Gerard's RFC proposes that we add statements to 'Category' items and then
Reasonator can use those statements to define a query and get a list of
wikidata items which correspond to that Category.
The justification is mostly that 'Category' items don't really have a
function in wikidata and this would give them one.
Personally I agree that Category items don't have a function in wikidata
but my solution is to merge all Category items with the corresponding Main
wikidata items. Both deal with different aspects of the same concept so
they should be the same item - just as a wikipedia article and a wikivoyage
article about the same thing link to the same item. The problem is that, at
the moment, the software doesn't allow this.
I just created bug
for
this.
Joe
On Sat, Mar 8, 2014 at 12:58 PM, David Cuenca <dacuetu(a)gmail.com> wrote:
On Fri, Mar 7, 2014 at 6:37 PM, Denny Vrandečić
<vrandecic(a)gmail.com>wrote;wrote:
I still disagree - let me explain why. I think
that trying to express a
query definition into a single statement is very hard. Having a specific
Query namespace allows us to create a completely new UI for them, allows us
to use a different data model for Queries than for items, and allows us to
treat Query pages very different (e.g. for caching) than, e.g. item pages.
Maybe I'm wrong, but "to create a completely new UI for [queries]" in my
head translates as "more barriers for users to understand and navigate
Wikidata".
The query doesn't need to be a single statement, but a single property for
queries (I called it "same as query" but it could be named differently).
You should be able to combine two statements of such property, or more, to
create a complex query. What I think it also matters is to reuse the
concepts the users are familiar with as much as possible.
For example, the different data model would allow us to restrict the
number of queries on a page. If they were just a statement, what would stop
a contributor from creating several such statements on one page?
On the mockup I intended to represent that the property "same as query"
can have several statements, but they combine into a single query.
What happens when someone removes the "same
as query" statement?
Same as now happens with other statements transcluded into wikipedia
pages, when somebody notices the deletion, it gets restored.
What happens if someone adds it to the page for
USA (e.g. "same as query"
"instance of"->"country", "continent"->"North
America",
"population"->>300M)?
It will get corrected, it is a wiki, right? :)
Would this page suddenly be treated differently?
Not necessarily.
Also, you already show in your mock up that the
"same as query" statement
requires plenty of special code (e.g. for the different visualizations,
etc.)
Same requirements as having queries as independent pages.
One option would be to have them as item pages, but then treat them
continuously different. This would mean more and more exceptions and
special casing in the code. I think that Queries and Items are sufficiently
different to deserve their own treatment. In my personal opinion, this is
provides sufficient reasons for Query pages and Item pages being distinct.
No need for different treatment, or exceptions. Just a property that acts
as a query descriptor with as many modifiers as needed.
What would be the advantage of having Queries being expressed in the
Items? Less entities?
Yes!
Less confusion about what these "list
of"- and "category"-items mean?
Yes!
Both reasons I don't find sufficiently
enticing to change my opinion on
this.
And better navigation, and more simplicity to create a query, and no need
of creating artificial divisions between pages that represent the same
concept...
Cheers,
Micru
_______________________________________________
Wikidata-l mailing list
Wikidata-l(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata-l