new monolingual language code requests for Wikidata (tracking)
Open, NormalPublic

Description

This aims to simplify viewing tickets for currently requested languages for monolingual text properties at Wikidata.

Candidates might be found with Q28368760.

Requests as comments here are unlikely to be handled.

NOTE: This is relates to Wikidata monolingual text language only. To request a new language for labels, descriptions, and aliases support, please first request enabling your language for localisation on translatewiki.net (see https://translatewiki.net/wiki/Translatewiki.net_languages), then translate MediaWiki (most important messages)(MUST BE 100% DONE), Wikibase - Client (at least one message), and Wikibase - Repo (at least one message). To request a Wikimedia project in your language, please use https://meta.wikimedia.org/wiki/Requests_for_new_languages.

Related Objects

StatusAssignedTask
OpenNone
OpenNone
Resolved Jonas
Resolvedjhsoby
Resolvedthiemowmde
Resolvedthiemowmde
Resolvedjhsoby
StalledNone
ResolvedAddshore
ResolvedNone
Resolvedthiemowmde
ResolvedLydia_Pintscher
Resolvedjhsoby
OpenNone
Resolvedaude
ResolvedMbch331
OpenLea_Lacroix_WMDE
OpenNone
OpenNone
Resolvedaude
Resolvedaude
Resolvedaude
ResolvedMbch331
Resolvedaude
Resolvedjhsoby
Resolvedaude
Resolvedaude
ResolvedLydia_Pintscher
ResolvedLydia_Pintscher
Resolvedjhsoby
OpenNone
Resolvedjhsoby
Resolvedjhsoby
ResolvedMbch331
OpenMbch331
ResolvedMbch331
Resolvedjhsoby
ResolvedMbch331
Resolvedaude
Resolvedjhsoby
ResolvedMbch331
DeclinedNone
ResolvedMbch331
ResolvedLadsgroup
ResolvedLadsgroup
DuplicateNone
OpenNone
OpenNone
ResolvedLadsgroup
ResolvedMbch331
ResolvedMbch331
ResolvedZoranzoki21
ResolvedZoranzoki21
ResolvedZoranzoki21
ResolvedZoranzoki21
ResolvedZoranzoki21
DeclinedNone
ResolvedZoranzoki21
ResolvedZoranzoki21
ResolvedZoranzoki21
ResolvedZoranzoki21
OpenNone
OpenNone
ResolvedMbch331
ResolvedMbch331
OpenMbch331
ResolvedMbch331
Resolvedjhsoby-WMNO
Resolvedjhsoby-WMNO
Resolvedjhsoby-WMNO
Resolvedjhsoby-WMNO
OpenNone
OpenNone
OpenNone
OpenNone
OpenXaris333
There are a very large number of changes, so older changes are hidden. Show Older Changes
Amire80 added a subscriber: Amire80.Jun 7 2017, 7:34 AM

I have a general question here: Are these languages codes provided in WikibaseRepo.php in addition to anything?

I mean, where does the basic list come from? Core Names.php? ULS's langdb? A wikidata.org query? A defined subset of CLDR? Something else?

I couldn't find an answer to this at any of the pages linked from the task description.

I feel quite uncomfortable about a yet-another growing list of language codes in the MediaWiki+extensions codebase. Special codes like mul and und can be in a special place, but there's nothing very special about brx or kjh.

Nikki added a subscriber: Nikki.Jun 9 2017, 3:59 PM

I have a general question here: Are these languages codes provided in WikibaseRepo.php in addition to anything?

I mean, where does the basic list come from? Core Names.php? ULS's langdb? A wikidata.org query? A defined subset of CLDR? Something else?

I may have misunderstood something and things might have changed since I last looked into it, but from what I previously managed to piece together, the list available for monolingual text consists of:

excluding the languages in the second array in getMonolingualTextLanguages.

OK, I still have my question (T144272#3322282): Why are languages still added in PHP code, such as https://gerrit.wikimedia.org/r/#/c/370308/2/repo/includes/WikibaseRepo.php ?

This is data. This is not supposed to be code. It's a really bad practice. We already have too many repositories of language codes.

I strongly suggest scrapping getMonolingualTextLanguages() and adding these codes to the same langdb that is used by the UniversalLanguageSelector extension. Recently it was separated to its own repository: https://github.com/santhoshtr/language-data (it's about to be moved under the wikimedia account). Languages are added to a YAML file, which is also autoconverted to JSON, and it's usable on backend and on frontend.

Please, let's have less places for language codes, not more.

Hoi,
I asked in the past why codes were added. In principle, we could include
them all but there is no clarity why codes are proposed. As a consequence I
lost interest.

