On Tuesday 21 January 2003 01:53 pm, Antoine Leca wrote:
> Jungshik Shin wrote:
> >>strongly prefered that locale names do not use a country name at all,
> >>unless it is necessary to distinguish between countries.
> >
> > LC_COLLATE is sometimes region/country dependent.
>
> Agreed for "region". "Country" here is only a corner case, that happens
> to distinguish correctly in some cases (for example, Korea), but does
> not adequately in other cases. As a practical example of something
> impractical, there are currently two widely used collating order for
> Spanish ("Castilian" as we say here): the "traditional" order, which is
> by far prefered in Spain (even if it is not any more the "official"
> one according to the RAE), so is likely to be tagged "es_ES". Great so
> far. But the "other" is used just everywhere, including as I just say
> in Spain. So there is no really good POSIX designation to used then :-(.
>
> Similar pains comes with English. OTOH, you may encounter variations
> _inside_ a country, which may ask for a finer grain. (An example is
> Occitan inside France, which is not standardized. There are surely
> others.)
>
> > In addition, differences between zh_* in LC_MESSAGES are not
> > trivial.
>
> AFAIK, Hong Kong is now part of CN. Still, they use "Traditional
> Chinese". So what are we doing then? ;-)
Traditional Chinese with a substantial number of characters unique to HK, for
which there is no inclusive formal standard. Some of them are in Unicode 3.2,
and many more will be in Unicode 4.0.
I have an intense dislike of locales, in part because I have had to use
British English while writing in the U.S., and vice versa, and because I have
to use both Traditional and Simplified Chinese characters in other writings.
I also use prerevolutionary Russian occasionally.
I like the idea of choosing my basic date, time, number, and money formats
independently of each other and of any spelling or grammar rules, which
should be selectable by ruleset, not by country.
My personal time and date preferences are
24-hour clock with leading zeroes referred to UTC
yyyy/mm/dd with leading zeroes (sortable!)
which as far as I know doesn't correspond to anywhere.
There are also languages written in two or three writing systems within the
same country, including Serbo-Croation before Yugoslavia broke up and still
to some extent in Bosnia (Cyrillic and Latin), Hindi/Urdu in both India and
Pakistan (Devanagari and Arabic), Mongolian (Mongolian and Cyrillic), and
various Central Asian languages in several former Soviet Republics (Arabic,
Cyrillic, and Latin).
> Antoine
--
Edward Cherlin
Generalist & activist--Linux, languages, literacy and more
"A knot! Oh, do let me help to undo it!"
--Alice in Wonderland
--
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/linux-utf8/
