We rolled out usage tracking on the first two wikis (French Wikisource and Dutch Wikipedia). Users should not notice anything. More wikis will follow in the next weeks. This is the remaining step for enabling arbitrary access on wikis other than Commons.
The students team is working hard to get a first release of the improved constraint reports and checks against 3rd party databases out.
Ricordisamoa fixed the issue with long descriptions being cut off.
We fixed the focus flow in the property selector.
We improved the messages on Special:EntityData to make it more understandable.
Past: Andy Mabbett (User:Pigsonthewing) spoke about authority control in Wikidata (and sister projects) at the World Digital Library's Arab Peninsula Regional Group Symposium, organised in conjunction with the Qatar National Library, in Doha on 13 May. His slides, with an Arabic translation, will be online soon.
Past: Wikidata editing workshop at the Wikimedia Conference in Berlin
Past: Wikidata workshop for archivists of Catalan City of Justice
Upcoming: MediaWiki hackathon in Lyon
Other Noteworthy Stuff
Elections for the board of the Wikimedia Foundation are ongoing. Wikidata should be well represented among the voters. Go and vote!
Dutch Wikipedia and French Wikisource now have the arbitrary access feature. Persian Wikipedia, English Wikivoyage and Hebrew Wikipedia are following today. Italian Wikipedia and all remaining Wikisource projects are following on June 1st.
Italian Wikipedia and all remaining Wikisource projects now have arbitrary access. The rollout will continue. The schedule for the next projects is at d:Wikidata:Arbitrary access.
The Content Translation tool now automatically connects translated articles to Wikidata. (Thanks Content Translation developers!) Previously translated but unconnected articles have been connected by a bot. (Thanks Amir!)
Code review of the extensions written by a team of students to improve the constraint reports and make it possible to automatically check our data against other databases. A first version will go live soon pending further codereview and fixes.
A challenge was raised to calculate the degrees of separation to Kevin Bacon. Already one routine has been produced, the next step is to use live data so that we can work on reducing the number of intermediary steps.
More work on getting the Wikibase Quality extensions ready for deployment. They'll improve constraint checking and provide checks against 3rd party databases.
Special:UnconnectedPages had to be reworked (thanks Amir!) because of huge performance issues. Because of this it loses some functionality (being able to specify where the list starts and limiting the list to pages that have local interwiki links). We're working on bringing functionality back without the performance penalty.
Enabled arbitrary access on arwiki, cawiki, eswiki, huwiki, kowiki, rowiki, ukwiki, viwiki, and usage tracking on dewiki, ruwiki, cswiki and all s3 wikis. (ruwiki and cswiki get arbitrary access on June 23)
Finishing work on improved handling “invalid” values and statements with deleted properties.
Small breaking DOM change: Wikibase no longer adds its own h1 tag but uses MediaWiki's default .firstHeading element. (phabricator:T93534)
Michael, could you look at the issue raised at User talk:Revent/Archive 1#Ignorant threats and continued on my talk page User talk:Colin#Fair use content on Wikimedia Commons. There are two issues I see. Firstly that our licensing and fair use policy applies to our media files only. Other rules, mostly undocumented on Commons, apply to our talk page, forum and guideline text. Secondly that we are permitted to quote copyright text on such talk/forum pages, and my understanding is that this is "fair use". User:Revent seems to think that since fair use is not allowed on Commons, the posting of even small amounts of copyright text here is a blockable offence, and has threatened Jee. -- Colin (talk)
And note that all descriptions and tags in Flickr pages are copyrighted; only the media is free licensed. Jee 09:27, 22 June 2015 (UTC)
If true, that has consequences for our bots that scrape the text and insert it into our file pages. Such copy/paste could not, imo, be regarded as fair use. That's probably a separate issue from the dispute by Revent, but I would be interested also in what Michael understands about that too. If true, it would severely limit our ability to make use of Flickr images on an automated basis (each uploader would have to reword the description text much like a Wikipedia author has to reword their sources). And what about descriptions in EXIF data? -- Colin (talk) 09:41, 22 June 2015 (UTC)
In Flickr, we can post a photo, screenshot or artwork/illustration and the license is limited to it. Any one can add tags (like identification tags) without giving a license that means "copyrighted to that person". There is no provision to specify a license for description provided by the uploader. Most pages contain third party copyrighted texts (like in my Flickr pages) that the uploader can no way release the license. Jee 09:49, 22 June 2015 (UTC)
Interesting, and troubling. I agree that the description on Flickr could certainly contain text from a third party that can't be re-licensed. Many people quote Wikipedia on their descriptions, which is ironic. I wasn't aware people apart from the author could add a tag (ignoring the autotag thing for now). I'd like to get this Revent/fair-use thing out of the way first. Then it may be a good idea to open up a Village Pump / Copyright discussion on this issue. It has serious consequences for out bot users in particular, but also anyone uploading an image from Flickr. And it applies to other sites too, where the image is CC but the description text may be added by another person. -- Colin (talk) 09:59, 22 June 2015 (UTC)
The tagging can be enabled or disabled in user preferences. It can be limited to friends, contacts or even allow anyone to tag. I usually allow anyone as I need to get the IDs of unidentified organisms. Jee 10:04, 22 June 2015 (UTC)
More work on getting the Wikibase Quality extensions ready for release. The part doing constraint violation checks should go to test.wikidata.org next week. The part doing checks against 3rd party databases will still take a bit longer but is also shaping up nicely.
Made it possible to create items even when another user is mass-creating items (phabricator:T103796)
Bene* worked more on making Wikidata work nicely on mobile