From: Florian Weimer <[EMAIL PROTECTED]>
> I don't think it's worth the trouble. Collation will always be
> confusing if you're mixing two or more languages.
I realize that, but we should minimize confusion. If there is no national
standard (formal or informal) on sorting of certain characters, then it
should be consistent with they way they're used where they're used, while
being efficent at the same time. I see no reason for ja and ko to sort Latin
characters differently, just like I see no reason for de, eo and en to sort
Japanese and Chinese characters differently. If I go to Germany and start
using de_DE, I see no reason for non-Latin characters to sort differently,
and it would rather annoy me to have to mess with it.
> You certainly cannot sort a dictionary based on a locale collation
> order, or an index for a book.
Why not? Maybe the Oxford English Dictionary (THE English dictionary) can't,
but anything without such a critical audience can. I looked through the
GURPS Japan index (GURPS books are renown in the roleplaying industry for
their quality indexes), and found nothing that couldn't be done use
LC_COLLATE=en_US rules -- nothing that would even push the edges of the
rules.
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/linux-utf8/
