IAB comments addressed my main concern about the document: a too large part given to "Conservation" but I still have minor comments: - 4.1.3.1 Verification of Utilization it is not reasonable to ask that all assignments to end-sites must be registered in the public database: ISPs need some kind of confidentiality in their business (cf Ruediger Volk). - same sub^3 section ping doesn't work well with anycast destinations - 4.2.1 Assignments to End-users I'd like to see conditions of assignment for: * an address (128 bit "prefix") * a subnet (64 bit prefix) * a site (48 bit prefix) Today one of more nasty problems is for a dial-up user it is hard (or expensive) to get more than one (dynamic) address then many people install NATs for that not technical reason. Of course we don't want to see the same thing with IPv6 then this point must be clear and there will be a FAQ about the "connected dentist office" (technically easy, it is a case where RIPng is well suit!) - I can't understand second opinion requirements (a must for users and a should for ISPs). I am impatient to read the next version of the document... Francis.Dupont@inria.fr