Language committee/May 2011 meeting

Active members of the language committee are going to meet for the first time during the Berlin Hackathon on 13-15 May 2011.

Attendees edit

  • Amir E. Aharoni
  • Antony D. Green
  • Gerard Meijssen
  • Michael Everson
  • Miloš Rančić
  • Oliver Stegen (via Skype)
  • Robin Pepermans
  • Santhosh Thottingal

Schedule and logistics edit

  • Place:
    Prinzessinnenstraße 19-20, 10969 Berlin

Schedule draft:

Friday edit

12:00 - 17:00 (or more if needed)

Saturday edit

12:00 - 17:00 (or more if needed)

Sunday edit

12:00 - 17:00 (or more if needed)

Agenda edit

What should the language committee discuss? Please edit this section and its subsections and propose topics that should be discussed in the meeting.

Technical edit

  • Automated archivalpublic archival of committee discussion is done manually and has suffered as a result; design a semi-automated archival process?
  • Automated request tracking — requests are manually tracked using status pages, which are inefficient and rarely updated. Should we invest time into creating an automated tool? Possible features:
    • automate some of the checks (e.g., localization completion);
    • view all requests by status for prompt processing;
    • admin UI to edit manual details (e.g., approval progress or links to archived discussions);
    • allow users to submit comments or questions (currently posted to Talk:Language committee)?
    • I once made a small tool to facilitate submitting a request [1] SPQRobin
  • Incubator usability — The incubator currently has several usability problems. What are they and how should they be solved?
    • 18788 enable the Incubator extension SPQRobin
    • Incubator as "virtualization server" for small Wikimedia projects (for example, => Millosh
  • Internationalization bugs — There are many internationalization bugs in MediaWiki, MediaWiki extensions and other software used by the WMF. How can they be solved more quickly? Which bugs need to be solved most urgently to advance the WMF strategic plan?
    • 19986 Wikis waiting to be renamed for years (e.g. als -> gsw) SPQRobin
    • 14407 CentralAuth global session doesn't include SPQRobin
    • 6100 Different directionality (rtl/ltr) for interface and content depending on language preference and &uselang SPQRobin
    • 2399 Unicode normalization "sorts" Hebrew/Arabic/Myanmar vowels wrongly - "RESOLVED LATER", but still pertinent Amir E. Aharoni
  • Localization testing - When programming, it's not enough to write the code - it must be tested, too. The same goes for localization - it's not just about translating a list of messages, they must be also tested in context. The difference is that there are plenty of methodologies and tools for testing code, as well as books and courses that teach them, but it seems like there are almost no set methodologies for testing localization. See this and this threads on Translators-l; i received almost no replies there. I sent similar questions to many of my friends in the software industry, some of them are experienced testing tools developers and certified software testing expert and none of them had an answer. Since Wikimedia projects are massively multilingual and need quality localization, the WMF can must develop a testing strategy for localization, as well as methodologies and tools. The results can be reused by other projects, as free software and free books. The Langcom can lead this effort.
  • Private Wiki or not: discussion. Millosh

Policy edit

  • How should Proposals for closing projects be handled? Should langcom take over? SPQRobin
  • Committee transparency & efficiency (in general) SPQRobin
  • Discussion about membership in Committee: At least, introduction of yearly acknowledgment per member will she or he be willing to continue to be a member of LangCom. Millosh 09:19, 23 March 2011 (UTC)[reply]
  • How to deal with macrolanguages? For example, on the Quechua (qu:) Wikipedia not all Quechua variants can be allowed, but you cannot start a Wikipedia for each code within this macrolanguage. There is a Wikipedia request for w:Kichwa language, which has a unified orthography (Kichwa Unificado) but has no ISO code (variants within this language, do). SPQRobin
  • 2010-07-01 note about Western Punjabi (pnb): pnb: Wikipedia is written in the same language (Eastern Punjabi) as pa: Wikipedia, except the script is different. "pnb" is reserved for Western Punjabi. What should we do? There is also a new request for a Wiktionary and a Wikiquote in the same languages as pnb: Wikipedia . SPQRobin
  • Approving new projects. Millosh

Projects and plans edit

  • Webfont extension. Millosh
  • Creating free fonts for languages which don't have them. Millosh Adding additional missing characters to fonts for Latin, Greek, Cyrillic. Creating free fonts for ancient scripts because already these are cited in articles. Evertype 18:28, 12 April 2011 (UTC)[reply]
  • Creating written forms for languages which don't have them. Millosh This needs linguistic expertise as well as expertise in orthography design. There are not so many people in the world with such expertise. It also needs buy-in from a willing user community, hopefully one which is in a position to influence things back home. Evertype 18:28, 12 April 2011 (UTC)[reply]
  • List of "missing Wikipedias": Create a list of languages which don't have Wikimedia projects. (Limit it on 100k or 10k speakers.) Millosh
  • A report on "what is needed" to improve language availability and use throughout all the communities. Evertype

See also edit