[Ubuntu] Re: new language team

[Danilo Segan]

Hi Bardaqani,

Данас у 16:25, bardaqani bardaqani написа:

> English (Australia) en_au
> English (Canada)  en_ca
> English (United Kingdom) en_uk

AU, CA, UK are ISO 3166 codes for these countries.

>
> en chinese
> Chinese (China)
> Chinese (Hong Kong)
> Chinese (Taiwan)

CH, HK, TW are ISO 3166 codes for these countries/territories.

This is no accident, it's a rule imposed by GNU libc locale naming
policy.  And since we are providing translations for, among other
things, Ubuntu, which is GNU libc-based system, we have to follow such
naming scheme.  Another reason to follow it is that you won't be able
to push your translation to different projects with non-standard names
(such as GNOME, KDE, GNU).

As for "political" issues, all Chinese Traditional translations used
to use zh_TW, even if it wasn't correct.  It was just a work-around,
and it caused no problems.

> Language (Geographical Location/country)
> In case of Kurdish; Kurdish(Turkey/Iran/Iraq/Syria). However such a
> geographical indication is emotionally and politically loaded so we can
> avoid it by using the locally used terms:
> Kurdish (North) ku_no
> Kurdish (West) ku_we
> Kurdish (South) ku_so
> Kurdish (East) ku_ea

As explained above, this would not work.  For example, "ku_NO" would
mean "Kurdish in Norway".

> so:
> ku at no
> ku at we
> ku at so
> ku at ea

Again, if you try to push this into GNU libc proper, you'll have to
make those modifiers full English words, so you'd end up with
'ku at north', 'ku at west'...

However, this is problematic for Launchpad, since we don't yet support
variants, and you'd be unable to carry on your translation if we used
these identifiers.  So, until ISO 639-3 gets official, I'd rather
recommend using 'ku_CC' where CC is ISO 3166 country code for wherever
is this variant spoken the most.

Just like Chinese Traditional is commonly using 'politically loaded'
'zh_TW'.

> I think the solution is found somewhere in the above mentioned urgument.
>
> From my point of view the three-letter iso639-3 code represents nothing in
> case of Kurdish. Please be aware of the fact that there is no language
> or sub-language called Central Kurdish.

Just for the reason that everybody's personal opinion is going to
differ, I strongly believe we should trust such decision making to
international standards organisation such as ISO, Unicode, IETF, etc.
(IETF is managing language-tags scheme; does anyone know what they
recommend for this variant in their RFC?)

Cheers,
Danilo