[arin-ppml] Some observations on the differences in the various transfer policy proposals
Geoff Huston
gih at apnic.net
Mon Oct 20 04:00:17 EDT 2008
Hi Scott,
I'm pretty sure that what I was saying was not working for you, and
maybe rephrasing my thoughts might possibly help.
What I'm trying to say is that attempting to place the registry on a
course of saving the Internet from all kinds of dire perils in
routing, address speculation, unfairness, and such, with only a
registry function to work with to achieve this is pretty high risk
path to take, in my opinion.
[Press delete now if you're already drowning with the volume on the
ppml list! The rest of this note just expands upon this argument.]
When there are no more allocations to be made there is no de facto
monopoly of the RIRs. The registry function is not one that is the
unique prerogative of the RIRs to operate - quite the opposite, and,
as we've already seen with the proliferation of IRRs, when you don't
like the policies or practices associated with using one registry you
can always use another! Yes, I appreciate that these are heretical
thoughts. It would be good to believe that the RIRs will continue to
have some enforcement authority in this space that could allow the
inclusion of a broader agenda related to mitigating some of the worst
risks in address transfers and markets, but, frankly, I personally
don't find such assertions of enforcement authority at all comforting
or even credible.
If access to the registry has onerous constraints and conditions, the
registry runs the risk of annoying the registry's customers. And while
the address allocation function is a de facto monopoly, registries are
not. Registries can come and go in the blink of an eye. If a registry
manages to alienate its customers then the most obvious response is
that other registries, with more neutral stances, will emerge and at
that point address chaos is only a short step further down that track.
thanks,
Geoff
Disclaimer: These are all my opinions, totally.
More information about the ARIN-PPML
mailing list