> >> perhaps, but one might also reasonably expect 2^0 networks to be >> insufficient. > > > At the risk of repeating myself, I respectfully disagree. Given that you > can reasonably build a flat subnet of 1000 hosts today, it does > not seem like an unreasonable entry point. Mom & Pop 6-pack > have one computer. George & Jane Jetson only have 6. Why > does the entry point have to be more than two orders of > magnitude above the common case? you seem to be stuck in the 20th century. where do you get the idea that IPv6 is only going to be used (or even, mostly going to be used) for things that resemble desktop PCs? desktop PCs are dinosaurs nearing extinction, and early indications are that homes are going to have multiple kinds of wireless networks. bridges between dissimilar media don't always work so well. and given the range of some of those media, it should not be assumed that everything within a "home" will be in the same layer 2 segment. IPv6 needs to last a lot longer than IPv4. it shouldn't be constrained to fit yesterday's notions of what's reasonable. >> so are prefix allocations to users of less size than a /48. but if ARIN >> can't follow the specifications that everyone else uses, this should >> call their competence into serious question, and maybe IANA should find >> someone else to dole out IPv6 prefixes for that part of the world. > Once again, that's the MSO's doing that, not ARIN. whoever is doing it needs to stop, or be stopped. _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf