Re: [idn] Planning ahead for good IDNs

[Marc Tamsky <tamsky@www.tv>]

>>>>> On 13 Jan 2001 00:38:53 -0000, "D. J. Bernstein" <djb@cr.yp.to> said:

>> There still seem to be at least three credible camps:  ACE only, ACE
>> transition, and UTF-8 only.

>  I don't see any serious support for the ACE-now-and-forever position.

Agreed.  As a final product, it would be the ugliest hack next to IMAP
to come out of the IETF lately.


>  The rest of us, the UTF-8 people and the ACE-now-UTF-8-later people, all
>  agree that some programs are going to have to be fixed. Sendmail's
>  address handling should be made 8-bit-clean, for example, and cc:Mail
>  should allow 255-byte local parts and 255-byte domain names.

>  The IDN WG can and should issue an immediate warning for implementors.

Agreed.

>From RFC-2277:

   But it is the responsibility of the WGs to ensure that whenever
   responsibility for internationalization is left to "another layer",
   those responsible for that layer are in fact aware that they HAVE
   that responsibility.

I think this applies to the local parts issue that was recently
brushed off as "another layer problem" -- those responsible for local
parts need notice as well.


In no way should readers of either warning be led to believe that by
implementing the necessary changes to support ACE, that they are done
upgrading their programs to support internationalized DNS.

Without a clear set of transition elements, guidelines for
implementation, and test suites, ACE-now-UTF-8-later will create a
credible demand for UTF-8 in other systems.

All those who face changes to their systems, also face the questions:
"Do I update my programs twice to work with ACE-now-UTF-8-later, or
just do the work once for UTF-8?"

-- 
Marc Tamsky <tamsky@www.tv>