I think this subject should also be discussed on the Commons mailing list,
as this plan is to demolish the navigational structure of Commons.

2015-08-27 15:03 GMT+02:00 Romaine Wiki <[email protected]>:

> No we have not a clear policy on only linking sitelinks to categories if
> the item itself is about a category. So not let's not break that.
>
> You suggest to break down almost the complete navigational structure
> Commons has in relationship with Wikipedia, and makes it possible to find
> articles that are about the same subject as the category. Without it
> becomes almost impossible to identify a category on Commons to be related
> to an article in Wikipedia.
> Sorry, but your proposal is insane and making the navigational situation a
> thousand times worse. And does it make anything better? No, totally not.
> Only the opposite: worse.
>
> Wikidata is currently heavily used to connect categories on Commons to
> articles on Wikipedia. This so that interwikilinks are shown on the
> category on Commons to the related Wikipedia article. This for navigational
> purposes but also to uniquely identify categories on Commons to articles on
> Wikipedia and items on Wikidata.
>
> How nice Commons galleries are giving an overview, they are crap in
> speaking of navigational purposes. For every subject a category on Commons
> is created and used and the Commons categories form the backbone to media
> categories.
>
> It has been pointed out for a long time that the linking situation on
> Commons is problematic and this is a software issue, not a user side issue.
> This consists out of:
> * There can only be added one sitelink to an item.
> * If no sitelink added (but only added as property), a Commons category
> can't show the interwikilinks.
> * If a category and an article on Wikipedia/etc exist for a subject, only
> one of them can be shown on the Commons category.
>
> The annoying part is that some large wikis, especially the English
> Wikipedia, creates too many categories that are not created on other
> Wikipedias. This causes that categories on Commons are only linked to a
> category on Wikipedia, which is useless for most other wikis and on Commons
> we miss an interwikilink to the related article.
>
> A gallery on Commons is a great way as alternative to show images, but is
> not suitable for navigational purposes, as that requires a much higher
> coverage and being a backbone everything relies on. On Commons only
> categories have that function. A counter proposal makes more sense: no
> Commons galleries as sitelinks any more and having Commons galleries only
> as property added.
>
> But this only solves a part of the problem: on Commons I would like to see
> somehow that both the related category as the related article are shown.
> Example: on the Commons category for a specific country both the country
> category on Wikipedia is linked as the article on Wikipedia is linked.
>
> Something I have been wondering about for a long time is why there are 2
> places on an item where a Commonscat is added. I understand the development
> and technical behind it, but this should not be needed.
>
> So the developers of Wikidata should try to find a way to show both groups
> of interwikilinks on categories on Commons.
>
> As long as this is not resolved in software, this problem of 2 items both
> strongly related to a Commons category keeps an issue.
>
> Romaine
>
>
>
>
>
> 2015-08-27 11:29 GMT+02:00 James Heald <[email protected]>:
>
>> A few days ago I made the following post to Project Chat, looking at how
>> people are linking from Wikidata items to Commons categories and galleries
>> compared to a year ago, that some people on the list may have seen, which
>> has now been archived:
>>
>>
>> https://www.wikidata.org/wiki/Wikidata:Project_chat/Archive/2015/08#Trends_in_links_from_items_to_Commons
>>
>>
>> A couple of headlines:
>>
>> * Category <-> commonscat identifications :
>>
>> ** There was a net increase of 61,784 Commons categories that can now be
>> identified with category-like items, to 323,825 Commons categories in all
>>
>> **  96.4% of category <-> commonscat identifications (312,266 items) now
>> have sitelinks.  This represents a rise in sitelinks (60,463 items)
>> amounting to 97.8% of the increase in identifications
>>
>> **  80.0% of category <-> commonscat identifications (259,164 items) now
>> have P373 statements.  This represents a rise in P373 statements (8,774
>> items) amounting to 14.2% of the increase in identifications
>>
>>
>> *  Article <-> commonscat identifications :
>>
>> ** There was a net increase of 176,382 Commons categories that can now be
>> identified with article-like items, to 884,439 Commons categories in all
>>
>> ** 23.4% of article <-> commonscat identifications (207,494 items) now
>> have (deprecated) sitelinks. This represents a rise in sitelinks (112,595
>> items) amounting to 63.8% of the increase in identifications.
>>
>> ** 91.3% of article <-> commonscat identifications (807,776 items) now
>> have P373 statements. This represents a rise in P373 statements (110,727
>> items) amounting to 62.8% of the increase in identifications
>>
>>
>> *  In addition, a recent RfC showed considerable confusion as to what
>> actually was the current operational Wikidata policy on sitelinks to
>> Commons:
>>
>>
>> https://www.wikidata.org/wiki/Wikidata:Requests_for_comment/Category_commons_P373_and_%22Other_sites%22
>>
>>
>> In view of the trends above; and the need for predictability and
>> consistency for queries and templates and scripts to depend on; and
>> particularly in view of the apparent confusion as to what the operational
>> policy currently actually is, can I suggest that the time has come for a
>> bot to monitor all new sitelinks to Commons categories,
>> *  adding a corresponding P373 statement if there is not one already, and
>> *  removing the sitelink if it is from an article-like item to a
>> commonscat.
>>
>>
>> I believe we have clear policy on only sitelinking commons categories to
>> category-like items, and commons galleries to article-like items; but there
>> is currently confusion and unpredictability being caused because these
>> relationships are not being enforced -- breaking scripts and queries.
>>
>> It's time to fix this.
>>
>>
>> All best,
>>
>>   James.
>>
>>
>> _______________________________________________
>> Wikidata mailing list
>> [email protected]
>> https://lists.wikimedia.org/mailman/listinfo/wikidata
>>
>
>
_______________________________________________
Commons-l mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/commons-l

Reply via email to