To start from the bottom up. The Ottoman Turkish language has been
approved and will shortly be added to Wikidata; it will happen with the
next release of functionality for Wikimedia projects. As to supporting the
Incubator with Wikidata, that is technically not really possible. At this
time you can have only associate one article of a Wiki with one item in
Wikidata. Given that Incubator hosts multiple languages that all can have
articles associated with the same item and given that Incubator does not
have a high priority it is not going to happen any time soon. There is work
planned for the next two years already if I understand things well.
The 47930 is likely to happen maybe even in the near future. However,
building info boxes based on this for use in Incubator will just make these
boxes more complex to create, for this reason it will be expedient when a
language moves more quickly out of the Incubator.... At this moment we do
not even have proper info boxes for use in existing Wikipedias.. For your
information there are some designs for really slick looking info-boxes
My proposal seeks to ensure that we gain more quickly a situation where a
language will provide meaningful information to its readers. When you state
"there is no such thing as a complete labelling of wikidata items" you are
in my opinion so wrong. When all the statements used on an items have
labels in a language, it is completely labelled for that language. These
labels are likely to be used on many other items and that is EXACTLY why
using Wikidata makes sense; it quickly has an effect on many more items. An
example, both "human" and "sex" are used on something like a million
the labels have only to be added once. 250 well chosen items will not only
have a big effect on the availability of searchable items in a language, it
will also provide many labels that are used on most other items.
The objective is that we provide a service, in my opinion the combination
of quickly created Wikidata information provided in quality visualisations
like can be seen in the Reasonator info for Cambridge or Johan Sebastian
Bach makes it more obvious that there is no need for the current crop of
sub-stubs we find in the Incubator.
The Wikidata information can be used to provide search results and, it can
already be used as the front end to Commons (by way of the categories).
I do not mind that you are against, but I do want to understand your
arguments and, I do not. So I have tried to be even more clear why I think
we need to go this way.
On 23 November 2013 20:50, MF-Warburg <mfwarburg(a)googlemail.com> wrote:
Am 19.11.2013 09:58 schrieb "Gerard Meijssen"
You may have missed it in the mail I send yesterday, but I want to
change the requirements for a new Wikipedia.
My proposal is to ask for the current number of full articles or provide
articles and the complete labelling of a selected set of 250
The rationale is that it takes much less effort to work on Wikidata and
quickly amount to providing additional results for other items. An
example: the property "sex" should be in use on all humans and the
qualifiers "male" and "female" should be in use on all of these as
I oppose this in its entirety.
First of all, there is no definition for what is a "full article" and
neither do I think we need to spend time with creating a definition or
later checking if tests have in fact managed to write 50 of such articles
that conform to such a definition.
Second, there is no such thing as a complete labelling of wikidata items.
Not all items use all labels, and even graver: it is irrelevant how many
labels are used on an item, since they do not require translation or so -
they just refer to other items which have translations of their content.
In this regard, it would be much more helpful for tests to push for the
fixing of bugs 47930 and 52971 to actually allow incubator tests (and all
langs with ISO codes) to use wikidata.
Langcom mailing list