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 ↓
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/2022/10.

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:

# 💭 Title 💬 👥 🙋 Last editor 🕒 (UTC)
1 Permissions 10 4 Kognos 2022-10-21 14:20
2 AI-generated works 24 8 Whym 2022-10-18 10:51
3 UploadWizard and location metadata exclusion 23 8 JWilz12345 2022-10-19 17:03
4 A tool/shortcut to make upload categorization faster? 7 4 Danidamiobi 2022-10-17 09:17
5 Updated copyright law of Vietnam and no more FOP? 18 8 Violetbonmua 2022-10-23 17:30
6 How to make a structural solution for not empty disambigious pages? 20 6 Auntof6 2022-10-20 06:32
7 What should be in a description field (and what not)? 15 6 RTG 2022-10-20 09:20
8 How to use Flickr images with All Rights Reserved (C) on Commons. 3 3 LPfi 2022-10-22 13:34
9 Wikistories on Commons: a proposed visual content tool in WikiCommons 8 5 RZuo 2022-10-19 12:52
10 Commons:File renaming 1 1 Speravir 2022-10-17 23:15
11 Illustrations that are pure speculation named after the subject "depicted" 7 5 RZuo 2022-10-19 12:52
12 ImageAnnotator sometimes does not work 2 2 Jmabel 2022-10-18 15:18
13 Delink dead "Leave feedback" at Special:UploadWizard 3 2 Mateusz Konieczny 2022-10-18 18:38
14 Special:UploadWizard - better for new users 5 3 Speravir 2022-10-18 22:59
15 CC Template for files from Salt Research 6 3 Brainulator9 2022-10-19 18:16
16 Files from the Australian Human Rights Commission Flickr stream 4 2 Animalparty 2022-10-23 01:06
17 INVITE: Queering Wikipedia 2022 Meeting on Friday and Saturday (October 21-22) 1 1 Zblace 2022-10-20 06:08
18 UCoC EG Community review period closed 1 1 Zuz (WMF) 2022-10-20 11:37
19 What is the copyright status of West African CFA franc and Central African CFA franc? 4 2 Jeff G. 2022-10-20 17:28
20 File deleted with unclear reason. User in charge not available 1 1 Achim55 2022-10-22 10:23
21 emoji cats 6 2 Jmabel 2022-10-24 14:45
22 Videos whose components are hard to verify 2 2 Ruslik0 2022-10-22 19:03
23 Can someone download a hires version of this image? 6 4 Richard Arthur Norton (1958- ) 2022-10-24 21:28
24 dark mode? 3 3 Jeff G. 2022-10-24 05:41
  • 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 29[edit]

View it! Tool
Upcoming View it! tool discussion - for those interested in Wikimedia Commons & structured data - August 31, 2022

A project has been funded by the Wikimedia Foundation as part of the Structured Data Across Wikimedia Work to create a tool called View it! The tool aims to increase the discoverability of images on Commons, give readers and editors access to more images, and encourage contributors to utilize Commons & structured data. Please visit the Meta page if you are interested in trying out the prototype. We are having a demo and feedback session on August 31st at 16:00 UTC, please join us if you wish!

We hope to see you there! Sincerely, Dominic, Kevin, & Jamie

October 07[edit]


I have uploaded a lot of my image files to Commons without thinking much about permissions, but I probably should think! On this file: Panorama of Bryggen, Bergen I have stated: "Acknowledgement to Wikimedia Commons is sufficient. No author acknowledgement is required when re-using this file". Is there a template or standard form of words for this? Thanks. Kognos (talk) 19:49, 7 October 2022 (UTC)Reply[reply]

@Kognos: What you appear to seek is a {{Cc-sa-1.0}} license, but it is deprecated, retired, and not recommended. If you don't think even your username deserves credit for your work, please seek professional help.   — Jeff G. please ping or talk to me 20:25, 7 October 2022 (UTC)Reply[reply]
@Kognos and Jeff G.: Deserves is one thing, needs is another. Why wouldn't somebody want not to ask for credit? I thought seriously about why I'd use CC-BY-SA and not CC-BY, and many people even use CC-zero. For the original question, you can specify an attribution line in the {{Information}} template (which is used for the file description in most cases): other_fields={{credit line|Other=Wikimedia Commons|License=CC-BY-SA-3.0}}, choosing not to specify your (user) name there. That line dictates the attribution for the CC-BY licenses, I think. Specifying just the name and no other words makes it easier for those that want to reuse the file in the context of other languages. There may be other caveats. –LPfi (talk) 15:46, 10 October 2022 (UTC)Reply[reply]
You can also use something like {{self|cc-by-sa-4.0|author=YOUR PREFERRED ACKNOWLEDGEMENT}}. - Jmabel ! talk 17:34, 10 October 2022 (UTC)Reply[reply]
@Jmabel and Jeff G.: To give a bit of context, here's a page that uses two of my photos: The credit to Commons, with a link, seems appropriate. I'm not sure what incuding my username would add for readers who are not familiar with Commons. Anyone interested can folllow the link and find out more. The cc-by-sa-4.0 license is fine, I was just wondering if there was a standard form of words for this form of acknowledgement. Thanks for the tips. Kognos (talk) 12:52, 13 October 2022 (UTC)Reply[reply]
@Kognos: Those two photos are not "©Wikimedia Commons 2019", you hold the copyright (unless you transfer the copyright in writing).   — Jeff G. please ping or talk to me 14:49, 13 October 2022 (UTC)Reply[reply]
@Jeff G.: Quite right, it should be a credit to Commons as the source, not a copyright notice. Thanks.Kognos (talk) 15:31, 13 October 2022 (UTC)Reply[reply]
@Jeff G.: I contacted the site, and the credit has been amended. Kognos (talk) 16:40, 19 October 2022 (UTC)Reply[reply]
@Kognos They still don't link to,_Meldreth_05.jpg or,_Meldreth_09.jpg .   — Jeff G. please ping or talk to me 23:08, 19 October 2022 (UTC)Reply[reply]
@Jeff G.: They link to the Commons category containing the two images as well as others of the cement works. Looks reasonable to me. Kognos (talk) 14:20, 21 October 2022 (UTC)Reply[reply]

October 08[edit]

AI-generated works[edit]

How should Commons treat AI-generated works? I recently saw some entries which were generated with a prompt asking specifically for images in the style of specific artists. My understanding is that the copyright status of such images is still uncertain, provided they are sufficiently distinct from existing works, but to me, the ethical issues of recreating someone's style, at scale, and then releasing it under CC0, is unacceptable. Thoughts? Ovinus (talk) 02:09, 8 October 2022 (UTC)Reply[reply]

"An astronaut riding a horse, in the style of Monet". Monet did not paint this image, and even if he were alive today, he is not the copyright holder of this work simply because of the brushstroke patterns.
There's three parts to this question: "Does AI-generated art copy existing images?", "can art styles, compositions and motifs be copyrighted?", and "is it ethical?", so it's a bit more nuanced than a simple one-line answer.
  • Does AI-generated art copy existing images? No, a diffusion model that generates images from text prompts does not take bits and pieces of existing images and paste them together, contrary to common misunderstandings of how AI art works. Diffusion models generate images based on how the algorithm has been taught by the dataset it has learned from. These AI models do not store the images themselves, and they merely generate an image based on how a particular prompt is parsed and weighted. Thus, generated images cannot be considered derivative works, since there is no image to derive from. That said, if an AI is prompted to generate a visually indisputably copyrighted object, for example the Eiffel Tower, Spiderman, or the Taco Bell logo, then those visuals would be protected by copyright.
  • Can art styles, compositions and motifs be copyrighted? At least not under United States copyright law. An artist cannot copyright a brushstroke style, a photographer cannot copyright a particular shooting angle, and a musician cannot copyright a sequence of four chords; there has been no legal precedent where mere motifs can be protected under copyright. If nebulous visual ideas such as motifs can be copyrighted, then no artist on the planet would be able to create art any more, since anything that is made today would inevitably slightly resemble something else made 20 years ago, 40 years ago, or 60 years ago, by any one of thousands of artists. In fact, there is an entire category of artists who have the sole gimmick of imitating the style of famous painters, and they are freely able to sell their works commercially.
  • Is it ethical? This is a subjective question where the answer will differ depending on who you ask. If you ask me, then it's probably not ethical. But Commons is a repository of free content, and not ethical content. There are images of exhumed war graves on Commons that some might find objectionable, there are images of the Prophet Mohammed on Commons, among other things. I don't believe ethics should be a consideration point when it comes to AI art, as neither Commons scope nor Commons policy outright forbids "unethical" content.
Regards, --benlisquareTalkContribs 03:02, 8 October 2022 (UTC)Reply[reply]
Your first two arguments, together implying that most Stable Diffusion–generated images are not copyright infringement per se, are compelling, although I would like to eventually see some US case law, and I'm still not sure how you can verify that all the images are sufficiently different from existing pictures (not necessarily by the artists in the prompt). The connection between emulating a person's work through digital processing seems far more direct than the standard fare of artists taking motifs and styles from others—direct in a way that a court might find germane to a finding of infringement.
I simply disagree with your third statement about Commons being a repository of copyright-free content, without regard to ethics. Commons:Photographs of identifiable people is a simple example; there is no copyright concern in uploading candid pictures of people in public, but there are certainly ethical concerns. Luke Plunkett in Kotaku puts it better than I could: If you train an AI to perfectly match a specific artist’s style, I think that obviously violates the artist’s rights somehow, if not their copyright. I just don’t know how to legally enforce that. ([1]). Ovinus (talk) 03:32, 8 October 2022 (UTC)Reply[reply]
This would be a case where the law would need to catch up with the technology. US Congress moves slowly, but until a decision is made that explicitly says "no, you can't do that", I don't believe Commons should act upon "what ifs".

"I'm still not sure how you can verify that all the images are sufficiently different from existing pictures" - Let's put it this way; had the prompt not been added to the file description, how would you even know that the image emulates a particular artist's style? I have my doubts that you'd simply be able to eyeball it and go, "yep, that's Greg's painting style". AI is groundbreaking, edge-cutting technology, but it's not perfect either, and makes errors all the time; what the AI might think is Sophie Anderson's art style might just be something else completely. When I lay out which prompts I use, and what configuration settings used to get to the final product, as I'm tinkering around with Stable Diffusion, I'm doing so because I feel like I have the obligation to be transparent about how I got there; imagine a scenario where someone less idealistic just plops down a piece of art with no context - how would anyone even tell with certainty?

Finally, regarding ethics on Commons: Was it ethical when we deprived a photographer of his livelihood, all to prove a legal technicality regarding animals? It's not a clear-cut "yes we are, no we aren't" as one might hope; yes, we do have guidelines regarding the privacy of identifiable people, but that is only one facet out of many different ethical issues regarding the kinds of images hosted here. If we are to come up with a decision on AI art that imitates certain styles, it would be Commons consensus-based, and not based purely on what is ethically correct. --benlisquareTalkContribs 03:37, 8 October 2022 (UTC)Reply[reply]

To further illustrate what I mean in regards to the imperfections of AI: Here, just a few seconds ago, I punched "A Toyota 86 in anime style" into Stable Diffusion, and obtained this image. This is clearly not a Toyota 86; rather, it appears to be an unbadged car with Ford vents and the side elements of a Mercedes Benz AMG. Even if an AI text-to-image generation prompt requested for particular art styles, how would you know with certainty that the AI's output is correct, and that the art indeed emulates Van Gogh, or Picasso, or Da Vinci? --benlisquareTalkContribs 04:10, 8 October 2022 (UTC)Reply[reply]
I don't know all the legal or ideological background, or if it's even valid, but we do have {{PD-algorithm}}. Some recent discussion on the subject is here and here. Addendum: it looks like the PD-algorithm template was conceived here. --Animalparty (talk) 03:19, 8 October 2022 (UTC)Reply[reply]
Hm, interesting. {{PD-algorithm}} definitely makes sense for some things, including many AI-generated works, but there are still three potential copyright holders whose roles must be considered: programmer, user, and artist emulated. Those discussions seem to mostly consider the first two. Ovinus (talk) 03:42, 8 October 2022 (UTC)Reply[reply]
Seems to me that the issue may not be so much infringing rights of the artist whose style is copied so much as whether the image may derive from one or more other images. E.g. if I have an AI make a portrait of person A in the style of artist B, it might infringe photographer C's photo of person A. - Jmabel ! talk 17:03, 8 October 2022 (UTC)Reply[reply]
If photographer C's photo of person A was never used as any component of the generated image, would it still infringe upon the copyright of that photo? As an example, when I last went to the Art Gallery of New South Wales, there were numerous visitors huddled around various marble busts drawing notebook sketches; would drawing influence without outright copying it still fall foul of copyright? Because while the portrait of person A might draw influence from motifs found in photographer C's photo, it wouldn't copy portions of photographer C's photo, because it outright would be incapable of doing so. The model for Stable Diffusion is 3.97GB, and is sitting on my hard drive right now; the training dataset used to create the Stable Diffusion model via machine learning consists of over 600 million images, spanning over 200+ terabytes. There is physically no way the original image dataset can be stored within the AI diffusion model; the model only contains a set of algorithms that it can use to generate instruction-based images, and is merely trained by the photo from photographer C. This is no different to an art student looking at a marble bust of Augustus, and then sketching out a pencil sketch of Arnold Schwarzenegger based on that marble bust. --benlisquareTalkContribs 18:46, 8 October 2022 (UTC)Reply[reply]
"Does AI-generated art copy existing images? No" - actually sometimes it created clearly derivative works or nearly exact copies. See from Left is what prompt “Girl with a Pearl Earring, by Johannes Vermeer” to Stable Diffusion produces. On right: original image. In some cases specific images are recorded and restored. In some cases part of images are exact copy of other works. For example sometimes watermarks are appearing Mateusz Konieczny (talk) 12:15, 11 October 2022 (UTC)Reply[reply]
This image contains a Getty watermark. However, no such image exists within the Getty library.
From what I understand, Getty watermarks appear when you use prompts such as "flowers" or "warzone", and signatures appear when you use prompts such as "masterpiece" or "artstation", because from the hundreds of thousands of training images the AI model has churned through that have image-caption pairs containing those keywords, it associates the phrase "masterpiece" with adding a signature to the image, because that's what is taught from the original training set (many original images with these keywords will contain an artist's personal signature). In these cases, the AI is inventing a signature and putting it on the image, it's not actually taking a real signature. Same goes for the Getty watermarks, it isn't getting the watermark from an existing image, it's generating a watermark because it associates those keywords with watermarks, and thinks that's what you (the user) actually wants, based on the prompt. As for replicating existing works, of course it will try its best to replicate well-known works if you put the title of the work in the prompt as Alexander Wales has done (his prompt was literally "Girl with a Pearl Earring, by Johannes Vermeer"). In this case, it would 100% be a copyvio and a derivative work, yes. But, again, this copyvio comes from the AI model constructing Girl with a Pearl Earring by scratch, it is not fetching an image of Girl with a Pearl Earring from its storage compartment somewhere and then photobashing together an image; both are copyvios, but the journey is different. --benlisquareTalkContribs 12:46, 11 October 2022 (UTC)Reply[reply]
Exact method of storing image does not really matter. For copyright purposes it does not matter whether I stored image data as JPEG, weights in image generator, as text description of each pixel group, wool construction in Minecraft or by encoding it by arrangement of cages with hamsters. Image data was still stored. In case of watermark: in many cases watermark was copied independently from other image parts, and as usual it is not a proof that image is copyrighted or made by whoever watermarked it. But it is clear evidence that AI generation sometimes copies image parts and rearranges them to create new one. Mateusz Konieczny (talk) 10:51, 13 October 2022 (UTC)Reply[reply]
"copyvio comes from the AI model constructing Girl with a Pearl Earring by scratch, it is not fetching an image of Girl with a Pearl Earring from its storage" - actually it depends on that. As I understand, copyright status depends on whether it is an independent creation or derivative work - even if outcome would be the same. So in case of reproducing copyrighted work (or something extremely close to it) it would matter that it is derivative work. Mateusz Konieczny (talk) 10:53, 13 October 2022 (UTC)Reply[reply]
Well yes, no one is disputing that the Girl with a Pearl Earring example is a derivative work and a copyvio. If someone uploaded an imitation work of an existing work using AI software onto Commons, it would most definitely fall under copyvio. The debate here is whether it's a copyvio to copy the style of a specific artist, for example, Vincent van Gogh's uniquely textured brushstrokes (but for an artist who is still alive). "But it is clear evidence that AI generation sometimes copies image parts and rearranges them to create new one" - I'm sorry, but you're going to have to provide a citation for that, it's not clear evidence at all given the fundamentals of how latent diffusion models work. The AI model knows what a Getty watermark looks like, because it is trained from a Common Crawl dataset containing 600 million+ images scraped from all across the internet. Just because the training dataset contains copyrighted images, does not mean that the AI's output image contains those same images photobashed together. Whether or not a user intentionally recreates a copyrighted object such as the Eiffel Tower or Spiderman is up to a discretionary decision by deletion review on Commons; however, it is a fundamental misunderstanding of how the technology works to assume that all images are merely photobashed together from existing copyrighted images. --benlisquareTalkContribs 13:48, 13 October 2022 (UTC)Reply[reply]
In summary of what I am trying to say here: Asking the AI to create for you "please draw me the Mona Lisa" would be a copyvio, as it creates a derivative work. Asking the AI to create for you "please draw me a Tesla Cybertruck, but using Vincent van Gogh's style of brushstrokes" is not a copyvio, because works can be copyrighted, while compositions and motifs cannot be copyrighted (under US copyright law, where the Wikimedia servers are located). --benlisquareTalkContribs 14:02, 13 October 2022 (UTC)Reply[reply]
I agree that copying style is not against current copyright law. "I'm sorry, but you're going to have to provide a citation for that, it's not clear evidence at all" - I have already given example of watermark. Due to it being really common in the exactly same format it is especially likely to end in this way but it clearly proves that there is at least one case where element is a direct copy of what is appearing on images. More powerful models may start doing it also with more elements. "assume that all images are merely photobashed" - I am NOT claiming this. I am claiming that at least some elements may be direct copy of other elements and we have no real way to detect this. Mateusz Konieczny (talk) 18:23, 14 October 2022 (UTC)Reply[reply]
Pictogram voting comment.svg Comment in our (Philippine) context, AI works' wligibility for copyright protection is subject to academic discourse. Without existing court cases internationally, AI works' copyright statuses will remain shrouded in academic circles. See the October 2021 edition of the Copyright Bulletin by our Intellectual Property Office of the Philippines (IPOPHL), pages 9–13. JWilz12345 (Talk|Contrib's.) 13:00, 11 October 2022 (UTC)Reply[reply]
Pictogram voting comment.svg Comment no, copies (even very literal copies) of "Girl with a Pearl Earring" or the "Mona Lisa" are not copyvios, because these works are in the public domain, but equivalent reproductions of copyrighted works would be problematic. - Jmabel ! talk 15:12, 13 October 2022 (UTC)Reply[reply]
Yes, I agree (and I think that I managed to not describe this as copyvio). I have used this image because there was a direct linkable example, and it was sufficient to demonstrate that sometimes individual images are stored and reproduced Mateusz Konieczny (talk) 18:26, 14 October 2022 (UTC)Reply[reply]
After/before copyright issues and ethics issues are sorted out (and I don't think they are yet), you need to show that an artwork is in scope. Published works by famous artists in the past are easy in this regard - they are used to illustrate articles about the artists (or sometimes the works) themselves. As for AI art, the same argument could apply, that is, to illustrate articles about the engines. However, I'm not sure if that alone could justify a large number of uploads. You don't need nor want to add hundreds of images to one Wikipedia article. While he number of modern amateur artworks hosted on random subject matters may not be zero on Commons, I believe it is limited (unless they serve practical purposes similar to charts and diagrams). COM:HOST is often cited in deletion requests of amateur works, although they are not always deleted. I don't see why we want to be more accepting towards AI art than towards amateur art. whym (talk) 08:21, 16 October 2022 (UTC)Reply[reply]
Of course. As long as there is a valid educational use in various Wikimedia projects, and there is a sufficient threshold of quality being met, there shouldn't be any problem in that regard. Given that this kind of software makes it too easy to generate large quantities of images, often with varying usefulness, relevance, or quality, is is likely that in future we'll see more dubiously useful content being uploaded, and such content would certainly need to be dealt with at COM:DR accordingly, similarly to how we deal with dubiously useful photographs. --benlisquareTalkContribs 08:52, 16 October 2022 (UTC)Reply[reply]
Just to make the discussion more concrete, I think the image of "An astronaut riding a horse, in the style of Monet" linked above would not be useful in relation to Monet, nor astronaut, nor horse, while it might be useful in illustrating an article about the software used to generate it. whym (talk) 10:51, 18 October 2022 (UTC)Reply[reply]
Pictogram voting comment.svg Comment Copyrighting the style specifically is extremely problematic. All art is created with a certain method. The end result is certainly copyrightable on an individual basis, but imagine if "cubism" in general were licensed to one particular individual... Childrens art would be unsellable. Undisplayable even, without consent or commission. Artists have experienced spurts of fame for dripping paint, for screen printing and stamping, for gluing things together, tearing them apart. Art styles have included inviting violence from the audience, blowing into molten glass through a tube... operating a camera... ringing a bell. However, the input being debated really falls down to a sequence of words. A sequence of words in itself, such as a song or poem, is certainly copyrightable. However, does that copyright extend to the output of a machine to which that sequence of words is input? The point of copyright is to protect the sales and use of specific unique creations. Style has never, to my knowledge, been copyrighted. Harvard does not even mention the word "style" in its guide to Copyright and Fair Use and other sources that do seem to claim style is not copyrightable, to which I would strongly agree. The only valid debate in my opinion, is around the use of a copyrighted sequence of words input to the machine, and the input of words which refer to a specific copyrightable work. If the words of a song are input to a machine, does the output fall under the same copyright as the song? And when the sequence of input words refers to a specific copyrighted work, producing a derivative of a specific copyrightable work... But as to style or the use of a particular tool, in general, it would not seem protective of one specific work piece to extend a copyright, and the copyright should be grounded in the specific, similar to the point behind patent laws. ~ R.T.G 09:48, 16 October 2022 (UTC)Reply[reply]

