Right. Exactly. But UNTIL such time as easy renumbering for everyone is a reality,
People have known that renumbering technology was extremely important since the first CIDR drafts came out. What has happened in the intervening time? Precious little -- there has been little to no incentive for the R&D necessary as it was always easier and cheaper to get a big block from the registries. You might look at SprintLink's filtering as something of an incentive. Now that there is an incentive for people to investigate renumbering technologies (e.g., provider based addressing), architecturally heretical and impure NATs and ALGs are popping up like the vile weeds that they are (:-)). Heck, there is even an IETF working group looking into the issue. Why has it taken 7(?) years for the creation of that working group?
I'm doomed to renumbering in the event that provider can't provide me with all the service I need, and I can't subject my customers to that.
Um, have you asked them? From ancedotal evidence, it seems that only a small minority of customers go into convulsions when you whisper the word 'renumber'. Don't force it on them, make it worth their while, e.g., offer them a discount or something.
I forsee a large number of new /19's filled with fictitious entities, followed immediately by the application for an additional, larger, block...
As if this was new...
and if we're concerned about the efficient use of address space (and not just routing table size) that's a serious waste.
That was my original contribution to this thread. This isn't surprising, it is simply the result of the "tragedy of the Internet commons". Regards, -drc