Re: [wg-c] IAB Technical Comment on the Unique DNS Root

[Karl Auerbach <karl@CaveBear.com>]

> [The IAB has sent the following note to comments@icann.org]
> 
> IAB Technical Comment on the Unique DNS Root
> ============================================

That's pretty old news. (From the the September time frame.)

And like the old claim that bumblebees can't possibly fly, the fact of the
matter is that competitive roots do work.  (I've been using 'em for more
than two years.)

    http://www.cavebear.com/cavebear/growl/issue_2.htm#multiple_roots

If the IAB were right that the current design is susceptable to damage
from something that any kid can do with any Linux box without violating
anyone's security then one would have to question the viability of the
design.

Further, the IAB's statement begs the question that if there is but one
root system, why should it be ICANN's?

If one is concerned about "stability" of the net, then one's prime focus
ought to be those sources of actual daily failures - misrouting,
inter-carrier exchange point congestion, and inter-implementation
interoperability - rather than focusing on some hypothetical problems that
might occur if somebody publishes a competitve internet book of names
without ICANN's blessing.

I am happly using a competitive root system, and as you may note, I am
having no difficulty communicating with everyone else.  Nor is anyone
having trouble communicating with me.  And I haven't seen a single DNS
root service outage in more than two years of operation.

As for the costs of our Ostrich-like head-in-the-sand approach to
competitive roots: It is an approach that is precluding useful, indeed
valuable, directions of network innovation.  The imposition of one
catholic root removes an element of self-choice from those who wish to
limit their view of the Internet name landscape.  And by preventing
proximate location of roots and net users, the single-rootists impose upon
carriers a significant packet burden and cost (measured both in dollars
and in waiting, unsatisfied users) that otherwise need not be borne.

		--karl--