Re: [sig-policy] IPv4 transfer proposals {Security = Unclassified}
> Most ISPs look to RIR databases to confirm authorization prior to accepting
> a prefix from someone.
They only do so because the RIR databases are the most comprehensive and/or accurate. If the RIRs take active steps to reduce the validity of their data, it isn't clear to me ISPs will continue to use those databases (particularly if a more accurate/comprehensive alternative is available).
> RIRs hold the keys to in-addr.arp as well.
Actually, http://in-addr-transition.icann.org/
However, ignoring that, it would be very interesting to see what would happen if the RIRs were to disallow an in-addr.arpa update because the administrator of the space was using a different registry. I suspect the results would largely depend on the anti-trust/anti-cartel laws in the RIR's (or ICANN's) legal venue.
> I think that those forces will drive the market towards transfers that
> can be recognized by the RIRs even with policy enforcement.
The question is at what point will 'policy enforcement' result in folks abandoning the existing system. The jury is still out as far as I can tell.
> A market with limited supply and no rules is a market begging to
> be manipulated for arbitrage opportunities. Doing this to the IP
> address pool will be a great disservice to the community.
The implication of manipulation/arbitrage is that IPv4 address space will become too expensive for (some) folks to obtain it. However, you might have heard the recent announcements about the end of the IPv4 free pool with the numerous statements that people "will be unable to obtain IPv4 addresses". Pragmatically speaking, I'm not sure what the difference is (other than some folks who have large stores of allocated-but-unused address space may get a significant windfall: sort of like waking up and finding an oil under your family's ancestral property).
I figure increasing the cost of obtaining IPv4 address space will result in folks taking a more serious look at the cost/benefit ratios of deploying IPv6 and tools like NAT64/NAT46. Seems like this is something all the RIRs should get behind.
However, I understood quite some time ago that these discussions had passed beyond the realm of rationality and moved into religion.
Regards,
-drc