Categories are not error prone, it are humans. And this means that as result this applies also to queries. Both depend on the quality of the edits users make, and both are as problematic.

I am 100% sure that wikibase technology can be helpful and improve the search for and use of images, but is not a replacement for stable categories and their pages. That it will be available in many languages is nice, but still a stable page system to categorise images is needed. I consider it much more likely that categories are converted in such way that the images in a category are added to an ID, instead of a language depended category name, and yes, that sounds much like a wikibase system, but is not exactly the same.

One of the most important usage of categories on Commons is navigation. Looking to the way how Wikidata is, navigation through groups of pages is hopeless. On Commons this type of navigation is essential and Commons can't be without.

But still, this subject is irrelevant for this discussion, as the problem situation still remains: from Commons to Wikidata there are two groups of pages to link to: articles and categories.

Romaine

2015-08-30 20:52 GMT+02:00 Gerard Meijssen <gerard.meijssen@gmail.com>:

Hoi,

A category includes images that have shared characteristics.  The problem with categories is that they are error prone, mistakes are made or they are just not added or understood. A category like "Frisian stallions" has obviously stallions of the Frisian breed in them.  When an image is known to have a "Friese stamboekhengst" in Dutch, you do not need to add a text for "Frisian stallion". Because one is the translation of the other. You can search for all the paintings by 文森·梵谷 and tind the same paintings I would when I seek them for Vincent van Gogh.

There is nothing wrong with having categories at this time. It is just that once we add statements, the same information will be available in every other language. We cannot do it now. The whole current issue with Wikidata is not that relevant as it will change when development time is put into place and have Wikidata type annotations and queries in a way that is not only but also English.

Thanks,

      GerardM

On 30 August 2015 at 20:37, Steinsplitter Wiki <steinsplitter-wiki@live.com> wrote:

>>The problem with Commons is that its categories will be largely redundant once Wikidata type technology will replace them

Can you elaborate please? What is replacing commons category system?

...

---

From: gerard.meijssen@gmail.com
Date: Sun, 30 Aug 2015 20:30:36 +0200
To: commons-l@lists.wikimedia.org
Subject: Re: [Commons-l] [Wikidata] Trends in links from Wikidata items to Commons

Hoi.

The problem with Commons is that its categories will be largely redundant once Wikidata type technology will replace them, It will be just a question of having a query for the result that you want.. If you want to have fences in mauve, by all means, query for it but there may be the surprise that there are none.

Creating items for categories for Commons is imho an exercise in futility.. What is the point after all ? Having those categories may mean that we have a clue what queries are of interest..So when people add those categories, it is current best practice.

Thanks,

     GerardM

On 30 August 2015 at 16:51, Gnangarra <gnangarra@gmail.com> wrote:

the problem I see is that commons will always have more  categories than wikipedia can have  articles take fences, commons has wooden fences this broken into many cats including wooden fences in a country, this then grows and then gets broken into sub national entities while the number of articles on wikipedia remains at one commons now has 196 country articles with anything between 5 and 50 sub national entities, then some idiot paints his fence now we have wooden fences by colour in a little over 3000 pantone colours....

What I'm seeing here is solution that has the horse pushing the cart problem lies not in linking commons cats to wikipedia articles wikidata but in ensuring wikidata articles are linked to the full range of categories available on commons and that those links can be easily adjusted as necessary

On 30 August 2015 at 18:42, Federico Leva (Nemo) <nemowiki@gmail.com> wrote:

Luca Martinelli, 30/08/2015 12:03:

Am I the only one that thinks that jheald's .js is a temporary solution?
Am I the only one that actually appreciate his attempt at solving a
*practical* problem by providing a *practical* solution,

It might be a practical solution, but I don't understand what it solves: what's the practical problem?
Quoting from the project chat, the problem to me seems this: «2.4 millions categories are not connected to corresponding Wikipedia articles. [...] — Ivan A. Krestinin (talk) 20:22, 18 August 2015 (UTC)».

Nemo

_______________________________________________
Commons-l mailing list
Commons-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/commons-l

--

GN.
Vice President Wikimedia Australia
WMAU: http://www.wikimedia.org.au/wiki/User:Gnangarra
Photo Gallery: http://gnangarra.redbubble.com

_______________________________________________
Commons-l mailing list
Commons-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/commons-l

_______________________________________________ Commons-l mailing list Commons-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/commons-l

_______________________________________________
Commons-l mailing list
Commons-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/commons-l

_______________________________________________
Commons-l mailing list
Commons-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/commons-l

_______________________________________________
Commons-l mailing list
Commons-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/commons-l