They came out for a solution for themselves, or do you happen to have the link to their github repo? Otherwise you may want to donate a few million dollars so they can come out with a similar solution.
My point
Wikipedia is an "encyclopedia anyone can edit".
Google aims to "index all the world's information".
These are different but complementary goals.
That the Wikimedia Foundation hasn't improved a user interface that was slapped together in the early 2000s is not somehow a mathematical, universal constraint on the growth of the subjects that are covered. It just isn't.
I am prepared to bet honest folding money that almost nobody uses the internal mechanisms of search and navigation that Wikipedia provides. Apart from clicking an in-text link referring to another subject, I am prepared to bet that traffic to Wikipedia is dominated by search engine referrals by at least one order of magnitude. Probably two.
Google so dominates the actual usage of Wikipedia that it is ridiculous to advance the poor UI of the Wikipedia platform as some kind of serious argument in favour of deletionism.
It'd be like walking into a library circa 1950 and saying "All these index cards are a schlep, let's start throwing away books".
You'd be committed to a loony bin. And now we have unlimited search and retrieval capability and you suppose that a poor interface is the killer blow for deletionism?
I believe I am having a Mugatu Moment here.
Most people don't return to Google after the initial access to the website, they keep using the Wikipedia interface when they need to search related subjects.
And I will surely throw away some books from the library if random people were allowed to put their books in there.
And the killer blow is the second part.
Edit: since you added this --
> Most people don't return to Google after the initial access to the website, they keep using the Wikipedia interface when they need to search related subjects.
I'll bet you $100, AUD or USD, to be donated to the charity of your choice, that this is not so, by a factor of at least 10 to 1.
As I said, this does not include following links in articles. I am talking about the wikipedia search engine and category pages. I am talking about search engines generally, though I expect google to be the dominant one.
And you overall do have a point, but is not as strong as you think it is.
Perhaps is related to the kind of subjects you are searching for; or it may just personalization by google.
Or maybe we are both biased for ultimately silly reasons and the truth lies somewhere in the middle but the lack of available data reduces everything to mere speculation.
Turn off 'instant' so you see the top 10 autocompletions, and watch over all your queries. You will very often see "wiki" as a suggested suffix.
Or better yet, just add " w" to the end of any of your own tests: " wiki" will be the top suggestion, which demonstrates that 'many' people add it as a suffix on Google.
Everything on this question is not reduced to bias and 'mere speculation'. I've observed many peoples' search behavior, not just my own, and habitual recourse to browser-based search boxes or always-requery-at-Google is growing over time (especially with the rise of Chrome and its 'onebox').
Wikipedia also did usability studies in the 2009-2010 timeframe, from which Wikimedia director/developer Erik Moeller reported: "our test subjects tended to resort to common web search engines to navigate Wikipedia instead of using the site’s own search" [1]. (Wikipedia has since moved the site search box to help it be found, and I suspect that's boosted its use, but it's still subtle compared to the always-available, always-familiar in-browser Google-powered search.)
[1] http://blog.wikimedia.org/2010/06/15/usability-why-did-we-mo...