Kurdish Language Name Problem in v.2.25

I think the underlying problem is with the base method which relies of locales present in Java (http://www.oracle.com/technetwork/java/javase/javase7locales-334809.html)

The list does not include all locales, including Kurdish Sorani. I have created an issue in our issue tracker here (https://jira.dhis2.org/browse/DHIS2-342) if you would like to track the progress.

Chris, I think we need to explore a bit and see if the proposed technical solution solves the problem but if you have connections with the ICU project (http://site.icu-project.org/) you could check with them and see if they support this locale. Thanks,

Best regards,

Jason

···

On Tue, Dec 6, 2016 at 11:37 PM, Antonia Bezenchek - INFORMApro antonia.bezenchek@informa.pro wrote:

Hi Chris,
thanks for your help to make the name of the Kurdish Sorani language stable in DHIS2.

Let me know if I can help in some way.

Regards,

Antonia

2016-12-06 23:07 GMT+01:00 Chris Leonard cjlhomeaddress@gmail.com:

Jason,

Do you know the exact source of this language list?

I know the Debian project maintains a localized set of the various isocode standards for language name, country, currency, etc. If you could help me track down the code source, I’m more than willing to look into bugging them about representing Kurdish Sorani properly in their code.

Chris Leonard

On Mon, Dec 5, 2016 at 5:59 AM, Jason Pickering jason.p.pickering@gmail.com wrote:

Hi Antonia,

I will discuss with the development team, but as you point out, its more of a problem of the underlying libraries themselves, which do not have a corresponding translation of the official ISO code for Kudish Sorani. We would need to add an enhancement I think to use the equivalent icu4j classes to get the language names for these codes which are not part of the core Java Locale classes.

Regards,

Jason


Mailing list: https://launchpad.net/~dhis2-devs

Post to : dhis2-devs@lists.launchpad.net

Unsubscribe : https://launchpad.net/~dhis2-devs

More help : https://help.launchpad.net/ListHelp

On Mon, Dec 5, 2016 at 11:26 AM, Antonia Bezenchek - INFORMApro antonia.bezenchek@informa.pro wrote:

Dear developers and translation team,

in the upgrade process from v.2.22 to v.2.25 we encountered the fallowng 2 problems with the translation.

  • the the language Kurdish, created especially for our implementation, is not more visible as Kurdish in the User’s Interface language, but with its ISO code “ckb”. All users with assigned Kurdish language have not assigned any language and and the system default language appears. If I assign manually the “ckb” language to the user, the Kurdish translation of the interface appears.

But for the final users “ckb” is not comprehensible name.

Could you change the name to Kurdish or Kurdish Sorani the language code (ckb - ISO 639-3)

for the v. 2.25 and all new versions.

I know that if not fixed in a stable mode in the source, the problem will appears in any new version, because it is not part of the underlying libraries (GTK and the Java libraries) which provides these names, as specified by Jason Pickering in a mail (May 1st 2016) and we could not make it change for the moment.

Thanks in advance for your help.

Antonia

www.informa.pro - *Seguici su *LinkedIn

INFORMAPRO S.r.l. - via Luigi Rava 43, Rome 00149 , Italy
tel. +39 06 5758926, fax +39 06 62207168, P.IVA IT08095781004

Skype: informa_antonia.bezenchek
Eng. Antonia Bezenchek - CIO - ICT Engineer

Jason P. Pickering
email: jason.p.pickering@gmail.com
tel:+46764147049

Jason P. Pickering
email: jason.p.pickering@gmail.com
tel:+46764147049

Interestingly, the nearly current version CLDRv30 of the CLDR locale data used by ICU https://cldr.unicode.org , (technically most current version is two months more current at CLDR v30.0.3) contains the attached supplementalMetadata.xml file which references ckb_IQ in the last line (some type of “sort order” statement).

The other file supplementalData.xml contains the following two lines:

I am not sure exactly if the following lines in languageInfo.xml suggest (I’m not 100% sure how this file is parsed in use)

<!-- variants of Kurdish --is

It is possible that they default to the ku macrolanguage code when they encounter the ckb (or ckb_IQ) locale setting.

