Zdałem dziś pytanie podczas dyżuru zespołu Wikidata:
17:06 < Lydia_WMDE> ok last question i have then: [18:57:57] <saper> My
question: Suppose I have a pretty good, normalized SQL database like registry of
monuments in my country (
http://wikilubizabytki.centrumcyfrowe.pl/) or
good database of species. All are already imported into MySQL
tables. Can I use wikidata to refer to those data sets on my own
MediaWiki installation? How would that done? If so, could similar
process be applied to WMF wikis?
17:08 < Denny_WMDE1> saper: now to your question
17:08 < Daniel_WMDE_> Lydia_WMDE: in reply to saper: no, that'S not what
wikidata is for. use the DataTransclusion extension instead.
17:08 < Denny_WMDE1> i would have phrased it differently, but as Daniel_WMDE_ says
:)
17:09 < Lydia_WMDE> ok that is all the questions i have and we have about 20 minutes
left
17:09 < Lydia_WMDE> any more questions or things to doscuss?
17:09 < Denny_WMDE1> saper: follow up questions?
17:09 < Lydia_WMDE> how did you like the demo system so far?
17:09 < Denny_WMDE1> saper: my suggestion would be to import the data into Wikidata,
and then use it directly from there.
17:09 < Daniel_WMDE_> Lydia_WMDE re saper: the main reason being that wikidata has a
very specific and complex data model. superimposing this on a "normal"
relational database would be a pain and quite useless
17:10 < Daniel_WMDE_> Denny_WMDE1: but "bulk imports are problematic" :)
17:10 < Lydia_WMDE> ;-)
17:10 < Denny_WMDE1> exactly
17:10 < Daniel_WMDE_> also, wikidata's features are pretty useless for
single-source simple-value bulk data
17:10 < Denny_WMDE1> but less problematic than having an external data table that we
use instead :)
17:10 < Daniel_WMDE_> it just makes things complex and slow
17:10 < Denny_WMDE1> true
17:11 < Denny_WMDE1> we have history, references, qualifiers, etc.
17:11 < Daniel_WMDE_> if you have data that fites the relational model nicely,
wikidata is probably not the best way to manage it :)(
17:11 < Denny_WMDE1> most of which are probably not available in the data table that
one might already have
17:11 < Denny_WMDE1> good point :)
17:12 < Denny_WMDE1> i have the feeling we are making people very unhappy with our
answers today