On Wednesday, July 2, 2014, Adam Baso <abaso@wikimedia.org> wrote:
I've made further updates to the Etherpad. Lots of good feedback.
Anyone have insight on the Chinese language thing? I was wondering if it has to do with the default soft keyboard being one thing but the display Wikipedia being a different variant.
So Chinese language is .... long story generally, but script-wise, you basically have two main variants:
* Traditional Chinese as used in Hong Kong, Taiwan, and generally outside PRC
* Simplified Chinese as used in People's Republic of China
Simplified Chinese makes various modifications to characters to reduce the number of brush/penstrokes to deal with etc.
Typically, software translators just do separate versions for traditional and simplified, and the appropriate version is selected based on the locale and country.
Wikipedia/MediaWiki handles Chinese a little differently, by allowing use of both on input, and applying a conversion on output to one of Simplified or Traditional (or several slight variants) based on user selection or configured language. There are some markup features for overriding the default conversions between the script variants.
(This system is also used for some Eastern European and Central Asian languages that have, for instance, Cyrillic/Latin forms or Arabic/Latin forms. Most notable is Serbian, with Cyrillic and Latin.)
Currently in the app we don't handle the variants for page content -- you can only select the base Chinese language, and you end up reading the default which may or may not convert the way you want. (That's probably why you see both people asking for Traditional and people asking for Simplified...)
Basically we have to pass a 'variant=zh-hans' or 'variant=zh-hant' parameter with the API content request, based on some override.
Either we can separate out zh-hans and zh-hant intp two language entries, or we can add he concept of variant selection as a second switch, and initialize it by default based on the locale.