Re: [idn] Using a new class for IDN

["Edmon Chung" <edmon@neteka.com>]

----- Original Message -----
From: "John C Klensin" <klensin@jck.com>
> > Oh sure, it lets registrars *sell* domains quickly, but its
> > success with the user community hinges on every application in
> > the world being upgraded to perform conversion. Furthermore,
> > it is known that this will be a disruptive process which will
> > absolutely cause interoperability failures. There is nothing
> > fast or resolute about this. The only thing going for it is
> > that it is backwards compatible. While that is a necessary
> > attribute of any solution, it is not by itself a "quick fix"
> > for anybody other than domain resellers.
>
> Again, I agree.  But domain resellers, others who believe that
> having "multilingual names" --somehow-- quickly are important,
> and the fear that, if the WG doesn't do something... Right
> Now... various sorts of interoperability catastrophies would
> occur, have been the sources of huge pressures on the WG to try
> to do _something_ that can be deployed quickly.   And "deployed"
> has been defined in that context by ability to put names in
> tables and get them back out.

I think it is more important for registrars as well as the more critical
registrant to have multilingual domain names that are usable by the general
audience quickly.  That is the key I see the market demanding.  Whether we
use IDNA or EDNS or new class the transition is going to be as slow because
the major problem is getting the new client propagated.  The only real to
addresss market demands is a good "Transitional Strategy".  One which has to
include a brute-force approach whereby the name server resolving a request
containing a non-ascii domain request.

While I find it kind of ironic, though not surprising that the so heralded
"quick fix" method is not so "quick" afterall, perhaps our effort should now
(that the IDNA thing is more or less ready) move towards a solution that is
truly more "satisfactory" as you have mentioned.

By the way, just to make it clear, this is not intended to block or stall
the current IDNA effort, but pointing out the plain fact about a real world
demand and a real world transition path that has to take place for
multilingual domain names to become usable.

Edmon