Re: [idn] time to move

[Eric Brunner-Williams in Portland Maine <brunner@nic-naa.net>]

Adam,

In your reply to Dan Oscarsson you wrote:

> Please remember that DNS is not the problem.  DNS is just one of many
> protocols that embed domain names.  Others include message headers,

How does dns embed domain names? Is the resolution question (assuming it
still exists) equivalent to the transparent to code-set-dependent apps
question?

I'd the impression it wasn't, and that most of this group's focus was on
constraints for, or elimination of, the transparency question.

I fail to see the point of your second note.

The issue of encoding efficiency is interesting. A small cabal (the IE6
P3P implementors, myself and some of the P3P Spec WG considered just how
"compact" to make XML privacy policies on cookies. Two other groups, the
Mozillia and Amaya developers, and quite a few IETF I-Ds hopefully set on
a standards track trajectory, seem genuinely committed to XML. If your
view is the best one, then most of these people, and their products in the
56k-bandwidth-defined mass market are very poorly advised.

As today is "call me an idiot" day, would you do me the kindness of letting
me know where the i18n architectures of AIX 3.2, Solaris 2.6, and HP-UX 10
and their successors in interests, all of which use, or anticipated use of
utf8 for internal encoding. I've a lasting interest in the strengths, and
weaknesses, of "code set independent architecture" conception, implementation
and deployment in mid-90's POSIX operating systems.

I've obviously failed to grasp something deep and profound, either about
l10n-capable edge devices (other than cell phones), or last-mile network
characteristics, and so I've probably made a muddle of the dns registry
provisioning protocol, which a) uses XML, b) allows for utf8 in the
encoding of technical and social data, c) declines to impose heroic hardware
on its nodes, and d) also declines to impose heroic bandwidth on its links.

Tia,
Eric