Re: [idn] Comparisons of the proposals

[Paul Hoffman / IMC <phoffman@imc.org>]

At 03:12 PM 3/23/00 +0900, Martin J. Duerst wrote:
>Sorry, but this doesn't work.

Yes, it does. It is just not as simple as straight UTF-8 or UCS-16 encoding 
in zone files.

>  Nobody in Europe or Asia wants to use
>an ASCII text editor and raw CIDNUC or UTF-5.

And we don't want to use them in the US, either. Please stop trying to make 
this into a "US vs. the rest of the world" argument; we don't need that 
kind of ugliness here. US firms with foreign subsidiaries that want correct 
internationalized names will have exactly the same problems and non-US firms.

>  If you can't imagine
>why, try asking yourself whether you would like to edit ASCII files
>in octal dump mode.

Exactly right.

>So the above should be changed to:
>
>UTF-8 for 8&down can be edited in an off-the-shelf editor
>(take for example MS Word or emacs/mule).

Um, how many people who edit zone files do so with Microsoft Word? Or 
emacs/mule in a graphics environment that handles more than just basic 
European characters? Probably next to none.

It is likely that the vast majority of zone files are edited over telnet 
sessions on VT-100 emulators, which will not translate UTF-8 correctly. 
Whatever the WG goes with, whether it is UTF-8 or an ASCII encoding, is 
going to need translation tools for editing the zone files. The translation 
tools for UTF-8 will be easier and more widespread because it is a format 
that is already in use.

Regardless of the tool, the zone administrator is still going to want to 
view the results in a non-ASCII environment to verify that what she or he 
just typed in looks right. Those tools are not widely available today (and 
certainly not available for many characters that we will appear in IDN 
names), but the IDN work will probably cause them to come into existence 
sooner than they might have.

--Paul Hoffman, Director
--Internet Mail Consortium