Chinese language codes

l10n
#1

Hi folks,

please excuse me for chiming in into what is totally not my business, but:
it looks like the two Chinese teams are favoring two different locale naming styles (zh-Hans for simplified vs zh-TW for traditional). Do we have any plans to favor one of those schemes in future? I think we should. To me, it doesn't seem to make much sense to keep using two different naming styles for those two locales.

As far as I know, the LibO package itself currently uses zh-cn and zh-tw as locale codes. That's the traditional way, which is also employed by Linux. The modern way is to use script codes as main modifiers instead of country codes (because both simplified and traditional variants are spoken in more than one country, but there often exists only one localization of each type). These codes are already used by OS X, Haiku OS, and latest versions of .NET on Windows.

It seems that the lead of Simplified Chinese is already in favor of the modern locale code. However, I don't think it would make sense to only change one of those two codes, so IMO either both teams should start using new codes, or both should keep the old names. Current situation (as e.g. the Wiki reflects it) doesn't look reasonable, at least to me.

Can we come up with a consistent strategy here?
Rimas

#2

It seems that the lead of Simplified Chinese is already in favor of the
modern locale code. However, I don't think it would make sense to only

No, one of our important team member (Dean Lee) proposed to use
zh-Hans, but I myself tend to use zh-CN for locale.

His reason for preferring zh-Hans was like Rimas had said on this list
before, that the new style is a fashion and some companies are
starting to using it in some of their products.

My reason for preferring zh-CN is basiclly we have been using it for
years and it will be very familiar to Linux users - which platform I
think we have most of our users.

Another important issue is zh-Hans/Hant does not equal to zh-CN/zh-TW:
primarily, zh-Hans includes zh-CN and zh-SG (still does not exist in
our project, though), zh-Hant includes zh-TW and zh-HK.

Rimas had raised a solution that we can use modifiers like zh-CN@Hans
(IIRC). But I think it just adds more confusion for translation teams
if we define it in this way.

change one of those two codes, so IMO either both teams should start using
new codes, or both should keep the old names. Current situation (as e.g. the
Wiki reflects it) doesn't look reasonable, at least to me.

Can we come up with a consistent strategy here?
Rimas

Personally I think we can use zh-Hans and zh-Hant for public (users),
but use zh-CN/TW/HK internally in our project - we can ship two
language packs named zh-Hans and zh-Hant, which includes the
corresponding languages, if possible and/or needed.

#3

Hi,

I just want to point out that a variation of Catalan known as «Catalan
(Valencian)» (both variants spoken in Spain) has a modifier assigned
by the IANA on 2005 (-valencia). Right now and because of the OOo
development, we're using the ca_XV code for this variant (ca_ES for
Catalan).

Something similar happens with Serbian and other variants. I don't
really known what is the state of this question on OOo or LibO, but it
should be interesting to implement the BCP47 standard
(http://www.rfc-editor.org/bcp/bcp47.txt).

Regards,

Pau Iranzo

#4

Hi,

Hi,

I just want to point out that a variation of Catalan known as «Catalan
(Valencian)» (both variants spoken in Spain) has a modifier assigned
by the IANA on 2005 (-valencia). Right now and because of the OOo
development, we're using the ca_XV code for this variant (ca_ES for
Catalan).

Thanks for you information.

There has no modifier ever assigned to Chinese languages, because they
are not commonly said "variants". Traditional and Simplified Chinese
are different in characters, vocabularies, and common usages. They are
both Chinese in linguistic, but really have many differences.

Something similar happens with Serbian and other variants. I don't
really known what is the state of this question on OOo or LibO, but it
should be interesting to implement the BCP47 standard
(http://www.rfc-editor.org/bcp/bcp47.txt).

Regards,

Pau Iranzo

I know LO would be interested in BCP47, so I think it's OK to use it
in public usages (to users). But it shouldn't make translators get
confused (zh-Hans@CN or zh-CN@Hans or whatever in this way).