Lovely work. On Mar 11, 2016 5:32 PM, "Dan Garry" <dgarry(a)wikimedia.org> wrote: wrote: dream? Perhaps you could do this w two queries, one to a composite index that is only updated weekly. results, Make this opt-in, add a different background color for results from the all-language index, & divide their search-relevance by a language-prominence factor... +SJ

Let me elaborate my case a little bit: Here is the scenario: *I spent a lot of time luring fresh internet users into the terrains of free knowledge, principally WM projects, and specifically WP, in a bustling Global South region, interlaced with several local language projects. *I have always treated Wikipedia, first for the readers, next for the builders. Without more people becoming readers, there is no way more builders turn in. *Most of these people are aware that Google will fetch them needed information quickly. *A nominal percentage of them are aware that there is Wikipedia, in English. *A thin percentage knows that there is wikipedia (however thin they may be), in their native language too. If only they knew, they would even be eager to explore more and venture into enriching it further (as new editors). (Editing English Wikipedia is too complicated, and bureaucratic, let's admit it!. esp. if a topic is only of local interest and notability) *Many among them are quite good in handling their native language. Some are good in reading and comprehending English. But thin is the population who can use English as a creative writing language. *Most people use a PC or Mobile device which is not necessarily equipped with their native language IME and/or rendering features. Most often, they can read a web page in their language but can't type in (for want of a suitable IME). But they can copy paste a string into a search box! *Many articles have a rather better expanded version in a local wiki than EN, if the topic is more notable in their community. But sometimes, it can be the other way too. *People gets turned off from a web site they have encountered for the first time, if it gives them results not that encouraging. They might never visit it at least for a long time to come. They have no time to waste. Now, how this may be helped off: 1. I know a portal wikipedia.org or even en.wikipedia.org 2. I see a news item in a local language web page about a local entity. I want to know more facts about this entity. 3. I get to the portal and inject this string (or type in this string) into the search box. 4. There comes a few links: The first among them is an article in my language about the entity. Following are versions of the same article (if they exist) in other language wikis. 5. I directly choose, click and get to read what I wanted. I am happy and decides to visit this portal again as a regular routine. What, possibly, mediawiki / wikipedia can do behind the walls: 1. Gets a search string. 2. Identify the language. a) Pick up the index for that language and process. b)Goto Wikidata links and find an equivalent link / article in other language projects for this string. pick index and seek. c)Harvest and show case the results (optimized in certain ranks say first article titles, then inline occurrences, then media etc.) I can vouch positively that this itself will boost our counts to a great extend, at least from regions where WP is yet to make a significant impact. computing resources should get strained for this simple extra steps. -Viswaprabha On 12 March 2016 at 05:40, Carlos Mora <lcharles.mora(a)gmail.com> wrote:

On Mar 11, 2016 19:32, "Dan Garry" <dgarry(a)wikimedia.org> wrote: Fair enough! Great once more to see this update. It makes this a much more engaging page to set as a browser default on public machines. Sj

Hi there - speaking to one thing I'm familiar with, with respect to image selection, we believe https://phabricator.wikimedia.org/T124225 should address fair use ("non-free") images, although page reparses will happen gradually (pages are cached for up to 30 days or so). -Adam

Thank you to the Discovery team -- it seems to me that your work has been largely overshadowed by political concerns in recent months (which may have been necessary, but not pleasant). I'm delighted to see working and useful software emerge, in spite of the challenging environment that has existed around your work. I'm delighted to see your substantive engagement (e.g., Dan and Adam, above) with feedback about the implementation. Kudos! -Pete On Fri, Mar 11, 2016 at 2:05 PM, Dan Garry <dgarry(a)wikimedia.org> wrote: