Re: [council] Discussion draft on unique, authoritative root

["Milton Mueller" <Mueller@syr.edu>]

>>> "M. Stuart Lynn" <lynn@icann.org> 05/29/01 12:42PM >>>
Thank you for your feedback. I know you are an enthusiastic proponent 
for abandoning a single root and I read your comments with great 
interest. 

>>> Brian E Carpenter <brian@hursley.ibm.com> 05/29/01 09:51AM >>>
  A hierarchical naming structure without a uniquely rooted tree
  is ambiguous (*) and therefore will balkanize the network. 

Stuart and Brian:
I'm afraid your comments show that you have a misconception
about the issue the DNSO is considering.

The problem was are concerned with is NOT "is the DNS designed to 
have a coordinated root?" Everyone knows that it was. The issue is, 
"if alternate roots or TLDs gain critical mass, what should be done 
about it?" and "how should ICANN deal with conflicts between
its own name space and others?" Those are policy issues and 
they are new ones.

No one that I know of is a proponent of "abandoning" a coordinated
root. The issue is: what coordination method will be used, and whose
interests will that method serve?

To help you both understand this distinction, I have forwarded some comments I have written about RFC 2826, "IAB Technical Comment on the Unique DNS Root." 

—--------
RFC 2826 is often cited in the policy debates over alternate roots. 
According to RFC 2826, the DNS protocol was designed 
with the assumption that there would be only one 
authoritative root zone file. The statement goes on to 
describe some of the difficulties that might occur if 
computers attempting to resolve domain names are 
confused about the contents of the root zone file. The 
fundamental conclusion of RFC 2826 is this:

"a degree of cooperation and agreed technical rules 
are required in order to guarantee the uniqueness of 
names. In the DNS, these rules are established 
independently for each part of the naming hierarchy, 
and the root domain is no exception. Thus, there must 
be a generally agreed single set of rules for 
[assigning the top-level domain names listed in] the 
root."

I agree with this part of the statement. As far as I know,
everyone does. Unfortunately, the IAB did not stop there but
attempted to influence policy debates by concluding that "it is not 
technically feasible for there to be more than one 
root in the public DNS." 

It is a rather strange claim. 
There ARE different root server systems in operation. 
These alternate root systems use the same DNS protocol 
and the same software implementation (BIND) as the 
ICANN root servers. Most, if not all, of them are 
capable of resolving all names under the IANA-
delegated top-level domains. So it cannot be argued 
that they are not an implementation of the Domain Name 
System protocol. Nor it is technically correct to say 
that they are "private" rather than "public" name 
spaces. All of the alternate root systems are open to 
any ISP or end user who wishes to point resolvers or 
name servers in their direction.

Even if alternate roots did not exist now, nothing in 
the DNS protocol prevents a subset of the world's 
Internet service providers or end users from 
redirecting their name servers to some place other 
than the ICANN-administered root, if they wished to do 
so. Thus, the IAB statement is wrong: it IS possible for there 
to be "more than one root in the public DNS." 

At worst, the IAB statement is just an empty tautology
that says that the authors will refuse to call another root
a DNS root. 

At best, the IAB really wants to say that the existence of
multiple roots may lead to various compatibility problems in 
resolving names. 

But RFC doesn't say anything useful about what to do
if an alternate TLD system such as New.net gains critical
mass. How should those compatibility issues be resolved? 
Furthermore, the document does not address whether the 
value added by such competition is worth the price. 
Indeed, that is not a technical matter and is therefore beyond 
the purview of the IAB.

It is also worth noting that one can agree with the 
assertion that "there must be a generally agreed 
single set of rules for the root" without necessarily 
agreeing that ICANN is the sole or proper source of 
those rules. Nor does the general need for a single
set of rules eliminate the legitimacy and benefit of
debate over what those rules should be. 

There is, for example, a need for governmental 
laws to be promulgated by a single source; few of us
(except for a few anarcho-capitalists) would like to 
see competing governments issuing different and
incompatible laws. However, all (except for a few
advocates of dictatorship) would agree that competition
through the political process over what the rules
should be is a good thing. 

So in general, I think that RFC 2826 contributes
very little of value to the POLICY debate over 
alternate roots. 

And its contribution to the technical
understanding of what to do about competing roots
when and if they arise is minimal.

—--------

In the future, if you wish to contribute productively to this
dialogue, please refrain from yet another assertion of 
the need for coordination at the root. Please try to explain
what methods can be used to preserve a coordinated root
when competitive and technological forces might lead in 
another direction. 

Milton Mueller