I have been somewhat concerned about the CLDR tendency to “roll up” or default to the ISO-639-1 macrolanguage or ISO-639-2 language collections codes from the more specific ISO-639-3 code set where ckb is defined as Kurdish Sorani (aka Central Kurdish) and is written in an Arabic script (one of the defining characteristics that distinguishes ckb from other Kurdish variants under the ISO-639-1 macrolanguage code ku

https://en.wikipedia.org/wiki/Central_Kurdish
https://en.wikipedia.org/wiki/Kurdish_languages

I’m encountering issues with this macrolanguage default issue in Peruvian indigenous languages (e.g. Aymara, Quechua) that are currently supported on Peru’s OLPX XO laptops because the Linux glibc locale project is essentially deferring judgement to CLDR (glibc is otherwise a bunch of C hackers, not really language experts).

Let me dig into it some more, if it is in fact another example of a “roll up to macrolanguage code” issue, I would appreciate your assistance in documenting this additional use case where the current policy fails and should be replaced with a policy that respects ISO-639-3 codes and eschews ISO-639-1 macrolanguage codes as being not only lossy encoding and sloppy implementation of a standard, but essentially a violation of various cultural self-determination principals enshrined in the UNDRIP United Nations Declaration on the Rights of Indigenous Peoples.

https://en.wikipedia.org/wiki/Declaration_on_the_Rights_of_Indigenous_Peoples

https://en.wikipedia.org/wiki/ISO_639_macrolanguage

kur

supplementalMetadata.xml (109 KB)

supplementalData.xml (381 KB)

languageInfo.xml (14.5 KB)

···

On Wed, Dec 7, 2016 at 3:16 AM, Jason Pickering jason.p.pickering@gmail.com wrote:

I think the underlying problem is with the base method which relies of locales present in Java (http://www.oracle.com/technetwork/java/javase/javase7locales-334809.html)

The list does not include all locales, including Kurdish Sorani. I have created an issue in our issue tracker here (https://jira.dhis2.org/browse/DHIS2-342) if you would like to track the progress.

Chris, I think we need to explore a bit and see if the proposed technical solution solves the problem but if you have connections with the ICU project (http://site.icu-project.org/) you could check with them and see if they support this locale. Thanks,

Best regards,

Jason

On Tue, Dec 6, 2016 at 11:37 PM, Antonia Bezenchek - INFORMApro antonia.bezenchek@informa.pro wrote:

Hi Chris,
thanks for your help to make the name of the Kurdish Sorani language stable in DHIS2.


Jason P. Pickering
email: jason.p.pickering@gmail.com
tel:+46764147049

Let me know if I can help in some way.

Regards,

Antonia

2016-12-06 23:07 GMT+01:00 Chris Leonard cjlhomeaddress@gmail.com:

Jason,

Do you know the exact source of this language list?

I know the Debian project maintains a localized set of the various isocode standards for language name, country, currency, etc. If you could help me track down the code source, I’m more than willing to look into bugging them about representing Kurdish Sorani properly in their code.

Chris Leonard

On Mon, Dec 5, 2016 at 5:59 AM, Jason Pickering jason.p.pickering@gmail.com wrote:

Hi Antonia,

I will discuss with the development team, but as you point out, its more of a problem of the underlying libraries themselves, which do not have a corresponding translation of the official ISO code for Kudish Sorani. We would need to add an enhancement I think to use the equivalent icu4j classes to get the language names for these codes which are not part of the core Java Locale classes.

Regards,

Jason


Mailing list: https://launchpad.net/~dhis2-devs

Post to : dhis2-devs@lists.launchpad.net

Unsubscribe : https://launchpad.net/~dhis2-devs

More help : https://help.launchpad.net/ListHelp

On Mon, Dec 5, 2016 at 11:26 AM, Antonia Bezenchek - INFORMApro antonia.bezenchek@informa.pro wrote:

Dear developers and translation team,

in the upgrade process from v.2.22 to v.2.25 we encountered the fallowng 2 problems with the translation.

  • the the language Kurdish, created especially for our implementation, is not more visible as Kurdish in the User’s Interface language, but with its ISO code “ckb”. All users with assigned Kurdish language have not assigned any language and and the system default language appears. If I assign manually the “ckb” language to the user, the Kurdish translation of the interface appears.

But for the final users “ckb” is not comprehensible name.

Could you change the name to Kurdish or Kurdish Sorani the language code (ckb - ISO 639-3)

for the v. 2.25 and all new versions.

I know that if not fixed in a stable mode in the source, the problem will appears in any new version, because it is not part of the underlying libraries (GTK and the Java libraries) which provides these names, as specified by Jason Pickering in a mail (May 1st 2016) and we could not make it change for the moment.

