> > Tony, we understand that we don't have a > > completely connected network. We've understood that for years. > > But we're not willing to adopt your delusions that this is > > the same problem as having ambiguous addresses. > > I am focusing on your 4/18 assertion: introduction of scoped addresses > causes far more problems than it solves. > > I am not talking at all about ambiguity. ambiguity is certainly one class of problems that site locals cause. now whether you define scoped addresses as "names for locations in the network whose meanings are relative to a particular subset of the network, and which may have other meanings in other subsets of the network", or "addresses are only reachable from a particular portion of the network" is, I suppose, fairly arbitrary. I've been using the former definition. Keith