I just want to reiterate something that Tony Li had mentioned (and I hope Tony will correct me if I've lost something in the translation):
Given that if the mechanism exists we know people will develop NAT functionality in order to isolate enterprises from IP address changes, what is the benefit of going forward at all with IP version 6? A large address space is useless if you only need a small one. We already have that.
It is true that enterprises could choose NOT to do what I wrote, above. But I think that's unlikely for the very reasons that many people who voted "no" have mentioned in their arguments. People understand a particular paradigm, today. It takes effort for them to learn a new one. The IETF doesn't make laws, so all it takes is some amount of interoperability -- not even a whole lot.
Nobody "markets" such a principle from a "for profit" perspective, and certainly no company has a business plan that sells the principle. That task is left to a few of us religious zealots.
This is it, folks. Here is an opportunity to market the end to end principle by not codifying the easy out that breaks it. Let's make a clean case for doing things right. We can relax that (admittedly) strident view later, but coming out the gate, remember the effort required for administrators to use site-locals as far thinking people like Tony Hain have envisioned.
Eliot
-------------------------------------------------------------------- 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] --------------------------------------------------------------------