So I agree with Amir that some sanity would be in order.
Thanks,

GerardM

@Lydia_Pintscher What's going to happen to the older request where there's no langcom approval?

Hoi,
The consideration has been only is it a valid code. The question asked is a
different one. What is the point to all these codes that probably do not
bring anything to Wikidata.
Thanks,

GerardM

I think that we should stop with these stamp collections there is no argument why we want to do this, we have seen that the meaning given often differs substantially from what the standard is. It makes more sense to study the effect of the introduction of Wiktionary content.

Are you implying that those monolingual language code I'm submitting does not represent anything useful? nan/cdo/hak-Hant/hans are language-script combinations being used to write wikipedia articles, and vi-hani, ko-kore, ja-Kyujitai are used to name people and things in respective countries. How do you write the name of "Ho Chi Minh City" in Vietnamese Han nom? The only place providing this info in wikidata for now is in the Japanese alias for the entry name. How about "Kim Jong-Il" in ko-Kore? Look at the Slovak alias. Is it better than having labels for each of these script variants?

Hoi,
What I find is totally unreadable.
Thanks,

GerardM

Please motivate.
Thanks,

GerardM

Op vr 1 dec. 2017 om 18:48 schreef VIGNERON <
no-reply@phabricator.wikimedia.org>

VIGNERON added a subtask: T181823: Add monolingual language code fro and
frm in Wikdiata.
*TASK DETAIL*
https://phabricator.wikimedia.org/T144272

*EMAIL PREFERENCES*
https://phabricator.wikimedia.org/settings/panel/emailpreferences/

*To: *VIGNERON

*Cc: *C933103, Nikerabbit, Mbch331, Lydia_Pintscher, Zoranzoki21, Nikki,
Amire80, jhsoby, Liuxinyu970226, GerardM, thiemowmde, Aklapper, Esc3300,
Lahi, Gq86, GoranSMilovanovic, QZanden, Wikidata-bugs, aude

Hoi,
The consideration has been only is it a valid code. The question asked is a
different one. What is the point to all these codes that probably do not
bring anything to Wikidata.
Thanks,

GerardM

I believe we should let the community decide that.

Hoi,
In the past we have been where we are now with Wikidata and new language
codes with new Wikipedias. The result was mayhem, and everything was just
as the "community" decided. The result is the many Wikipedias that have
no community. As a result the Language Committee took the place of "the
community"and from that moment new Wikipedias do conform to standards.

I have seen too many requests that are problematic that will in my opinion
not bring anything to Wikidata. I find a lack of purpose in what is
proposed. I do understand a purpose for texts as they were originally but
where is the purpose in adding labels that have arbitrarily transliterated
in other scripts.

There is no point. It is not the "community" supporting this it hidden in
phabricator.
Thanks,

GerardM

This is not about new languages for labels. This ticket and its subtickets are (almost all) explicitly about new languages for monolingual text values in statements. And Léa and I get asked a lot why this is being held up and we don't have a good answer. I don't want to do this but if this process isn't working I have to find another way in the interest of the project.

As I suggested above, I recommend reusing the language-data database for this. It is also used in the UniversalLanguageSelector, which is a dependency for Wikibase, so it's safe to assume that it's installed on Wikidata.

The criteria for adding a language to this database are pretty much the same as the general Wikimedia criteria. If anybody finds a discrepancy in the criteria, let me know.

Nikki added a comment.Jul 5 2018, 11:36 AM

The criteria for adding a language to this database are pretty much the same as the general Wikimedia criteria. If anybody finds a discrepancy in the criteria, let me know.

What are the general Wikimedia criteria?

The criteria for adding a language to this database are pretty much the same as the general Wikimedia criteria. If anybody finds a discrepancy in the criteria, let me know.

What are the general Wikimedia criteria?

Now that I think of it, I'm not sure whether they are written anywhere :)

But a valid ISO 639 3 code is definitely a criterion. Is there anything else that Wikidata requires?

Nikki added a comment.Jul 5 2018, 12:20 PM

Now that I think of it, I'm not sure whether they are written anywhere :)

But a valid ISO 639 3 code is definitely a criterion. Is there anything else that Wikidata requires?

Wikidata's own criteria (from https://www.wikidata.org/wiki/Help:Monolingual_text_languages#Requirements_for_a_new_language_code) suggest that anything with a valid IETF language tag should be allowed. The requests so far have included ancient languages, artificial languages, country variants, script variants, other variants with a registered subtag and codes involving macrolanguages.

Nikki added a comment.Aug 23 2018, 7:27 PM

@Amire80 Would all of what Wikidata wants (see previous comment) be allowed in the language-data database?

@jhsoby I see you're also interested in using the language-data database instead, what do you think about the different types of codes in my previous comment?