>I believe that if companies rearchitect their networks so that they: use private
network numbers with NAT >tools with leaf networks, and use public network numbers
for routing between networks, and in transit >routes, etc...This is already occurring,
thanks to the high cost of IP addresses. The only place you will see unique
IP addresses being delegated to individual workstations is at older companies and
.edu sites, where this was the way it was set up and they haven't spent the effort
to fix it. The problem with NAT (ipmasqing, subnetting, etc all included in
that) is that it is very hard to provide a server through that -- so IP addresses
are still being used up at a pretty good clip. NAT/ipmasqing/subnetting/supernetting
is the reason that we haven't reached the end of the available addresses yet -- but
that time is coming. And now people are wanting static or near-static IPs again
(DSL/cable users, etc) because they are beginning to provide services on their own
(gaming servers, small web servers, Freenet nodes, etc).
It boils down to this:
due to innovative technology, the problem has merely been delayed, and we have enough
time to do it right. Let's not screw up the next system, huh? That's
why we're here at ICANN.