Category talk:Commons category Wikidata tracking categories

Request for comment: Removing Commons Category link
As of late has been removing commonscat links from scores, maybe hundreds of articles. I have had several dozen articles I watch affected. Case in point Commons:Category:Berkshire should really have been, Commons:Category:History of Berkshire, but it simply gets removed. I'm not sure under what policy or authority this is being done, but these edits are disruptive and should stop. A couple of times, discussions have started, User_talk:Mike_Peel/Archive_56 and User_talk:Mike_Peel. --evrik (talk) 03:40, 19 September 2020 (UTC)
 * The cases I have been removing are where the Commons category does not match the article in question, but may be slightly related to it. The example everik mentions is Berkshire Record Society, where there is *no* media on Commons about that article, so there is no need to link to Commons. Another common case is articles like William Button (1526–1591) linking to commons:Category:Marlborough - how is that useful? You can see the other cases in my contributions, I'm happy to double-check any of them and/or explain why they aren't relevant if anyone wants.
 * For context, this is part of the larger clean-up/synchronisation of Commons category links by using the Commons sitelinks on Wikidata (the ones I'm removing are from Category:Commons category link is locally defined). I much prefer adding the links to Commons to Wikidata, or changing them to better links - I only remove the link if it is mismatched and there isn't an appropriate one to link to instead. Unfortunately this is both the least pleasant and the more controversial part of the cleanup work. Thanks. Mike Peel (talk) 10:25, 19 September 2020 (UTC)
 * As of the last two years I am not contributing much prose; I don't know about Wikidata, and don't want to. When I occasionally edit, much of it is drama, and I notice on Watchlist things such as this deletion by Mike Peel, without changing it to c:Category:BSA motorcycles. Unhelpful to the project, without a replacement in real time?--Rocknrollmancer (talk) 11:51, 19 September 2020 (UTC)
 * Thanks for pointing that one out, I've added the correct commons category link to BSA motorcycles now (it was linking to the parent company category). Thanks. Mike Peel (talk) 12:06, 19 September 2020 (UTC)


 * Support automated category removal; would consider more info about problem I am ready to recognize a problem if one exists. There is limited info here. I am coming to this RfC through an automated request from the Feedback request service. I edit Wikipedia, Wikimedia Commons, and Wikidata, and my editing includes some reconciliation of categories among them. Because of my interest, I have noticed that Mike Peel has developed automated processes for managing these categories. Perhaps the scale of these automated edits is in the range of 100s of 1000s. In various places at various times, Mike has documented what he planned to do and called for comments. I have participated in several of these discussions and asked several questions. Before I started looking into this myself I imagined that matching categories ought to be an obvious process, and I guess that it is in 90%+ of cases amounting to at least 10s of 1000s of edits, but there are many odd cases. I generally support more automated tools in Wikipedia especially to perform tasks which are noncontroversial. When there is some friction, I support talking it through to establish good boundaries and precedent for what to automate. In this case Mike says that he is removing category connections when Wikipedia and Wikimedia Commons seem to not match, and to me, this seems like a reasonable edit to try and a reasonable conversation to have. One possible next step in this conversation is to make a subsection below listing 1-3 problematic link removals along with a sentence describing why the removal is problematic. If anyone has other ideas about advancing this conversation I would participate wherever the conversation goes.  Blue Rasberry   (talk)  20:00, 19 September 2020 (UTC)
 * Automated processes are fine when there are no variations, however that is definitely not the case here. For instance the aviation and aircraft group projects generally provide wikimedia links to the manufacturer of an aircraft pending the specific type having its own wikidata page, as means of tracking what images are needed. Simply deleting these is not only not helpful, but it is actively disruptive. Additionally in several cases, the appropriate directory existed with images, and only needed to be updated rather than pointlessly deleted.&#32;- NiD.29 (talk) 19:23, 25 September 2020 (UTC)
 * This isn't an automated process (although it would be a lot easier for me if it was!), it's semi-automated edits that I'm only making only after correcting as many categories as I can, for cases where the current commons category link doesn't match the article, and it's only for cases where I can't see an appropriate replacement category to link to. For aviation articles, it makes no sense to link to a Commons category for the manufacturer where there are zero photos of the aircraft (I think would agree with me here, as they regularly thank me for the commons category removals) - but if we can get photos of them or if I haven't spotted the correct category or individual photos, then that's great, and you should revert the edit to link to a specific Commons category (like you did here). But what good are edits like this where we have no photos of the aircraft? Thanks. Mike Peel (talk) 19:56, 25 September 2020 (UTC)
 * I do agree with your removals in aircraft articles. These are cases where some editor has added a link to the Commons page for the manufacturer, where there are no photos of the aircraft in question. It must be confusing for readers to click on the link and not find a photo of the aircraft, but a bunch of different aircraft. - Ahunt (talk) 00:06, 26 September 2020 (UTC)
 * Often we don't know we have images because many are unsorted or misidentified, particularly for early or rare aircraft, and in many cases, people will recategorize a type to put it under the manufacturer, but not create the appropriate subdirectory (or appropriate wikidata), so until someone has done that particular bit of maintenance, there is no link, but there is an image - and once it is seen (better there), then it is a trigger to create the rest. When all there is are subdirectories for the manufacturer, people incorrectly categorize them, rather than leaving them for someone with better knowledge of that specific sub-type and it is much more work to dig through finding what has been misidentified. Also, many aircraft are closely related and there are distinct pages for what constitute very slight differences, so it makes sense to allow the link to those differences. Aircraft are not people, or places that exist in isolation.&#32;- NiD.29 (talk) 21:10, 25 September 2020 (UTC)
 * I replied on your talk page, but figure I should also reply here. I think a lot of that work has to take place on Commons, and I don't think the Commons links to the aircraft manufacturers help. I'm guilty of doing the kind of thing you describe myself - if I photograph an aircraft then I will upload it to Commons under a general category and hope that others will come along and improve the categorisation later. I won't start at a Wikipedia article about a specific aircraft and then upload to the manufacturer category, though, since I wouldn't know the specific aircraft in the first place (or even worse, e.g., File:At Tenerife 2020 809.jpg, any help identifying that one would be appreciated!). Slight differences are more tricky, but I still think there can be a simple match between articles here and Commons categories (remodeling one or the other to match if needed). Perhaps in the past it made sense to have redundant links here to Commons, which presumably people checked every so often? But now there's much more motivation on Commons to link to Wikidata and then Wikipedia (because of the infoboxes), so there's less need now for those links. Thanks. Mike Peel (talk) 18:34, 26 September 2020 (UTC)


