Please read Local Network Protection for IPv6 [RFC 4864].

You've obviously never tried to build a router in hardware.

Running code will get you halfway there, but if you're hoping to set a protocol standard with it, you'll need to get a lot better at politics.

I agree, and further assert that it is not too late to go back and do that.

I await the submission of your Internet Draft with bated breath.

A lot of people don't realize that going from unstructured IPv4 addresses to structured IPv6 addresses means that the 128-bit fixed address size places an upper bound on the aggregate size of all the structured subfields encoded in an address. For an example of how this pressure is already beginning to arise, have a look at how ARIN and NANOG are reacting to the 6RD proposal.

This would mean updates to all hosts and all NAT devices...

And that's why what you're talking about wouldn't have been any better than IPv6. It requires updates to almost every unmanaged residential host and gateway before anybody would be able to rely on it, which is the main problem with IPv6. Only, your proposal doesn't fix any of the other problems in IPv4 in the process. it just adds address space, and in that sense, isn't really better or different than Realm-specific IP [RFC 3103], which ended in precisely the same ignominious failure that your idea would have met. QED.

I hear he's really fat.

I assert that the migration would already have happened (and seamlessly) if we had just extended the address space and left everything else the way it was.

That "solution" would have still presented as "ships in the night" and uptake would have been resisted for all the reasons that uptake of IPv6 has been resisted: no forward or backward compatibility. The fixed address size in IPv4 is the cause of the compatibility problems, but it's very badly exacerbated by the ubiquity of network address translators.

At some point, the fixed address size in IPv6 may turn out to cause similar related problems, but we should be in a better position to avoid them as long as we don't succumb to the siren call of adding back network address translation on IPv6.

If I had a dollar for every time I've been proven wrong after saying exactly that, it would make my preparations for retirement a lot easier.

Probably around the same time we stop pretending that a toothless regulatory agency staffed via a revolving door to the regulated industry is functionally equivalent to a Stalinist command and control economy.

or perhaps, demanding higher prices from those customers who refuse to be NATed ...or perhaps just refusing to assign public addresses to anybody. "Don't like it? Tough. Call your congressman."

Should I use a ridiculous hack like upnp to beg the nat device for a forward?

The large-scale NAT that ISPs are deploying now do not implement any kind of port forwarding, so you can just forget all about that now.

What makes you think they would want to sell into what is obviously about to become a marketplace shocked by sudden scarcity? What, were you planning to slap them with a punitive "windfall asset" tax or something if they don't give up their legacy address grants? I can't WAIT to see the reaction *that* RFC.