IP: Something for IP on DNS squabbling

[Dave Farber <farber () cis upenn edu>]

To: farber () cis upenn edu
cc: king () wang-wei ics uci edu
Subject: Something for IP on DNS squabbling
Date: Mon, 18 May 1998 17:48:44 -0700
From: "John L. King" <king () wang-wei ics uci edu>




I have been surprised and amused by the postings from
intelligent people who seem to think that communications
addressing can be handled by adopting a lassiez faire
approach to the problem. The very fact that there is such a
fight over the "rights" to TLD assignment proves that
there is value in communications addresses.  In virtually
every market where valuable things are traded, there are
institutional constraints to govern trade.  Truly free
markets are idealized abstractions, and always will be.  The
question is not whether trade in domain names will be
regulated, but how it will be regulated.  Chances are pretty
good that we will go with what we know how to do, rather
than take a radical departure into the unknown.


Ashley Andeen and I have argued in our paper "Addressing and
the Future of Communications Competition: Lessons from
Telephony and the Internet" (in Brian Kahin and James
Keller, "Coordinating the Internet," Cambridge: MIT Press,
1997, pp. 208-257: a draft accessible on the web at
http://ksgwww.harvard.edu/iip/cai/king.html) that there are
important similarities between the issues of telephone
addressing (numbering) and internet addressing.  Both
require authoritative assignment of the top-level address
structures at each layer of the addressing strata, if for no
other reason than to provide effective dispute resolution.


The telepone addressing world, for all its faults, deals
effectively with a much larger addressing administration
problem than the DNS currently handles, through a set of
authoritative and hierarchical structures coming down from
the ITU's World Zone assignments (originating from the
original CCITT assignments before CCITT was absorbed into
the ITU), to regional administrators (in the case of the
US, this is World Zone 1, originally administered by Bell Labs,
then Bellcore, and now the North American Numbering Council
http://www.fcc.gov/ccb/Nanc/), and eventually to central
office code administration.  People who dislike hierarchy
and authority naturally object to such a scheme.  But it
works, and no one has yet offered a replacement scheme that
inspires sufficient confidence to start a reform effort.
I have yet to see a sound argument to the effect that Internet
addressing is fundamentally different in technological or
institutional dimensions from telephone addressing.  


This is not an argument to have the ITU or any particular
organization take over DNS TLD administration.  However, it is
an argument that SOME authoritative entity will hold the
vital agency role of assigning and adjudicating disputes at
the global DNS TLD level, and that similar hierarchical
authority will govern domain name administration as one
works down the tree.  The key question bedeviling the DNS
community is not the superficial haggle over who shall hold
that agency role, but the deeper and more important question
of who holds the principal power to assign such agency.
Indeed, this is the heart of much of the squabbling on the
issue at this time -- the question of who gets to decide how
it will be done.


John L. King
king () ics uci edu
http://www.ics.uci.edu/~king