Thanks in advance for your help.

Antonia

www.informa.pro - *Seguici su *LinkedIn

INFORMAPRO S.r.l. - via Luigi Rava 43, Rome 00149 , Italy
tel. +39 06 5758926, fax +39 06 62207168, P.IVA IT08095781004

Skype: informa_antonia.bezenchek
Eng. Antonia Bezenchek - CIO - ICT Engineer

Jason P. Pickering
email: jason.p.pickering@gmail.com
tel:+46764147049

The attached are also present in CLDR defining the ckb language and the ckb_IQ and ckb_IR locales.

cjl

ckb.xml (59.2 KB)

ckb_IQ.xml (536 Bytes)

ckb_IR.xml (1.19 KB)

···

On Wed, Dec 7, 2016 at 3:16 AM, Jason Pickering jason.p.pickering@gmail.com wrote:

I think the underlying problem is with the base method which relies of locales present in Java (http://www.oracle.com/technetwork/java/javase/javase7locales-334809.html)

The list does not include all locales, including Kurdish Sorani. I have created an issue in our issue tracker here (https://jira.dhis2.org/browse/DHIS2-342) if you would like to track the progress.

Chris, I think we need to explore a bit and see if the proposed technical solution solves the problem but if you have connections with the ICU project (http://site.icu-project.org/) you could check with them and see if they support this locale. Thanks,

Best regards,

Jason

On Tue, Dec 6, 2016 at 11:37 PM, Antonia Bezenchek - INFORMApro antonia.bezenchek@informa.pro wrote:

Hi Chris,
thanks for your help to make the name of the Kurdish Sorani language stable in DHIS2.


Jason P. Pickering
email: jason.p.pickering@gmail.com
tel:+46764147049

Let me know if I can help in some way.

Regards,

Antonia

2016-12-06 23:07 GMT+01:00 Chris Leonard cjlhomeaddress@gmail.com:

Jason,

Do you know the exact source of this language list?

I know the Debian project maintains a localized set of the various isocode standards for language name, country, currency, etc. If you could help me track down the code source, I’m more than willing to look into bugging them about representing Kurdish Sorani properly in their code.

Chris Leonard

On Mon, Dec 5, 2016 at 5:59 AM, Jason Pickering jason.p.pickering@gmail.com wrote:

Hi Antonia,

I will discuss with the development team, but as you point out, its more of a problem of the underlying libraries themselves, which do not have a corresponding translation of the official ISO code for Kudish Sorani. We would need to add an enhancement I think to use the equivalent icu4j classes to get the language names for these codes which are not part of the core Java Locale classes.

Regards,

Jason


Mailing list: https://launchpad.net/~dhis2-devs

Post to : dhis2-devs@lists.launchpad.net

Unsubscribe : https://launchpad.net/~dhis2-devs

More help : https://help.launchpad.net/ListHelp

On Mon, Dec 5, 2016 at 11:26 AM, Antonia Bezenchek - INFORMApro antonia.bezenchek@informa.pro wrote:

Dear developers and translation team,

in the upgrade process from v.2.22 to v.2.25 we encountered the fallowng 2 problems with the translation.

  • the the language Kurdish, created especially for our implementation, is not more visible as Kurdish in the User’s Interface language, but with its ISO code “ckb”. All users with assigned Kurdish language have not assigned any language and and the system default language appears. If I assign manually the “ckb” language to the user, the Kurdish translation of the interface appears.

But for the final users “ckb” is not comprehensible name.

Could you change the name to Kurdish or Kurdish Sorani the language code (ckb - ISO 639-3)

for the v. 2.25 and all new versions.

I know that if not fixed in a stable mode in the source, the problem will appears in any new version, because it is not part of the underlying libraries (GTK and the Java libraries) which provides these names, as specified by Jason Pickering in a mail (May 1st 2016) and we could not make it change for the moment.

Thanks in advance for your help.

Antonia

www.informa.pro - *Seguici su *LinkedIn

INFORMAPRO S.r.l. - via Luigi Rava 43, Rome 00149 , Italy
tel. +39 06 5758926, fax +39 06 62207168, P.IVA IT08095781004

Skype: informa_antonia.bezenchek
Eng. Antonia Bezenchek - CIO - ICT Engineer

Jason P. Pickering
email: jason.p.pickering@gmail.com
tel:+46764147049