Re: [idn] New protocol proposal: IDNRA

["Maynard Kang" <maynard@pobox.org.sg>]

I'd like to understand the basis of Section 3.3 "New Application, Old
Resolver", where it is recommended in this case that New Applications SHOULD
NOT do RACE conversion as "it is likely that applications would not fully
implement and rigorously test the name preparation logic ... and so on"

In my opinion if the application performs nameprep and race conversion ahead
of the resolver, problems associated with MTAs and SMTP gateways may be able
to be circumvented if the hostname that is presented to these guys are RACE
encoded (in ASCII) instead of UTF-8.

When the user enters an e-mail address containing an internationalized
hostname, there are more possible areas for the application to work on (such
as problems related to 821, representation of internationalized hostnames in
822 message headers, etc), as opposed to doing it only at the resolver,
where the only area to work on is the DNS packet itself.

maynard

----- Original Message -----
From: "Patrik Fältström" <paf@cisco.com>
To: "Paul Hoffman / IMC" <phoffman@imc.org>; <idn@ops.ietf.org>
Sent: Monday, 28 August, 2000 12:22 PM
Subject: Re: [idn] New protocol proposal: IDNRA


> At 14.05 -0700 00-08-27, Paul Hoffman / IMC wrote:
> >Where he and I agree is that, if the IDN protocol requires that many
> >DNS severs and caches need to be upgraded, getting to 90% might take
> >ten years or more.
>
> I am not so worried about DNS servers, but for intermediaries for
> applications, like SMTP MTAs and SMTP gateways to Lotus Notes, X.509
> certificate software etc.
>
>    paf