Very nice!
As a slight tweak on your query, here are some statues within 10km of
London's Trafalgar Square
http://tinyurl.com/htxqp5t
... or within 10km of Stockholm's Stortorget
http://tinyurl.com/jnv5qo3
(Warning: per a recent decision of the Swedish Supreme Court, the latter
search may be a copyright violation).
Probably quite a lot of items to add on both counts -- has anybody
scoured Commons for public art, that may not yet have an article in any
Wikipedia? -- but really exciting to be able to produce output like this
so easily.
One thing I'm a little nervous about is that the first search is taking
six and a half seconds to run, compared to 809 ms for every statue
within Greater London
http://tinyurl.com/hwf7pjg
which I would have thought ought to be a similar-sized query.
I haven't checked the optimiser output -- is this likely to be because
the join has been performed in a different order? Or is there a good
reason why the geo-search should be so much slower, even now it has
indexing built in ?
One other thing for the wishlist -- it would be really nice if the map
mouseovers could show thumbnails (and indeed if there was a toggle to
allow them for the table output, too). Also, should the Commons links
not point to the file information pages, or alternatively to
MediaViewer, rather than directly to the images? The latter makes it
very hard to see the attribution information (etc) for the images, which
could be a licence violation, eg of CC-BY.
But overall: wow! Fantastic!
All best,
James.
On 09/05/2016 21:23, Stas Malyshev wrote:
Hi!
After a number of difficulties and unexpected setbacks[1] I am happy to
announce that geospatial search for Wikidata Query Service is now
deployed and functional. You can now search for items within certain
radius of a point and within a box defined by two points - more detailed
instructions are in the User Manual[2]. See also query examples[3] such
as "airports within 100km of Berlin": http://tinyurl.com/zxy8o64
There are still a couple of things to complete, namely sorting by
distance (coming soon) and units support (maybe). Overall progress of
the task is tracked by T133566[4].
[1] https://lists.wikimedia.org/pipermail/wikidata/2016-May/008674.html
[2]
https://www.mediawiki.org/wiki/Wikidata_query_service/User_Manual#Geospatial_search
[3]
https://www.mediawiki.org/wiki/Wikibase/Indexing/SPARQL_Query_Examples#Airports_within_100km_of_Berlin
[4] https://phabricator.wikimedia.org/T123565
_______________________________________________
Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata
