Amir,

The most important thing to consider when you look at the burden to the
community is the "Bonnie & Clyde" problem where it gets to be the choice of
a few editors (maybe just one?) on any given Wikipedia. I noticed that the
English Wikipedia only has one article for both Q11629 the art of painting
and Q3305213 the physical object, whereas both probably need about 5
articles each (container article and subarticles; so eg. 1)
Q3305213 painting, 1.1)oil painting 1.2) miniature 1.3) watercolor, etc).
The enwiki article is currently linked to the Wikidata item for the art of
painting and the English Wikipedia has no article for the physical object.
This strange situation is probably caused by the fact that both meanings
are covered by the same word in English, which is of course not the case in
other languages. When I created an article in the English Wikipedia for
"painting (object)", it just got redirected to the other one, and I was
accused of "disruptive editing". Now this redirect is become one of the
many "hanging redirects" in Wikidata that should be deleted.

In the pre-Wikidata world, people tended to stuff articles full of every
lemma that was not encyclopedia worthy in and of itself. In the English
Wikipedia there is even a merge template for that. Today, in the
post-Wikidata world we are concentrating on getting the most info possible
into a mobile-sized screen, so we favor condensed chunks of twitter-style
info, rather than pages of text. I see this in my watchlist from the number
of changes to lead paragraphs. Everyone seems to just concentrate on the
lead these days, because that is what Google is serving to searchers.

In conclusion, though I share your dream, after considering this problem
from many angles I think the best interim solution is to "do nothing". A
cultural revolution is taking place in Wikipedia, whereby people are slowly
realizing that breadth (range of articles on any given subject) is more
important than length (depth of any given article on its own subject). Let
that revolution take place, one article at a time, at the pace granted by
the policies of any given Wikipedia.

Jane

On Thu, Feb 12, 2015 at 8:58 AM, Amir E. Aharoni <
amir.ahar...@mail.huji.ac.il> wrote:

> Yeah, looking into labels is certainly something that I considered, but
> that is by definition only a guess and not as bulletproof as Q numbers.
>
> We considered doing stuff like:
> * [[not-yet-written article about Douglas Adams|Douglas Adams]]<!-- wd:
> Q42 -->
> * [[not-yet-written article about Douglas Adams#Q42|Douglas Adams]]
>
> ... and this would kinda work, but would be leave a lot of mess to the
> community editors to clean up. The template way, suggested by Gerard, is
> similar and seems slightly less messy to me. But only slightly.
>
> (If anybody cares, the relevant task in ContentTranslation is
> https://phabricator.wikimedia.org/T88580 .)
>
>
> --
> Amir Elisha Aharoni · אָמִיר אֱלִישָׁע אַהֲרוֹנִי
> http://aharoni.wordpress.com
> ‪“We're living in pieces,
> I want to live in peace.” – T. Moore‬
>
> 2015-02-12 7:51 GMT+02:00 Maarten Dammers <maar...@mdammers.nl>:
>
>> Hi Amir,
>>
>> Amir E. Aharoni schreef op 11-2-2015 om 13:12:
>>
>>> If I may dream for a moment, this should be something that can be used
>>> in all Wikipedias, and without copying this template everywhere, but built
>>> into the site's software :)
>>>
>> Exactly, the template based approach doesn't scale at all. You have to
>> somehow make it automatic. One thing I thought about is adding suggested
>> sitelinks to Wikidata. The software would encounter a red link and would
>> look in Wikidata if it can find an item with a suggested sitelink of the
>> same title. Huge software overhaul so I don't see that happening.
>>
>> Another approach that is probably already possible right now:
>> * Take an article with a red link
>> * Look at the links in the article in other languages.
>> * If you find a link that points to another article which has the same
>> label as the red link in the same language, link to it
>>
>> I wonder how many good results that would give.
>>
>> Maarten
>>
>>
>> _______________________________________________
>> Wikidata-l mailing list
>> Wikidata-l@lists.wikimedia.org
>> https://lists.wikimedia.org/mailman/listinfo/wikidata-l
>>
>
>
> _______________________________________________
> Wikidata-l mailing list
> Wikidata-l@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikidata-l
>
>
_______________________________________________
Wikidata-l mailing list
Wikidata-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata-l

Reply via email to