This is really more of an administration bug than a deployment bug, but there's no category for that... Description of problem: Zanata is missing certain language codes that I need for my projects, specifically: ast bs nds ky la mn tl xh Some corresponding language-region combinations would be: ast-ES bs-BA nds-DE ky-KG mn-MN tl-PH xh-ZA "la" probably should not have any region code Version-Release number of selected component (if applicable): 1.3 How reproducible: 100% Steps to Reproduce: 1. Click Administration -> Manage languages -> Add new language 2. Try to find any of the above codes in the drop-down menu Actual results: Languages do not exist and cannot be added Expected results: Ability to add these languages Additional info:
Reassigning to Component-Logic, which is probably questionable too... The latest version of ICU4J (4.8) only seems to add one of the above languages as standard. I think we can add custom locales by coding them, but we should allow the administrator to enter any locale code which conforms to the appropriate format. I'm not sure how we'll handle not knowing the name of the locales in question. Where's a good reference to find the native name of a locale?
(In reply to comment #1) > The latest version of ICU4J (4.8) only seems to add one of the above languages > as standard. However, all of these languages have valid subtags in the IANA registry[1], which in turn is the list that IETF and W3C rely on. In each of these eight cases, the code corresponds to the code for this language in ISO 639-1 and ISO 639-2. The official list of ISO 639-2 codes is freely available, and includes many (most? all?) ISO 639-1 codes.[2] > Where's a good reference to find the native name of a locale? Unfortunately, the IANA registry only includes the English name, and ISO 639-2 only includes the English and French names. Wikipedia is by far the most readily accessible source for the native names. In every instance I've seen, the very first line in any article about a language includes its native name. Code -- English name -- native name ast -- Asturian -- Asturianu bs -- Bosnian -- Bosanski nds -- Low German -- Plattdüütsch ky -- Kyrgyz -- Кыргыз тили la -- Latin -- Latina mn -- Mongolian -- Монгол хэл tl -- Tagalog -- Tagalog xh -- Xhosa -- isiXhosa [1] http://www.iana.org/assignments/language-subtag-registry [2] http://www.loc.gov/standards/iso639-2/php/English_list.php
Sorry, I didn't mean to imply that ICU4J was the definitive list of locales we should follow. Our current implementation gets its list of locales there, so I thought I'd mention it. We obviously do need the ability to add extra locales. Thanks for looking up the names.
Created attachment 505167 [details] Script to add requested locales directly to database
All the above locales have now been added to the fedora test instance. (Disabled by default, admin must enable.) Not closing the bug, because we still need UI to let admin do this.
Reducing severity due to workaround, but let me know if you disagree.
See bug 730186 for a UI to add locales by their codes.