[Wikitech-l] Good things come in threes

It's time to remember one of those English proverbs I leaned in high school: "Good things come in threes!" 1. After being enabled on the German language Wikipedia a few weeks ago[1], Flagged revisions is a hit. Over 255.000 content pages have been reviewed (a third of all content pages) by almost 3.000 users[2]. The in my humble opinion Wikipedia with the highest quality content, now ensures that anonymous readers get to read checked versions of content, reducing chances of innocent bystanders being confronted with all kinds of pranks and filth. Great stuff. 2. Three days ago a shell user created a large number (15) of new MediaWiki wikis (Thanks Tim!)[3]. Many of those in languages that did not have a Wikimedia project yet (9). Thanks to everyone who has taken and is taking the time to make Wikimedia projects even more omnipresent. Imagine a world in which every single human being can freely share in the sum of all knowledge. 3. Last Tuesday, the illustrious "Bug 57" was finally closed[4]. Using multiple Wikimedia projects has never been this easy before. Yay! for meta projects like Wikimedia Commons. Now where's that WYSIWYG editor so that the learning curve does not have to be this steep[5]? I am putting a EUR 1000,00 bounty on the first to get one working properly and approved by Brion before the end of 2008. Contact me for details if you are serious about working on this. I am certain there are more people that would chip in for this[6] - and MediaWiki needs it to stay an interesting wiki engine outside of the WMF projects. 4. As you may not know (who am I anyway), I am heavily involved in the MediaWiki localisation project, which tries to make MediaWiki available in as many languages as possible. In the end of 2007 I formulated four ambitious goals for the project[7]. By the end of the year, 120 languages should have a minimal localisation (proper localisation for 'read only', back then 48 languages), 90 languages should have a localisation for at least 90% of all MediaWiki messages (then 50), 50 should have a 90% localisation of extension messages used by Wikimedia (then 11) and 20 should have a 65% localisation of all extension messages supposed in the MediaWiki localisation project (then 7)[8]. In December 2007, I thought all of those goals would be just or well beyond reach. Well, impossible really. Today the first of the forementioned four goals was reached: 20 languages now have a 65% or more localisation of the 3,700 messages of the extensions supported by Betawiki. Translators for Esperanto and Vietnamese completed the 20 languages. MediaWiki now has minimal localisation for 112 languages, an excellent localisation exists for 65 languages, and 16 languages have excellent support for the extensions that the WMF uses[9,10]. Brilliant if you ask me. And still we need more. There are Wikimedia projects in over 270 languages (counting Incubator projects) and MediaWiki supports about 313 languages. The least we should do is offer language communities a user interface in the language their parents taught them - nothing better to feel at home. Ouch. I have to learn how to count… And I was not even done yet. Just wanted to share my happiness :). Siebrand [1] http://lists.wikimedia.org/pipermail/foundation-l/2008-May/042705.html [2] http://toolserver.org/~aka/cgi-bin/reviewcnt.cgi?lang=english [3] http://lists.wikimedia.org/pipermail/wikitech-l/2008-May/038008.html [4] http://leuksman.com/log/2008/05/28/bug-57-laid-to-rest/ [5] http://en.wikipedia.org/wiki/Learning_curve [6] The bounty is made available through the 'MediaWiki accessibility project' of Stichting Open Progress (http://www.openprogress.org), subsidised by HIVOS (http://www.hivos.nl/). Acceptance criteria include a proper 'back and forth' conversion of 2,000 English language Wikipedia main namespace pages. [7] http://lists.wikimedia.org/pipermail/translators-l/2007-December/000571.html [8] http://translatewiki.net/wiki/Translating:Group_statistics_in_time [9] http://translatewiki.net/wiki/Translating:Group_statistics [10] the relatively low rise in localisation completeness for WMF extension is mainly due to the large increase in messages for CentralAuth and FlaggedRevs; translators need some time to catch up.