> -----Original Message-----
> From: Tomohiro KUBOTA [mailto:[EMAIL PROTECTED]]
...
> 3. I want a plain text of mixture of Japanese and Chinese. And more,
> I want foreign people who learn Japanese to learn correct Japanese
> characters. In short, this way works well as localization, but
> doesn't work as internationalization.
Well, I would not recommend plain text for that. Just like you cannot
switch between Roman and Fraktur in plain text.
> I prefer 'language tag' solution, though it is complex like ISO-2022.
> It seems the only way that plain text itself can have information on
> the text. (Plain text should have information on _character_, not
> _glyph_.
> For CJK people, difference between unified variants is
> regarded as difference of _characters_, not _glyph_.)
Not according to the East Asian experts on Han ideographs.
> Do you have any idea other than 'language tag'?
I would suggest using HTML text with language tags as per HTML 4
(font switching based on (HTML) language tag for Japanese/Chinese is
already available in popular (non-Linux...) browsers), or XHTML 1.0.
The "LANGUAGE TAG" 'characters' recently allocated (in 10646-2)
should never be used with HTML (or any XML based document),
and can (indeed should) be ignored or even filtered away
anywhere they occur.
/kent k
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/lists/
