that we need a Link identifier
> solution, fast RA, fast RS, optDAD and TSLLAO. it says with this
> we have a good DNA solution.
> 
> that doesnt seem like a "supporting structure" to me.
> 
> maybe I dont understand whats in the framework document.

This discussion, about the content of the document, is what I'd like to 
see. Thanks.

The current content of the document captures (captured) the authors 
ideas when it was written. It probably makes sense to rephrase things as 
the desirable properties.
For instance, I think we know we want a solution which uses router 
advertisements so that one can find the configuration of the new link at 
the same time as detecting whether or not one has moved to a new link.

And given that we have two different styles to detect the link (an 
explicit link identifier, compared to what I would characterize as a 
cookie or a landmark, where the host asks "does this landmark still 
exist" without an explicit linkid) we should express the need based on 
the property that a host should be able to determine whether it is on 
the same link or not from receiving a single message.

   Erik