Commons:Village pump

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

Shortcut: COM:VP

↓ Skip to table of contents ↓       ↓ Skip to discussions ↓       ↓ Skip to the last discussion ↓
🌐 Village pumps for other languages
COMMONS DISCUSSION PAGES
Welcome to the Village pump

This page is used for discussions of the operations, technical issues, and policies of Wikimedia Commons. Recent sections with no replies for 7 days and sections tagged with {{Section resolved|1=--~~~~}} may be archived; for old discussions, see the archives; the latest archive is Commons:Village pump/Archive/2024/09.

Please note:

  1. If you want to ask why unfree/non-commercial material is not allowed at Wikimedia Commons or if you want to suggest that allowing it would be a good thing, please do not comment here. It is probably pointless. One of Wikimedia Commons’ core principles is: "Only free content is allowed." This is a basic rule of the place, as inherent as the NPOV requirement on all Wikipedias.
  2. Have you read our FAQ?
  3. For changing the name of a file, see Commons:File renaming.
  4. Any answers you receive here are not legal advice and the responder cannot be held liable for them. If you have legal questions, we can try to help but our answers cannot replace those of a qualified professional (i.e. a lawyer).
  5. Your question will be answered here; please check back regularly. Please do not leave your email address or other contact information, as this page is widely visible across the internet and you are liable to receive spam.

Purposes which do not meet the scope of this page:

Search archives:

   

Start a new discussion

# 💭 Title 💬 👥 🙋 Last editor 🕒 (UTC)
1 Is renaming categories with an English name to local language names a good idea? 74 15 Jmabel 2024-09-10 06:20
2 {{Large image}} 7 5 ReneeWrites 2024-09-03 18:17
3 Massive backlogs are now the norm 33 16 Infrogmation 2024-09-09 02:26
4 Uncategorized categories, except infobox 18 3 Enhancing999 2024-09-07 11:33
5 Category descriptions 33 12 Adamant1 2024-09-10 00:58
6 When questioning "own work" 10 6 Multichill 2024-09-08 18:45
7 inconsistent database state 24 5 Herzi Pinki 2024-09-03 18:55
8 Disambiguating categories 7 3 Herzi Pinki 2024-09-04 13:49
9 Check copyright and remove if necessary 9 3 MGeog2022 2024-09-03 09:23
10 Speedy because "File has no source" but what they are saying is "I want to see a url" instead of what is listed 6 3 ReneeWrites 2024-09-05 23:31
11 Abusefilter to prevent uncategorised pages？ 7 4 Enhancing999 2024-09-10 06:42
12 Commons Gazette 2024-09 4 4 Enhancing999 2024-09-03 21:28
13 Japanese category 4 3 Immanuelle 2024-09-03 13:29
14 Have your say: Vote for the 2024 Board of Trustees! 2 2 Enhancing999 2024-09-03 13:02
15 Are amulets toys? 4 3 Jeff G. 2024-09-05 10:56
16 I have many photos of a ship, but I do not know which one it is 3 3 Jmabel 2024-09-03 18:07
17 Wikiview visual search 8 5 Jmabel 2024-09-07 14:15
18 CfD for categories of nonexistent WikiProjects 10 3 Adamant1 2024-09-04 19:56
19 Categories which should only contain media of a specific type 2 2 Jmabel 2024-09-05 12:42
20 Image Annotator not working 2 1 JotaCartas 2024-09-04 16:43
21 Wikitext weirdness in Commons:Editor's index to Commons 1 1 Jmabel 2024-09-05 12:39
22 Familytree 28 6 Jmabel 2024-09-10 06:40
23 The authorization information of two images is corrupted 2 2 RZuo 2024-09-06 12:46
24 Categorizing Newspapers by date (YYYY-MM-DD) 4 3 Enyavar 2024-09-10 05:38
25 Are court orders in scope for Commons? 10 4 DaxServer 2024-09-07 07:23
26 20@ Wikimedia COMMONS 2 2 Koavf 2024-09-07 15:35
27 Is there a way to click things in my or someone elses contributions and undo them? 6 3 Koavf 2024-09-07 15:34
28 Many images of book scans 28 13 Enhancing999 2024-09-10 06:50
29 Use of NoFoP-category template on broad categories 10 8 Jmabel 2024-09-10 06:52
30 Template:How to delete empty categories 6 4 Enhancing999 2024-09-10 07:01
Legend
  • In the last hour
  • In the last day
  • In the last week
  • In the last month
  • More than one month
Manual settings
When exceptions occur,
please check the setting first.
Women at the well, India, early 20th century. [add]
Centralized discussion
See also: Village pump/Proposals   ■ Archive
Template: View   ■ Discuss    ■ Edit   ■ Watch
SpBot archives all sections tagged with {{Section resolved|1=~~~~}} after 1 day and sections whose most recent comment is older than 7 days.

August 22

Is renaming categories with an English name to local language names a good idea?

Category:Heroes' Cemetery in the Philippines has been renamed to Category:Libingan ng mga Bayani, to "match Wikipedia and Wikidata", see history of this category. Though there are exceptions for English category names: "some proper names, ... and names for which the non-English name is most commonly used in the English language" (see Commons:Categories#Category names), I wonder whether such an exception applies to this category and whether this would be a good development for other category names that are now in English but might have other names in Wikipedia and Wikidata. Because I can understand the English name without a translation program, but not the name in Philippine (or the majority of other languages). @Seav: Can you give your opinion as well? JopkeB (talk) 05:43, 22 August 2024 (UTC)[reply]

If there is a Category Redirect, I see no problem. There is a problem with English names that are meaningless to local people. Like Our Lady of the Forsaken, that is a very common thing to find around my place, that means absolutely nothing to the Valencian local people, who are enterly familiar with either la Virgen de los Desamparados or la Mare de Déu dels Desemparats. I do use English category names, but I have to explain my wife once and again what is Our Lady of Good Health (la Virgen de la Salud) or Saint Anthony the Great (Sant Antoni del Porquet).
I think a more multilingual approach is required. B25es (talk) 05:55, 22 August 2024 (UTC)[reply]
Then I would prefer to have the redirect the other way around: let the English name be the category name and let the local name have the redirect. JopkeB (talk) 06:11, 22 August 2024 (UTC)[reply]
That's perfectly fine for us B25es (talk) 16:32, 23 August 2024 (UTC)[reply]
Hi! Thank you for bringing this topic up. I think the relevant policy is Commons:Language policy, which defers to a section: Commons:Categories#Category names. As you have stated, the important passage is:

Category names should generally be in English (see Commons:Language policy). However, there are exceptions such as some proper names, biological taxa and names for which the non-English name is most commonly used in the English language (or there is no evidence of usage of an English-language version).

I think my rename follows the "proper names [...] for which the non-English name is most commonly used in the English language". The English Wikipedia article title, Libingan ng mga Bayani, had been discussed and renamed a few times since 2009 between the English-translated name and the official native-language name (see the header on Talk:Libingan ng mga Bayani), with the latest move request in 2020 resolving to the native-language title. There is plenty of evidence that the native-language name is used in English-language sources (albeit sources in the Philippines, but then again, English is an official language of the country). Some examples of English reliable sources within the past year that use the native-language name: [1][2][3][4].
I consider this situation similar to categories like Category:Taoisigh which could be reasonably be actually named Category:Prime ministers of Ireland, but we're using the Irish name here in Commons (so far without any argument, I think?) and also in the English Wikipedia (w:Taoiseach). —seav (talk) 06:14, 22 August 2024 (UTC)[reply]
"Taoisigh" is apparently not a proper name of a person, like Mary Johnson, but the name of a function, in the rest of the world known as "Prime minister". So this is not part of the exception. AND it is clearly not conform the Universiality principle, which says that "Identical items should have identical names for all countries and at all levels of categorization." So, please make the redirect the other way around. JopkeB (talk) 08:43, 22 August 2024 (UTC)[reply]
"Taoisigh" seems a bit odd to me, I don't recall seeing that spelling (or hearing a pronunciation that matches that spelling), but as a native English-speaker, I'd consider "Taoiseach" entirely appropriate. The BBC, for example, pretty consistently use "Taoiseach". FWIW, Google gives 239,000 results for "Taoisigh", 5,820,000 for "Taoiseach", 545,000 for the phrase "prime minister of Ireland" and 419,000 for "Irish prime minister", which seems to confirm my instinct. - Jmabel ! talk 18:41, 22 August 2024 (UTC)[reply]
The plural of taoiseach is taoisigh, according to Wikipedia. --Geohakkeri (talk) 19:00, 22 August 2024 (UTC)[reply]
Interesting, I guess I'd never heard it in the plural, and certainly have no idea how to form Gaelic plurals. - Jmabel ! talk 04:06, 23 August 2024 (UTC)[reply]
So it is also OK to rename Category:Prime ministers of France to Category:Premiers ministres? And for all other countries alike? Then we might delete the Universiality principle as well. I am not a native English-speaker and I consider "Taoisigh" not appropriate because I have never heard it before, I am familiar with the name "Prime ministers" and I guess the same applies to the best part of non-native English-speakers. Why would there be an exception for Gaelic? JopkeB (talk) 04:30, 24 August 2024 (UTC)[reply]
Because in English when we talk about this person/office, we use the word "Taoiseach". Similar issue to Category:Tsars of Russia.- Jmabel ! talk 17:40, 24 August 2024 (UTC)[reply]
Is that true for other English-speaking countries as well, like the USA, Canada, Australia, New Zealand? And the word "tsar" is well known, also in other languages, and is about heads of state, not prime ministers. So I think this is not a good comparison. JopkeB (talk) 04:08, 25 August 2024 (UTC)[reply]
Definitely true for UK; for U.S. I've usually heard radio news reporters first use the word "Taoiseach", then define it once (e.g. "equivalent of a prime minister"), then use "Taoiseach". Again, I think that Google count speaks volumes: over ten times as many hits for "Taoiseach" as for "prime minister of Ireland". - Jmabel ! talk 04:51, 25 August 2024 (UTC)[reply]
"Chancellor" is also used for the German head of state. And "Teno" for the Japanese "king". Nakonana (talk) 11:48, 25 August 2024 (UTC)[reply]
So, it looks a little bit like a mess, for Germany even more than for Ireland (for Ireland there is at least a redirect, for Germany you have to figure out yourself what the category for federal prime ministers is). But luckily prime ministers of Japan‎ are still in Category:Prime ministers of Japan‎ (I could not find teno and japanese "king"). My concern are about (1) non-native English (and German) speakers AND (2) creators of templates and other technical solutions. Both groups need clear category names, with the same category name throughout the category tree. I think the Universiality principle is made for both groups. How can we apply this principle to categories for prime ministers of Ireland and the federal state of Germany (and perhaps other countries?)? JopkeB (talk) 16:27, 25 August 2024 (UTC)[reply]
I typed it wrong, of course, Tenno is a redirect. Maybe "Head of State" would work? Nakonana (talk) 17:16, 25 August 2024 (UTC)[reply]
Head of state is not the same as prime minister. Germany and Ireland both have a president as well as a prime minister of the (federal) government. JopkeB (talk) 05:02, 26 August 2024 (UTC)[reply]
Agreed that a prime minister is normally "head of government" but not "head of state". I can't think of a country with a prime ministerial system where the prime minister is considered head of state. - Jmabel ! talk 00:25, 27 August 2024 (UTC)[reply]
 Oppose Changing the name to the "native-language" for two reasons
  • 1. The idea that it's in the "native-language" now is totally ridiculous to begin with. According to Google 22.5 million people in the Philippines speak Tagalog. While 39.4 million speak English. Further the place being discussed here is a national cemetery within Fort Bonifacio. The national headquarters of the Philippine Army. So it's just more likely due to how many speak English then Tagalog nationally that they will speak English. Meaning the change will clearly reduce the number of people who will be able to find the category. I don't think it's simply enough to simply have a redirect in this case either. Otherwise you could justify renaming every category to the "native language" simply because redirects exists. That's not their purpose. Per the guidelines we have to go with the name whatever language has the most chance of being searched for and it's pretty clear that's English in this case.
  • 2. There's been multiple CfDs having to do with this exact issue in the last couple years and there was clearly no consensus from them to change the names of the categories to the "native-language" at the time. I highly doubt if Seav had pf started a CfD for this before changing the name that it would have gone anywhere. That's what they should have done instead of just unliterally changing the name based purely on how the place is named on Wikipedia. Regardless, it's pretty clear that there is no consensus to use "native-language" names for categories in cases like this one. I'm not sure what the circumstances around Category:Taoisigh versus Category:Prime ministers of Ireland, but "other stuff" isn't really a valid reason to make the change. Again, especially considering the outcome of prior CfDs, guideline, and fact that clearly more people speak English in the Philippines then Tagalog and this is the national headquarters of Philippine Army. It would be ridiculous to say that shouldn't matter "because Wikipedia article." --Adamant1 (talk) 06:37, 22 August 2024 (UTC)[reply]
To be fair, your first argument only works for this special case of a country where a lot of people speak English. It doesn't work for other categories. For example, sometimes I go through the Category:Media needing categories (Cyrillic names), and there I find a decent number of files that actually are properly categorized, except that all the categories are written in Russian. The Russian community has a bot that addresses the issue to some degree but automatically searching and replacing Russian-language categories for their English counterparts, but the process is not perfect and you still find a lot of "uncategorized" Russian media. The other issue with translated categories is that there are various ways to translate one and the same thing, and it's a pain to figure out what the English name of an already existing category is. Oftentimes I go to Ru Wiki, find the subject there, follow the link to the Wikidata item from the article, and then look up the Commons category on Wikidata. There are just too many ways to "translate" even something as simple as "площадь мира" (transc.: ploshad mira, lit.: square [of] peace). You can go by Category:Peace square, Kazan or Category:Peace Square, Krasnoyarsk with a capital S, or Category:Mira Square (Kaluga)‎ (like Google does), or Category:Mir Square (the word "peace" = "mir" without the genetive suffix "-a"; Mira Square is actually equivalent to "Peace's Square", and if you want to have "Peace Square", you need to drop the genetive in Russian too "Mir Square"*), or you could translate it as Category:Square of Peace, or you use the Russian word order Category:Square Mira, which is a construction that you can find in translations of "проспект мира" (prospekt mira / Peace Avenue), such as Category:Prospekt Mira (Kaliningrad)‎ — why even bother translating when you can transcribe instead? (Now we only have to agree whether it's prospekt or prospect — k vs. c, see Category:Prospect Mira in Lipetsk‎.) Or you can just translate it like in Category:Peace Avenue, Krasnoyarsk, or you can be like Moscow and create a grammatical language monster by keeping the Russian words while using English word order: Category:Mira Prospekt in Moscow — this one raises eye brows in English speakers and Russian speakers alike.

* This is done with "площадь Ленина" (transc.: ploshad Lenina, lit.: square [of] Lenin). While the suffixed "-a" is kept in "Mir-a" for some reason, it is dropped in case of "Lenin-a": Category:Lenin Square (Ufa)‎ instead of "Lenina Square". However if it's a street (улица Ленина / ulitsa Lenina), then Lenin can keep his suffixed "-a": Category:Lenina street (Irkutsk)‎... or not Category:Lenin Street (Gdov). Nakonana (talk) 20:51, 22 August 2024 (UTC)[reply]
* typo: "It doesn't work for other countries." Nakonana (talk) 20:54, 22 August 2024 (UTC)[reply]
It is a fine balance line between translating native names into English ones or keeping them in the original language (or as a transcription for languages that have other scripts than Latin). For me names of streets and squares may be kept in the original language, except perhaps for very well known ones, like Red Square in Moscow. So the same rule as for place names (where only names known in English should be in English in Commons categories). JopkeB (talk) 09:22, 24 August 2024 (UTC)[reply]
Do we have a reference for the English name or is it just a literal translation? File:2551Taguig City Landmarks 17.jpg is in English. Enhancing999 (talk) 06:43, 22 August 2024 (UTC)[reply]
In English the name literally translates to "Cemetery of (the) Heroes" (libingan = cemetery; mga bayani = heroes). I've seen that translation (with or without the definite article) and I've seen the variation "Heroes' Cemetery" as well. A couple more points: the official website of the cemetery is [5], which is in English but still uses the native-language name, and the law that established the cemetery is Proclamation No. 208, s. 1967, also in English, but the name is again the native-language name. —seav (talk) 07:03, 22 August 2024 (UTC)[reply]
 Oppose renaming but see meta:Community Wishlist/Wishes/Add machine translated category titles on WMC. Prototyperspective (talk) 09:19, 22 August 2024 (UTC)[reply]
I think in this case the Tagalog name is appropriate, with soft redirects from the maybe two most likely English-language names.
I'd also add: in practice, this varies a lot from country to country, and sometimes by region within a country. For example, for Catalonia, we pretty consistently favor Catalan names; for Romania, I've seen English-language names for things that it is hard to imagine anyone referring to that way, e.g. "Roman Square" for Piața Romană; it's like a Spanish-speaker calling New York's Times Square "La Plaza del Tiempo" or (even worse) "La Plaza de los Tiempos". - Jmabel ! talk 18:50, 22 August 2024 (UTC)[reply]
One example of that is how they call pizza "pie" in the New York City area. I'd prefer keeping Category:Pies for images of actual pies though. --Adamant1 (talk) 15:04, 25 August 2024 (UTC)[reply]
I think in English the name is w:Rio de Janeiro so it would not make sense to translate the name of the city. I do not know if there is an official or established name in English for this statue but Category:Statue of the Little Mermaid (Copenhagen) is in English and so is Category:Eiffel Tower. And Category:Denmark and Category:Brazil are also English. Should that be changed to local names too? I think the rule on Wikipedia is that articles is named by the most used name. So I think it would make sense to do the same on Commons. Just like it is Category:Bill Clinton and not William Jefferson Clinton. --MGA73 (talk) 07:08, 24 August 2024 (UTC)[reply]
@MGA73 the statue is "Christ the Redeemer", many know that. Unsure if using English language as the precedence will lead to the category being moved to "Category:Christ the Redeemer (Rio de Janeiro)". JWilz12345 (Talk|Contrib's.) 07:39, 24 August 2024 (UTC)[reply]
JWilz12345 I think that if there is no clear name for something then just leave the name as it is and make some redirects if needed. I would just hate to see if someone get the idea to rename hundreds of thousands of categories just to use local names in categories. --MGA73 (talk) 07:52, 24 August 2024 (UTC)[reply]
"Rio de Janeiro" also has a literal translation to English. Enhancing999 (talk) 07:55, 24 August 2024 (UTC)[reply]
@Enhancing999@MGA73 we are talking about the statue's name in English. Everyone in English-speaking world calls it "Christ the Redeemer". Little to none use "Cristo Redentor". It is understood, but I am surprised about claims here that there is no clear name of the statue in English. It is crystal clear: Christ the Redeemer. Of course, English-speaking world calls the city "Rio de Janeiro"! So: "Category:Christ the Redeemer (Rio de Janeiro)" is the most-fitting name. JWilz12345 (Talk|Contrib's.) 09:04, 24 August 2024 (UTC)[reply]
The concern is not so much this statue, but that people use the equivalent of "River of January" as it occasionally happens. Enhancing999 (talk) 09:09, 24 August 2024 (UTC)[reply]
Should that also be renamed to Category:Mga sementeryo ng militar sa Pilipinas? I think this is a strawman argument. That category name is not the official name or proper name of an actual entity unlike "Libingan ng mga Bayani" and so would run afoul of Commons:Language policy. —seav (talk) 08:47, 24 August 2024 (UTC)[reply]
The argument from MGA73 seems strange, but it's true that "Heroes' Cemetery in the Philippines" could suggest that it isn't a specific cemetery, but a class.
"in the Philippines" as a disambiguator is unusual, possibly incorrect. Enhancing999 (talk) 08:55, 24 August 2024 (UTC)[reply]
The question was "Is renaming categories with an English name to local language names a good idea?" And I think not. I think in general it is best to use English names if there is one. For example "Denmark" not "Danmark" and "Statue of the Little Mermaid (Copenhagen)" not "Statuen af Den Lille Havfrue (København)". You can always find cases where the name can be discussed and in these cases just leave the name the creator have chosen. --MGA73 (talk) 09:14, 24 August 2024 (UTC)[reply]
(1) The problem was the sample didn't quite match the question. If there is an actual name that can be referenced and that is in use, not a "river of january"-thing. (2) Your argument is about classes of objects, not specific objects. (3) there are casses where we don't use English names even for classes: Category:Betula pendula etc. Enhancing999 (talk) 09:23, 24 August 2024 (UTC)[reply]

Conclusions + proposals + questions

  • The question is: Is it allowed to rename categories with an English name to local language names?
  • Answers:
    • Local language names are OK for category names if:
      • there is no English name, or
      • there is a redirect from the local language name to the English name or
      • it meets the criteria of Commons:Categories#Category names (names for which the non-English name is most commonly used in the English language).
    • Categories with local language names are doubtful if:
      • it is not about a proper name, but about something else, like a function: Category:Taoisigh (prime ministers of Ireland) and Category:Chancellors of Germany (prime ministers of Germany) are in the rest of the world known as prime ministers. Though these local words may be well known in the UK, they are not in most of the rest of the world and they certainly do not meet the Universiality principle, which says that "Identical items should have identical names for all countries and at all levels of categorization."
      • names are hard to translate into English (then they should at least be transcribed to Latin script).
    • It is not OK if:
      • there was no CfD for (there was no CfD for this case)
      • the name is not in the language that has the most chance of being searched for (in this case that is English)
      • the name is not in Latin script.
    • Another solution is to put alternative names, in however many languages, into Wikidata.
  • Category:Heroes' Cemetery in the Philippines should not have been renamed to Category:Libingan ng mga Bayani.

Proposals

  1. Revert the renaming of Category:Heroes' Cemetery in the Philippines and let the local category have a redirect to the English one.
  2. Try to include the conclusions of this discussion in Commons:Categories.

@B25es, Seav, Geohakkeri, Jmabel, Nakonana, Adamant1, Enhancing999, Prototyperspective, MGA73, JWilz12345, Broichmore, and Immanuelle:

  • Do you agree with the conclusions?
  • Do you agree with the proposals?

--JopkeB (talk) 15:29, 30 August 2024 (UTC)[reply]

I thought we don't want "River of January" type of translations (easy to do)? The question for a reference for the past name of the category seems still unanswered. We do have it for the current category name (even in the category). We do seem to agree the "Category:Cemeteries in the Philippines"-category should remain in English.
 ∞∞ Enhancing999 (talk) 15:35, 30 August 2024 (UTC)[reply]
  • I think there is a big difference between "allowed" to rename and "a good idea" to rename. I understand "a good idea" as "we should" rename. I'm from Denmark and we have Category:Copenhagen and I think it should be left alone because there are more readers that know the city as Copenhagen than København. But if someone had named the category "Buy a harbour" then I would ofcourse agree that the original name was better just like I think "River of January" should be renamed.
I think both the conclusions and the proposals are fine. But I also live happily with the new name. Just don't start a mass rename of thousands and thousands of categories just to get local names. --MGA73 (talk) 16:01, 30 August 2024 (UTC)[reply]
  • Both conclusions and proposals make sense.
    Some understanding has to be used regarding those of us that not being proficient enough in English ignore how to say some things in English. Some categories written the best we can will pop up.
    Also when things may have an English name but it is not widely known even among native English speakers themselves, category names may be not as perfect as desired.
    I do appreciate when categories I have created are properly Anglisized, but category redirects from other languages are of great help.
    A long term plan/idea/intention to have something like Wikidata's Qs, where somebody writes iron, my wife reads hierro, and someone in Dodoma reads chuma, would be great. B25es (talk) 17:25, 30 August 2024 (UTC)[reply]
I agree Immanuelle ❤️💚💙 (please tag me) 10:22, 5 September 2024 (UTC)[reply]

I certainly disagree with the conclusion that we should call the Chancellor of Germany the "prime minister of Germany". The term "Chancellor" dates back at least to the 1870s (and may have been used in Prussia before that, I'm not sure). It is simply the correct word in English (the German being Kanzler). The Germans do not use the word Kanzler to refer to comparable positions in other countries (e.g. they would not refer to the Prime Minister of the UK as a Kanzler, they would use Premierminister. It is simply not the same title, even if the role is similar. - 19:11, 30 August 2024 (UTC) — Preceding unsigned comment added by Jmabel (talk • contribs) ‎‎ (UTC)

Maybe chancellor would be an appropriate term for historical figures who traditionally called that, but I don't anyone outside of Germany uses the term these days. Like I know Hitler was called a Chancellor in Germany during World War 2 but if I do a search for him Google there's plenty of sources, including Google's information panel that calls him a Prime Minister. But then conversely the Wikipedia article for the role is called "Chancellor of Germany." So...There should be some standard and at least IMO Prime Minister makes more sense for our purposes if for no other reason then universality. I don't think it would make sense or work if he had diifferent terms for political appointees based purely what they are called locally regardless of if there's a universally recognized term for the role or not. Categories mainly (if not completely) exist as a way to organize images and naming them based on universality is clearly the best to do that regardless of if its 100% acccurate in every single instance. --Adamant1 (talk) 20:41, 30 August 2024 (UTC)[reply]
"Chancellor Angela Merkel" (in quotes) gets 1,880,000 Google hits. "Prime Minister Angela Merkel" gets a mere 5,940. Rather more recent than Adolph Hitler (also held the office rather longer), I'd say a 30:1 ratio suggests that the term is quite current. And since both "Chancellor" and "Prime minister" are specifically English-language, this should be specific to English. - Jmabel ! talk 21:59, 30 August 2024 (UTC)[reply]
@Jmabel: With Hitler it's 60,600 results for "Chancellor Hitler" versus 6,550,000 for "Prime Minister Hitler" and as I've already said that's even with him being called Chancellor in Germany at the time. So it clearly depends. You'd have to agree it would be pointless to argue about which term should be used used based on who the particular head of state was at the time. "Well, my German head of state has 6 million results on Google for Chancellor and yours only has 60 thousand for Prime Minister. So clearly the category should be named 'Chancellor'!" Not to say you'd do that, but I do think it's a case where it would just be better to go with the university principle instead of trying to determine how to name the category based on a game of one upmanship over who's favorite term for a head of state gets more results on Google Search. --Adamant1 (talk) 22:47, 30 August 2024 (UTC)[reply]
@Adamant1: I'm pretty sure you didn't use the quotation marks in doing that Google search. Obviously, Hitler would often be mentioned on the same page as some prime minister. But with quotation marks I get and 59,700 for "Chancellor Hitler" and only 9,440 for "prime minister Hitler", and even the latter include some more-or-less false positives: right near the top I see things like, "To the British Prime Minister, Hitler complained…" or "On May 13, 1940, Churchill gave his first speech as. Prime Minister. Hitler invaded France…" - Jmabel ! talk 23:23, 30 August 2024 (UTC)[reply]
@Jmabel: Weird. I thought I had. It looks like I probably left out the leading quotation mark though. My bad. --Adamant1 (talk) 23:26, 30 August 2024 (UTC)[reply]
(Reaction to the disagreement of Jmabel, 19:11, 30 August 2024) The point is not "that we should call the Chancellor of Germany the "prime minister of Germany""; of coarse you may call anyone the way you or the other person likes it. Nor whether what has more hits in Google or any other search machine. The point is: how do we name (sub)categories in a structural and consistent way. The real question here may be: What is more important: the consistancy of the category structure (like the Universality principle prescribes) or the titles of the prime ministers of Germany and Ireland? And if you choose the latter: what other exceptions exist to the Universality principle? --JopkeB (talk) 09:12, 9 September 2024 (UTC)[reply]
@JopkeB: as long as there are appropriate redirects and appropriate descriptions on the category page, ultimately I can deal with it. In the specific case of German chancellors, though, perhaps then we want a generic Category:Prime ministers of Germany that exists only to be a parent category of Category:Chancellors of Germany, much as Category:Monarchs of France includes (directly) the Kings of France and has subcategories for the emperors, the monarchs of areas within present-day France, etc.? - Jmabel ! talk 06:20, 10 September 2024 (UTC)[reply]
If we are insisting that we use strict English for the cemetery in question (despite the Tagalog name being extensively used in English sources [see above]), then there is still the question of which English translation should be used. Should it be the original "Heroes' Cemetery", or "Cemetery of Heroes" or "Cemetery of the Heroes"? All three are acceptable English translations of the official name, but I prefer any of the two latter names as these more closely resemble the official name (this is like preferring "River of January" instead of "January River" for Rio de Janeiro). As for disambiguation, do we append "in the Philippines" as with the original category name, or use a parenthetical "(Philippines)" which seems to be the convention? —seav (talk) 22:32, 30 August 2024 (UTC)[reply]
At least in regards to the first part of your message, probably the best way forward at this point would be to revert the edit that changed the name and then someone can open a CfD to figure out what the best option is from there. It's pretty clear from this that the name shouldn't have been changed without there being a CfD first though and at least IMO figuring out what is to go with instead is probably out of scope of this conversation. --Adamant1 (talk) 22:55, 30 August 2024 (UTC)[reply]
I agree with Adamant1 about reverting. And, @seav, I intended "River of January" as a reductio ad absurdum, not a serious proposal. In English, the city in question is universally called either "Rio de Janeiro" or just "Rio". This is never turned into English words. - Jmabel ! talk 23:29, 30 August 2024 (UTC)[reply]
Yes, I know that "River of January" wasn't a serious suggestion, but I simply used it as an example of which of several possible English translations ought to be used if we are translating a name. — seav (talk) 00:39, 31 August 2024 (UTC)[reply]
I don't think anybody tried to do a literal translation of "Rio do J.", but occasionally I come across similar ones, where it's likely that Commons is the sole source for that name. I could name one or the other, but I don't want to embarrass the user who did try to do so. If you find a bad translation of mine, feel free to cite is as a negative example.
Obviously, I don't think there is an issue to provide a literal translations in the description.
For the cemetery, I agree we should have a proper discussion on CfD, especially as the previous name was the result of one.
 ∞∞ Enhancing999 (talk) 11:51, 31 August 2024 (UTC)[reply]
In the UK, the German prime minister is always referred to as the chancellor, as is his Irish equivalent the Taoiseach. Broichmore (talk) 13:18, 31 August 2024 (UTC)[reply]
In passing, I've just looked at Category:General Wali Kothi, who would believe this is actually a building aka Star house, Tara Khothe, Tara Khothee, andd Lucknow Observatory. There will be other names and spellings too. Never mind the Hindi variants, and counting. You can bet your bottom dollar, that the locals call it Star House, or the old Observatory, terms not officially used since 1947.
Almost every Indian category suffers from the same malaise, and exemplify reasons why, English is the standard here. Broichmore (talk) 13:26, 31 August 2024 (UTC)[reply]
  •  Comment, personally, I think that the solution is really simple, we need to reform the way category redirects and title displays work on a technical level. We simply change Wikidata's notability guidelines to include any and all Wikimedia Commons categories, then integrate each and every Commonswiki category with Structured Data on Wikimedia Commons (SDC), then have a robot automatically create redirects based on alternative names in different languages based on what is listed in Wikidata, then make title display names based on the settings of the viewer. Congratulations, now the Wikimedia Commons is truly a multilingual project rather than an Anglocentric one.
How do we prevent vandals from deliberately adding wrong terms and / or insulting words and phrases in translations? Simple, we limit who can add these translations, it's a special but very easily acquired new right at Wikidata that any administrator could bestow upon someone based on either a local (at Wikidata) or Commonswiki endorsement.
How do we integrate this at every level? Simply upgrade the MediaWiki Upload Wizard to properly integrate category redirects like Cat-a-lot already does, this way redirects become useful and someone can theoretically only ever contribute using French or Japanese without ever seeing a single word in English and not even notice that the Wikimedia Commons is in a language that they don't understand.
These changes would solve a lot of issues overnight, all we need is for Wikidata to slightly adjust its notability guidelines (something user "Mike Peel" and I have been advocating for for years), then the rest can be 90% (ninety percent) done by robots, and humans can simply do quality checks and focus on more important things like finding educational content and checking if that educational content is licensed in a free way. --Donald Trung 『徵國單』 (No Fake News 💬) (WikiProject Numismatics 💴) (Articles 📚) 07:40, 5 September 2024 (UTC)[reply]
This is probably a hot take but I've been a fan of how people use redirects for essentially everything on here. There's a point where it just kind of defeats their purpose and usefulness if everything is redirected multiple times and with multiple different variations of the word, concept, or whatever. Like create a redirect for something that clearly needs one, but do we really need redirects for every random misspelling of a name or redirects for every name in every niche language on the planet? Probably not. --Adamant1 (talk) 07:58, 5 September 2024 (UTC)[reply]
Every misspelling rather not, but "reasonable" alternative yes. That would definitely solve most category issues with "Lenin Street", "Lenina Street", "Ulitsa Lenina" and "улица Ленина". Wikidata entries don't usually have all that many alternative spellings listed. Even if we would just use an English category name and one native language name for a category, that would already help a lot. Nakonana (talk) 18:47, 5 September 2024 (UTC)[reply]
Insofar as additional spellings are valid, adding them to Wikidata is helpful all around. - Jmabel ! talk 18:55, 5 September 2024 (UTC)[reply]
Where do I sign? Nakonana (talk) 18:48, 5 September 2024 (UTC)[reply]
Donald Trung: That looks like a fine solution and proposal. Would you please add it to the Community Wishlist? JopkeB (talk) 08:48, 9 September 2024 (UTC)[reply]

Overall conclusions

We agree on:

  1. The renaming of Category:Heroes' Cemetery in the Philippines to Category:Libingan ng mga Bayani should be reverted.  Action JopkeB ✓ Done
  2. After the reverting, whoever wants a different name for this category can start a CfD (discussion).
  3. (For future cases:) Whoever wants to rename an English category name to a local name should start a CfD before the renaming.
  4. The proposal of Donald Trung to solve the problem with category names in different languages with a technical solution, like in Wikidata.  Action Donald Trung (to bring it to the Community Wishlist)

We do not agree on: Whether it is allowed to have exceptions to the Universality principle, for example: Should category names of the German and Irish prime ministers (now in German and Irish) be renamed to the general (English) name for categories about prime ministers? I suggest to move this discussion to Commons talk:Categories. If you agree:  Action (perhaps Jmabel?)
--JopkeB (talk) 09:44, 9 September 2024 (UTC)[reply]

Why should we start a category discussion to fix "River of January" problems?
 ∞∞ Enhancing999 (talk) 09:55, 9 September 2024 (UTC)[reply]
Because there are cases that might be obvious for you and some others, but not for everyone. JopkeB (talk) 10:18, 9 September 2024 (UTC)[reply]
Well, that's just your POV. I think everybody agreed we shouldn't have "River of January".
 ∞∞ Enhancing999 (talk) 10:20, 9 September 2024 (UTC)[reply]
Started the discussion now for "Cristo Redentor" (which the English world has a fitting name: "Statue of Christ the Redeemer"). Commons:Categories for discussion/2024/09/Category:Cristo Redentor (Rio de Janeiro). JWilz12345 (Talk|Contrib's.) 11:20, 9 September 2024 (UTC)[reply]
I would differentiate between category names for object classes and for individual objects. Class names, descriptive names and categories for universal concepts should always be in English. Category names for specific things (persons, official titles, places, institutions, etc.), on the other hand, should be permitted in the local language. There may be exceptions where the English names are well established and immediately understood in the language area concerned, but as a rule, arbitrary translations bring few advantages for international users and significant disadvantages not only for local users: For example, unambiguity is often lost in the case of political parties or parliamentary names; in the case of buildings the alphabetical sorting may change; in the case of squares, streets or churches, the assignment in the city map becomes a guessing game. And the systematic information is provided in English by the parent categories anyway, so this does not need to be repeated in the individual object category. Rudolph Buch (talk) 16:27, 9 September 2024 (UTC)[reply]

August 23

{{Large image}}

What file size on commons is now regarded as being very ‘’high-resolution’’, and should, therefore, be marked with {{Large image}}? Does this vary by file type. Broichmore (talk) 13:08, 23 August 2024 (UTC)[reply]

This is a good question... Some years ago, files with 50 megapixels were considered large. Today, I would see large images as images with higher resolutions than the common ones (70 megapixels and more), of course with respective level of details. AFAIK "large" refers to the amount of pixels, not to the file sizes in MB, GB or whatever --PantheraLeo1359531 😺 (talk) 17:42, 23 August 2024 (UTC)[reply]
Would that advice apply, anywhere in the world? This is the image that raised the question. It's less than 21 megapixels. I find it, just a tad slow if you fully open it up, where I am (UK). However it presents no problems on wikipedia, or the front end of commons. Broichmore (talk) 10:30, 24 August 2024 (UTC)[reply]
I would say, yes. When we compare this 21 MP image to the largest 5 or 1 %, then this image is of course not large (I would not consider it as large), and the largest 5 % are. As it is a JPEG, the filesize is also not large. If it is a TIFF with 32 bits per channel, then it may qualify as large image. Of course, having a low internet connection, it may take long to load, but the larger ones take longer of course. --PantheraLeo1359531 😺 (talk) 16:16, 24 August 2024 (UTC)[reply]
I'd put the line on photographs at over 50 MP or over 65 MP. Modern phones, even cheap ones, frequently have 50 MP cameras, and 100 or 200 MP sensors in smartphones seem to work best when taking 50 MP or smaller pictures.. Even professional cameras top out at 60 MP, with the exception of $4K Fujifilm cameras or $35K Hassleblads. So that's the line between normal photo and multi-shot merges or extremely high end equipment, IMO.--Prosfilaes (talk) 21:19, 25 August 2024 (UTC)[reply]
I wouldn't take the 50MP thing at face value. My phone has a 50MP camera and it's impossible to take sharp pictures with. I suspect they're basically just upscaled versions of the default 12MP camera, which does take sharp pictures. ReneeWrites (talk) 18:17, 3 September 2024 (UTC)[reply]
Isn't this about browsers freezing not actually image size? I dont think it should matter what the average image size is, only what does or does not freeze a browser. I guess someone should test various sizes and see. Bawolff (talk) 17:15, 28 August 2024 (UTC)[reply]

August 27

Massive backlogs are now the norm

Looking at Commons:Deletion requests, there are currently nearly three thousand deletion discussions that are past due for closure. Commons is teetering on being a failed project due in part to the lack of administrators. There simply is too much work for the available pool of administrators. This problem isn't getting better. It's getting worse.

So, why are we allowing new uploads to the project from anyone which just exacerbates this problem? The status quo is failing. --Hammersoft (talk) 14:03, 27 August 2024 (UTC)[reply]

Category:Items with disputed copyright information is another example of backlogs. And it's hard to tell how well new uploads are reviewed. --EugeneZelenko (talk) 15:14, 27 August 2024 (UTC)[reply]
This is a real problem. Aside from more administrators, what do you propose? What should be done to limit bad uploads (for copyrights or other reasons)? JopkeB (talk) 15:37, 27 August 2024 (UTC)[reply]
Maybe more agressive and pro-active use of CheckUser? Trade (talk) 17:18, 27 August 2024 (UTC)[reply]
I forgot where I read it, but I think CheckUser has to be extremely rare because of the nature of the thing. --Adamant1 (talk) 18:03, 27 August 2024 (UTC)[reply]
Then almost every other Wikiproject should have the same high rejection rate. Except they dont. Its just Commons Trade (talk) 20:44, 27 August 2024 (UTC)[reply]
I haven't participated in Wikipedia for at least a few years but from what I remember they have pretty high standards to. But then they also have way more users and bad actors because of that. Plus it's just seems easier to come up with evidence of socking on there. So it just makes sense that they would have a higher CheckUser rate. I know from my own experience that there's been at least a couple of times where I wanted to file a CheckUser report on here but ended up not doing one because there just wasn't enough of a paper trail in both cases to justify it. I feel like there probably would have been if it had happened on Wikipedia either time though. --Adamant1 (talk) 20:50, 27 August 2024 (UTC)[reply]
The larger Wikipedia projects are in my experience much more strict in almost all areas compared to Commons. ReneeWrites (talk) 18:12, 3 September 2024 (UTC)[reply]
There's been at least a couple of discussions lately about expanding and/or improving the criteria for speedy deletion. It seemed like none of them went anywhere at the time, but I think that would a lot with the backlog. As would making it easier for people to nominate images for speedy deletion to begin. As I guess its not an option on the side panel right now outside of clear copyright violations when it really should be. We could also use more admins, but I the standards are currently to high and I doubt they would work in that area anyway. Really you could argue current admins not dealing with deletion request is as much of or more of an issue then us just not having enough of them. There's no point in having more admins if they aren't going to work in the areas where they are most needed to begin with though. --Adamant1 (talk) 15:54, 27 August 2024 (UTC)[reply]
The backlog was much lager some month ago. If we need an urgent action we should block IP edits to prevent their spam and get time to work on deletion requests. GPSLeo (talk) 15:56, 27 August 2024 (UTC)[reply]
I had thought about that but it doesn't seem like there's any support to block IP edits on here. I totally agree that they should be blocked though. IP editors are a massive time suck in general. --Adamant1 (talk) 16:23, 27 August 2024 (UTC)[reply]
How exactly is blocking IP's supposed to help in any way? The issue is with people uploading copyvio. IP's obviously cannot upload anything in the first place Trade (talk) 17:16, 27 August 2024 (UTC)[reply]
But if we need 6 admin hours per day to check and clean up IP edits we can not use these 6 hours to handle deletion requests. It is simply a question of admin time availability. GPSLeo (talk) 17:50, 27 August 2024 (UTC)[reply]
Aside from what GPSLeo says I work in the area a lot and IP editors tend to either create a lot of meritless deletion requests or troll DR discussions. Obviously it would be a lot easier to deal with the back log if there weren't as many clearly pointless deletion requests being created to begin with. --Adamant1 (talk) 17:53, 27 August 2024 (UTC)[reply]
It's a bit odd for random IPs new to Commons to be so familiar with the DR procedure Trade (talk) 20:46, 27 August 2024 (UTC)[reply]
Most of the time their comments amount to something along the lines of "not own work?" Otherwise I'd agree with you. There are some times where there's clearly socking involved or the IP editor is probably a previously blocked user though. Both of which is all the more reason to block them IMO. It's purely conjecture on my part, but I highly doubt most IP editors on here are participating in the project that way for legitimate reasons. --Adamant1 (talk) 21:13, 27 August 2024 (UTC)[reply]
Agree with Adamant1. I also see too often violation and other non-constructive contributions by IP adresses, also in discussions. It takes a lot of time to fix them. Why do IP's have nearly the same rights on Commons as accounts have? My proposal: Restrict their rights. They should just be allowed to read things on Commons and have no possibility for violation. They are already excluded from uploading, why not from editing as well? If they want to start or join discussions, nominate files for deletion and so on, then they should use an account, like the rest of us. JopkeB (talk) 06:36, 28 August 2024 (UTC)[reply]
The deletion requests would be much easier to close if more people participated. Trade (talk) 17:19, 27 August 2024 (UTC)[reply]
Less than 3.000? Good job then, just a few years ago it used to be over 6.000. --Rosenzweig τ 15:59, 27 August 2024 (UTC)[reply]
3000 sounds like what @Yann does in a blink. ;) Thanks btw.
 ∞∞ Enhancing999 (talk) 20:48, 27 August 2024 (UTC)[reply]
Is the problem really getting worse? As in, are the various backlogs growing faster than they're shrinking? This argument came up on enwiki with regard to unsourced articles recently and it turns out that though there are an enormous number of those, they are being taken care of faster than new ones are being created.
I suspect something similar is going on here, especially since there is over a decade's worth of files to go through. To use another backlog as a comparison point, there are roughly 128,000 uncategorized files uploaded in 2024. Which is bad, but more than 128,000 files have been removed from the uncategorized backlog from 2021 and 2023 alone. Gnomingstuff (talk) 00:24, 28 August 2024 (UTC)[reply]

" Commons is teetering on being a failed project" if backlogs mean a project is teetering on the edge of failure, the virtually all Wikimedia projects have been teetering on the edge of failure since their inception. Yes, it would be good to do better. No, this is not a crisis, let alone a death knell.

Do not forget you need a native English speaker who is at the same time a railway fan (not necessarily a railway expert, but has knowledge above the average) to close this discussion. I just looked at it and I can not close it. I can not say a difference between "close" and "closed" in this case. Ymblanter (talk) 21:02, 27 August 2024 (UTC)[reply]
More admins would help, the issue is really active admins. Among the 183 admins, only 129 have done any admin action during the last month, and only 64 have done more than 20 admin actions during the same period. Yann (talk) 21:10, 27 August 2024 (UTC)[reply]
A CFD is not a Deletion request, so I think it is out of scope of this discussion. CFD's cannot be closed just by administrators, but every experienced editor can do so. JopkeB (talk) 06:07, 28 August 2024 (UTC)[reply]
When deletion discussions remain open essentially forever, when backlogs remain in the thousands across a number of types, when copyright violations are allowed to sit on the project indefinitely, then yes I would call this a failed project. We are not able to effectively police ourselves, and the incoming level of problematic files overwhelms existing ability to deal with it. --Hammersoft (talk) 17:36, 1 September 2024 (UTC)[reply]
maybe you can become a com:sysop and help? would you accept a nomination? RZuo (talk) 18:55, 1 September 2024 (UTC)[reply]
Frankly, I'm not active enough. I've seen various requests here, and noted concerns about activity levels. I doubt I'd meet expectations. Regardless, adding another administrator to the mix, especially at my activity levels, isn't going to help. I have other reasons as well, but this is really getting out of the scope of this discussion. --Hammersoft (talk) 19:18, 1 September 2024 (UTC)[reply]

I don't think we're teetering on anything, but we should all be able to agree the backlog is too big. Here's another reason it exists: DRs are often hard. Most admins aren't IP lawyers in one country, let alone every country, and there are a lot of really rough calls. Scope-based and people-based debates are also often hotly contested and difficult to assess. Beyond that, AfD on enwp is contentious and results in a lot of grief for closers, especially from newer users who don't understand the process. On commons, even many of the experienced users (or experienced users from other projects) routinely misunderstand the basics and make life sadder for commons admins. Back to the main point, however, there are definitely a lot of DRs that aren't that complicated which wait a long time to be closed, too, so more admins would help, at least. — Rhododendrites talk |  17:59, 1 September 2024 (UTC)[reply]

The same thing applies to CfDs. It just took me 2 days to deal with 4 of them because they all involved moving thousands of files and deleting hundreds of categories. Then there's ones that are extremely simple and can be closed in a couple of minutes. But its near impossible to deal with the later to any meaningful degree because of the time it takes deal with the former. Then there's a bunch that are clearly just complex to deal with for whatever reason. But its such a big mishmosh. Anyway, these things don't really scale. For them to be managable you'd need to have everyone on here participate and inforce clear time limits on how long the discussions can be open for. As well as cnsistantly keep the back logs down to a week or two. Otherwise its just going to be an unworkable mess. Its hard to say something should be closed a certain way after a week or two when it has a low turn out and an unclear outcome though. Sometimes it takes years just for someone to comment. So I don't know. There really is no single, easy solution. --Adamant1 (talk) 20:46, 1 September 2024 (UTC)[reply]
Just adding an example to go with my "DRs are often hard" comment above. I started looking through the oldest DRs and saw Commons:Deletion requests/Files in public domain featuring various Indian film artists post-1945. Taking action on that DR would involve a few dozen images and an understanding not just Indian copyright or US copyright but the interplay between the two. Even if we get more admins, the number of people who would feel confident weighing in on something like that are very few. Let's not have a discussion here about the correct outcome, though -- I'm just highlighting that these are hard, and while the act of closing a discussion may not take much time, attaining the level of understanding required to parse such a case would be very time consuming. It's no wonder nobody wants to do it. I wonder if there's a potential project for the WMF to undertake that would effectively create charts like COM:HIRTLE for media published outside the US and uploaded to Commons. Maybe even an interactive tool. — Rhododendrites talk |  20:56, 1 September 2024 (UTC)[reply]
I think for reusers and uploaders the open request signals there is a complex issue. Highly problematic uploads get nuked within hours.
 ∞∞ Enhancing999 (talk) 22:15, 1 September 2024 (UTC)[reply]

August 28

Uncategorized categories, except infobox

Report Special:UncategorizedCategories provides additional information to Special:UncategorizedCategories. Ideally it would be updated daily. Regular updates seem to make it easier to maintain it.

Report UncategorizedCategories with infobox has categories that don't appear there as they are in Category:Uses_of_Wikidata_Infobox_with_no_item, meaning, there is only an empty {{Wikidata Infobox}} on the category, but no other parent categories. Eventually the two reports might be merged.

For both Intro Special:UncategorizedCategories lists ways to take care of these categories.
 ∞∞ Enhancing999 (talk) 11:28, 28 August 2024 (UTC)[reply]

