> Dave Crocker wrote: > By way of noting one possible scanario that builds on today's > reality and leads down a path that never adopts IPv6, As of today this _is_ the scenario. IPv6 is not even a buzzword anymore. > I'll ask: What if users turned all leaf networks into private > address space, so that public IPv4 numbers were needed only at > the level of roughly 1 per public interface? I don't think the question if "what if", the question is when. NAT will be improved, procotols that don't cross NAT nicely will suffer to the expense of those that do cross NAT nicely. > This is, of course, the ultimate breakage of the end-to-end > addressing model, but we tend to forget that the model is > something rather esoteric to non-protocol geeks. Indeed. Non-protocol geeks care about features, not about elegance. > At the least, it means we had better have an end-to-end > reference (identification) construct that is a) separate > from IP Address, This might be too ambitious, as the infrastructure required to deliver the ID-to-LOC mapping might represent more trouble that accommodating the identifier being in the same name space as the locator. > And now we try to characterize the minimal deployment of IPv6 as > if it represented success. Or perhaps the real problem was that > what has been happening with NATs, et al, is fine but we that > have preferred to tout our idealized solution and ignored > market pragmatics. I could not agree more. People have deployed NAT it solved problems, not because of address scarcity. Instead of acknowledging their needs we have repeatedly told them that their solution sucks. It has not prevented them to do more NAT, and it works better and better. I wonder how many years of double-NAT IPv4 we will have to support before we acknowledge the failure of IPv6 and begin the work on a replacement that actually addresses people's needs at a cost they can afford, not protocol designer dreams. Michel. _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf