While I completely agree with the general message of this memo, that IANA works fine so don't break it, I have a few questions and a concern: At the top of page 8, it refers to disputes about policy. To me the question is ambiguous: does it mean disputes within the IETF, or disputes between the IETF and IANA, e.g., "we can't implement that"? As far as I know, there's never been a significant policy dispute with IANA, so you might as well say so for anyone who was wondering about that question. At the top page 11, it claims that the MOU is "global in nature." While that is surely the intention, the MOU is in practice a contract between ICANN, a California corporation, and the IETF which to the extent it exists, is a Virginia trust. So if push came to shove and one side or the other had to enforce some provision of the contract against the other, US law would apply and it'd happen in US courts. So say that -- the IETF operates globally, but it is domiciled in the US, and the current MOU with ICANN is a US agreement. Following that, there is a discussion of all the stuff we don't want to change, all of which is fine. But it doesn't say other than sort of implicitly in #3 on page 13 that the IETF needs a binding agreement with the IANA operator that has protections for the IETF community that are substantially the same as those in the MOU in RFC 2860. It really needs to say that explicitly. If IANA stays with ICANN and ICANN reaffirms the MOU, we're fine. But if it were someone else (wave hands about free-floating non-US things that keep coming up in ALAC and NCUC discussions) who was less interested in spending $1m/yr to provide a free service to what they saw as a bunch of nerds from rich countries who keep getting in the way of their core mission*, things could get very unpleasant. R's, John * - providing an ever increasing revenue stream to contracted registries and registrars