Awesome! This is very useful. It implements one part of the two I requested here. Could you also create a report for categories with only redcategories? Then all categories practically without categorization would show up on some report. A next step thereafter I think would be to have some bot identify cats missing cats and making e.g. Suggested Edits category suggestions, for example using data of the corresponding WP article in a way that makes these suggestions constructive. Such bots would be especially useful for experienced users adding categories to the categories on these reports and greatly reduce the time required for that and the backlog size.
Some change proposals:
  • could you rename "top users" to something like "users creating most uncategorized categories"? (this feature also ties in somewhat with the gamification-like feedback mechanisms discussed here)
  • it would be good to link to these reports from here and other places where Special:UncategorizedCategories is linked from
  • it would be good to link to these reports from Special:UncategorizedCategories itself
Prototyperspective (talk) 12:46, 28 August 2024 (UTC)[reply]
Thanks for the input. Feel free to link the reports from places you think it's useful. I made an edit request to add it to Special:UncategorizedCategories at MediaWiki talk:Uncategorizedcategories-summary.
For actual implementation of additional reports, you might want to use Commons:Bots/Work_requests. Special:WantedCategories exists and seems to be long (5000 categories wanted 14 and more times). It also has the advantage that it gets updated when a category is created.
Maybe Category:Pages with coordinates also includes categories that aren't categorized otherwise. Possibly a few other categories are similarly suboptimal.
 ∞∞ Enhancing999 (talk) 13:01, 28 August 2024 (UTC)[reply]
Briefly, as for Special:WantedCategories those only show the redcats but not categories with only redcats. This is very different and doesn't really address the same problem but if it did then in a very different way. For example, many of these cats only have a nonexisting category that is used only once (namely in that category) while this shows the top XYZ by number of use and currently only shows categories with at least 14 items. Prototyperspective (talk) 13:14, 28 August 2024 (UTC)[reply]
Oh, something like this (except that 4 of 13 exist).
 ∞∞ Enhancing999 (talk) 13:21, 28 August 2024 (UTC)[reply]
That's a separate subject and not necessarily a problem. Prototyperspective (talk) 13:24, 28 August 2024 (UTC)[reply]
If it's something else, do you have a sample?
 ∞∞ Enhancing999 (talk) 13:26, 28 August 2024 (UTC)[reply]
It's categories that have nothing but nonexisting categories. I don't have an example now but could add one if I find one but I'd suggest simply categorizing a category missing categories in such a way for a test-case. There's many of these. Also note that categories having nothing but nonexisting categories and Wikidata infobox meta categories should also show up in some report. The two examples I listed here (this and this) apparently are not included in your report so I think it still needs further work so it also includes these. Prototyperspective (talk) 13:34, 28 August 2024 (UTC)[reply]
Oh, it seems with "nonexisting categories" you mean categories that have the "hidden category" attribute set (visible on the second line). Generally all categories that are not topical categories. Category:Sinauli has four such categories (Uses of Wikidata Infobox, Uses of Wikidata Infobox with no image, Uses of Wikidata Infobox with maps, Pages with coordinates). The infobox sets 3 of them, MediaWiki the last one.
 ∞∞ Enhancing999 (talk) 13:40, 28 August 2024 (UTC)[reply]
No I don't!
The categories it has are all just meta categories set by the Wikidata Infobox and I thought that's what your report is about. Prototyperspective (talk) 13:43, 28 August 2024 (UTC)[reply]
The second report is about empty infoboxes only.
 ∞∞ Enhancing999 (talk) 13:47, 28 August 2024 (UTC)[reply]
Thanks for the info, all the samples I looked at didn't have any normal categories so can this report also be used for that and if not the title of the report is flawed and doesn't match what the report is about. It could be named "Report UncategorizedCategories with empty infobox" but I'd suggest changing it so it also shows items with non-empty infoboxes or creating and additional report "Report UncategorizedCategories with non-empty infobox". Prototyperspective (talk) 13:54, 28 August 2024 (UTC)[reply]
The intro defines the exact scope. For your request, maybe the infobox itself can determine if there are any other categories available. It's conceivable that a category with an infobox is correctly only in "hidden cats", so checking those isn't sufficient.
 ∞∞ Enhancing999 (talk) 14:00, 28 August 2024 (UTC)[reply]
What do you mean with "isn't sufficient"? There simply are many cases like the two examples that are practically without categories / missing categories but don't show up in reports because they have a few meta-categories set by the Wikidata Infobox. If there are cases where this is fine these are very rare and one could think about what to there once there are some examples of that where one can see how such can occur and maybe how they could be excluded from the report. The infobox itself can't determine that, it can at most add or suggest a few more categories but that's basically a separate subject and doesn't solve anything. Prototyperspective (talk) 22:37, 28 August 2024 (UTC)[reply]
BTW the "last user who edited the category" may not be the one who created it. Also, for the second report, it may have been disconnected from Wikidata since the last edit.
The default sort is by page_id, meaning newer categories appear first. The categories can be sorted by last edit date.
 ∞∞ Enhancing999 (talk) 13:14, 28 August 2024 (UTC)[reply]
I'd say that the report is quite useful. Few (if any) false positives, so pretty much anything that shows up there needs some work. - Jmabel ! talk 18:48, 29 August 2024 (UTC)[reply]
Thanks. Updates are a bit complicated as infoboxes don't necessarily update when a category is connected to an item.
Commons:Report Special:UncategorizedCategories is now down to reasonable length (469 categories):
  • only 52 categories with 10 files or more (cat_size>9)
  • only recent empty categories (cat_size=0)
  • most users with many creations on the list have been made aware of the issue
∞∞ Enhancing999 (talk) 10:03, 31 August 2024 (UTC)[reply]
Special:UncategorizedCategories is now at 8, including 5 categories that can be deleted once all files in it are (likely) deleted. Interestingly, it gets new entries every day. either by people who omit parent categories, people who don't know how to request speedy deletion or people who create categories that need (English) descriptions to be of any use to a more general public.
Commons:Report_UncategorizedCategories_with_infobox has progressed as well. Few larger categories (cat size >9) remainin, all empty ones cleared.
 ∞∞ Enhancing999 (talk) 11:33, 7 September 2024 (UTC)[reply]

August 29

Category descriptions

I'd like to establish some community consensus on what constitutes an appropriate amount/use of category descriptions. Categories that have to do with North Brabant often have large, self-referential descriptions with all manner of interwiki linking and use of external linkage that is not appropriate, for example Category:Geertruidenberg or Category:Heusden, North Brabant, or any of the places linked in the template above. Compare that to famous cities like Category:London or Category:Chicago, which have minimal descriptions by comparison.

Then there's subcategories. Category:Van Goghkerkje mentions it's at the Vincent van Goghplein 1, 4881 DG Zundert in the municipality of Zundert in the province of North Brabant in the south of the Netherlands. This information is then repeated in almost every single subcategory, including the metacat Category:Van Goghkerkje by year.

Category:Streets in Geertruidenberg mentions in the category description that this subcategory is for pictures of streets in the municipality Geertruidenberg in the province of North Brabant in the south of in the Netherlands. Category:Nature of Bergen op Zoom mentions in the category description that this subcategory is for pictures of nature and nature reserves in and around Bergen op Zoom in the province of North Brabant in the south of in the Netherlands. The category Category:Geography of Moerdijk mentions in the category description that this subcategory is for pictures of the geography in de gemeente Moerdijk in the province of North Brabant in the south of in the Netherlands.

Thousands of categories are like this, almost exclusively those in North Brabant. I'd like to start cleaning these up, but before I do I want to establish community consensus on where the line is drawn when something turns excessive so I know what to trim and what to keep.

My opinion is this:

  • No repeating of information that can be found in the category name ("The category "Streets of Geertruidenberg" is for pictures of streets of the municipality of Geertruidenberg...")
  • No repeating of information that can be found in an infobox
  • No repeating of information that can be found in a parent category
  • No external linking to personal websites, or other sites that would not be considered reliable sources on other Wikiprojects

ReneeWrites (talk) 07:36, 29 August 2024 (UTC)[reply]

I guess most categories where the description would be relevant have {{Wikidata Infobox}}, which has all necessary information. Ymblanter (talk) 09:11, 29 August 2024 (UTC)[reply]
Maybe we miss something, but both Category:London and Category:Chicago currently have tons of category description, most can be found elsewhere, compared to Category:Van Goghkerkje. The later mentions an essential point that seems odd to mention without including a basic description before including stating the category name in the description itself.
Not sure how Category:Geography of Moerdijk can be compared negatively to Category:Chicago except when there is some bias involved. Stating the topic in local language(s) is fairly important.
Category:Chicago seems relatively bad as it has a large seal image in the center. It mentions "largest city in Illinois," which is marginally helpful.
A nice thing about Category:London is that it has that collapsed list that helps sorting. I find such Commons specific pointers fairly important.
 ∞∞ Enhancing999 (talk) 12:02, 29 August 2024 (UTC)[reply]
I was comparing the categories of Geertruidenberg and Heusden with those of London and Chicago, the latter of which have very succinct descriptions. London's is just one line of text. This is not about the navigation templates.
My critique of Category:Van Goghkerkje is that this information is repeated in most subcategories like Category:Van Goghkerkje by year and Category:Van Goghkerkje in 1969, etc.
And I was not comparing Category:Geography of Moerdijk to any of the above categories. I was making a stand-alone critique (that applies to Category:Streets in Geertruidenberg and others) that the category description just says what's in the category name, and that this is superfluous. ReneeWrites (talk) 12:26, 29 August 2024 (UTC)[reply]
Sorry about that then. Contentwise Category:Geertruidenberg or Category:Heusden, North Brabant don't compare that badly, though I don't think there should be two descriptions in Dutch and English.
If there are Wikipedia articles on the topic, references would generally not be needed. There was some discussion about this at Commons:Categories_for_discussion/2024/07/Category:Harp_Guitar_Form_3a.
The repetition at Category:Van Goghkerkje by year seems suboptimal, but that applies to the entire "by year" tree as well.
 ∞∞ Enhancing999 (talk) 12:52, 29 August 2024 (UTC)[reply]
There seem to be two types of descriptions (maybe one more popular than others), for a sample Category: "ABC":
  • 1. "ABC" is a ..
  • 2. This category is about media/photos/images related to/about "ABC", a ..
Personally, I prefer (1.), but 2. is not uncommon.
Content-wise, there are three types of descriptions about "ABC":
  • 1. no description
  • 2. Just repeat the title literally ("ABC")
  • 3. State something about ABC
Personally, I prefer 1 or 3.
 ∞∞ Enhancing999 (talk) 12:38, 29 August 2024 (UTC)[reply]
I agree with Ymblanter, wikidata supplies this info.
In the case, of all things North Brabant wikidata may be insufficient. I more than suspect, what you have here is an editor set in their ways since 2008, who finds Wikidata impenetrable, or even superfluous and persists without it. The cats are cluttered, I agree, and few (if any), will use these links. However, there’s no harm done here.
After all, this is supposed to be a project that anyone can edit.
I've beaten this drum about over complication, before, and I contribute to Wikidata (4000 edits), and it falls on deaf ears.
There's too much of high tech people here, making subject matter complicated and difficult to edit. I'm still waiting for Category:Gartenlaube (Magazine)'s open architecture to be restored. All new uploads there, are incompatible with the established closed off format.
Even today I discovered a source template for a major library that is no longer viable, because the library changed its catalog system rendering our source template into a link to an enormous pile of link rot. I fail to understand how people, make templates or such, then not put in place the mechanisms for continuous maintenance, or just walk away from them. Broichmore (talk) 13:21, 29 August 2024 (UTC)[reply]
I think I disagree with some premises here. Respectively, to the bullet list of opinions given by User:ReneeWrites:
  1. Don't forget the multilingual aspect of this. Repeating the category name in other languages can be very useful.
  2. I get what you say about the infobox, but sometimes a tight piece of prose is easier to skim quickly.
  3. Not everyone will be navigating down the hierarchy. I think it would be ridiculous, for example, for someone navigating up from a village to have to navigate many, many layers up the hierarchy to find out what country it is in.
  4. I could perfectly well accept the last point, with one possible exception: on topics where there is a serious limitation on commercially available material, it can be useful to have a link to a trove of NC content on (for example) Flickr. Not sure whether that is even an exception, maybe more of a clarification.
I'd also add: I think somewhat longer than usual descriptions are often useful for topics not likely to get a Wikipedia article. Example: Category:Court in the Square.
Further: the main problem with the subcats of Category:Van Goghkerkje is that there are this big batch of "by year" categories, mostly with one or two photos, and with little or no differences from year to year to suggest that is a useful way to subcat this. - Jmabel ! talk 14:03, 29 August 2024 (UTC)[reply]
Re. "the multilingual aspect" - surely that should be handled by Wikidata? Wikidata already has extensive support for multilingual desriptions of entities, and those descriptions should get pulled in by the Wikidata infobox. Writing those descriptions locally on Commons shouldn't be necessary. Omphalographer (talk) 19:16, 29 August 2024 (UTC)[reply]
Maybe you could explain how it should be happening for the samples where Renee writes that [Dutch] is super-flous.
 ∞∞ Enhancing999 (talk) 19:25, 29 August 2024 (UTC)[reply]
Regarding 4, I think this is a good point and a good way to add nuance to the topic. I'm not against removing all external links, though I didn't really clarify what I meant with "personal websites", and I think the exception you mention is reasonable. The website in particular I was thinking of was https://breda-en-omgeving.nl/ which you can find linked numerous times in the categories above (6 times in Geertruidenberg, 7 times in Heusden, as well as many other categories linked in the navigation template at the top). I'll also ping @Prototyperspective: because this is relevant to a thread below. ReneeWrites (talk) 11:45, 30 August 2024 (UTC)[reply]
  •  Support ReneeWrites opinion about it. Although I think Jmabel makes some good points, but I don't see them as mutually exclusive or applicable in every situation. Just to give an example from Jmabel's comment, don't forget the multilingual aspect of this, I'd say it depends on the topic of the category and what level down it is. For instance if we're talking about shoes and the person is browsing through Category:Shoes to find images of them then I don't think it's necessary or helpful to have the word "shoes" translated into every single language possible just because this is a multilingual project or whatever. It's a pretty good bet that most users know what a shoe looks like and has heard the word in English before. So there's zero reason to translate it or to have a description, in English or any other language. "Shoes are things you wear on your feet." No really? We don't need that in English, let alone 10 other languages.
Maybe you'll say that's a bad example since the category for shoes doesn't have a description to begin with though. But there's plenty of categories for extremely obvious universally known subjects out there that have totally pointless descriptions in multiple languages. So essentially I agree with ReneeWrites opinion about it. Although I think Jmabel makes a few good points, but multilingual thing shouldn't be a free pass to create descriptions that are otherwise totally pointless for whatever reason. --Adamant1 (talk) 17:22, 29 August 2024 (UTC)[reply]
@Adamant1: Again, though, you are assuming they got there by following the category tree from a more abstract category. They could just as easily have arrived from a category on an individual photo. - Jmabel ! talk 18:51, 29 August 2024 (UTC)[reply]
@Jmabel: Sure, but if someone gets to Category:Shoes from an individual photo of something that's not a shoe then there's other problems involved that categories having descriptions or not has nothing to do with. The only way to deal with people confused about where they are in the category structure at any given time is to categorize images properly. You can't make up for or fix that by putting "chaussure" at the top of a category. --Adamant1 (talk) 18:59, 29 August 2024 (UTC)[reply]
@Adamant1: "other problems involved" Well, yes. And if they are, for example, a Chinese-speaking editor, it would be useful if they don't have to to running way up the category hierarchy to work out that it was an error.
Some of this can be solved if there is a corresponding Wikidata item, but the more narrow the category, the less likely to have a Wikidata item, so the more important mere translation of the category name may be. - Jmabel ! talk 19:53, 29 August 2024 (UTC)[reply]
@Jmabel: I don't even disagree with that, which is why the used an extremely general topic like shoes as an exmple. It's certainly a different thing altogether the further down in the categories some goes. I'm not sure what the solution to that is but I still think ReneeWrites' proposal still makes sense with broader topics. Maybe there could be a cavit to it about how to handle categories for more obscure topics though. --Adamant1 (talk) 20:00, 29 August 2024 (UTC)[reply]
 Comment Generally agree except for the point 4. For example a link to a github repo may not be a reliable source at WP but still relevant and useful in a category about the software without a Wikidata infobox. Another problem is that several links in the Wikidata item don't show up in the infobox and something should be done about it there instead of adding a link also to official website (example: links to mediawiki.org in the "Multilingual sites"). Another caveat is that I don't know how Web search engine indexing works and that "No repeating of information that can be found in a parent category" is ambiguous – if you mean useful info that is not self-explanatory should be excluded in a category just because it's already in the parent category then I disagree (e.g. people may have gone directly to that cat from search results or a file instead of having seen the parent cat earlier). Prototyperspective (talk) 19:46, 29 August 2024 (UTC)[reply]
I don't think Github is a very good example. A Github page isn't a personal website, and it's perfectly acceptable to use as a source on Wikipedia (there's even a template for it: Template:GitHub). Github can't be used as a source to establish notability however. ReneeWrites (talk) 08:03, 30 August 2024 (UTC)[reply]
I think it's a good example because it illustrates well how websites considered nonreliable can be very appropriate and useful to have there. You agree that it can be appropriate so it's evidently a good example. That Wikipedia template is about external links, not references. I think it's used sometimes as a reference when the article subject is the software but it's not generally considered reliable. There are more possible examples like a category about some small organization linking to the organization's website and so on. There are many more cases where some website that would be considered unreliable would be very useful so at a minimum one would need to change that part, especially be considered reliable sources on other Wikiprojects, but I think it may be best if this point was removed or turned into a recommendation about what is usually (instead of always) the case. I'm still unsure what is meant with repeating info already in parent cat. However, I haven't seen any cases of such anyway but it doesn't seem like a good requirement. Prototyperspective (talk) 10:22, 30 August 2024 (UTC)[reply]
 Support Most of what has been said before. Thanks for starting this discussion. Maybe the four points of ReneeWrites just need some nuance. My remarks:
  • This does not concern only categories about the province of North Brabant, but for many location categories of the Netherlands. I have seen many of those descriptions elsewhere too.
  • For me they are redundant and annoying, but I am not the target group: perhaps to occasional visitors they might give useful information, especially if they are foreigners.
  • To limit obvious redundant information: No repeating of information that can be found in an infobox. But then there should be a Wikidata infobox. For subcategories like "in" categories (for countries, "streets in" and so on) and years, there usually are no, and should not be either (except for when there is a Wikipedia article).
  • I prefer "ABC" is a .. as well. The shorter, the better.
  • I agree with Jmabel: Repeating the category name in other languages can be very useful. Ever since I discovered in an English discussion about a Dutch subject that a Dutch participant had trouble discussing in English, I translate categories about the Netherlands into Dutch (unless there is a Wikidata item or it is a "by" category). And sometimes there is more explaning to do because the English word has no equivalent in Dutch (see for instance Category:Estates in the Netherlands) or the English word looks like a Dutch one, but means something else.
So in general:
  1. Every topic category should have either a Wikidata item or a description. Exceptions: if the category name is about a universally known subject (like "shoes"). In case of doubt: add a description in English, other languages are allowed too.
  2. No repeating of information that can be found in a corresponding Wikidata infobox.
  3. If there is no Wikidata item: Add descriptions to categories for countries that are not native-English speaking countries, preferably in their native language.
  4. Keep descriptions as short as possible. Add a link to the corresponding Wikipedia article in the native language if that is appropriate (especially in case of country categories). You may add links to other relevant websites that you trust if there is no Wikidata item and the website is within the scope of Commons (no avertising, and so on).
  5. Repeating of information that can be found in the category name ("The category "Streets of Geertruidenberg" is for pictures of streets of the municipality of Geertruidenberg...") should be avoided as much as possible. But sometimes it can be useful, even in this case: Geertruidenberg is a municipality AND a populated place. It is good to know for which one the category is (though it would be better to have it in the category name).
--JopkeB (talk) 07:41, 30 August 2024 (UTC)[reply]
What do you think of the samples Chicago and London?
 ∞∞ Enhancing999 (talk) 10:21, 30 August 2024 (UTC)[reply]

