Re: [idn] Please do not pass an IDN document which you know will cause conflicts and chaos

["Soobok Lee" <lsb@postel.co.kr>]

----- Original Message -----
From: "Dave Crocker" <dhc@dcrocker.net>
> Until then, please consider that further delay means that there is no
> standardized DNS support at all for any characters other than ASCII.
>
> Please consider whether it is better to wait years for a perfect, complete
> solution,

Sorry. Premature IDN solutions ,if ever issued, will last for several decades .
If just a few years of more work could give us perfect IDN solutions including TSCONV,
I'm willing to be patient living only with IDN testbed registrations or
alternative directory/search solutions which address the need for
internationalized *ACCESS* to LDH domain names. that's enough to today's
needs.

> or whether it is better to do something useful sooner.
>

Solving unresolved IDN issues using Registratation policies above IDN
doesn't remove the apparent serious problem of application-side matching failures.
Policies can guarantee the ownership by the same registrant, but cannot guarantee
the universal identifer equivalence across the entire internet and applications.

Moreover, it is not scalable in general.
disputes resolution processes or zone admin's inspections for IDN-vs-IDN and IDN-vs-LDH RR
conflicts will explode and they will consume very precious time and resources
from zone admins and end users and therefore, they are not scalable to huge zones
and the general public. Dynamic RR insertion/updates features should be turned off
for the fear of introducing new conflicting RR.

The concept of IDN itself has been challenged repeatedly.
When end users - most of them is unilingual or at most bilingual if gifted -
are exposed to exotic IDN labels, the level of confusions and difficulties
will be great but are not measured and researched sufficiently.
Even after more than one century from the invention of telephone networks,
we still have digits-only phone numbers.

Enen though this may looks somewhat drastic and ironic,
If domain names are to be truely *international*, the basic lexical atoms of the domain labels
should be remained as LDH set, the greatest common denominator of characters and scripts used by modern civilized language groups.
LDH is the true IDN, and the "IDN" that we discuss  had better be renamed to LDN (localized domain names) that have requirements to
be internationally interoperable by
machines and human beings but not satisfied.

Soobok Lee