 * Question for Mike Peel, and an example It seems like your objection to these categories is when the category does not exactly match the topic of the article. But my question is, why does the template commonscategory even have a parameter "catname" that allows it to be linked to a differently named category? One case where I ran into one of these deletions by was Dr. Mary B. Moody House. The history of that house, and part of the reason it even is listed on the National Register, is because of the significance of Dr. Mary Moody who lived there. Pictures of the house and of Mary Moody herself are cross-used in the article on the house and in her bio at Mary Blair Moody. As a result when I uploaded some of those photos, I grouped them all in a category "Mary Blair Moody" which seemed logical, and linked the commonscategory template in both articles to this category. It seemed to me people reading about her, might be interested in seeing pictures of the house, and folks who were reading about the house might want to see pictures of its namesake. Also if you searched on the woman's name on commons, you'd find the category with all the pictures. Wins all around.


 * I assumed this was why the template had a category name in the first place, was for uses like this where interest in things overlap. Along comes Mike Peel and says no, by his standards the category must match the article exactly one-to-one. I don't know where in policy he has derived this, but I will confess that categorization is not an area of Wikipedia that I pay super close attention to, so please point it out if desired.


 * At first he deleted the template on the house article entirely, which I reverted. Then his response was to revamp the categorization of the photos so that the pictures of Moody and the pictures of the house are now disjoint sets. The end result is, now if you click the commons link on the biography, you ONLY get pictures of the woman. If you click the commons link on the house, you ONLY get pictures of the house. I think this is a non-optimal result for users of Wikipedia, and I don't see why Mike Peel wants to enforce this narrow view of how categories should be used on everyone else, when the template is clearly designed with other uses in mind. --Krelnik (talk) 22:36, 19 September 2020 (UTC)
 * The situation now where Mary Blair Moody links to commons:Category:Mary Blair Moody and Dr. Mary B. Moody House links to commons:Category:Dr. Mary B. Moody House seems fine. If they are as closely entwined a topic as you say, then why do we have separate articles here on enwp? On Commons, having the separate categories means that the categorisation and the infoboxes make sense, rather than mixing a human and a building. In the long run I would like to see the "catname" parameter deprecated, but that's a separate discussion (there was a related one I started on VPP a while back). Thanks. Mike Peel (talk) 09:08, 20 September 2020 (UTC)
 * I don't think you understood my point at all. I really think you are getting too lost in the mechanics of categories and Wikidata and forgetting that the goal here is to build an encyclopedia that is read by people. It's great to have categories that are consistent. But the world is a messy wonderful place and you're not going to be able to jam everything into perfectly categorized siloes in every single case. I think having that "catname" parameter on that template isn't just a naive mistake that undermines the category system, it's a deliberate realization of this fact. I think trying to make categories PRECISELY line up on Commons and Wikipedia, and match them to Wikidata items one-for-one in every single case is an obvious violation of one of the Five Pillars. --Krelnik (talk) 14:33, 20 September 2020 (UTC)
 * I understood your point (the topics are closely connected), however I don't agree with it for the reasons I said before. In general, while the different projects definitely have different levels of granularity, I think enwp is actually the least granular (there are more Commons categories and Wikidata entries than enwp articles, by far), and in most of the cases I've seen there can easily be a Commons category matching the enwp article, if there actually *is* relevant material on Commons to link to (which there isn't in most cases I've been removing). I don't think my edits go against the five pillars at all. I think the problem here (not you specifically, but collectively with those raising objections) is sadly WP:OWN. Thanks. Mike Peel (talk) 18:09, 21 September 2020 (UTC)
 * MP, you go at this task with a machete, when you should use a scalpel. --evrik (talk) 19:20, 21 September 2020 (UTC)
 * I'm trying to use pruning shears. I've already cleaned up the grounds (added as many sitelinks between Wikidata and Commons as I can - a few million so far, mostly by bot - these show up in the sidebar here), and redirected the good growth that was going in the wrong direction (changing links to redirects, deleted categories, or general categories where there was a more precise category available). I'm now trying to remove the dead growth (removing misplaced links to unrelated categories) in the hope that it will spur new growth (the creation of new categories, and reducing the backlog sufficiently so that we can potentially bot-add many new links). It's as precise a surgical action as I can make it, but you can't cut a branch with a scalpel. Thanks. Mike Peel (talk) 20:33, 21 September 2020 (UTC)
