W3C home > Mailing lists > Public > ietf-discuss@w3.org > June 2001

Re: Are international characters allowed in email addresses?

From: Eric Brunner-Williams <wampum@maine.rr.com>
Date: Thu, 28 Jun 2001 13:56:08 -0400
Message-Id: <5.1.0.14.0.20010628133155.037ca6d0@pop-server>
To: Keith Moore <moore@cs.utk.edu>, Eric Brunner-Williams in Portland Maine <brunner@nic-naa.net>
Cc: Keith Moore <moore@cs.utk.edu>, discuss@apps.ietf.org

>etc.  the notion that an application that actually uses domain names
>as protocol elements can be fixed by simply making it 8bit clean has
>been soundly and repeatedly refuted.

Was that claim ever offered by anyone? As rhetoric it may suffice to
preclude work clearly identified in rfc2825 and elsewhere, but rhetoric
isn't a substitute for a restatement of the problem as infrastructural.

As Fred Baker remarked when opening his for-and-against discussion
(unique dns root) with Karl Auerbach at the ISOC-MINC Workshop in
Stockholm earlier this month, unfortunate rhetoric drives discourse to
less than productive ends.

I do make the claim that there are several places where sub-8-bit
processing takes place, e.g., sendmail header processing, and each
can be fixed, e.g., 8-bit clean plus a transition mechanism (EHLO) for
8-to-7 down-conversion, as well as dns label processing. These can
be viewed as "infrastructure" or as "applications".

Can you suggest how these two statements can be reconcilled?
1. the IDN requirements memo is irrelevant.
and
2. the issues ... will be revisited in due time ... but that time is after the
IDN group has finished its work.

Requirements Considered Optional is not something I look forward to.

Eric

At 6/28/01 01:05 PM, Keith Moore wrote:
> > > and you've been in the IDN discussion so you should also be familiar
> > > with those issues.
> >
> > The IDN discussion has produced a requirements memo that states, inter
> > alia, that
>
>the IDN requirements memo is irrelevant.  the discussions on the IDN
>list have repeatedly focused on the problems of names in multiple
>character sets, translation of those names to a common format,
>canonicalization of those names, encoding of those names in ASCII to
>make them safe for existing applications, comparison of those names,
>etc.  the notion that an application that actually uses domain names
>as protocol elements can be fixed by simply making it 8bit clean has
>been soundly and repeatedly refuted.
>
>the issues for email addresses are not identical, but are quite similar.
>I'm sure they will be revisited in due time by applications area groups,
>but that time is after the IDN group has finished its work.
>we'll have to have those discussions then anyway, just to bring everybody
>up to speed.  so there's no sense in our having them now.  folks who
>are interested are welcome to read the IDN list archives or join that
>discussion.
>
>Keith
Received on Thursday, 28 June 2001 14:02:48 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:08:12 UTC