[IP] more on Ending ICANN, U.N. overreaching through technical means: a proposal
Begin forwarded message:
From: Joe Touch <touch@xxxxxxx>
Date: July 16, 2005 3:42:42 PM EDT
To: dave@xxxxxxxxxx
Subject: Re: [IP] Ending ICANN, U.N. overreaching through technical
means: a proposal
Dave,
Comments below:
David Farber wrote:
Begin forwarded message:
From: Declan McCullagh <declan@xxxxxxxx>
Date: July 14, 2005 12:39:32 AM EDT
To: politech@xxxxxxxxxxxxxxx
Subject: [Politech] Ending ICANN, U.N. overreaching through technical
means: a proposal
Previous Politech message:
http://www.politechbot.com/2005/07/13/will-the-un/
---
POST ANONYMOUSLY
ICANN's most useful contributions are the assignment of IP addresses,
DNS domains, and other numbers as specified by the IETF (Internet
Engineering Task Force) in its RFCs. When Jon Postel ran it from his
office, everything was fine. Now ICANN is an unregulated bureaucracy
focusing on increasing its status and revenues by approving more
expensive top level domains and charging more for existing domains.
Because they share so many goals, a hostile takeover of ICANN by the
U.N. is a perfect move, and sure to doom all of us to even worse
governance. (I won't bother to go through the complete failure of
ICANN
to stick with its promised democratic reforms, or its inability to
respect its own elected officials).
The IETF can put a stop to much of this through technical means. For
example, develop an alternate domain resolution algorithm (much like
Google's "I'm feeling lucky"), so that when you type in Mary, the
browser may go to Mary Kay, the Catholic Church, or your friend Mary's
home page, or when you copy links from your browser to your e-mail
program to send to your friends, the domain name is displayed however
the webmaster may want. Once domain names become invisible bits
like IP
addresses, then there will be no point in fighting about them.
There have been numerous attempts at this - e.g., new.net
For every new root, large organizations will just purchase names in all
such roots. Users, who can't assume that's been done, must search all
such roots to find what they want. It's a lose-lose proposition.
Domain strings are meant to be recalled by humans, but not guessed by
humans.
The DNS is not a search engine. We have those, and they work fine.
Also, the IETF should end concerns about IP shortages by working to
expand NAT implementations and the use of IPv6, with its effectively
infinite number of IP addresses, into the real world (perhaps IPv4
with
NAT at the client end where needed with a mostly IPv6 Internet core).
Finally, the IETF should stop giving IANA and ICANN new duties in its
RFCs. Where unique numbers must be assigned (e.g. well known TCP and
UDP ports), technical means should be used to assign those numbers
(e.g.
a simple web based registration form), rather than going through IANA
and ICANN.
-- Anonymous
Technical solutions are great - when they work. NATs don't - or at least
not the way many think; by making the NAT'd network look like a single
host, some protocols (e.g., VoIP, web servers, etc.) don't work or
require extraordinary efforts to compensate.
And a web page to hand out a limited resource is a fine way to deplete
it quickly, nothing more. The issue with assigned numbers is exactly the
need to avoid competition for preferred numbers (e.g., for ports,
400-405, 411, etc.) and preserve the limited resource (1024 system ports
and 48K of user ports).
A web page is not a technical solution; such a solution might be a
protocol for dynamically finding the port for a protocol, ala the RPC
portmapper, but we're a long way from a global portmapper.
Joe
-------------------------------------
You are subscribed as roessler@xxxxxxxxxxxxxxxxxx
To manage your subscription, go to
http://v2.listbox.com/member/?listname=ip
Archives at: http://www.interesting-people.org/archives/interesting-people/