Case study, using dead artists[edit]

I just generated this X/Y plot grouped by text prompt and CFG value, using Stable Diffusion V1-4. The artists that I've specifically chosen for this experiment, namely Alphonse Mucha (1860 - 1939), Albert Lynch (1860 - 1950), and Sophie Anderson (1823 - 1903) are all long dead. How much of the visual difference is genuinely an art style, and how much of it is mathematical noise? You can't copyright noise.

An X/Y plot of algorithmically-generated AI art, using text prompts requesting for the images to be generated in imitation style of various famous deceased painters.

The notation [a|b|c] and the notation "a AND b AND c" are functionally the same, the difference is that the latter adds yet even more pseudorandom noise, thereby changing the output. Even slightly adjusting the word order, the punctuation, or any of the notation will slightly alter the output image, either slightly or dramatically. The last few columns involve the AI model attempting to create a rough amalgamation that's somewhere in between the art styles of Mucha, Lynch and Anderson, but I'm sure as you've noticed, it isn't exactly perfect.

Now, the Wikipedia articles for each of these artists contain many examples of their works. Can you honestly say that there is a definite match in art style? Or would it be that there might be some influence, but much of it is lost to noise? What if I never told you who the artists were, would you be able to eyeball it and tell me, "Yep, that's Albert Lynch bro, easy peasy"? At least at its current state, AI image generation is still flawed; but even if AI was perfect, I'd still like to remind everyone that you can copyright works; you cannot copyright motifs and themes and vague nebulous ideas, I honestly can't stress that enough. --benlisquareTalkContribs 11:34, 11 October 2022 (UTC)Reply[reply]

October 11[edit]

UploadWizard and location metadata exclusion[edit]

I already noticed this for sometime, since around mid-part of December 2021, that my photos containing location data in their metadata are being stripped off of their location info in their EXIF metadata upon uploading via UploadWizard. It somehow bothers and disturbs me, as a user who prefers to have complete location data in his photos of roads, public places, and sceneries. Sample affected files: File:N. Roxas Street QCjwilz.jpg and File:Pulilan Church interior 20220925jwilz.jpg.

It appears similar concerns were raised at a different forum (COM:Upload Wizard feedback#Location precision information loss by Upload Wizard and COM:Upload Wizard feedback#Geotag data not transferring from image exif.) but were aren't given any concrete response (like a Phabricator thread). I hope this issue is fixed soon. Ping the participants of those two threads in COM:Upload Wizard feedback for their awareness: @ClemRutter, Djm-leighpark, LPfi, and Pigsonthewing: . JWilz12345 (Talk|Contrib's.) 02:14, 11 October 2022 (UTC)Reply[reply]

Two questions leap to mind:
  1. Could you give a specific example of a photo you've uploaded that had location data that got stripped from the EXIF, and could you be specific about the names of the fields that got stripped?
  2. Is it really stripped, or is it just not displayed? If you download the same file you uploaded, is the file itself actually changed, with this data stripped out? - Jmabel ! talk 03:19, 11 October 2022 (UTC)Reply[reply]
@Jmabel: I already gave two examples above. As for the second question, downloading the Commons file of my photo gives a photo with no location data from its EXIF metadata, apparently removed during the UploadWizard process. JWilz12345 (Talk|Contrib's.) 06:39, 11 October 2022 (UTC)Reply[reply]
Addition: the last photo file of mine still bearing location information in metadata even after upload was File:National Museum PH 2021 jwilz.jpg. The first photo of mine that was uploaded here without including location data in the EXIF metadata of the upload is File:Pulilan–Calumpit Road 20211210jwilz.jpg, uploaded in December 2021. JWilz12345 (Talk|Contrib's.) 06:49, 11 October 2022 (UTC)Reply[reply]
@JWilz12345: Of the several questions I asked, your remark above answers part of one. You haven't said exactly what fields got stripped ("location" is presumably not the property name, or is it?) and you have not answered my question: "If you download the same file you uploaded, is the file itself actually changed, with this data stripped out?" - Jmabel ! talk 14:54, 11 October 2022 (UTC)Reply[reply]
@Jmabel: I already answered your second question. I now highlight that answer above. For the first question, it is common sense: I'm talking about the location data in EXIF metadata that got removed upon UploadWizard: both the latitude and longitude. JWilz12345 (Talk|Contrib's.) 15:55, 11 October 2022 (UTC)Reply[reply]
Is the problem Phab:T223051? Glrx (talk) 05:40, 11 October 2022 (UTC)Reply[reply]
I don't know, Glrx. It was still possible to have location data in the EXIF metadata of my file included in the Upload Wizard upload in August last year. That changed in December last year. JWilz12345 (Talk|Contrib's.) 06:51, 11 October 2022 (UTC)Reply[reply]
I think the EXIF data, and especially the location, is a big privacy issue, and I thought it is stripped for that reason, but I cannot remember any specific thread about that. I think a phabricator thread should be started, but before that we should know what we want.
  • I think there should be a checkbox for including the coordinates, unchecked by default. If checked, the location data should be added to the location fields of the UploadWizard, otherwise stripped also from the EXIF.
  • There is now a general warning about EXIF and privacy, but it is shown (with the same wording) also in the case where no EXIF data is included, and links to Commons:Exif, which doesn't say a single word about the privacy issues. Adding a section on privacy, with links to suitable Wikipedia articles, to that page should be a prime priority, but the wizard should probably check whether especially sensitive fields (such as location) are present, perhaps also do something sensible about the copyright fields.
LPfi (talk) 08:59, 11 October 2022 (UTC)Reply[reply]
(edit conflict with LPfi's replay) Excuse me if I've saying things you've already checked, but did GPS location get turned off or cease to function in your device? Probably about the only device I can enable GPS on are my phones (I thought I might have it on a camera but I'm fairly sure I don't). The recording of the GPS is great if out is out in public, I would have loved it for my recent pictures of Arundel, but not so useful if you've take photographs of objects in a private home. I did have a problem when I was on holiday and turned on by E7i GPS in Wales and it happily added EXIF of the wrong location and was even using that Welsh location over 12 hours to EXIF stamp images taken in Dublin Ireland something like 200 km away, though the E7i eventually connected properly to roaming after I'd been in Ireland for about 12 hours but was OK on Wi-fi on trains and buses! I am fairly sure Flickr strips EXIF and then displays it on the web page and I think Google may do some stripping also. I don't think upload does. I've been using Mounmental to load listing nuilding information recently which save me a lot of trouble having to look up object co-ordinates on google maps, though monument loads the object location as the (camera) location. The key checkpoints to confirm is that GPS / location in on on your device and permission given to add to photographs and that the images taken have EXIF embedded by checking those details on your device (or on e.g. Microsoft Windows) before upload. (this is a non-authoritative reply and treat as such) - Thankyou -- User:Djm-leighpark(a)talk 09:04, 11 October 2022 (UTC)Reply[reply]
@Djm-leighpark: thanks for reply. The location settings for my w:en:Samsung A20's camera is always on. I manually remove location info within my phone's gallery in the first place when the photos to be uploaded are not of public space or place (like from what you said, private homes), but when my photos are of churches, streets/roads, highways, and other objects that are outside of private homes, I retain the location data. Regards, JWilz12345 (Talk|Contrib's.) 09:09, 11 October 2022 (UTC)Reply[reply]

@JWilz12345: I have absolutely no experience with the Upload Wizard. Did you ever try to upload your photos through Special:Upload? To my experience, all EXIF data including location information remain unharmed through this upload interface but no attempt is done to use the location information. I usually add {{GPS EXIF}} in the hope that DschwenBot will create the {{Location}} template from it but unfortunately that bot is inactive most of the time. --AFBorchert (talk) 15:17, 11 October 2022 (UTC)Reply[reply]

