[136599] in North American Network Operators' Group
Re: And so it ends...
daemon@ATHENA.MIT.EDU (Ernie Rubi)
Thu Feb 3 16:23:24 2011
From: Ernie Rubi <ernesto@cs.fiu.edu>
In-Reply-To: <76CBE1FA-FDFF-446D-8ACA-C304D9CA2DE2@virtualized.org>
Date: Thu, 3 Feb 2011 16:08:50 -0500
To: David Conrad <drc@virtualized.org>
Cc: NANOG list <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
Way off topic here...and into the legal arena:
As to the monopoly classification, do you think, at least with ARIN =
(since it is a US/Virginia corporation) that Sherman Act =A72 (i.e. =
antitrust) principles could be applied to require that it relinquish =
some of the control over said IP space/database and act in a more =
competitive manner? What about the other RIRs worldwide? I'm not an =
antitrust lawyer, but there may be an issue there.
There was a paper a while back from a UMiami (Michael Froomkin) =
professor talking about ICANN and Antitrust. =
http://arxiv.org/pdf/cs/0109075 - This is a legal paper, not an =
engineering paper.
I wonder if those same principles could be applied here. =20
On Feb 3, 2011, at 3:42 PM, David Conrad wrote:
> On Feb 3, 2011, at 8:59 AM, Owen DeLong wrote:
>> That remains to be seen. If they give up their space, it is unclear =
that they have any right to transfer it to another
>> organization rather than return it to the successor registry. There =
is no precedent established showing that
>> this is allowed.
>=20
> Right. Like Compaq returned 16/8 when they acquired Digital (and HP =
returned 16/8 when they acquired Compaq).=20
>=20
>> That remains to be seen. IANA has declared them the successor =
registries
>=20
> No. First, "IANA" does not exist. The term "IANA" now refers to a =
series of functions currently performed under contract from the US Dept. =
of Commerce, NTIA by ICANN. As such it can't declare anything.
>=20
> Second, neither ICANN nor the USG has (to my knowledge) declared the =
RIRs to be "successor registries" (whatever they are). The IPv4 =
registry continues to exist and will undoubtedly be maintained as it =
always has been. The only real difference is that there aren't any more =
IPv4 /8s tagged with "UNALLOCATED".
>=20
>> The other thing to consider is that the RIR doesn't really need to =
"reclaim" the block, per se. They can simply stop providing uniqueness =
to the organizations that don't have a contract with them and issue =
those numbers to some other organization that has a contract. The other =
organization would know that their uniqueness is limited to those =
cooperating in the registry system.
>>=20
>> Does an organization that has no contract with an RIR have a right to =
expect that RIR to continue to provide them a unique registration?
>=20
> The RIRs are self-defined geographical monopolies that provide a set =
of public infrastructure services to the Internet community at large. =
It's an interesting question whether that service is limited to only =
those folks who pay -- my guess if the RIRs took this stance, they'd be =
looking down the barrel of numerous governmental =
anti-monopoly/anti-cartel agencies.
>=20
> However, pragmatically speaking, the folks who matter in any of this =
are the ISPs. The RIRs exist primarily as a means by which ISPs can =
avoid doing a myriad set of bilateral agreements as to who "owns" what =
address space to ensure uniqueness. If the RIRs reduce their value by =
no longer providing that service in an effective way (e.g., by doing =
what you suggest), I suspect the ISPs would find other entities to =
provide global uniqueness services.
>=20
> Regards,
> -drc
>=20
Ernesto M. Rubi
Sr. Network Engineer
AMPATH/CIARA
Florida International Univ, Miami
Reply-to: ernesto@cs.fiu.edu
Cell: 786-282-6783
