Re: [DNS] the road ahead

Re: [DNS] the road ahead

From: Geoff Huston <gih§>
Date: Tue, 23 Jun 1998 19:45:06 +1000
These are all wonderful assertions, but they bear little detailed scrutiny. 

You have put together a number of assertions cloaked in some emotive
wrappers and proffer it as some single way forward. If it were all so easy
to work through then all this would never have got to where it is today.
But, oddly enough, there are other views which are quite contrary to the
ones espoused in your posting.

The issue is that mistakes in a name structure are hideously difficult to
undo. Like it or not, names are heavily overloaded on the Internet and they
have to serve a multiplicity of roles. In this light the assertions you've
made do not withstand more leisurely analysis. Pulling out the old emotive
catch cries of the past 5 years is simply not anywhere near enough to make
progress here. Perhaps we all need to spend some time to think through
these issues before pulling out the trusty keyboard and blasting away.

Maybe the following will assist here - to quote from an earlier draft I
wrote on the subject...

However this proposal, of increased levels of supply of SLDs is, in its
present form, as functionally flawed as the current environment of
processing bottlenecks within the current set of SLDs in that neither
process can definitively ensure the continued value and utility of the .au
Domain Name environment itself, at least in its current form. 

It is noted that in increasing the number of second level domain names
there is a distinct risk of over compensation and a distinct risk to the
overall perceived integrity of the environment. The integrity perception
issue will be problematical with the likely entry of operators into the
environment who see this is an opportunity to generate income without
undertaking any service obligations associated with the task. Irrespective
of the administrative preconditions imposed on DAs there will be incidents
of scam-based trading practices without doubt. The over compensation risk
lies in the proliferation of SLDs, and the consequent devaluation the
inherent utility of the name system. For example, if there are some 1,000
second level domain names and a named entity of "xyz" a search for the
"right" xxx where the "xyz" of interest is named as in may be
prohibitively lengthy and may not yield a definitive result. The
alternative is that entity "xyz" request a name in every available SLD
category - an unwieldy and probably expensive course of action which will
not be successful in every case, and if commonly attempted simply negates
the value of having SLDs in the first place (the comparable observation is
that immediately increasing the number of distinct categories within the
Yellow Pages directory listing by one or more orders of magnitude would
make the directory functionally useless). This proposal therefore is not a
solution to the expectation of name functionality that many users of the
Internet implicitly expect with the name system, as the equation of the
name system to a defacto Internet directory service is a strongly
entrenched belief.

The second factor is that the basic issue of liability of the SLD DA
administrative role is not definitely addressed, and the ultimate
derivation of authority to make decisions regarding the overall
administration of the domain name system is not addressed. Accordingly in
either environment any SLA DA must operate in a way that minimises
potential liability rather than in a rational way that maximises the value
and utility of the Internet domain name system. 

The fundamental cause of this flawed environment is the implicit assumption
that the Internet Domain Name environment is not considered an
authoritative name source, and litigants can cite other name authorities as
some form of overriding legitimacy of exclusive right to a name and claim
damages against the operators of the domain name system. This is a flawed
environment, and can only be addressed by a fundamental shift within the
public regulatory environment to allow a managed framework of registration
of Internet domain Names, where due adherence to the defined framework will
ensure that the process creates a name authority for the a name which is
functionally independent of any other means of citation of authority to use
a name. Thus the process of assigning a domain name of, say, to
an entity (on the basis that the name
was not assigned to any other party, that the bona fides of the applicant
had been established, the appropriate fees (if stipulated) had been paid
and the relevant administrative policy and process for the SLD had
been followed) should be an act which does not incur a subsequent liability
on the part of the DA or the registry operators. This does require
some form of public instrument to effect this critical change in
recognising the authenticity of the Internet Domain Name environment. It is
noted that such public regulation of the activity may be accompanied by
operation of the process by a duly created operating entity as a component
of government administration, or in other cases it may be accompanied by a
process of passing the responsibility for the operation to duly licensed
bodies. Either approach would be a significant step forward in preserving
the value and utility of the .au name domain over the current situation, as
it would add a visible line of derivation of authority
within a process where the ultimate functionality of the Internet domain
name system is the foremost concern and the integrity of that functionality
would be best safeguarded through such measures.


At 19:26 23/06/98 +1000, vicc&#167; wrote:
>a) stakeholders not allowed to participate.

Received on Tue Jun 23 1998 - 17:49:44 UTC

This archive was generated by hypermail 2.3.0 : Sat Sep 09 2017 - 22:00:03 UTC