On Sun, Oct 08, 2000 at 06:08:53PM +0200, Byrial Jensen wrote:
> On Sun, Oct 08, 2000 at 17:46:10 +0200, Byrial Jensen wrote:
>
> > > A Dane user will expect � -> aa transliteration while in other parts
> > > of the world the � will simply be dropped: � -> a. Similar for � in
> > > Swedish which probably should be transliterated � -> y. In German it
> > > must be � -> ue.
>
> Or to take your example: How should a user who wishes to have both
> Swedish and German translations (LANGUAGE=se:de), and who have not
> � in his charset, get right translitterations for both Swedish and
> German messages?
There is always the rule that the data should be presented
according to the expectations of the user in question.
We do assume here that the user wants to have the � presented
in one way always, and not according to the language he thinks
that the � is a part of. So maybe what we have is too simple a model.
Or we may use it to say that the user expects transliterations to occur in
the language that is the origin, and then change transliterarations
according to language tags, eg in a text document.
ISO TR 14652 will be changed to taht transliterations will occur in
its own category, LC_TRANSLIT.
Kind regards
Keld
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/lists/
