I certainly don't want to cookie-lick this usecase for Wikidata / Wikibase, but I think that using the Wikibase extension for this might be easier than it looks. The one major point would be a bit of coding to allow claims for the media namespace. But indeed, right now we do not have the bandwidth to do this, and if we are to work on it, it would be deferred to the Summer (the design can be started earlier, obviously).
On the other hand, as said, we don't want to cookie-lick the usecase: if there is a quick way to implement it, it is also obviously clear that moving well-defined structured data from one system to the other is far easier than the original creation of that structured data, and thus inefficient use of human work would be minimized anyway. As long as we are creating such well-defined structured data, I am happy either way.
2013/3/10 Yuvi Panda yuvipanda@gmail.com
While I understand that wikibase would be 'ideal' (allowing reads *and* writes), I do not know if the WikiData people have the personell bandwidth to do that. The separate extension outlined in this RFC would be a bit hackish, but still faaar more efficient than the current solution of having to retrieve the HTML and parse it again (or worse, parse the Wikitext itself). Would also be incredibly more efficient to get the data for multiple images, since you could feed a generator into this.
+1 for implementing this now :)
-- Yuvi Panda T http://yuvi.in/blog _______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l