 * Yeah, no. The only reason I am here on this page is because most of your edits on pages I follow are not productive. Please stop. 20:49, 21 September 2020 (UTC)
 * On the contrary; these are good edits. Mike: please continue. Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 21:19, 25 September 2020 (UTC)
 * I think maybe I understand better what you're going for, but let me pursue some other examples based on articles I've worked on. Consider Bazoline Estelle Usher who has a handful of photos on commons categorized under a category named Bazoline Usher. She has a elementary school named after her here in Atlanta, I've been meaning to go over and take a photo of the sign or the front of the school to help illustrate her article. However, as you may know local elementary schools are very rarely considered notable enough to rate their own article. If I upload a photo of the school, if I understand you correctly, I should NOT put it in the category for the woman. So does that mean I must create a category for the school, which will likely NEVER merit a Wikidata item nor a Wikipedia article of its own, just to give me a way to connect the photo to her category on Commons? Keep in mind this category would likely end up having ONE photo in it, which might attract attention from editors doing cleanup of another kind.
 * Here's another example I just discovered this week, which is kind of the reverse of the above. Mining some old biographical dictionaries for photos, I found photos of Susan Denkmann Hauberg and John Henry Hauberg that might be worth uploading. Frankly, I think they are both fairly marginal on the notability scale and might not merit their own articles, however there is an article Denkmann-Hauberg House for which their pictures might be useful. It has a category on commons with several photos in it. Should I create separate categories for each of the people, again just to put a single photo in each, just so that they people don't end up in the category for the house? It seems like a whole ton of extra effort that's not really, in the end, worth it. --Krelnik (talk) 14:34, 26 September 2020 (UTC)
 * Thanks for the examples. They are a bit different than the first case, since they don't both have Wikipedia articles, so notability is a bit more of an issue. On Wikidata I don't think there's a problem with either example - you should create a Wikidata item for the school and the people, and link that to the photos (using on Wikidata, and  in structured data on Commons). I don't have a problem with you putting them in the existing Commons categories, but it might be a little better if you do create Commons categories for them, as then we can use the infobox on Commons to add context, and they can be subcategories as appropriate - however if there's only a few photos then it makes less sense (please try to take photos of the school building itself, and also be sure to check the copyright issues around photographing signs). Here on enwp, though, I still think you'll only need one commons category link per article, to the matching commons category, which is what this discussion is about. I hope that helps, I'm happy to clarify anything. Thanks. Mike Peel (talk) 18:25, 26 September 2020 (UTC)
 * Are you saying that is now policy that ANYTHING we upload a photo of on Commons must have a Wikidata item that uniquely identifies it? So in order to upload a photo, which itself is a multistep process sometimes involving copyright claims and so on, I have to go over to Wikidata and create items there too?  This seems onerous to me. Is this policy or just your opinion of how things should work? --Krelnik (talk) 19:04, 26 September 2020 (UTC)
 * No, it is not policy at all. I think it's a good way to work, particularly given how much Commons uses Wikidata information now (and this is only increasing), but it's entirely optional. TBH most of the photos I upload aren't linked to Wikidata entries (but in the long term I hope they will be). We're getting quite off-topic, though. Thanks. Mike Peel (talk) 19:09, 26 September 2020 (UTC)
 * UGH. Well, I think I'll go back to not thinking about categories anymore. This hurts my brain. --Krelnik (talk) 19:57, 26 September 2020 (UTC)


 * you're still doing this. Please stop. I feel this is not constructive editing. --evrik (talk) 21:05, 15 January 2021 (UTC)
 * Re ? The links I'm removing aren't useful - but perhaps new Commons categories could be created that would better match the articles? Thanks. Mike Peel (talk) 21:08, 15 January 2021 (UTC)
 * No, there is not enough content on the commons to justify a category. Looking at this page, there is not a lot of support for what you're doing. Remind me again what authorizes you to make hundreds of changes like this? --evrik (talk) 21:21, 15 January 2021 (UTC)
 * So if there's not enough relevant content on Commons, why link to it? Thanks. Mike Peel (talk) 21:24, 15 January 2021 (UTC)