@AFBorchert: I used to upload my photos through the standard Upload field (Special:Upload), during the time I was using a second-hand w:en:Samsung Galaxy Star. Since using new phone (2019 onwards), I tend to use UploadWizard more often. With relatively busy college works, it is easier to upload via UploadWizard than Special:Upload. JWilz12345 (Talk|Contrib's.) 15:58, 11 October 2022 (UTC)Reply[reply]
@JWilz12345: Do you have tried it with the given link, i.e. with experienced option? To my knowledge this is the quickest direct upload method available. You can even load prefilled forms using the wpUploadDescription parameter. Otherwise you can use external upload utilities (see COM:UT) but I have never tried them either. --AFBorchert (talk) 16:29, 11 October 2022 (UTC)Reply[reply]
@AfBorchert: I will find a time in the future to overwrite UploadWizard-uploaded photos of mine with the same photos on my phone but still having location in EXIF intact. As for future uploads, I will handle the uploading accordingly, if I may not use Upload Wizard anytime soon if it continues to exclude latitude/longitude information, perhaps on our semestral break (All Saints'/All Souls' Days) or on Christmas break. JWilz12345 (Talk|Contrib's.) 16:49, 11 October 2022 (UTC)Reply[reply]
@AFBorchert: I think I can no longer overwrite those affected files. The warning "The upload is an exact duplicate of the current version of <FILE>" constantly appears when I try to overwrite. Perhaps the damage of Upload Wizard's removal of latitude and longitude info in EXIF is already done. JWilz12345 (Talk|Contrib's.) 15:12, 12 October 2022 (UTC)Reply[reply]
From what I know the UploadWizard cannot be responsible for missing GPS-EXIF. All uploads (UW, Special:Upload, upload tools, Serverside upload) undergo the same processes on the server: Upload, Assembling, Publication. Some (user written) tools may show the transmitted EXIF after assembling and before publicaton. You can also use the API to read out the EXIF from the database table in MediaWiki Software.
Keep in mind that the EXIF will only be included in downloads, if you choose "original fle" (by any of the options to get the original file). Thumbs will ever only have EXIF fields connected to copyrihght, everything else stripped. C.Suthorn (talk) 17:35, 12 October 2022 (UTC)Reply[reply]
@Jmabel, C.Suthorn, Djm-leighpark, LPfi, AFBorchert, and Glrx: I think there is no hope for my future uploads taken via my w:en:Samsung A20 to have coordinate exif info still intact upon upload here. I tried using the standard form, then the result: File:AH26 Cutcot Bridge 1 Pulilanjwilz.jpg. No latitude and longitude fields. I also tried via Commons Android app, but I cancelled the upload after the app notified me that "no location found." Perhaps the latitude and longitude exif info were made into an unreadable format upon update of my phone's software in early December 2021. Perhaps it is an issue created by Samsung itself that I cannot fix. Perhaps as long as I still use my phone for taking photos to be uploaded here, no location exif info will be present. JWilz12345 (Talk|Contrib's.) 10:52, 19 October 2022 (UTC)Reply[reply]
@JWilz12345: I, too, have a Samsung A20. Mine got the recommended software update on 14 April 2022, as follows:
Current version : A205USQSBBVC1 / A205UOYNBBVC1 / A205USQSBBVC1
Security patch level: March 1, 2022
I just used the stock Google Chrome on it to upload File:Vaping at the mall.jpg (photographed 17 August 2022, 11:39:13), and the GPS came out just fine. The file is just as I photographed it, uploaded with the Upload Wizard. Maybe I'll crop it later.   — Jeff G. please ping or talk to me 13:29, 19 October 2022 (UTC)Reply[reply]
@Jeff G.: thanks for input. But mine seems to be a strange and somehow disappointing case. Mine has the latest software recommended: "Current version: A205GNDXSACVC1/A205GNOLOACVC1/A205GNDXSACVC1," and the security patch level was dated March 1, 2022. In any case I won't mind my uploads here not having location exif info despite having so before uploading; the other fields of exif metadata matter (camera info, date and time, etc..). I treat the date and time as the most important exif field. I will store backups of my photos of roads and places on a flash drive. JWilz12345 (Talk|Contrib's.) 14:15, 19 October 2022 (UTC)Reply[reply]
Addition: those backups still bear the location exif metadata. JWilz12345 (Talk|Contrib's.) 14:17, 19 October 2022 (UTC)Reply[reply]
@JWilz12345: Do you have access to another device like a computer that can read the flash drive and upload a photo as a test?   — Jeff G. please ping or talk to me 14:26, 19 October 2022 (UTC)Reply[reply]
@Jeff G.: I have a laptop, but I rarely use it in contributing Commons as I am a slow typist (not used to QWERTY keyboarding). I already accepted the technical limitations on my phone (with regards to non-inclusion of location exif info in uploads), and I will continue to upload via UploadWizard. The most important fields of exif for me matters: camera brand and model, and date and time (of generation). JWilz12345 (Talk|Contrib's.) 17:03, 19 October 2022 (UTC)Reply[reply]

October 14[edit]

Updated copyright law of Vietnam and no more FOP?[edit]

Vietnam just updated their intellectual property law this year (June 2022), through Law No.07/2022/QH15 (read: Vietnam’s Intellectual Property Law: New Amendments by ASEAN Briefing). Apparently much of the copyright provisions were heavily altered, including FOP provision.

The amendment to FOP provision at Article 25(1/h) turned out negative for Wikimedia:

Original Vietnamese text as per [2]

"Điều 25. Các trường hợp ngoại lệ không xâm phạm quyền tác giả

1. Các trường hợp sử dụng tác phẩm đã công bố không phải xin phép, không phải trả tiền bản quyền nhưng phải thông tin về tên tác giả và nguồn gốc, xuất xứ của tác phẩm bao gồm:

h) Chụp ảnh, truyền hình tác phẩm mỹ thuật, kiến trúc, nhiếp ảnh, mỹ thuật ứng dụng được trưng bày tại nơi công cộng nhằm giới thiệu hình ảnh của tác phẩm đó, không nhằm mục đích thương mại;"

Google translation, but some fixes to harmonize sentence structure:

"Article 25. Exceptions for non-infringement of copyrights

1. The following cases of use of published works do not require permission, not requiring remuneration, but the name of the author and the origin of the work must be cited, including:

h) taking pictures and broadcasting works of fine art, architecture, photography, or applied art displayed in public places for presentation of such images, not for commercial purposes;"

Vietnamese Wikimedians, be noted about the abolition of freedom of panorama in your country thanks to the new amendments. Non-commercial use is explicitly against COM:L and the Definition of Free Cultural Works which Commons enshrines.

Pinging people who participated in the second part of deletion request against {{FoP-Vietnam}}, @Aymatth2, Buiquangtu, Liuxinyu970226, Mxn, Violetbonmua, MGA73, and Clindberg: . Also pinging Vietnamese Wikimedians and vietnamwiki admins mentioned by Liuxinyu before, @Alphama, Avia, Bộ lọc sai phạm, Conbo, DHN, Ping, Dung005, Hoang Dat, P.T.Đ, Prenn, Quenhitran, ThiênĐế98, Thái Nhi, Trungda, Trần Nguyễn Minh Huy, Tttrung, TuanUt, Tuanminh01, and Viethavvh:

To Vietnamese Wikimedians, if there is a way to convince the Vietnamese legislature to restore Vietnamese FOP. _ JWilz12345 (Talk|Contrib's.) 03:38, 14 October 2022 (UTC)Reply[reply]

Pictogram voting info.svg Info Article 3(1) of this amendment law states that the amended law will take effect on January 1, 2023, "except for the cases specified in Clauses 2 and 3 of this Article."("Luật này có hiệu lực thi hành từ ngày 01 tháng 01 năm 2023, trừ trường hợp quy định tại khoản 2 và khoản 3 Điều này.") JWilz12345 (Talk|Contrib's.) 04:24, 14 October 2022 (UTC)Reply[reply]

Thank you for the info JWilz12345. I'm not sure how to understand article 4 about Transitional provisions. Google translate says the new law will apply if there was some protection after the old law? So what happens with FOP for existing works when there was no protection after the old law? But perhaps it would be better to discuss on COM:FOP Vietnam.
The other question is more political. Should we actively look for files we can delete or should we only act if someone starts a DR. As I understand it we have no obligation to check for files. So it's a bit like URAA. We know there is a risk but we do not delete files unless we have to. Personally I think we should focus on photos where there is a high probability of commercial use (statues) and ignore photos like File:563 Das Volk der Nung ist bekannt für seine Schmiedekünste.JPG even if you could argue that someone designed the bulding (hut) and therefore have the copyright etc. --MGA73 (talk) 08:45, 15 October 2022 (UTC)Reply[reply]
I would assume that a law is not retroactive unless it says so, especially one that indicates a future date when it comes into effect. After all, print publication cannot be withdrawn in any case.
Our usual approach, though, is that any violation of a country's current copyright law (or where commercial reproduction would violate that law) is a reason for deletion, so presumably any photo of a copyrighted building, sculpture, etc. in Vietnam taken after January 1, 2023 would be deleted as aggressively as any other copyvio or NC-licensed photo. - Jmabel ! talk 15:49, 15 October 2022 (UTC)Reply[reply]
Pictogram voting comment.svg Comment Probably a notification on viwiki should be happened too. --Liuxinyu970226 (talk) 08:02, 16 October 2022 (UTC)Reply[reply]
Hello, I don't know much about the law, let me pass the discussion to Vietnamese community. Thank you!  A l p h a m a  Talk 10:43, 18 October 2022 (UTC)Reply[reply]
@Alphama: I already posted it on the talk page of your Wikimedia user group at Meta wiki. Basically, the copyright law of your country was recently amended (in June 2022), and the amendments will be effective on January 1, 2023. Unfortunately the freedom of panorama provision was altered to restrict commercial uses of photos of public art and architecture still in copyright. Such non-commercial condition makes it not suitable for Wikimedia Commons; COM:Licensing explicitly states that Commons only accepts commercial licenses, and does not accept non-commercial content. JWilz12345 (Talk|Contrib's.) 15:18, 18 October 2022 (UTC)Reply[reply]
From Commons:Licensing, it is weird to me this statement "commercial use of the work must be allowed". Can you show a use case for why Commons needs to accept commercial licenses? I thought "không nhằm mục đích thương mại" --> "not for commercial purposes" fit our freedom of use. My best guess is any work taken from Commons can freely use for any purpose, including commercial uses. In that case, such a nightmare for our project to take photos from public works and show them for the purpose of demonstration in the future.  A l p h a m a  Talk 16:44, 18 October 2022 (UTC)Reply[reply]
@Alphama: examples of commercial uses are the following: use of photos of copyrighted architecture and monuments in tourism souvenirs like postcards or T-shirt prints, in content creators' vlogs, travel portal sites, calendar designs, covers of books or novels, and stamps. In fact, the free Creative Commons licenses only permitted on WikiCommons (as well as the rest of Wikimedia) are commercial. JWilz12345 (Talk|Contrib's.) 17:59, 18 October 2022 (UTC)Reply[reply]
Anything non-commercial is no longer free for Commons. The freedom of panorama legal rights of northern countries like France, Slovenia, Montenegro, Kazakhstan, and South Korea are all non-commercial and treated as not OK by the whole WikiCommons project. For example, COM:FOP Bulgaria. JWilz12345 (Talk|Contrib's.) 18:01, 18 October 2022 (UTC)Reply[reply]
There are a couple of reasons. (1) The definition of what is not "commercial" can be very narrow. For example, an NC license would not allow use by a newspaper that has ads, or in a book that is able to be sold. (2) There was a decision early on that drawing a sharp line -- no NC licenses, period -- we would encourage more individuals and organizations to offer more generous licenses.
But, definitely, some of the interaction with laws like this was a bit of an unforeseen consequence. The government of Vietnam presumably does not care what Commons (or any other particular website) does: they make their FOP law, and the fact that it will really limit certain types of photos from their country from appearing on our site is probably of little or no concern to them. - Jmabel ! talk 18:02, 18 October 2022 (UTC)Reply[reply]
@JWilz12345: So, should {{FoP-Vietnam}} be changed to {{NoFoP-Vietnam}} after January 1, 2023? --Ox1997cow (talk) 17:09, 19 October 2022 (UTC)Reply[reply]
@Ox1997cow: I think we must wait first. The Vietnamese Wikipedians are going to discuss the effects of the restricted Vietnamese FOP in the area of Vietnamese Wikimedian community, see Alphama's comment above. JWilz12345 (Talk|Contrib's.) 17:20, 19 October 2022 (UTC)Reply[reply]
I'm not sure if they will intervene to have the original FOP wording restored (removal of "not for commercial purposes" or "không nhằm mục đích thương mại"). JWilz12345 (Talk|Contrib's.) 17:23, 19 October 2022 (UTC)Reply[reply]

