Re: [idn] Comparisons of the proposals

[Harald Alvestrand <Harald@Alvestrand.no>]

At 18:16 26.03.00 +0800, James Seng wrote:
>The question is why not? We could easily say allow any charset registered with
>IANA and tag it according to the number assigned by IANA. This allows both
>localized encodings and also UTF-8. We are also not forcing people to use
>Unicode (for those who really really hate Unicode).

The big problem with using multiple character sets is always what set of
character sets one requires a client to be able to handle intelligently.

I would prefer that I can tell implementors "if you do intelligent handling 
of ONE character set, you don't need to worry about others".
If I have to tell them "you have to do intelligent handling of 8859-2 in 
Poland, 8859-7 in Greece, EUC-KR in Korea, Shift-JIS in Japan, UTF-8 in 
Trondheim, and all of the above for the Internet", I don't feel happy.

Labelled noninteroperability is an improvement over just-send-8, but not a 
big one.

                      Harald

--
Harald Tveit Alvestrand, EDB Maxware, Norway
Harald.Alvestrand@edb.maxware.no