BTW, if anyone wants to see a good example of major overkill when it comes to a category description check out Category:Louvre Museum - Room 185 and scroll down past the infobox. I assume everyone here would agree that something like that is way to excessive. --Adamant1 (talk) 05:49, 31 August 2024 (UTC)[reply]

Not only is that wildly excessive, it seems like an unreliable basis for categorization. The museum can reorganize their collection; Commons shouldn't have to shuffle a bunch of categories around (or argue about what objects used to be where) when that happens. Omphalographer (talk) 06:15, 31 August 2024 (UTC)[reply]

Category instructions

There are descriptions, but sometimes instructions are also usefull. Examples: Category:Rail vehicle doors: Please place images of closed train doors only if they are prominent. All passenger trains photografed from the side have train doors. There are subcategories for open doors. Other example: Category:Rail vehicle sliding-plug doors: The door slides outside the vehicle by closing is pulled in at the last moment inside (plug). There is no supporting structure outside the dooropening and a closed door is flush with the vehicle surface. animation and rail door systems. This category is sorted by country. (Spain and Sweden under the letter 'S'). Train material door types can be determined as soon as there are enough good examples of 'open' doors to classify. In the Commons the definition of train doors is broader than in the NL article nl:Treindeur. Outside doors not used by passengers are excluded (for example drivers or loading doors) or aisle doors.Smiley.toerist (talk) 13:45, 6 September 2024 (UTC)[reply]

Coordinates

I just want to add one aspect or make it clear as it is already contained in No repeating of information that can be found in an infobox. Object coordinates should not be added to categories / should be consolidated and removed, if any. --Herzi Pinki (talk) 22:32, 6 September 2024 (UTC)[reply]

@Herzi Pinki: can you point me to anywhere that was decided as policy? Last I remember this being discussed (a few years ago) the opposite conclusion was reached: that it was useful to have Commons and Wikidata both contain location information, because the duplication was very useful in identifying possible vandalism, since a bot would notice the discrepancy, allowing for easy review. There could well have been a change, but if so I missed it. - Jmabel ! talk 14:04, 7 September 2024 (UTC)[reply]
No, nothing has changed. We don't trust. But I just wanted to point out that as a consequence of the above No repeating of information that can be found in an infobox, this will also affect the policy on coordinates. best --Herzi Pinki (talk) 14:40, 7 September 2024 (UTC)[reply]
The above is discussion is a bit odd, because it says that and people agree that including samples that do exactly the opposite. Go figure. Except when it's translated into Dutch. So local languages are ok, except Dutch.
Actually, it's a mistaken assumption about coordinates that the infobox offers the same. Some users might find some parts of it, most don't that see that. I think we discussed it before. Obviously, we could integrate coordinates better in the interface.
 ∞∞ Enhancing999 (talk) 20:03, 7 September 2024 (UTC)[reply]

Family trees (see below )

There is a discussion below if family trees should cover up the category pages or not, see #Familytree.
 ∞∞ Enhancing999 (talk) 20:03, 7 September 2024 (UTC)[reply]

Example images for subcategories

I would appreciate a typical sample image for each subcategory to ease correct assignment to subcategories. Of course not for standard subcategories like 'Views of', 'Nature of', 'Buildings in', etc. This should be done by marking an image of the subcategory in a suitable way (by a dedicated cat, by QI?) and done automatically when generating the category pages. best --Herzi Pinki (talk) 18:13, 9 September 2024 (UTC)[reply]

Maybe there could be a gadget that shows an image when hovering the subcat (for longer than 0.2 s). However, I think showing four images instead of just one would be best. For categories that are linked to a Wikidata item, selecting a typical representative fairly high-quality is simple: the one(s) that are set in the image property of the item. However there are many categories without Wikidata item (or with item but no image set there). In regards to your use-case however, the category title should be fairly self-explanatory so I think the use of that would be quite limited (but this and other potential usecases could still be useful enough for having some gadget that enables that). Prototyperspective (talk) 20:25, 9 September 2024 (UTC)[reply]
Don't take this the wrong way, but what's the ultimate end game here and how exactly does it relate to the topic of the discussion, "category descriptions"? Because I really fail to see the connection between that and your suggestion for a gadget that shows an image when hovering the subcat. I don't think anything ReneeWrites said in their original comment would be negated or effected by your idea. --Adamant1 (talk) 00:58, 10 September 2024 (UTC)[reply]

When questioning "own work"

I see images tagged with {{No permission since}} when clearly the issue is "I don't believe this is 'own work'." Shouldn't we have a distinct tag for that? - Jmabel ! talk 21:51, 29 August 2024 (UTC)[reply]

+1 I've noticed this as well. It is currently the closest to correct tag that we have, but it's still not really correct. -- King of ♥ ♦ ♣ ♠ 21:58, 29 August 2024 (UTC)[reply]
How about {{No source since}}? An obviously invalid claim of own work fits the description of "...appears to have an invalid source listed". I agree that a more specific template would be helpful, though - something along the lines of "please update with the correct source/license, or send permission to VRT if it's really your own work". Omphalographer (talk) 01:03, 30 August 2024 (UTC)[reply]
I view {{No source since}} as having more or less the same problem (though per the phrase you cite, a bit less so). I still think a more specific template would be better, since this particular situation comes up so often. - Jmabel ! talk 04:52, 30 August 2024 (UTC)[reply]
Perhaps some template like "dubious source" could be even better, since there could also be misattributed files to other sources (source present in the file page, but it isn't the true source), not only for own works. MGeog2022 (talk) 12:24, 31 August 2024 (UTC)[reply]
I'd be OK with that; no reason for the template to be specific to "own work". - Jmabel ! talk 17:04, 31 August 2024 (UTC)[reply]
That sounds good -- DaxServer (talk) 17:57, 31 August 2024 (UTC)[reply]
The semi-speedy deletion templates should only be used for obvious cases that are recent uploads. For the "I don't believe this is 'own work'.", a regular deletion request should be used. Multichill (talk) 12:03, 1 September 2024 (UTC)[reply]
@Multichill: you are probably right, which makes my original proposal moot. So do you think the answer is simply that {{No permission since}} should never be used this way if "own work" is claimed? Presumably it can still be a speedy if you actually have the original source to reference (or we'd be bogged down enormously), but lacking that, if you doubt "own work" you should always need a DR. Is that correctly understood? - Jmabel ! talk 14:08, 7 September 2024 (UTC)[reply]
@Jmabel: Not much space between {{No permission since}} and obvious {{Copyvio}} I guess? Because if someone just rips an image of the internet and tags it as own work, we'll just use {{Copyvio}}, right?
All the semi-speedy deletion templates are things that can either be fixed by the uploader or just end up being deleted. These are not suitable at all for cases where discussion might happen. I would be conservative. Multichill (talk) 18:45, 8 September 2024 (UTC)[reply]

August 30

inconsistent database state

Hi, source Category:Files with coordinates missing SDC location of creation (49° N, 11°E) was moved (in May 2024) to target Category:Files with coordinates missing SDC location of creation (49° N, 11° E) by User:Multichill, but there are still >50000 files shown in the source category. Strange enough, this is not reflected by the file description, where the target category is shown, nor is a move reflected in the version history, e.g. [6]. A miracle I don't understand. A touch / nulledit helps to remove misscategorized files. Is there a way to fix this in a bulk operation for all files? Can someone fix this issue? best --Herzi Pinki (talk) 08:10, 30 August 2024 (UTC)[reply]

It's usual for categories added by templates or mediawiki itself.
 ∞∞ Enhancing999 (talk) 10:18, 30 August 2024 (UTC)[reply]
Ok, the behaviour is caused by {{Location}}. And it's usual, that the cache invalidation will happen within 4 months when changing templates. But it didn't. I prefer a solution over an explanation. --Herzi Pinki (talk) 11:33, 30 August 2024 (UTC)[reply]
It's a question for WMF.
 ∞∞ Enhancing999 (talk) 11:35, 30 August 2024 (UTC)[reply]
@Herzi Pinki: Working on 50,796 files with COM:AWB applying {{subst:void}} at ~15-30/min (with time off for RL)...   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 14:31, 30 August 2024 (UTC)[reply]
There isn't just one such category: Special:WantedCategories.
 ∞∞ Enhancing999 (talk) 14:48, 30 August 2024 (UTC)[reply]
@Enhancing999: Can anyone do this with touch.py?   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 15:28, 30 August 2024 (UTC)[reply]
Not sure if this is the solution Herzi Pinki is asking for, as they state it in their question.
 ∞∞ Enhancing999 (talk) 15:39, 30 August 2024 (UTC)[reply]
@Enhancing999: Either is "a bulk operation for all files".   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 15:45, 30 August 2024 (UTC)[reply]
I can also write a phab-ticket to the database admins, not sure, whether this will help. Jeff G., thanks for your try but even with AWB this will take days for all categories that Enhancing999 mentioned. I will try touch.py. Herzi Pinki (talk) 16:06, 30 August 2024 (UTC)[reply]
I managed to run touch.py (via pwb.py touch), but I'm slow (sleeping 10 sec. between each touch). Will take approx. 14 hours for Category:Files with coordinates missing SDC location of creation (53° N, 4°E). anything that can be done better? --Herzi Pinki (talk) 16:10, 30 August 2024 (UTC)[reply]
ok, pwb.py touch -pt:1 makes it faster. --Herzi Pinki (talk) 16:25, 30 August 2024 (UTC)[reply]
@Enhancing999: All such categories have now been addressed. VFC is faster than AWB for things it can do, but only on files.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 10:10, 3 September 2024 (UTC)[reply]
@Herzi Pinki: ✓ Done, finally. Along the way, Windows decided to reboot and AWB displayed "stopped" and had to be killed. I finished with VFC, but the 100 limit is painful to get around and it no longer displays a visual indication while it is processing and no longer auto-activates "more" at the bottom.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 22:10, 1 September 2024 (UTC)[reply]

one touch per second works, even two jobs in parallel (meaning 2 touches / sec). We have about 1390000 entries in categories listed in Special:WantedCategories. Root cause is the change by User:Multichill on 8 May 2024 [7]. Touching all the 1.4 millions files will take 1400000/60/60/24 = ~16.2 days (1 per sec). --Herzi Pinki (talk) 21:49, 30 August 2024 (UTC)[reply]

Tracked in Phabricator
Task T373722
It seems that recurring changes to Module:Coordinates do not trigger the invalidation of cache. But a change in {{Location}} might (as a workaround). Anybody with admin rights who dares to make a nulledit on this template? best --Herzi Pinki (talk) 22:17, 30 August 2024 (UTC)[reply]
@Herzi Pinki: That template is currently transcluded 28,638,450 times. Template Editors are also allowed to edit it, but "edits should be kept to a bare minimum" after discussion on the talk page Template talk:Location.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 01:08, 31 August 2024 (UTC)[reply]
@Jeff G.: yes I know all that, but if we don't fix problems, because of such performance guidelines, we are doomed. btw, I think something is wrong with the statement of the category, that SDC location is missing. I made a few samples, and all had SDC coordinates. So maybe there is just a flaw in the implementation of these maintenance categories. --Herzi Pinki (talk) 07:27, 31 August 2024 (UTC)[reply]
Did you check that coordinates was in same property (location of creation (P1071)) than tracking cat function expects? Zache (talk) 08:31, 31 August 2024 (UTC)[reply]
No, never heard of location of creation (P1071). What is the difference between the location of creation (of a photo) and the camera location? Ah, it is not just coordinates. So it only makes sense, if this means the location of creation of the depicted object, a property for the vast majority of depicted objects we don't know (think of persons, where they have been created in most cases not even the parents know for sure; a work of art was created in some atelier, but not on the location of final placement). Just another maintenance category that wise men have created and nobody will do the necessary maintenance for. --Herzi Pinki (talk) 08:56, 31 August 2024 (UTC)[reply]
AFaik P1071 is most specific item of place where image has been taken (ie. position of photographer). Ie. it can be administrative area, island, building etc. It is useful expecialy when there is no known coordinates. --Zache (talk) 12:49, 31 August 2024 (UTC)[reply]
@Herzi Pinki: usually it takes a while for cache to pick up, but this is quite extreme. I'll have a bot flush it out. That will be fairly quick.
As for the usefulness: Please assume good faith for the things you haven't heard of before. We're talking about photographs here. Photographs have been taken from some location. See for example File:Grote Kerk an Vleeshal building Harlem.jpg: Here we have coordinates and location of creation (P1071) -> Grote Markt (Q1083850).
In the future this could be used for faceted search: You search for a topic, for example "church", and the search engine will give you options, like location to drill down. In this case by location Europe (Q46) -> Netherlands (Q55) -> North Holland (Q701) -> Haarlem (Q9920)
To break the chicken and the egg problem, I'm adding the location of creation (P1071) based on several sources one being reverse geocoding so that hopefully, we can do a faceted search pilot in the future. Multichill (talk) 11:50, 1 September 2024 (UTC)[reply]
@Multichill: thanks for caring - it seems to me that your quota is much better than mine. Thanks for the explanation, was expecting something like this. As for AGF, I'll try to trust everybody here (without much individual difference in weigth btw.), but if things get contradicting, I'm lost. For me still it is not clear, that location of creation (P1071) (I also trust the property description) is about the photo itself and not the object depicted on the photo. Maybe you can make that clearer in the description of the categories {{Missing SDC location of creation}} mentioned here and maybe we should add something to location of creation (P1071) like most accurate place a photo was taken. May I mention, that openstreetmap nominatim can do this mapping of coordinates to place names out of the box. best --Herzi Pinki (talk) 12:59, 1 September 2024 (UTC)[reply]

mostly done by collective endeavour. --Herzi Pinki (talk) 18:55, 3 September 2024 (UTC)[reply]

August 31

Disambiguating categories

From what I've seen there appears to be two or three main ways that categories are disambiguated at this point. Usually it's done by using brackets or a comma, but I've also seen them disambiguated using other symbols. For instance the Russia symbol for "<" and "<." Having multiple ways of doing it seems to go against the universality principle. Plus at least with commas they aren't super intuitive since plenty of normal names contain commas. Nor are they that easy to see in the first place. "John Malkovich (actor)" makes way more sense and is much easier to see then "John Malkovich, actor", "John Malkovich <<actor>>", or any of the other alternatives I've seen. So at least IMO round brackets should be the de-facto way to name disambiguated categories. Thoughts? Adamant1 (talk) 01:21, 31 August 2024 (UTC)[reply]

Round brackets (parentheses) are the normal way to do this, but U.S. and Canadian place names (and, I suspect, some others) are rather universally written with commas when including the state/province. I think anyone from those countries would find it very odd to see, for example, "Columbus (Ohio)" or "Stratford (Ontario)" rather than "Columbus, Ohio" and "Stratford, Ontario". - Jmabel ! talk 01:54, 31 August 2024 (UTC)[reply]
@Jmabel: That's probably one the few places where a comma is fine, if not expected. There's a lot of categories for streets in Germany its like "X street name, Y location" Which I don't think is a good way to do it. I don't have a problem with a comma being used for something like "city, state" as long as its the norm for similar categories. Its annoying to see a common in cases where the norm is circular brackets, visa versa, or no disambiguating of similar categories to begin with though (Its tangential, but a lot of disambiguation seems to be totally pointless in the first place). --Adamant1 (talk) 02:22, 31 August 2024 (UTC)[reply]
Pretty sure Mexico also shares that North American convention. - Jmabel ! talk 01:57, 31 August 2024 (UTC)[reply]
@Jmabel: What do you think about something like the subcategories for streets in Category:Streets in Lutsk? Would you disambiguate them using a comma, circular brackets, not disambiguate them at all, or just leave them as is with "in Lutsk" at the end of the street names? --Adamant1 (talk) 05:11, 1 September 2024 (UTC)[reply]
I'd certainly want to see them consistent with one another, whatever is chosen. "in Lutsk" seems awkward, but it's also apparently what the most active categorizer here has chosen. I don't much care between comma and parentheses (circular brackets) as long as it's consistent. I'd have no objection to a site-wide standard, but we should at least be consistent within a city. Without a site-wide standard, if you really wanted to work on this for a particular city, I'd make sure to have consensus from the (presumably few) people who most work on images of that city. - Jmabel ! talk 16:18, 1 September 2024 (UTC)[reply]
I changed my mind from ()-notation to comma-notation for Category:Streets in Vienna for two reasons: 1) to make names uniform and 2) acc. to Commons:Language policy category names (except proper names) have to be in English. In English for me does not only mean to translate words, but also to use the structure as it is used in :en:WP. Which is for geographical names (I consider street names as such) "Columbus, Ohio" or "Stratford, Ontario". But uniform name schemes are prior to that, and I try to follow naming schemes already found, sometimes change to the more widely used scheme. best --Herzi Pinki (talk) 13:49, 4 September 2024 (UTC)[reply]

September 01

Hello, I'm sorry. I believe this image is not under permission and I'm just now realizing it. Mário NET (talk) 02:38, 1 September 2024 (UTC)[reply]

@Mário NET: it looks like that file page is a mess in terms of what it states, but if the image is from Adalbert Seitz (died 1938, which means PD in Germany) and was published in 1927 (which means PD in the U.S.) it should just be a matter of cleaning up the claims on the file page. And if it does come originally from the 1845 work mentioned there, then the case is even clearer. What has you concerned? - Jmabel ! talk 02:52, 1 September 2024 (UTC)[reply]
I thought it would say "free copyright" in "Show Info" and it said "Copyright & Usage, Due Diligence"; so I couldn't interpret what it was. 03:18, 1 September 2024 (UTC) Mário NET (talk) 03:18, 1 September 2024 (UTC)[reply]
@Mário NET: "free copyright" makes no sense at all. Perhaps you are thinking "copyright-free" (meaning that there is no copyright)?
I'll try to fix the page myself, because I'm pretty sure your recent edit (which I reverted) only made this worse. - Jmabel ! talk 16:23, 1 September 2024 (UTC)[reply]
Please correct me if I am wrong, but, it's not the case for public domain works such as the one here, but, for CC-licensed works, I believe that there is copyright in place, and there is a free license one (the requirement to cite the original author, or to redistribute the work under the same license, in the case of CC-BY-SA, is due to copyright law: the work's author has copyright ownership, and licenses the work to others under a CC license). So in some cases, not for public domain works such as this, but "free copyright" could make sense as a phrase. MGeog2022 (talk) 09:15, 2 September 2024 (UTC)[reply]
@MGeog2022: I couldn't fully follow some of your English, but if I follow you correctly you are describing "copyrighted free-licensed work", and I stand by my original statement. If I've misunderstood you, feel free to write in Spanish, which I read without difficulty. - Jmabel ! talk 22:25, 2 September 2024 (UTC)[reply]
@Jmabel, it's probably because how I chained one sentence after another, perhaps I did it in a too Spanish-like way. Yes, I'm talking about copyrighted free-licensed works, and, so, you're completely right, "free copyright" perhaps may never make sense, since there is this another way to say it. Sorry, but, some time away, I saw a misunderstanding in a Wikipedia talk page, where someone seemed to say that the mere presence of a copyright tag implied that a work could not be freely licensed, and I think it's a dangerous error, since it may lead to the removal of legitimate content, so I wanted to draw attention to it (not so much because you thought so, which seemed very unlikely, but because of others who might read it). MGeog2022 (talk) 09:23, 3 September 2024 (UTC)[reply]
Wait, I'm confused, too. Where did the mention of an 1845 work come from? What is the source for that information? - Jmabel ! talk 16:35, 1 September 2024 (UTC)[reply]
This work is not from 1845, what happens is that I used a text from another image to construct this one and I didn't realize that the "1845" from the previous text had remained, I will correct that. Mário NET (talk) 01:55, 2 September 2024 (UTC)[reply]

Speedy because "File has no source" but what they are saying is "I want to see a url" instead of what is listed

@Winmyint998: See: Special:Contributions/Winmyint998 where they are sending images to speedy because the "File has no source" but what they are saying is "the source listed is inadequate". Should these images be going to speedy? A few years ago there was a problem with another editor using Speedy with: "File has no license" but what they were really saying was "I don't think this image has the correct license". Should these be going to speedy for auto-deletion and no further research, or through the regular process where the community does more research if needed? Should the normal deletion nomination process be so easy to circumvent? Should the "Speedy" be removed? RAN (talk) 07:39, 1 September 2024 (UTC)[reply]

Yes, this is an inappropriate use of the speedy deletion tag and they should be converted to regular deletion requests. Speedy deletions should only be used for clear and unamibiguous copyright violations. Checking a few files that they tagged, despite lacking in sourcing the files appear to be old enough for their copyright to have lapsed in Myanmar. ReneeWrites (talk) 07:57, 1 September 2024 (UTC)[reply]
Agree with RAN and ReneeWrites, but things are certainly not helped when (for example) [[User:History of the Burmese}} says source is {{Own work}}, date is three days before the upload, and the license is {{PD-Myanmar}}. Presumably it's the source and date that are wrong here rather than the license, but the source and date are not even plausible. - Jmabel ! talk 16:57, 1 September 2024 (UTC)[reply]

Abusefilter to prevent uncategorised pages？

are all pages in main and category namespaces supposed to be categorised? if so then perhaps an abusefilter can be created to prevent creation of such pages without any "Category" or "{{" in source? (pages are categorised either directly or by transcluding a template.) RZuo (talk) 11:12, 1 September 2024 (UTC)[reply]

The ones on Commons:Report UncategorizedCategories with infobox all have "{{".
To update the list I have to purge recent creations to ensure those connected to Wikidata, but without edits since don't end up there.
For files, everything just ends up in Category:All_media_needing_categories_as_of_2024. Do users still get notified when they don't categorize?
 ∞∞ Enhancing999 (talk) 11:38, 1 September 2024 (UTC)[reply]
My experience is that especially some bulk bot uploads from external sources (like flickr) are not properly categorized or not categorized at all. While inexperienced users fail to categorize quite often, they do not upload such huge numbers of media to commons. So coping with lacking categories by inexperienced users is manageable. But bulk uploaders can work around such an abusefilter by adding some not helpful category like Category:Files from xxxx stream on flickr needing categories and not taking care for proper categorization. Even if bulk uploaders do proper categorization, they will do this in a second step after uploading - an abusefilter will just disable their usual procedures.
for the cat mentioned above, https://wikimap.toolforge.org/?cat=All_media_needing_categories_as_of_2024 will show (after a while) images on the map and then it is easy to care for your area of interest. best --Herzi Pinki (talk) 21:54, 6 September 2024 (UTC)[reply]
"pages in main and category namespaces". not files. RZuo (talk) 09:29, 9 September 2024 (UTC)[reply]
Yes, but the approach used for files could be applied to categories as well. There already is a (mostly empty) Category:Uncategorized categories.
 ∞∞ Enhancing999 (talk) 09:35, 9 September 2024 (UTC)[reply]
Category:Uncategorized categories is "mostly empty" at the moment, but only because it was recently winnowed down from over 1500 categories. A year ago, it had ballooned to about 8000, and those past 5000 in alphabetical order couldn't even be seen in the list. - Jmabel ! talk 06:27, 10 September 2024 (UTC)[reply]
You mean Special:UncategorizedCategories, not Category:Uncategorized categories, isn't it?
 ∞∞ Enhancing999 (talk) 06:42, 10 September 2024 (UTC)[reply]

Commons Gazette 2024-09

Volunteer staff changes

In August 2024, 1 sysop was removed. Currently, there are 183 sysops.

We thank him for his service.

Other news

Edited by RZuo (talk).

Commons Gazette is a monthly newsletter of the latest important news about Wikimedia Commons, edited by volunteers. You can also help with editing!

--RZuo (talk) 23:04, 1 September 2024 (UTC)[reply]

I suggest that the Commons community should take action in the event that Cat-a-lot is slowed down (effectively to the rate of 1 edit per second). I don't want to disagree with SRE-team's conclusions, but this unilateral change to the "coolest tool" has affected many users' workflows (including mine). — Draceane talkcontrib. 20:11, 3 September 2024 (UTC)[reply]
Maybe you can propose a better implementation ? Or sprinkle some pixie dust ? As long as it doesn't cause all of Wikimedia to got down multiple times a week, any of it is welcome. —TheDJ (talk • contribs) 21:21, 3 September 2024 (UTC)[reply]
I think the bug was identified two months ago. Something in relation to a change WMF made to Special:Search.
 ∞∞ Enhancing999 (talk) 21:28, 3 September 2024 (UTC)[reply]

September 02

Japanese category

A Japanese-speaking editor is need to describe and categorise Category:Nagoya Location Navi, please.

As an aside, do we have a forum to find editors with certain language skills; or cultural understanding? On Wikipedia I would use a wikiproject, or WP:Embassy. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:32, 2 September 2024 (UTC)[reply]

Could you use the Wikipedia embassy for this as well? That's what I would do, even if it's not 100% the right avenue. There are two editors listed for the Japanese language there, and BorgQueen is highly active. ReneeWrites (talk) 21:19, 2 September 2024 (UTC)[reply]
Yes, but my question was whether we have a local forum. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:35, 3 September 2024 (UTC)[reply]
@Pigsonthewing I described it. Immanuelle ❤️💚💙 (please tag me) 13:29, 3 September 2024 (UTC)[reply]

September 03

Have your say: Vote for the 2024 Board of Trustees!

Hello all,

The voting period for the 2024 Board of Trustees election is now open. There are twelve (12) candidates running for four (4) seats on the Board.

Learn more about the candidates by reading their statements and their answers to community questions.

When you are ready, go to the SecurePoll voting page to vote. The vote is open from September 3rd at 00:00 UTC to September 17th at 23:59 UTC.

To check your voter eligibility, please visit the voter eligibility page.

Best regards,

The Elections Committee and Board Selection Working Group

MediaWiki message delivery (talk) 12:13, 3 September 2024 (UTC)[reply]

In case you were wondering:

 ∞∞ Enhancing999 (talk) 13:02, 3 September 2024 (UTC)[reply]

Are amulets toys?

I am having some of the files I uploaded being proposed for deletion as toys, to my knowledge they are amulets, not toys, considered to have religious significance. Does that change things? Commons:Deletion requests/File:Okimono omikuji.jpg This overall applies to the entire category Category:Doubutsu mikuji Immanuelle ❤️💚💙 (please tag me) 13:15, 3 September 2024 (UTC)[reply]

They are presumably not "toys", but uhey are certainly copyrightable in the U.S. I can't speak for Japan. - Jmabel ! talk 18:06, 3 September 2024 (UTC)[reply]
@Jmabel toys are not copyrightable in Japan. But would that mean we should delete the entire category? Immanuelle ❤️💚💙 (please tag me) 10:24, 5 September 2024 (UTC)[reply]
@Immanuelle: Yes, unless the toys are licensed freely in the US.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 10:56, 5 September 2024 (UTC)[reply]

I have many photos of a ship, but I do not know which one it is

In this category Category:Kyushu Yusen there are six operational ships. Of these ships, two are of one class, and 4 are of another class. I have many photos taken of and on each class, but I am not sure how to categorize them. Immanuelle ❤️💚💙 (please tag me) 13:18, 3 September 2024 (UTC)[reply]

You could add them in the categories for classes and Category:Unidentified ships.
 ∞∞ Enhancing999 (talk) 13:32, 3 September 2024 (UTC)[reply]
And it is certainly OK to create a category for each class. - Jmabel ! talk 18:07, 3 September 2024 (UTC)[reply]

Hi there! A colleague of mine has been working for many years on the topic of image searching. His focus is not just on finding images that are tagged with keywords, but also finding similar images for various definitions of similar. He managed to obtain 6 million images from Commons to demonstrate his technology, it is freely available at WikiView. He has tried to find someone at Commons who knows if an API for Commons exists and if it does, if it is possible to just pull new pictures and not have to pull a dump at regular intervals. It would be cool to offer this visual search (quite a different way of searching, try it!) from the home page and also to have it stay up-to-date. If you are such a person, please contact me in-wiki and I'll put you in touch with my colleague. Thanks! --WiseWoman (talk) 18:43, 3 September 2024 (UTC)[reply]

"He managed to obtain 6 million images" ... sigh. I trust they didn't clear this with anyone ? —TheDJ (talk • contribs) 21:18, 3 September 2024 (UTC)[reply]
I was thinking about doing a similar thing to train an AI image generator but who knows how it work with all the different kinds of licenses on here. At least in that case the generated images would be different from the originals. It still seems questionable though to just take millions of images with different licenses and use them in a single product in mass though. I don't there's any way to follows the licensing terms for every image. Especially an in instance like this one. Let alone with an AI image generator. --Adamant1 (talk) 02:07, 4 September 2024 (UTC)[reply]
I wasn't part of the project, but they did speak to someone. The problem is, that the people change and this is so difficult for people not in the Wikiworld to sort out who they have to speak with. But there is a link back to Commons when an image is found the license is clearly given in Wikiview. I believe they only used images that were usable according to the license, and that was the problem, being able to obtain new pictures but only those ones with a fitting license. The WikiWorld seems so difficult to people not active here, often because the tone is a bit rough. --WiseWoman (talk) 11:42, 4 September 2024 (UTC)[reply]
Template:WiseWoman After I commented I noticed that they have a way to sort the images by license and that there's links back to the images on Commons. Both of which helps. I can understand why this would be difficult for people who aren't active in the project already. It can be a little rough and the sheer amount of different licenses with multiple variations per license based on the personal preferences of uploaders really doesn't help. I'd love to see them at least get rid personalized licensing terms. If not just the different defaults down to 1 or 2. There's like what, 7 different versions of the Creative Commons license on here at this point though? The whole things totally ridiculous. --Adamant1 (talk) 20:41, 4 September 2024 (UTC)[reply]
Pinging @WiseWoman, since Adamant1's ping is ill-formed and won't have worked. - Jmabel ! talk 14:15, 7 September 2024 (UTC)[reply]
Sounds amazing! Have a look at m:Community Wishlist/Wishes/Search Wikipedia with image or sketch search and maybe also comment on the talk page there. It didn't work so well with search terms but one can use WMC search for that anyway and it seems like the focus there is searching by image. I would suggest the asap it also shows the WMC categories by number of images in the results they contain, this way one can find relevant categories that contain more images. Also see my proposal about data dumps (and the link at the top there) which may make it possible to just pull new pictures which would also be great to improve WMC performance and reduce server load. I don't think it will be linked on the main page if it's not searching across all WMC images and is an external site but probably it could be linked at some other page not far from it – I don't know of use-cases where one would want to search by image and not via search terms or by browsing. Prototyperspective (talk) 22:21, 3 September 2024 (UTC)[reply]
They also have a tool that kind of sort of works with sketching (but used a different image collection for this one). It would be so important for the researchers in this area to speak with people at Wikimedia to sort out how to move forward on this. --WiseWoman (talk) 11:42, 4 September 2024 (UTC)[reply]

September 04

CfD for categories of nonexistent WikiProjects

Hi. I got in a little tiff with a another user earlier because I nominated some categories they created for nonexistent WikiProjects for speedy deletion. I don't really feel like relitigating it here. Except to draw people's attention to Commons:Categories for discussion/2024/09/Category:WikiProject Iran by city. Personally, I feel like it clearly violates the guidelines and consensus to create categories as part of a personal pet project for things that clearly don't exist. Apparently the user who created the categories thinks it's totally fine though. So I started a CfD to see what other people think about it. The categories seem totally pointless and unhelpful to me, but who knows. Maybe I'm missing something or there's a valid reason to keep them that I'm just not aware of. Adamant1 (talk) 08:43, 4 September 2024 (UTC)[reply]

To repeat it for fourth (!) time: these categories are not intended for existing WikiProjects, but for separating huge WikiProject Iran by particular city, although there are some individual projects like WikiProject Tehran on Persian Wiki. They serve as maintenance categories and exist for years. The issue here is Adamant1 yesterday planned to delete everything, without any proposal or discussion, making an utter mess with hundreds of related categories, so I reverted him. --Orijentolog (talk) 09:04, 4 September 2024 (UTC)[reply]
The CfD is a better place to discuss the particulars. I will point out though that Commons:Categories states "we should not classify items which are related to different subjects in the same category. There should be one category per topic; multi-subject categories should be avoided. The category name should be unambiguous and not homonymous." Maybe it's just me, but I really don't see how it isn't ambagious to have a category called "WikiProject Jahrom" when there's WikiProject Jahrom and it's just a maintenance category. I certainly thought it was a category for a real WikiProject at first. I certainly didn't know it was maintenance category for WikiProject Iran until you told me. So there's clearly some ambiguity here about the purpose of the categories and why exactly they exist. Just because it's clear to you as the person who created them that doesn't mean other people will understand it and they should. Otherwise the categories should be deleted. Period. There's absolutely zero justification what-so-ever to have categories that only make sense to the person who created them. --Adamant1 (talk) 09:13, 4 September 2024 (UTC)[reply]
WikiProject Jahrom serves as a part of the WikiProject Iran and contains things related to the city of Jahrom. If it doesn't make sense to you, that's not my problem. --Orijentolog (talk) 09:27, 4 September 2024 (UTC)[reply]
Then the category should be called something like "Category:WikiProject Iran in Jahrom" or "Category:WikiProject Iran (Jahrom)"; if you make categories that do not conform to our naming standards, that will very quickly become "your problem"". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:14, 4 September 2024 (UTC)[reply]
@Pigsonthewing: don't think I'm scared Andy, making 1,000 or 10,000 edits to correct something ain't an issue for me, and I'm a collaborative person. Only the messiness and lack of unification scares me. As I said to Adamant1 today, if there's a consensus to rename everything, I'll do it personally. I prefer keeping the status quo (yes, it's laziness!), but I would accept your proposal if others agree. Generally it's fine and I have only minor issues with it. Another thing, I don't see that strict naming standards apply to maintenance categories, there are a lot of uneven, non-English and even illiterate ones around. --Orijentolog (talk) 17:37, 4 September 2024 (UTC)[reply]
I don't think you're scared at all. Why on earth would you say that? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:58, 4 September 2024 (UTC)[reply]
@Pigsonthewing: please don't take me literally, it was a jesting message that I'm always ready to massively correct my own errors. --Orijentolog (talk) 18:12, 4 September 2024 (UTC)[reply]
There are a lot of uneven, non-English and even illiterate ones around. The guideline for naming categories was improved last year (after an extremely long CfD that had more then enough support at the time BTW). So there's naturally going to be some uneven, illiterate categories at this point. It takes time to cleanup this stuff up once a guideline is changed and that's not helped or made quicker by the needless edit waring and discussion over it on your end.
More to the point, something like "Category:WikiProject Iran in Jahrom" or "Category:WikiProject Iran (Jahrom)" would be better. Although I still think the categories would be mostly pointless and not make sense. As a category like "Category:WikiProject Iran in Jahrom" still insinuates that there's a WikiProject in Jahrom or that Wikiproject Iran has a satellite group when neither one is true. If someone creates a category called "X business in Y location" then that business should have some dealings in that location or the category is fundamentally wrong. --Adamant1 (talk) 19:16, 4 September 2024 (UTC)[reply]

Categories which should only contain media of a specific type

what have i just found 🧐

Category:Categories which should only contain media of a specific type. RZuo (talk) 11:37, 4 September 2024 (UTC)[reply]

@RZuo: This sentence no verb. (I'm not sure what you meant to discuss in this thread.) - Jmabel ! talk 12:42, 5 September 2024 (UTC)[reply]

Image Annotator not working

Hi, is there a problem with ImageAnnotator ?--JotaCartas (talk) 16:05, 4 September 2024 (UTC)[reply]

OK, solved, I had "ImageAnnotator" disabled in Preferences-Gadgets, thanks and sorry for trouble JotaCartas (talk) 16:43, 4 September 2024 (UTC)[reply]

September 05

Wikitext weirdness in Commons:Editor's index to Commons

In Commons:Editor's index to Commons (a.k.a. COM:Glossary), there is some weirdness I don't quickly see how to fix. The display where the wikitext says Bots: <span id="Catbot"></span>{{Shortcut2|COM:EIC#Catbot}} surely cannot be what we want, and also can be really confusing about the alphabetical order when trying to look up anything else (looks like you are already to "COM", but actually is part of "Categories" and comes before "Censorship". Currently, I'm really tired and not at my sharpest after some travel SNAFUs. Could someone else please take a look at this? Thanks. - Jmabel ! talk 12:39, 5 September 2024 (UTC)[reply]

Familytree

@Enhancing999: See: Category:Byington Ford A change was just made to collapse the family tree function in categories as the default. Now it isn't obvious that the navigation device is there. What are people's preferences? Which is the most useful? RAN (talk) 13:14, 5 September 2024 (UTC)[reply]

I don't recall ever seeing this before and don't see the purpose of it here. If I want to know genealogical information, Wikidata and Wikipedia are appropriate for that. I guess that it's useful for navigating within the same family, but that is pretty niche and probably not something I would ever use. That said, I generally don't like auto-collapsing any actual content and the default should be for these to be uncollapsed. —Justin (koavf)❤T☮C☺M☯ 13:16, 5 September 2024 (UTC)[reply]
There was a bug in the classes used by the template, see Template_talk:Wikidata/FamilyTree#class="collapsible_autocollapse"_not_working. This was fixed today.
 ∞∞ Enhancing999 (talk) 13:19, 5 September 2024 (UTC)[reply]
I don't see how that responds to what I wrote, but it does seem to respond to the original comment. How is what you wrote relevant to what I wrote? —Justin (koavf)❤T☮C☺M☯ 13:47, 5 September 2024 (UTC)[reply]
You write that you prefer it in an uncollapsed form. If you read my comment and its link to additional discussion then you discover that your preference leads to category pages being replaced by family trees. I think this response both to the original comment and yours.
 ∞∞ Enhancing999 (talk) 13:50, 5 September 2024 (UTC)[reply]
Thanks. —Justin (koavf)❤T☮C☺M☯ 14:06, 5 September 2024 (UTC)[reply]
I think it should be collapsed if it is to be included in category pages. That's not really what category pages are for, not how or where people look for such data and it hides category contents (mainly subcats and files) as well as cause UI issues. Another example. Prototyperspective (talk) 13:44, 5 September 2024 (UTC)[reply]
  • The new template has been around since 2019, replacing the hand built trees that been around for a decade or more. You only see them where the person is a member of a large family such as in royal/noble lines. It isn't decorative, it is a navigation device. It sometimes is the only way to see that there is a disconnection in the family, caused by incorrect merges or the deletion of someone in the family when someone doesn't realize it has a structural need. The problem is that in the current collapsed state it is not obvious that a tree is present. --RAN (talk) 16:59, 5 September 2024 (UTC)[reply]
    The problem is that in the current collapsed state it is not obvious that a tree is present. Because the way it's wrapped into a collapsed box is having it show only the text "Byington Ford (Q5004096) [Expand]" (in this example) instead of e.g. "Show family tree [Expand]". The name is already in the cat so redundant and the Q number is not useful to the reader and also already on the page via the infobox. Prototyperspective (talk) 17:07, 5 September 2024 (UTC)[reply]
    I think "autocollapse" is the old version for collapsed state. It must have been designed to be collapsed.
    The header of the template should be improved. Merely using the name of a person and a quid doesn't really explain what it's meant to be. This isn't really better when it's expanded and covers the entire screen.
     ∞∞ Enhancing999 (talk) 17:08, 5 September 2024 (UTC)[reply]
  • See: Category:Thomas_Jefferson where consensus was to have the tree collapsed so a collapse function was wrapped around the tree, now it is doubly collapsed. --RAN (talk) 17:15, 5 September 2024 (UTC)[reply]
    Seems the problem was known, but incorrectly fixed.
     ∞∞ Enhancing999 (talk) 17:19, 5 September 2024 (UTC)[reply]
One thing I like about the wrapper on Category:Thomas Jefferson is that it actually says "Family Tree". If that text got added in the template header it would already tell people a lot more on what the information is about before/without having to expand it. ReneeWrites (talk) 18:20, 5 September 2024 (UTC)[reply]
Having it say "Family Tree" is great, having it doubly collapsed is probably something we should fix. - Jmabel ! talk 18:58, 5 September 2024 (UTC)[reply]
Looking at the Thomas Jefferson example: do I understand correctly that the tree reads right-to-left, and that for the central figure there is no indication who is the other parent of their various children, nor is there any indication of legitimate vs. illegitimate offspring? - Jmabel ! talk 19:03, 5 September 2024 (UTC)[reply]
Oh my goodness, this is heinous: now there are two collapsed boxes and you have to vertically scroll within a div?!?!? This is inaccessible and poor web design, folks. —Justin (koavf)❤T☮C☺M☯ 19:11, 5 September 2024 (UTC)[reply]
  • About the label issue, it's seems that the default text is "Family Tree (tree of ancestors and descendants)", but somehow this is overwritten sometimes by the QID. It seems that the template is just not used as designed: Sample at Category:Ulrica Christina Wellingk, the default is overwritten with |title={{Q|Q104549892}}.
     ∞∞ Enhancing999 (talk) 19:15, 5 September 2024 (UTC)[reply]
    • Jeez. At least that could be |title={{Q|Q104549892}} family tree. - Jmabel ! talk 21:14, 5 September 2024 (UTC)[reply]
      For translation purposes, it would be better if that label came directly from the template. In the sample, |entityId=Q104549892|title={{Q|Q104549892}} could be limited to |entityId=Q104549892, if the QID isn't determined directly from the category.
      To sum it up: the following maintenance seems to be needed:
      • (1) fix the label in the template
      • (2) remove |title={{Q|Qnnnn}}
      • (3) remove the duplicate collapsing (as on Category:Thomas Jefferson)

       ∞∞ Enhancing999 (talk) 08:28, 6 September 2024 (UTC)[reply]
1 only have to add in Q number once, instead of twice and have title be "Q family tree" appended. --RAN (talk) 15:58, 7 September 2024 (UTC)[reply]
2 have collapsed be default but allow that to be changed manually to open. We have Categories called X family or X noble family, here you expect to see the tree by default. --RAN (talk) 15:58, 7 September 2024 (UTC)[reply]
3 if using the year function, take out the line space between the two so it is more compact and put the years in parentheses to match our style guide. --RAN (talk) 15:58, 7 September 2024 (UTC)[reply]
I would say always collapsed. This is a media repository; family trees are metadata that we may keep here for convenience, but they should not dominate the page. It's easy enough to open it if that is what you want. - Jmabel ! talk 23:03, 7 September 2024 (UTC)[reply]
I don't see how to make any individual tree not-collapsed for the display when a category is opened. When I am working on a noble family, it would nice to have the tree open while I am working on it. RAN (talk) 02:09, 9 September 2024 (UTC)[reply]
@Richard Arthur Norton (1958- ): So for your particular purpose, you can click once to open it. But most people go to a category looking for media. - Jmabel ! talk 06:40, 10 September 2024 (UTC)[reply]

September 06

The authorization information of two images is corrupted

The licensing information of these two images of the Moldovan Parliament Hall are damaged, but free licensing information can be found based on the image sources. Please help manually repair the licensing information, thank you!

Tracked in Phabricator
Task T374235
RZuo (talk) 12:46, 6 September 2024 (UTC)[reply]

Categorizing Newspapers by date (YYYY-MM-DD)

Hi, there are a few newspaper collections here on Commons that are categorized by date, for example the Abilene Daily Reporter. Other newspapers are not categorized this way, which makes it harder to find contemporary news. With some newspapers, the amount of files is large enough that I think this work could be delegated to a bot. For starters, there is The Ohio Sentinel, with conveniently labelled files like File:Ohio Sentinel 1952-06-28 - DPLA.... Older files of the same newspaper from 1951 are less convenient: File:Ohio Sentinel August 11-August 18, 1951 - DPLA.... Then there is the Galveston Tribune with file names like File:Galveston Tribune. (Galveston, Tex.), Vol. 17, No. 100, Ed. 1 Tuesday, March 16, 1897 - DPLA.... More examples: File:Gazeta săteanului 1886-05-05, nr. 07.pdf, File:Epoca 1886-05-18, nr. 146.pdf, File:Bukarester Tagblatt 1886-05-02, nr. 095.pdf, File:Томские губернские ведомости, 1886 № 19 (1886-05-15).pdf, Journal de La Haye 03-01-1847 (IA ddd 010258178 mpeg21).pdf.

Is bot-processing feasible? --Enyavar (talk) 16:50, 6 September 2024 (UTC)[reply]

There seems to be {{Book}} (most likely) with {{{publication date}}} set to the date. Assuming it is correctly done, or corrected beforehand, a bot can easily categorize that info -- DaxServer (talk) 17:22, 6 September 2024 (UTC)[reply]
Why not use Wikidata en SD? Newspaper issues can be itemized in Wikidata with publication date (P577). Examples d:Q45747062 (Category:Journal de Bruxelles nr 76), d:Q46834135 (Category:Journal de Bruxelles nr 83), d:Q62015763 (Category:Journal de Bruxelles nr 90) Smiley.toerist (talk) 08:55, 9 September 2024 (UTC)[reply]
Hi, whatever "en SD" is, it seems to not help at all with search results. When I search for "1799-12-17", the Journal de Bruxelles does not appear in the search results (with quote marks not at all, without quote marks not before literally tens of thousands of other manuscripts from different dates, some of them decades/centuries earlier/later than 1799. Furthermore, the Category:1799-12-17 is so empty that it appears as if Commons does not have any media regarding that specific date. Only your link proves the contrary. So I think that WD-en-SD could also benefit from some categorization.
What good are the vast newspaper archives of Commons if they are buried and inaccessible unless you know the title of an accessible newspaper? Please tell me what was in the news around the world, on 1844-02-20 or another random date. --Enyavar (talk) 05:38, 10 September 2024 (UTC)[reply]
It is in the category 1799 newspapers. I have my doubts that a specific publishing date is of any use in researching a specific event. At that time there was no instant communication, and all news travelled slowly. So in practice the newspaper got dispatches from different places and times. In the issue of 76 (12 december 1799), there are many dates and places: al in the year 1799:
  • Paris: 3-12 (2X), 29-11, 27-11
  • 20-11 (Marseille), 25-11 (Nice)
  • Genua 8-11, Florence 18-11
  • Lausanne 23-11, Bern 23-11, Basel 25-11, Zurich 25-11
  • Leiden 30-11
  • London 21-11
  • Frankfurt 24-11, Rastadt 29-11
  • Stockholm 15-11
  • Sint Petersburg (Russia) 9-11
  • Istanbul (Constantinople) 29-10
  • Ratisbonne 28-11 (latest incoming news in the morning)
If you looking for a specific event, you must look at all issues, months after. News across the Atlantic could even take more than a month. Smiley.toerist (talk) 09:13, 10 September 2024 (UTC)[reply]

Are court orders in scope for Commons?

Documents such as court orders like this are usually in PD (I hope everywhere). Are such docs in scope for us here in Commons? -- DaxServer (talk) 17:32, 6 September 2024 (UTC)[reply]

In the United States, all laws, proclamations, statutes, court decisions and opinions, etc. are all in the public domain. And in principle, any of them would be acceptable here at Commons. We need works to be appropriately licensed or in the public domain in both the United States and wherever it was originally created (if these are different). I assume that court decisions are in the public domain in India and our page on copyright rules states that the government can own a copyright (unlike in the United States, where all federal government publications are public domain), so it's not obvious to me that this is necessarily public domain in India. As a non-lawyer and non-India expert, I would guess that it's legit to uplaod here, but not for sure. —Justin (koavf)❤T☮C☺M☯ 17:55, 6 September 2024 (UTC)[reply]
That's actually not correct. Court orders and the like are matters of public record in the United States. That's different from them being in the public domain or not though. They are usually PD on the federal level, but every state has their own individual copyright terms for works created by the state government. So it really depends on if the court order was created by the federal government or which state it comes from. --Adamant1 (talk) 18:38, 6 September 2024 (UTC)[reply]
*emphasis added: "The Supreme Court held that the government edicts doctrine, which states that materials created by courts in the performance of their official duties belong to the public domain, also applies to “non-binding, explanatory legal material” that is created and published by a legislative body.". Cf. Edict_of_government and "Edicts of government, such as judicial opinions, administrative rulings, legislative enactments, public ordinances, and similar official legal documents are not copyrightable for reasons of public policy. This applies to such works whether they are Federal, State, or local as well as to those of foreign governments." —Justin (koavf)❤T☮C☺M☯ 19:04, 6 September 2024 (UTC)[reply]
@Koavf: First comment https://answers.justia.com/question/2011/01/19/are-court-documents-considered-public-do-7048. I'd say that pretty much matches my understanding. "the courts have not allowed an enforceable copyright." Although that's different then if the document or parts of it can actually be copyrighted or not. Obviously if the document contains otherwise copyrighted material then it doesn't magically become PD simply for being part of the court record. --Adamant1 (talk) 22:14, 6 September 2024 (UTC)[reply]
Actually, incorporation is not "obviously" not public domain and that's exactly what Carl Malamud has spent many years fighting in court, etc. The first comment there was about Harry Potter being used as evidence and thereby attached, which is different from (e.g.) incorporating ISO standards into a law. Either way, that's not germane in this case: court orders, laws, edicts, etc. cannot be copyrighted in the United States. —Justin (koavf)❤T☮C☺M☯ 22:42, 6 September 2024 (UTC)[reply]
This order is from an Indian court. US law isn't applicable. Omphalographer (talk) 22:27, 6 September 2024 (UTC)[reply]
Please re-read what I wrote: in the United States, all laws, edicts, court orders, etc. are in the public domain, even ones from India and works that are hosted on Commons must be properly licensed or in the public domain in the United States, so the fact that it's public domain in the United States is 100% relevant. —Justin (koavf)❤T☮C☺M☯ 22:42, 6 September 2024 (UTC)[reply]
Are such docs in scope for us here in Commons? That's really going to depend on the nature of the order. Some court orders will have historical significance, or may be useful as generic examples of court proceedings. However, most cases are not notable, and most court orders even within notable cases are routine procedural motions of no lasting significance.
As far as this particular order is concerned, this appears to be a notable case (cf. en:Asian News International#Lawsuit against Wikipedia), but this court order is entirely procedural in nature and is unlikely to be useful as a reference document, particularly given the substantial news coverage of the case. Omphalographer (talk) 20:10, 6 September 2024 (UTC)[reply]
Thanks for the input! -- DaxServer (talk) 07:23, 7 September 2024 (UTC)[reply]

September 07

20@ Wikimedia COMMONS

🎉 HAPPY #20 BIRTHDAY WIKIMEDIA COMMONS & COMMUNITY !

(not as widely popular or loved as Wikipedia or Wikidata,

but hopefully worth fixing, updating and making sustainable,

if not advancing where no media server has gone before)

What are you doing today to celebrate Commons:Birthday?

Zblace (talk) 06:14, 7 September 2024 (UTC)[reply]

We really should mention this on the homepage and we should have had birthday-related media spotlighted. This reminds me of when we hit 100 million files and no one really seemed to care. :/ —Justin (koavf)❤T☮C☺M☯ 15:35, 7 September 2024 (UTC)[reply]

Is there a way to click things in my or someone elses contributions and undo them?

I have recently started using the visual file editor. A few times I found myself accidentally making erroneous edits. Is there a plugin that could allow me to just click from a list of my contributions and undoing them? Immanuelle ❤️💚💙 (please tag me) 12:06, 7 September 2024 (UTC)[reply]

To clarify, I manually undid the edits in question, but I would like to know how to do this in the future Immanuelle ❤️💚💙 (please tag me) 12:07, 7 September 2024 (UTC)[reply]
@Immanuelle: I don't think anyone but admins can do a mass revert, and unlike Cat-a-Lot, VFC doesn't have a built-in reversion feature. Do preview at least one example before having VFC act on an enormous number of files! - Jmabel ! talk 14:21, 7 September 2024 (UTC)[reply]
@Jmabel that's unfortunate. Is there some kind of reason for a rule against it or just no tools available? I feel it would be helpful at lease for people reversing their own edits.
I'll try to be more careful. I am still new to using the tool but I think I figured out where I went wrong, and will try to use the preview more. Immanuelle ❤️💚💙 (please tag me) 14:25, 7 September 2024 (UTC)[reply]
@Immanuelle: It's built into Cat-a-Lot, so it's not a general principle. Yes, some sort of mass self-revert might be a good thing; I don't know the technical details. You wouldn't want people to be able to use it after more than an hour or so, though, because we occasionally get someone doing nasty stuff in a fit of pique, and imagine the chaos they could create with a year of self-reverts. - Jmabel ! talk 14:44, 7 September 2024 (UTC)[reply]
Everyone makes mistakes and if you do so on a large scale and do your best to fix it and ask for help, that's not a problem. Rollback allows you to undo your or someone else's edits with a single click. —Justin (koavf)❤T☮C☺M☯ 15:34, 7 September 2024 (UTC)[reply]

Many images of book scans

Is there some policy about how book scans are to be handled? I think it would be best if books and similar literature were uploaded as one PDF document rather than as many scattered separate image files.

This way one can easily download or read them / go through pages and they don't clutter search results or require subcategories for just one literature item. For scans already uploaded, a bot could convert them into PDF files. One can also embed single PDF pages like an image on Wikipedia so there's not really any disadvantage to converting them to document files. Example example. --Prototyperspective (talk) 17:22, 7 September 2024 (UTC)[reply]

How about we leave them as the projects that are using them are uploading them? One may be able to embed single PDF pages, but I don't know how, and it's hard to find the appropriate illustration pages. Not to mention that PDFs use compression, and if the originals were PNGs, converted to JPEG2000 for the PDF, and then converted to JPG for Commons, that makes worse final images than directly working with the original scans.--Prosfilaes (talk) 18:19, 7 September 2024 (UTC)[reply]
I described several reasons why not to keep them as people and projects upload them. Moreover, they could change how they upload them. Scans of one page of a book are nearly never used on a Wikimedia project and when they are used often having the full book available at the page would be much better and useful. How to embed a specific page of a PDF is described here: Help:PDF#Page. As for quality they could be converted without loss of quality and it doesn't have to be PDF files if there is something better. Prototyperspective (talk) 18:24, 7 September 2024 (UTC)[reply]
They could be converted without loss of quality how? You're adding another level of conversion.--Prosfilaes (talk) 06:04, 8 September 2024 (UTC)[reply]
With a command that converts losslessly. Maybe ImageMagick's convert ./page*.png ./output.pdf already converts losslessly and somebody should check if that is the case. Prototyperspective (talk) 13:20, 8 September 2024 (UTC)[reply]
  • I am not against merging into a pdf or djvu file, but as pointed out previously book illustrations can be used by other projects as well as the title page to illustrate the Wikidata entry. --RAN (talk) 20:51, 7 September 2024 (UTC)[reply]
    Can one not specify the page of a PDF document on Wikidata for the image/cover property? Maybe that functionality can be added. Then people could also upload the cover or a particular image separately rather than uploading a whopping 300 files for one book. That is also useful because the title can be more descriptive and here is an example of an image from a PDF file that is used on Wikidata...it's much better than having the image just as some whole-page generically titled document scan image. Prototyperspective (talk) 21:04, 7 September 2024 (UTC)[reply]
    As far as I'm aware, the default thumbnail for PDF/DjVu documents is always a thumbnail of the first page. Which is a shame; there are a lot of books where that page is a Google Books notice, or an image of a badly damaged cover. Being able to override that with a thumbnail of the title page would be a huge improvement. Omphalographer (talk) 21:33, 7 September 2024 (UTC)[reply]
@Prototyperspective: I feel like the benefits of something like this would depend on the subject of the book. Like if its mainly or only text, cool. Convert it into a PDF and get rid of the JPEGs. There's of books with photographs or illustrations of historical subjects and places where it makes sense to have individual images. And not just for other projects but in general. Its kind of redundant to have individual images for books that are purelu text though but I think they are easier to transcribe for other projects that way. But I'd still argue its pointless (if not against the guidelines and /or goals of the project) to have jpegs of individual pages that are purely text on our end. --Adamant1 (talk) 21:22, 7 September 2024 (UTC)[reply]
Strong  Support towards converting multi-page documents into multi-page document formats - PDF. (I also posted some thoughts about this once) ~TheImaCow (talk) 21:28, 7 September 2024 (UTC)[reply]
I tend to agree with the idea that it should just one file, but PDF support hasn't really gotten better. It's actually somewhat broken and even the WMF person in charge of Commons has no idea if or when it's going to be fixed.
 ∞∞ Enhancing999 (talk) 21:49, 7 September 2024 (UTC)[reply]
 Weak oppose Maybe for type-written works it makes sense to only have a PDF, but for hand-written it's often really useful to have higher-resolution files and if they were combined into PDFs the PDFs could be quite large. A small example could be something like this where it's often useful to be able to zoom right in on a letter to better see the pen strokes etc. It's also more likely that people preparing PDFs will end up with trusting whatever their scanner software gives them and by doing so be throwing away information that we'd rather keep (e.g. postmarks on letters). I wonder if more could be done with the search system and structured data to make it easier to exclude individual scans? Sam Wilson 04:47, 8 September 2024 (UTC)[reply]
Maybe a Wikidata property for scan quality or something that could be used to push bad quality PDFs and images down in the search results? I've been wanting something like that for awhile now to suppress crappy scans of postcards from the search results. --Adamant1 (talk) 05:02, 8 September 2024 (UTC)[reply]
@Adamant1: Yes, good point. Maybe DPI for original size (P10300) would suffice? I've not ever bothered adding that, but it would totally make sense. Sam Wilson 05:12, 8 September 2024 (UTC)[reply]
Good call. I didn't know that existed. Although the "for original size" thing seems a little needless, but whatever. I don't see why it wouldn't work. --Adamant1 (talk) 05:26, 8 September 2024 (UTC)[reply]
If you have poorly printed pages, where OCR does not work, pdf is not an usefull format. For the Category:Journal de Bruxelles, I scan most pages, two at a time folded open. I have a standard naming principle for the files. In Wikisource the pages have to be manualy typed over as the OCR does not work. A lot of work, but a usefull result.Smiley.toerist (talk) 11:16, 8 September 2024 (UTC)[reply]
 Comment Commons doesn't have an official policy on this, but I like to keep in mind what other projects might need these files for and how. For regular books, Wikisource needs an original .pdf or .djvu to be the source/reference material, and illustrations exported separately, usually retouched. See for example here: https://en.wikisource.org/wiki/Page:The_Osteology_of_the_Reptiles.pdf/23 ReneeWrites (talk) 10:39, 8 September 2024 (UTC)[reply]
Note that Wikisource can also work with individual image files as well as PDFs/DjVus. Sam Wilson 12:52, 8 September 2024 (UTC)[reply]
 Comment I've been working on Category:Helix (newspaper). If these had been PDFs, I'd have had to extract files for almost every page to document them decently. Yes, there are probably times whe PDFs are best, but it's not a universal. - Jmabel ! talk 12:15, 8 September 2024 (UTC)[reply]
 Comment something I find suboptimal with individual pages is that {{Book}} doesn't really seem to be adapted for it. Also, when actually using elements from a page, one generally needs to do another crop just for that part. In the end, that part and the original page might end up categorized for that content.
 ∞∞ Enhancing999 (talk) 13:23, 8 September 2024 (UTC)[reply]
  • Remember nothing actually gets deleted, so we are not saving any space. We also have had the case where a page was missing, a page was out of order, or a page was upside down in the pdf, and we had to recompile the file. Perhaps if all the individual pages were in a folder there would be less clutter once the pdf was created. --RAN (talk) 18:01, 8 September 2024 (UTC)[reply]
    It's not about space but about clutter. It also causes misleading upload stats and clutter the UploadedFiles page. The main issue is the clutter in the search results and from views that e.g. combine categories. The many subcategories are also clutter. Then it's also hard to download the file and a main problem is that it's hard to read a document if it's in dispersed pages instead of in one document. And I think this also applies to old 1800s drawings of organisms which are not really useful anymore, not what people look for, and for which there are now high-quality photos & illustrations. If an image is needed one can specify the page of the PDF or extract an image (that is different from a whole page) from it. Prototyperspective (talk) 18:21, 8 September 2024 (UTC)[reply]
    This also applies to old 1800s drawings of organisms which are not really useful anymore, not what people look for, and for which there are now high-quality photos & illustrations. I strongly disagree with this. You have made the common mistake of assuming that what you find useful or interesting is what other people also find useful or interesting, and what you find irrelevant or annoying is what other people also find irrelevant or annoying. For some users of the Commons, "old 1800s drawings of organisms" are in fact more useful than modern photographs, and individual page scans are more useful than PDFs, especially if PDFs of the same public domain documents are already widely available from Google Books, Internet Archive, Hathi Trust, and other repositories. We all use the resources here differently, and the specific resources you like to use and the way you like to use them are not universally shared. There may be good reasons to consider changing the way that multipage documents are handled, but generalizations about "what people look for" based solely on your own preferences are not among them. Crawdad Blues (talk) 20:00, 8 September 2024 (UTC)[reply]
    No, I did not make this mistake – maybe I have made this mistake of not clarifying which of my statements are a statement of opinion/argumentation. It's not surprising that there are users who think so since there are users who upload such images particularly as separate pages and with that sentence I was addressing an earlier comment arguing it would be good to do so for images that contain images. These users are very few and I have yet to see any actual usefulness case / application of such images. It's great that we have them on WMC, but I don't see why having them in separate scans would be reasonable. Please make a survey or look at pageviews or explain specific use-cases with examples if you think what I said is wrong. Prototyperspective (talk) 20:17, 8 September 2024 (UTC)[reply]
    (With such images I was referring to separate pages as scanned page image files, not whether or not these files are on WMC.) This proposal is mainly about documents with text that include images, not documents which contain only drawings as a page. I think an example of a WMC search that shows mostly scans of a few books and burying what most users are looking for is needed to illustrate what I mean. Here's a bad but maybe still useful example: if you search for "animal" there is an undue number of scans from Category:The animal kingdom, arranged after its organization, forming a natural history of animals, and an introduction to comparative anatomy (1834) at the top instead of higher-quality drawings/artworks and photos (like the collages at the very top). They can be useful but often inaccurate unreliable 1834 drawings are usually not what the user can use for any of the purpose such as adding them to a WP article or illustrating the species. These images (the book) are still useful and it's worse for other searches. Prototyperspective (talk) 20:35, 8 September 2024 (UTC)[reply]
    I agree with Crawdad Blues: images of images from books (and other documents), like drawings, prints and photographs, should stay and be images: one in a file. Then they are easy to reuse and then end users do not have to go through the upload process, but can directly use the file. For a type-written, printed book that primarely is about text: yes, then a PDF file is easier, for searching within the text and to leafing through it. JopkeB (talk) 05:17, 9 September 2024 (UTC)[reply]
It depends a lot on each specific case, but, if uploading a new book, I think PDF is the best option. If the book contains images that are interesting in their own right, they should be uploaded as separate files (or even be left for another user who is specially interested to upload them, if you don't have time for it). For already uploaded books, I think that existing images should not be deleted, but creating a new PDF book can be a good idea.
I think Commons, for some cases, can be a good book deposit/library, in addition to Wikisource (with OCR or some freely licensed or public domain original PDF books, you can even have machine-readable PDF books in Commons with relatively little effort, while migrating their content to Wikisource can be a really big work in many cases). MGeog2022 (talk) 10:51, 10 September 2024 (UTC)[reply]

Use of NoFoP-category template on broad categories

I'm in a bit of a disagreement with someone on this issue, so I thought I'd ask for clarification here: should Template:NoFoP-category be applied to broad categories in a NoFoP country (e.g., "Statues in X city" or slightly more specific ones like "Statues of animals in Y city")? My understanding is that it should be reserved for categories dedicated to specific structures or monuments, as some in those cities may be old enough to no longer be under copyright protection. What are your thoughts? — Golden talk 21:31, 7 September 2024 (UTC)[reply]

The (English) text of the template suggests to me that it's meant for specific works, not classes. Not entirely sure if we should have such a template.
 ∞∞ Enhancing999 (talk) 21:47, 7 September 2024 (UTC)[reply]
I would never put it on that broad a category. - Jmabel ! talk 23:07, 7 September 2024 (UTC)[reply]
I agree with Enhancing999 and Jmabel. This belongs as a category header for specific works in NoFoP countries, for the reason you stated. ReneeWrites (talk) 10:23, 8 September 2024 (UTC)[reply]
Totally agree with everyone else. It doesn't make sense to use the template on a general topic because works in the category (or that will be uploaded to it) will inevitable be PD due to age or other factors. --Adamant1 (talk) 10:30, 8 September 2024 (UTC)[reply]
  • Disagree I now accept that the template should not be used in broad categories. All such templates that I erroneously added to categories have been helpfully deleted by the nominator and his friends. Thanks for that guys. Moving to non-broad category usage, the example cited - "Statues of animals in Y city" - does not fit the broad criterion in my view. As I mentioned to @Golden: on my talk page, "In the case of the "Statues_of_animals_in Qəbələ" category, I see two pics of wooden eagles and one pic of dinosaurs. I conclude that it is not general, it is not broad, the statues are new, it is quite specific (a triple intersection of art (statue), subject(animals) and location (Qəbələ)). So as far as I'm concerned, it meets all of the criteria". Certainly where all the pics in a category contain nothing but images which, of themselves, would all be legitimate candidates for marking with the template "NoFoP-Azerbaijan", then I think that a categorical marking is merited and a more efficient use of editors' time. Laurel Lodged (talk) 13:23, 8 September 2024 (UTC)[reply]
    @Laurel Lodged or better, leave the generalized categories with no such templates? I was the one who proposed this template and its original purpose was for specific works. Leaving the generalized categories without such templates removes the unsightly clutter in the said categories. JWilz12345 (Talk|Contrib's.) 13:45, 8 September 2024 (UTC)[reply]
    TBH, I can see the merit of using this template in certain broad categories where virtually any photo will be a FoP violation. For instance, Qatar has strict FoP, and essentially no buildings or public artwork old enough to be in the public domain, so placing FoP warning templates on related categories might be warranted. Omphalographer (talk) 21:02, 8 September 2024 (UTC)[reply]
@Omphalographer: I'd be interested to know what exactly you think the merrits of doing it that way are. People are going to add images of copyrighted works to the categories regardless. The same could be said for the main categories for the subjects in the image, but at least users expect to see the template there. Plus at least in my experience broad categories tend to not be persistent anyway. So it's just encouraging people to add the template to the main subject level category instead of the one for the particular work and in an instance where the template will likely be removed or deleted at some point anyway. Categories for specific works at least have less chance of being deleted, and again, users expect the template to be there. --Adamant1 (talk) 00:12, 9 September 2024 (UTC)[reply]
@Omphalographer: I just fixed Adamant1's ping above, so it may not have pinged you. - Jmabel ! talk 06:52, 10 September 2024 (UTC)[reply]

September 08

Template:How to delete empty categories

When working on Commons:Report Special:UncategorizedCategories, I noticed that many categories are there, merely because users don't know how to have categories deleted with typos or other minor problems.

Template:How to delete empty categories (brought to my attention by @Jmabel) is meant to explain it to them.

I think it could be formulated better, though I'm not entirely sure how. Personally I prefer {{Badname}} or {{empty, parentless category}}. Also I noticed some users adding {{SD|C2}} including nowiki tags.
 ∞∞ Enhancing999 (talk) 12:50, 8 September 2024 (UTC)[reply]

Someone is not doing it right if they include those nowiki tags.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 03:18, 9 September 2024 (UTC)[reply]
Yeah, accordingly, I hope we can improve the template
 ∞∞ Enhancing999 (talk) 09:07, 9 September 2024 (UTC)[reply]
I usually just use C2 unless it's clearly a bad name. I actually didn't know C1 was an option until just now. So things could clearly be explained better. Heck if I have any idea how though. --Adamant1 (talk) 03:31, 9 September 2024 (UTC)[reply]
Whatever we do: keep the template short. If this is going to turn into a wall of text, it should be in a page linked from the template, not the template itself. - Jmabel ! talk 06:55, 10 September 2024 (UTC)[reply]
Yeah. It's an important template. Commons:Report Special:UncategorizedCategories is again full of categories that could have been tagged for speedy deletion.
 ∞∞ Enhancing999 (talk) 07:01, 10 September 2024 (UTC)[reply]

September 10

Munich metro type

I would normaly put these metrotrains into the Category:MVG C (U-Bahn München), However the windows are different to the other images in that category. They ont have the pull open part. Is this an early C type, where later the windows where replaced or is it a totaly different type? Smiley.toerist (talk) 11:20, 10 September 2024 (UTC)[reply]

Button for 3D models added on main page

A button that shows the 3D models was added on the main page, next to images and videos. It is a good point in time to highlight textured meshes that could be presented :) --PantheraLeo1359531 😺 (talk) 11:41, 10 September 2024 (UTC)[reply]

Retrieved from "https://commons.wikimedia.org/w/index.php?title=Commons:Village_pump&oldid=922270554"