Could the RIPE and Merit folks please analyze the current routing tables looking for the largest blocks of addresses that could be aggregated and determine who they belong to? This way we can look for maximum leverage. I know you guys don't want to hear this, but it is a good idea to look for similar leverage via renumbering. If every provider did a quick scan of their customners with pre-CIDR allocated class Bs and Cs, figured out how many hosts there are withing each of these network numbers, and contacted those who have fewer than (let's start low) 10 hosts and got them to renumber, we could get real savings. How many class Cs are out there to service a slip line going to someone's house and a few machines? We are in the middle of renumbering a couple of K-12 nets like this in New Mexico, placing them within a CIDR block (thanks to Phil Wood/LANL). The work takes a short time (small number of hosts). Net effect will be these routes turn into vapor. Think globally, act locally, peter