Bound, Jim wrote: > The fact is that the WG believes this is important > discussion.
It is an important discussion. There is a very critical architectural point here, and it is being glossed over by 'the sky is falling' claims that apps might fail, or have to do some work. The architectural point is that IPv6 nodes support simultaneous use of multiple dynamically acquired addresses per interface. This is a distinct value over the capabilities of IPv4, as many failure modes arise from the inflexibility of a single address at a time. The reality of the network is that different nodes will have different perspectives of reachability for any given address. Some consider this dysfunctional, others consider it a feature. In any case, as long as we provide a flag to indicate when there is a reachability discrepancy (intentional or an artifact of network dynamics), there is an opportunity to prioritize the list to achieve the app goals. > Whether Mr. Hain likes it or not. There are arguments that apps should not ever be aware of topology. I completely agree. Whether you like it or not, the current attachment point of the end systems is topology information. If the apps require input of, or choose to reach down and grab the attachment topology information to pass to a peer, they have violated the premise that they don't know anything about topology. This is not a function of my likes or dislikes, it is a simple matter of technical fact. Tony -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