I have opened our discussion here vi:Wikipedia:Thảo luận#Updated copyright law of Vietnam and no more FOP?. From @Vinhtantran: , we can keep all works according to article 4 before January 1, 2023. I believe our community can not do anything now and even in the future when the government has no interest in Wikipedia. It may be a RIP to many works.  A l p h a m a  Talk 17:38, 19 October 2022 (UTC)Reply[reply]

With regards to retroactivity, there are at least four possibilities, ranging from most lenient to most restrictive:
  1. All buildings completed prior to 2023 may be freely photographed, even after 2023.
  2. All photos taken prior to 2023 may be freely uploaded, even after 2023.
  3. All photos uploaded prior to 2023 may be kept, but new uploads are not permitted after 2023.
  4. All existing photos of copyrighted buildings in Vietnam must be deleted after 2023.
King of ♥ 20:14, 19 October 2022 (UTC)Reply[reply]
@King of Hearts: my input goes to the third, as the act of uploading is an action by itself. Two actions are involved here: taking photos of copyrighted Vietnamese public works and art, and uploading such photos. In the latter case, when one uploads, he/she chooses the commercial license upon uploading. This action already equates to violation of the upcoming version of the copyright law, regardless of the date when the photo was taken. The choice of free and commercial Creative Commons licenses occurs on upload, not on photography. But Vietnamese authorities or the intellectual property agency may lean to the fourth, as the existence of commercially-licensed photos here allows Vietnamese content creators or netizens to further do exploitations in 2023 (and possibly beyond this year, should Vietnamese Wikipedians concede to their worst nightmare), potentially violating the upcoming amended copyright law. Though I personally lean to the third than fourth, lest Vietnamese authorities and IP agencies contest this lenient perspective. JWilz12345 (Talk|Contrib's.) 08:57, 20 October 2022 (UTC)Reply[reply]
I think the new law and the old law on this matter have the same meaning. The old one said it's ok "for the purpose of presenting images of these works", but it's vague and debatable that commercial purpose is included or not. So now, the law makers realized it and rewrited it more clearly: "not for commercial purposes" (this happens a lot with other Vietnam laws). Anyway, I agree with No. 3 above: "All photos uploaded prior to 2023 may be kept, but new uploads are not permitted after 2023." ~ Ultraviolet (talk2me) ~ 17:30, 23 October 2022 (UTC)Reply[reply]

How to make a structural solution for not empty disambigious pages?[edit]

Category:Non-empty disambiguation categories has over 2.000 subcategories, while there should be zero. I have tried to diminish that number, but it's mopping with the tap open (Dutch expression for which I could not find a proper translation, I hope you'll understand the meaning). There are at least two problems:

  1. There are a lot of them with "(empty)". These are DAB's that once had one or more files and/or categories in them, thereafter that problem has been solved, but then you need to do an extra action to get the DAB category out of the Category:Non-empty disambiguation categories (click on "edit" and "Publish changes"). Not many people know or do that, very understandable, it should not have to be done at all to solve this problem.
    1. The solution might be a technical one: if a DAB is empty, do not show it on Category:Non-empty disambiguation categories anymore.
  2. DAB categories are not recognizable as such when you add a file or another category to one. And you do not get a notification when you add a file or category to a DAB. Solutions might be:
    1. (Techical solution) It would not be allowed/possible anymore to put files and categories into a DAB category. It is technical not possible to choose a DAB category for a file or subcategory and/or you get an error message.
    2. Until that technical solution has been implemented: It is allowed to have "(disambiguation)" in the name of the DAB category. It won't solve all the misunderstandings, but at least volunteers will profit. I have proposed to do so for Category:Decoration, but there was no consensus about this solution, see Commons:Categories for discussion/2022/06/Category:Decoration.

So my questions are:

  • Is the problem clear? Do others recognize it?
  • How to make a structural solution for not empty disambigious pages? Would the proposed solutions help? Do you know better solutions?
  • How can we implement such solutions? What would have to be done to realize them?

--JopkeB (talk) 06:53, 14 October 2022 (UTC)Reply[reply]

@JopkeB: I would suggest, "trying to mop up water under a running tap" or "like mopping up water under a running tap". Apologies for interrupting, but it's a good cliche. ~ R.T.G 16:01, 16 October 2022 (UTC)Reply[reply]
1. is a known bug (phab:T86492): Things should update automatically, but they don't. There's not much we can so about it apart from making developers more aware of it. Realistically, it's still going to be a low priority thing, though. It pretty much only affects maintenance categories, so while it's annoying and triggers our collective OCD, it is close to irrelevant to end users who come her looking for files.
2.1. probably cannot be fixed by a single technical solution. It won't be possible for direct editing of the wiki markup any time soon, as there is no mechanism in place to check categories (or any other part of the markup) in any kind of way. But people who edit that way are hopefully more likely to know what they're doing. Edit: this is incorrect, there is a popup warning you Highest priority would probably be HotCat. That could probably be solved somehow, it's already smart enough to resolve redirects. Maybe try to get some attention at MediaWiki talk:Gadget-HotCat.js.
2.2 might be worth discussing. There is Commons:Category disambiguation, but it currently is nothing more than an essay. It is not well-integrated with Commons:Categories, and seems to assume that the reader is familiar with or willing to read up on the procedures at the English language Wikipedia. El Grafo (talk) 08:34, 14 October 2022 (UTC)Reply[reply]
1 can be worked around by bot making null edits, for example once a day Mateusz Konieczny (talk) 08:44, 14 October 2022 (UTC)Reply[reply]
2 would benefit from clear marking such problematic categories in HotCat and special support to make disambiguation easier Mateusz Konieczny (talk) 08:45, 14 October 2022 (UTC)Reply[reply]
Thanks El Grafo and Mateusz Konieczny, for your aswers and information.
@1. @Mateusz Konieczny; I guess such a bot could do the trick, once a day would be enough for me. Who can make such a bot and implement it?
@2. I agree with Mateusz Konieczny: clear marking such problematic categories in HotCat could prevent a lot of troubles. I see now that a pop up warning appears when you try to add a DAB category via the edit tab. So the problem is mainly with HotCat editing AND during the Upload process.
@2.2. What would be a good starting point to discuss the question "Contrary to EN-WP, should it be allowed in Commons to have "(disambiguation)" in a category name, even when there is no category with the same name that is a primary topic?" It looks to me that the essay Commons:Category disambiguation only focusses on clear names for end users, I see no leads with maintenance reasons. And I think there should be a discussion about this question before putting it in such an essay.
--JopkeB (talk) 10:55, 14 October 2022 (UTC)Reply[reply]
A general discussion could be started at Commons talk:Categories. Once the whole thing has been fleshed out a bit more, it may be wise to present it at COM:VP/P for approval. If you want it to stick, I'd suggest aim for it to become part of Commons:Categories rather than that essay. El Grafo (talk) 11:11, 14 October 2022 (UTC)Reply[reply]
@El Grafo: - see Commons:Bots/Requests Mateusz Konieczny (talk) 18:13, 14 October 2022 (UTC)Reply[reply]
Huh? Suppose you meant to ping @JopkeB instead of me? El Grafo (talk) 10:46, 16 October 2022 (UTC)Reply[reply]
I have had similar thoughts. My comments:
  • I think item 1 is part of a larger issue: I think it takes a cache time to catch up when things are added to or removed from a category.
  • If it were impossible to put things into a dab category, we would have some people who search for correct categories, but we would also have some who just leave their media uncategorized. And that's just people: I think there are also bots that categorize, and they might not even recognize warnings. I think being uncategorized is worse than being in a dab category because a dab cat would at least be closer to a correct category. We have many more things that are completely uncategorized than we do things in dab categories.
  • I tried adding a file to a dab category using Cat-a-lot, and it worked without giving me a warning. Maybe this tool should be addressed in whatever solution comes out of this discussion.
-- Auntof6 (talk) 07:53, 15 October 2022 (UTC)Reply[reply]
  • "Cat-a-lot" - it is one more case where the next step would be notifying authors of a tool Mateusz Konieczny (talk) 08:11, 15 October 2022 (UTC)Reply[reply]
@User:Auntof6, Thanks for your contributions, but I have a questions about the first one:
=> What do you mean by "it takes a cache time to catch up when things are added to or removed from a category"? What are the implications for this problem?
Yes, you are right: files being uncategorized is worse than being in a dab category. I did not think of that.
Yes, Cat-a-lot is in the picture. JopkeB (talk) 04:16, 18 October 2022 (UTC)Reply[reply]
@JopkeB: About the cache: it is my understanding that when a dab category is emptied, it will eventually disappear from Category:Non-empty disambiguation categories, it just takes some amount of time.
I've seen similar things when a page is removed from or added to a category by some means other than editing the page. For example, if a category is added by a template and the template is changed so that it no longer adds the category, the page will still appear in the category until the system catches up somehow. Maybe "cache" is the wrong term, but I think I've seen it used before in this context. -- Auntof6 (talk) 05:45, 18 October 2022 (UTC)Reply[reply]
Thanks, User:Auntof6, so I just have to be patient. JopkeB (talk) 07:00, 19 October 2022 (UTC)Reply[reply]
@JopkeB: Well, yes and no. Yes, because the ones that show as empty will eventually disappear, but no, because new ones will show up. There's a joke in English that this kind of thing is "like painting a bridge:" when you get to the end, you have to start all over! -- Auntof6 (talk) 06:32, 20 October 2022 (UTC)Reply[reply]
As far as purging is concerned, you could go through each category and hard purge them with s:MediaWiki:Gadget-PurgeTab.js installed. That's what I've done. -BRAINULATOR9 (TALK) 22:24, 16 October 2022 (UTC)Reply[reply]
@User:Brainulator9, Thanks a lot, your solution clearly works for problem 1. But I do not understand at all what you have done. Could you please explain for people with only basic knowledge of computer programming like me? And should this action be done once, or day after day? And for which categories (for all DAB categories)? JopkeB (talk) 04:25, 18 October 2022 (UTC)Reply[reply]
Just copy the following line to Special:MyPage/common.js to install it: mw.loader.load('//');
When using it... admittedly, what I've done is go through each category I want to clear and click on the drop-down button that appears. Something like Special:ApiSandbox#action=purge&format=json&forcerecursivelinkupdate=1&generator=categorymembers&formatversion=latest&gcmtitle=Category:Non-empty disambiguation categories could also be used to do the same thing. Either way, it does require regular maintenance... maybe some day, I'll just create a bot to do this sort of thing automatically. -BRAINULATOR9 (TALK) 22:09, 18 October 2022 (UTC)Reply[reply]
Thanks for your explanation. I have installed the tool and it works (with "Hard purge" and "Null edit" in the category with no files). So this is a useful, but not a structural solution. I hope you will one day create a bot to do the trick automatically. For now I'll make a request on Commons:Bots/Requests. JopkeB (talk) 07:31, 19 October 2022 (UTC)Reply[reply]

Conclusions so far and action items[edit]

Thanks, everyone, for your contributions. My conclusions so far:
@1. (redundant empty DAB's in Category:Non-empty disambiguation categories):

@2. (DAB categories are not recognizable as such when you add a file or another category to one)
@2.1. (technical solution)

  • There are already warning popups when adding DAB cats via the edit tab.
  • There are several places where there aren't yet:
(1) Most important: HotCat. Advice: Try to get some attention at MediaWiki talk:Gadget-HotCat.js. [action item JopkeB] - request is at MediaWiki talk:Gadget-HotCat.js#Could a warning popup appear when adding a DAB cat while using HotCat?
(2) In the upload procedure.
(3) In bots which categorize automatically.
  • It is not a good idea totally blocking the possibility of putting items into a DAB category, because this may cause more uncategorized files.

@2.2. ("(disambiguation)" in the category name)

--JopkeB (talk) 07:53, 19 October 2022 (UTC)Reply[reply]

October 16[edit]

What should be in a description field (and what not)?[edit]

Hello, in this item, Tm and I have a conversation about whether some imported texts from Flickr files are descriptions according to Commons standards. Input from others is welcome. Best regards, Zijling (talk) 07:35, 16 October 2022 (UTC)Reply[reply]

Another example of a lot of text in the description where a short concise text is difficult. When the long text fits in a Wikipedia article, preferably there. Wouter (talk) 09:03, 16 October 2022 (UTC)Reply[reply]
I wouldn't say there's any difficulty here. This reduction has the result that visitors know at which event the photo was taken. If someone thinks: that descritption needs extension to be more informative, then in my view such an extension would mention the person you see speaking there, for instance. Not a list of demands and calls for visiting someone's social media accounts and supporting them. Also, such things should not appear in articles or any other Wikimedia space, since all those spaces are meant to be neutral. Zijling (talk) 09:46, 16 October 2022 (UTC)Reply[reply]
It is just as problematic to avoid mentioning what a person is trying to achieve as it is to go into too much detail. The issue here is about the political nature of the event. However, if you remove all words which may be deemed "political", you get an arbitrary result. You need to be much more specific, and somewhat inclusive, to remain neutral. Anti-politics is not the goal any more than it is anti-protests. The goal is neutrality, and your resulting method for the description in such a case should be based solely on achieving neutrality. For an instance of consideration, I would like you to compare the title of the image to the cut down description. Apparently the rally seeks to improve public health and safety, but a politically paranoid, cut down description, says it is no more than a demand for worker control. As long as the description was on topic, it doesn't matter if it is "political". What opinion is not political, and why should a list of demands not be mentioned, beyond the fact that they are a list of demands? Neutrality is not about silence.
Regarding the calls to visit the authors websites and support them... And is the objection based on the fact the purpose of the authorship is political? I am neutral to the fact that the author of this work has political goals, and I can see nothing beyond a 7 item list of short demands, specifically related to the event, and contact details for the author. The fact the subject of the particular image is not noted refers us only to the subjects notability, not the nature of politics. I totally object to any actions preventing details or mentions of politics being available, simply because they are political. As long as Wikimedia appears neutral, no further interference is warranted. Wikimedia does not author these images. It simply hosts them. ~ R.T.G 10:29, 16 October 2022 (UTC)Reply[reply]
I initially thought these were demands uttered by the photographed subject. In that case, it would make sense to keep them there as a description of what the subject said. However, the list appears on every photo from that series, so I would be inclined to leave it out, because it does not describe that particular photo. --HyperGaruda (talk) 10:36, 16 October 2022 (UTC)Reply[reply]
Cannot see how these changes relate to the number of images. The changes I see are: from a list of the goals of the event and the author details, to a mention of the nature of the event. That is a significant change. My argument is that these images should appear political. The idea that their political appearance is a reflection on the neutrality of Wikimedia is misguided. It is the removal of the demands which suggests that Wikimedia is anything more than neutral. It is not a goal of Wikimedia to avoid the demands of a political protest being available for viewing here. What would be an issue is for Wikimedians to use the sites to promote those goals, or to obscure them. We should do neither to remain neutral. ~ R.T.G 11:07, 16 October 2022 (UTC)Reply[reply]
External links should not appear in the description. They should only appear as the source if needed. Of course a short summary about what a protest is about and what they are demanding can and should be in the description. But this text must not be copied from the protest organizers website. This is only the source for writing the short summary. One example File:FridaysForFuture protest Berlin 2022-06-24 104.jpg "Bicycle demonstration of Fridays for Future addressing the G7 summit and demanding debt cancellation for states in the global south." This is a short explanation with the main demand of this event but not a copy of the website or press release text and without any opinion on the topic. GPSLeo (talk) 11:40, 16 October 2022 (UTC)Reply[reply]
There doesn't seem to be a specific guideline. I would suggest if there was a description guideline written which depreciated the idea of external links, on the basis say of avoiding link farms, I would suggest replacing the links with a note that there are some links if the user would like to navigate to the source. And I would agree that the Wikimedians own wording be preferential to the copy pasting. But it would make sense to preserve information on the spirit of the protest, beyond simply saying workers looking to influence leadership. There is no need for a full manifesto for instance. But it should perhaps note in the given example, the types of things the protestors are demanding, rather than simply saying they are workers protesting, which diminishes them, and therefore calls neutrality into question. Certainly however, the fears are valid. There should be considerations towards avoiding Wikimedia appearing particularly supportive of a protest, but at the same time it should not appear particularly against the protest. I don't see anything wrong from either perspective in the example given by GPSLeo. ~ R.T.G 15:45, 16 October 2022 (UTC)Reply[reply]
There is a specific guideline regarding neutrality: COM:NPOV, the section "Text" to be precise. Then there is our help guide on descriptions, in addition to some hints in the template documentation itself. By the way, a less verbose writing style would be of great benefit, not just to you, RTG, but also to other readers. EN-Wikipedia tried to drop a hint, but seems to have missed its mark. --HyperGaruda (talk) 17:58, 16 October 2022 (UTC)Reply[reply]
Excuse me, but if you think I have done something here which merits exclusion, I suggest you take it up with the appropriate channels. Threatening me against something which I, quite frankly do not understand, is unlikely to achieve anything desirable. ~ R.T.G 01:15, 17 October 2022 (UTC)Reply[reply]
@HyperGaruda: , ignoring for a moment the rest of your comment, the first steps guide does not seem to lead to its sub pages (i.e. "/Quality and description"). And although that is the best guide to descriptions I have seen, it doesn't quite cover this particular issue. So there is room to improve. And as far as I can see, the others contributing here, seem willing and able to provide that, to my satisfaction at least. So I would ask, @Zijling, El Grafo, GPSLeo, and Wouterhagens: , would you consider improving the links provided by HyperGaruda by covering this aspect and linking the resulting guide to the "First Steps" tutorial? And if you do so, would you inform the talk page of Commons:First steps/Quality and description#Good file descriptions and perhaps even hold the discussion on the details there? ~ R.T.G 09:20, 20 October 2022 (UTC)Reply[reply]
Yes, non-neutral descriptions and author's comments can be very useful (even necessary) for providing context. No, we should not just import them and use them as descriptions for the files they came with. That's not a contradiction, though. There's a simple solution to this disagreement if you just take a moment to step back a little bit:
  1. import the file
  2. give it a neutral description
  3. add the original, biased description as a clearly marked quote.
Handled this way, even the worst kind of Nazi propaganda can be educational. I think we should consider creating a general-purpose {{Imported description}} that a) clearly formats the imported text as a quote and b) has an informative text along the lines of "here's the original description/caption for reference, it comes from this person/institution/government, be aware that it might be biased". We've been doing it like this for ages with files imported from the German Bundesarchiv via {{BArch-biased}}.
On a related note, be aware that all text on Commons is licensed CC-BY-SA 3.0 (see footer). This is separate from licenses for media files. Similarly, the description of a freely licensed file you'll find on the web is not necessarily freely licensed as well. So be careful when importing non-trivial descriptions, they may be subject to copyright and/or incompatible with Commons' licensing.
--El Grafo (talk) 09:16, 17 October 2022 (UTC)Reply[reply]
Now that is a good idea, a disclaimer. I myself have imported thousands of Flickr files at a time and their descriptions would take days to rewrite. But it is probably wise to also have a policy or guideline that is specific for those who are rewriting the description to suit. ~ R.T.G 10:11, 17 October 2022 (UTC) ~ R.T.G 10:11, 17 October 2022 (UTC)Reply[reply]
A disclaimer would not be a bad idea. It can be helpful to reduce confusion. Zijling (talk) 16:45, 17 October 2022 (UTC)Reply[reply]

October 17[edit]

How to use Flickr images with All Rights Reserved (C) on Commons.[edit]

I had gone through Commons:Flickr_files but I still wish to understand: To use an image from Flickr with "Copyright.svg All rights reserved", do I just need to credit the user that uploaded it, the date and the source page and that's it? For example, I want to use this image on the article of the subject. Can I do this with the credit and attribution I stated above? — Preceding unsigned comment added by Danidamiobi (talk • contribs) 11:25, 17. Oct. 2022 (UTC)

You cannot upload these images to Wikimedia Commons at all. Only freely licensed images are accepted here, which these are not. See Commons:Licensing. Regards --Rosenzweig τ 09:28, 17 October 2022 (UTC)Reply[reply]
@Danidamiobi: You can ask the Flickr user, if the images are theirs, to change the licence. I've had several successes doing this; even though they are the minority of such requests I have made. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:13, 21 October 2022 (UTC)Reply[reply]
Yes. The "all rights" that are reserved include the right to make copies and to distribute the work (except what is permitted as fair use, which doesn't apply to Commons). You need permission from the copyright holder, or else you need to change copyright law. –LPfi (talk) 13:34, 22 October 2022 (UTC)Reply[reply]

Wikistories on Commons: a proposed visual content tool in WikiCommons[edit]

Greetings, everyone!

The WMF Inuka team is exploring having the Wikistories tool on Commons, so we want to get your feedback.

Our proposal

The Inuka team believes the Wikistories tool would significantly increase our projects' visual form of knowledge if we introduce a Wikistories workflow in WikiCommons. The above thinking is because the Commons is a repository of freely licensed media files that is discoverable across other projects and languages.

We have carefully outlined what we want to achieve with the tool on Commons, and we are optimistic about a good outcome. However, the Wikistories on Commons initiative is still a proposal that will require the WikiCommons community's feedback/ input on our design explorations before we commit to the work.

Our request

In the next three weeks, the Inuka team would want you as a member of the Commons community to:

  • Read more about the initiative.
  • Attend our proposed office hours, where we will present our design exploration and discuss our plan with this proposal.
  • Give feedback on this initiative and share your input about the design exploration on the talk page we have created.

We have provided three spots for office hours.

Please click on the date that works for you and register to attend the office hour. If you can't attend, that is fine. However, we will appreciate your feedback about the design exploration and the Wikistories on Commons proposal on the talk page.

We, the Inuka team, are excited about this work's possibilities, and we look forward to having you all onboard at every phase.

Thank you!

UOzurumba (WMF) (talk) 18:59, 17 October 2022 (UTC) On behalf of the Inuka team.Reply[reply]

I am bit confused. You are not part of Commons:Product and technical support for Commons 2022-23? So there are currently two WMF teams interdependently working on improvements on Commons? GPSLeo (talk) 19:24, 17 October 2022 (UTC)Reply[reply]
Well, if you really want to do wikistories, there are two things you need to work on first: 1) fix the upload so that the images for the stories can actually be uploaded. 2) Change the scope of Commons so that the images are not immediately deleted as oos. And then have fun with your tool. C.Suthorn (talk) 20:16, 17 October 2022 (UTC)Reply[reply]
What do you mean by "fix the upload so that the images for the stories can actually be uploaded"? Wikistories references images that are already on Commons. Take a look at Story:Wolf story on betawiki for a (non-interactive) example; alternatively, create an account and enable it in your beta features to see the full extension. Vahurzpu (talk) 23:24, 17 October 2022 (UTC)Reply[reply]
Wofür sollen diese Stories denn eigentlich sein? andere WM-sites haben Inhalte, und in Stories können Bilder aus Commons eingebunden sein. Commons hat überhaupt keinen Content (außer ein paar Galerien, die nicht in einem Galerien-Namensraum, sondern im Mainspace liegen) Commons ist eine Mediendatenbank. Stories mit Bilder aus Commons werden in wikinews gemacht und für die Präsentation von Bildern zu einem Thema gibt es Picsome. Wikistories auf Commons können eigentlich nur dazu gut sein, zu beschreiben, wie Bilder hochgeladen werden, wie sie kategorisiert werden, wie sie mit SDC versehen werden, wie sie auf anderen Wikis eingebunden werden und wie Bilder für die Einbindung ausgesucht werden.
Und wie eigentlch wird bewerkstelligt, dass ich eine Wikistory in deutsch lesen kann, ein Koreaner in Koreanisch und eine Portugiesin in Portugiesisch? C.Suthorn (talk) 06:45, 18 October 2022 (UTC)Reply[reply]
@UOzurumba (WMF) how does Wikistories handle i18n? Is it connected with deepl or does Wikistories come with its own ML engine? C.Suthorn (talk) 06:47, 18 October 2022 (UTC)Reply[reply]
@C.Suthorn We will use the multilingual support already in Commons, similar to how image descriptions can be available in many languages. We have not made plans for now, to incorporate automatic translation. Thank you. UOzurumba (WMF) (talk) 19:59, 18 October 2022 (UTC)Reply[reply]
i dont quite get the point of this. id:Story:Osaka -- a bunch of texts stacked on vertical photos?
please no tiktok/instagram/snapchat-style dumb stuff.--RZuo (talk) 12:52, 19 October 2022 (UTC)Reply[reply]

October 18[edit]

Illustrations that are pure speculation named after the subject "depicted"[edit]

Hei hei! I came over this illustration, File:Johanne Nielsdatter.jpg, when patrolling the Norwegian Wikipedia project. After some researching, it appears to me that the illustration is pure speculation, as there aren't sources available for Johanne Nielsdatter's description. I have also discussed this externally, and I was guided here. Isn't this file then inappropriately named and categorized? It seems logical to me that the connection to the historical person should be removed, and rather renamed something such as Woman being burned at a stake.jpg (as well as removed from the category of which it resides in solitude)? What is the right approach?

I am also made aware that this has been created as part of the French Wikipedia initiative Les sans imagEs, and while I appreciate well-researched depictions of subjects lacking CC-licensed images, I think the mentioned illustration is pure speculation about the woman's appearance. Although not relevant to this project, I also think it's wrong to use it to depict Nielsdatter/Nilsdatter in infoboxes/described as "Johanne Nielsdatter" across Wikipedia-projects, again, due to the lack of verifiability. EdoAug (talk) 03:57, 18 October 2022 (UTC)Reply[reply]

I just made Commons:Illustration as a future guide to those seeking to aid in such efforts; hopefully after a few additions it will be useful as a central place to put resources to help illustrators for Wikimedia projects. Arlo James Barnes 04:45, 18 October 2022 (UTC)Reply[reply]
Should there be something in this guide about copyrights as well? How do we know for certain that the uploader is really the creator? Should every creator send a mail to the Volunteer Response Team and should the VRT team then mark such a user as being a creator of such illustrations? JopkeB (talk) 04:54, 18 October 2022 (UTC)Reply[reply]
VRT has a backlog, right? It seems to me we should handle this type of file like we do all 'own work' files and assume good faith unless/until we find evidence to the contrary; at that point, opening a ticket can help clear (or confirm) any doubts. Arlo James Barnes 05:11, 18 October 2022 (UTC)Reply[reply]
@Arlo Barnes: Per {{Permissions backlog}}, "Current backlog (oldest unanswered mail or ticket) in permissions-commons queue is 2 days."   — Jeff G. please ping or talk to me 08:53, 18 October 2022 (UTC)Reply[reply]
For photos I look now at Exif data. But for illustration there are no Exif data. For recent paintings (= painter cannot be dead for at least 70 years) I often ask for a VRT ticket, even when the name of the uploader is similar to the painter, because anyone can make an account with any name. So I wonder how we'll treat illustrators. JopkeB (talk) 11:19, 18 October 2022 (UTC)Reply[reply]
i share the same concern. earlier discussion: Commons:Village_pump/Archive/2021/08#Artist's_impressions.--RZuo (talk) 12:52, 19 October 2022 (UTC)Reply[reply]

ImageAnnotator sometimes does not work[edit]

In a number of annotated images, neither the annotations nor the button "Add a note" show up, maybe because a process hangs. I also commented this problem to MediaWiki talk:Gadget-ImageAnnotator.js#No button "Add a note", but nobody replied yet. My impression is that it occurs more often in smaller and / or heavily JPG compressed images, but maybe this is just my fantasy. Who can help? --Kuhni74 (talk) 11:26, 18 October 2022 (UTC)Reply[reply]

Pinging @Kuhni74 - Jmabel ! talk 15:18, 18 October 2022 (UTC)Reply[reply]

Delink dead "Leave feedback" at Special:UploadWizard[edit]

Special:UploadWizard has "Leave feedback" leading to Commons:Upload Wizard feedback with

"This page is presently inactive and retained primarily for historical interest. (...) you may try using the talk page (though your voice may not be heard) or start a discussion at the Village pump."

How it can be removed? Mateusz Konieczny (talk) 15:22, 18 October 2022 (UTC)Reply[reply]

@Mateusz Konieczny: This is set as $wgUploadWizardConfig['feedbackPage'] in CommonSettings.php. If left out it defaults to linking to "UploadWizard's bug tracker", wherever that is[3]. Updating the link would require going through the procedure for Requesting wiki configuration changes, which usually starts with a proposal at Commons:Village pump/Proposals. bjh21 (talk) 16:30, 18 October 2022 (UTC)Reply[reply]
Posted Mateusz Konieczny (talk) 18:38, 18 October 2022 (UTC)Reply[reply]

Special:UploadWizard - better for new users[edit]

"ideally, we’d link to the user-friendly upload wizard; however, if you aren’t logged in, you get a login page without a good explanation of what Commons is good for, not a great user experience." -

Mateusz Konieczny (talk) 15:23, 18 October 2022 (UTC)Reply[reply]

@Mateusz Konieczny: The text that you get if you visit Special:Upload or Special:UploadWizard when not logged in comes from MediaWiki:Uploadnologintext. So it would be quite possible to make it a bit friendlier. --bjh21 (talk) 16:42, 18 October 2022 (UTC)Reply[reply]
Easy for an Interface admin. But this linked message is only for English, the other languages would have to be updated, as well, see Special:PrefixIndex/MediaWiki:Uploadnologintext/. Mateusz Konieczny, “a good explanation of what Commons is good for”: Is Commons:Welcome good enough? (I know the cited text is not yours.) — Speravir – 17:51, 18 October 2022 (UTC)Reply[reply]
@Speravir: seems fine, especially as the first iteration Mateusz Konieczny (talk) 18:37, 18 October 2022 (UTC)Reply[reply]
So, an {{Edit request}} should be filed on MediaWiki talk:Uploadnologintext with a phrasing suggestion for English, then it needs some people who add the according translation for the existing languages. For some languages it may be hard to find one. BTW: Interestingly the German translation has already now some more text with a link to the German start page of Commons. — Speravir – 22:59, 18 October 2022 (UTC)Reply[reply]

CC Template for files from Salt Research[edit]

Can someone make a custom template similar to:

w:en:Creative Commons
This file is licensed under the Creative Commons Attribution 4.0 International license.
Attribution: Salt Research, (Insert Name here) Archive
You are free:
  • to share – to copy, distribute and transmit the work
  • to remix – to adapt the work
Under the following conditions:
  • attribution – You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.


But it should instead say:

Salt is a partner of Flickr Commons. All visual materials within Flickr Commons and digitized by Salt Research which are accessible via are shared under the Creative Commons license:
This material is provided solely for the purpose of individual research. It can be used under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license. The terms specified hereunder:
• The creator or the licensor of the work should be stated in all copies,
• Copies of documents, or documents created from copies can not be used commercially,
• Documents can not be remixed, transformed or built upon.
For each use, credits should be given as stated in the record.
(e.g. Salt Research, Harika-Kemali Söylemezoğlu Archive)
For any rights requests outside of personal research purposes, please contact Salt Research via

So the only real diffrence is saying that Documents can not be remixed, transformed or built upon, instead of "You are free: to remix – to adapt the work" as this current template states.

This is the general rights statement of files from [4]. — Preceding unsigned comment added by Gazozlu (talk • contribs) 20:13, 18 October 2022 (UTC)Reply[reply]

No, because that violates COM:L. -BRAINULATOR9 (TALK) 22:00, 18 October 2022 (UTC)Reply[reply]
Whoops, then all these may need to be deleted. They are all under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0):

1974 article about Güzel İstanbul.jpg
Burhan Temel and Gürdal Duyar.jpg
Photo of Gürdal Duyar sitting.jpg
Gürdal Duyar at work.jpg
Sketch by Gürdal Duyar.jpg
Turkish artists.jpg
Fine Arts Union 90th anniversary exhibition invitation card.jpg
High Sculptors Society Exhibition 9-25 January 1973.jpg
Kuzgun Acar exhibition.jpg
Türkiye - Kuzgun Acar.jpg
A meeting held at the State Academy of Fine Arts in Istanbul.jpg
Kuzgun Acar.jpg
Kuzgun Acar 2.jpg
Ne yazık ki hep şaşırıyoruz.jpg
Kuzgun Acar, Retrospektif Sergi, İş Sanat Kültür Merkezi Kibele Sanat Galerisi, İstanbul.jpg
Article about first Military Museum Art Exhibition.pdf

Gazozlu (talk) 22:12, 18 October 2022 (UTC)Reply[reply]
@Gazozlu: I’ve converted the gallery into text links. I think the best way handling it would be if you’d go yourself to COM:AN. Copy the file links from above, tell that you’ve uploaded these files in error not knowing the license limitation, and ask yourself for deletion. — Speravir – 22:53, 18 October 2022 (UTC)Reply[reply]
Or just tag them for speedy deletion as COM:G7 and COM:F1. -BRAINULATOR9 (TALK) 23:31, 18 October 2022 (UTC)Reply[reply]
Allright, i'll do that. Is there any specific rules around newspaper clippings? I see for example this one Perfil de Gilberto de Carvalho, por Horacio Cardoso.jpg has been uploaded as "own work" although the author appears to be someone else and the person=/=the publisher of the newspaper either. Asking for this one: 1974 article about Güzel İstanbul.jpg Gazozlu (talk) 10:35, 19 October 2022 (UTC)Reply[reply]
Those should be deleted, most likely, unless you can prove that the file is free for whatever reason. The file you linked to dates to 2011, so a normal deletion request should be done. -BRAINULATOR9 (TALK) 18:16, 19 October 2022 (UTC)Reply[reply]

October 20[edit]

Files from the Australian Human Rights Commission Flickr stream[edit]

Some files from the commission's Flickr stream have third-party copyright notices in the EXIF, see this file for example. However, their Social Media copyright policy says:

The Australian Human Rights Commission may use social media channels for communicating with the public, including Facebook pages, Twitter accounts, YouTube channels and blogs. Subject to the terms of use of the particular channels, and except where otherwise stated:

1. you may use all material posted by us on these channels under the Creative Commons Attribution 4.0 International Licence, and

2. post 15 September 2015, material posted by any other person, such as public users, is deemed to be made available to you by that person under the Creative Commons Attribution 4.0 International Licence as a condition of use of that channel.

–– Copyright Policy, Australian Human Rights Commission

Does the policy cover us here? Joofjoof (talk) 05:08, 20 October 2022 (UTC)Reply[reply]

In general the original photographer (Matthew Syres in the example above) is the copyright holder unless contractually stated otherwise or the photographer is creating images for an employer, and only the copyright holder can determine the license of their work. Syres appears to have done commissioned work for the Australian Human Rights Commission, but a freelance contractor is generally not considered an employee for the purposes of corporate authorship (at least not in the US, Australian law may differ). COM:AUSTRALIA is unfortunately mum on the issue of commissioned/works-for-hire. The Copyright policy linked above states excludes CC-BY-4.0 license for "photographs, images and video" on its own website, and item 2 above seems like an illegal assumption of copyright from third party users (they can say they freely release a photo or comment I post in a reply on their Facebook page, but that doesn't make it so). Additional investigation, including contacting the author/agency, may be needed to clarify. --Animalparty (talk) 22:00, 20 October 2022 (UTC)Reply[reply]
Thanks Animalparty, this makes sense to me. Is there an Australia group here on Commons? Local users may be better-equipped to investigate. Joofjoof (talk) 01:02, 23 October 2022 (UTC)Reply[reply]
I don't know about an Australia group, but you might get more helpful advice at Commons:Village pump/Copyright. --Animalparty (talk) 01:06, 23 October 2022 (UTC)Reply[reply]

INVITE: Queering Wikipedia 2022 Meeting on Friday and Saturday (October 21-22)[edit]

Dear project members Queering Wikipedia 2022 takes place on: Friday, October 21, at 18:00 UTC and Saturday, October 22, at 14:00 UTC

Though the initial 2019 title centers Wikipedia, all Wiki projects, LGBT+ Wikimedians and allies are welcome, especially You :-)

Please register using the link on the page to the Zoom based system as soon as possible even if your attendance will be partial. Pardon the late announcement.

--Zblace (talk) 06:08, 20 October 2022 (UTC)Reply[reply]

UCoC EG Community review period closed[edit]

Dear Wikimedians,

Thank you for participating in the review of the Revised Enforcement Draft Guidelines for the Universal Code of Conduct (UCoC). The UCoC project team and the Revisions Committee appreciate you all taking the time to discuss the guidelines, suggest changes, and ask questions.

This community review period lasted from September 8 to October 8, 2022. Over the past four weeks, the UCoC project team has collected valuable community input from various channels, including three conversation hours sessions, where Wikimedians could get together to discuss the revised UCoC Enforcement Guidelines. The Revisions Committee will review community input when they reconvene in the second week of October 2022. The UCoC project team will support them in providing updates as they continue their work and will continue to inform the community about all important developments and milestones as the Committee prepares the final version of the UCoC Enforcement Guidelines that is currently scheduled for a community-wide vote in mid-January of 2023.

On behalf of the the UCoC project team,

Zuz (WMF) (talk) 11:37, 20 October 2022 (UTC)Reply[reply]

What is the copyright status of West African CFA franc and Central African CFA franc?[edit]

I would like to know the copyright status for the CFA franc currency, since I added light green dots to the countries that use the euro in File:Whether to allow uploading of currency World Map.svg.

What is the copyright status of West African CFA franc and Central African CFA franc?

-Ox1997cow (talk) 16:59, 20 October 2022 (UTC)Reply[reply]

@Ox1997cow: You may find guidance at COM:VPC and the COM:CRT pages relevant to the countries discussed at en:CFA franc.   — Jeff G. please ping or talk to me 17:15, 20 October 2022 (UTC)Reply[reply]
I see. Ox1997cow (talk) 17:25, 20 October 2022 (UTC)Reply[reply]
@Ox1997cow: It will also help to know where the coins are minted.   — Jeff G. please ping or talk to me 17:28, 20 October 2022 (UTC)Reply[reply]

October 21[edit]

File deleted with unclear reason. User in charge not available[edit]

Hey, was checking [5] to use for a project and it is appear that file was deleted with pretty much meaningless reason--as per COM:SPEEDY which of source can imply dozens of reasons. I cannot also find any requsets for deletion at relevant log.

User:Yann kindly semi blocked his/her Talk page so I cannot make any request.

Would somebody please kindly explain?

Disclaimer--I am not in any way connected with original work--just considered it for reuse. —Preceding unsigned comment was added by (talk) 07:27, 22 October 2022 (UTC)Reply[reply]

The reason for deletion was {{copyvio|1=Minecraft is a copyrighted video game}}. --Achim55 (talk) 10:23, 22 October 2022 (UTC)Reply[reply]

emoji cats[edit]

Since I notice categories of the form category:U+1F3FA popping up more often recently, could someone with the permissions add redirect cats from the emoji themselves (like category:🏺)? Since they are on the title banlist, not anybody can do it. Arlo James Barnes 06:28, 21 October 2022 (UTC)Reply[reply]

I suspect that someone with the privileges would be much more likely to move forward with this if someone provides a list of such redirects that are needed. - Jmabel ! talk 15:08, 22 October 2022 (UTC)Reply[reply]
Ah, I meant to link emoji/Table (warning, heavy page, probably best to open it directly in source mode) in my original message, but better late than never. My proposal would be that the contents of column 3 would redirect to the contents of column 1. Arlo James Barnes 18:02, 22 October 2022 (UTC)Reply[reply]
With over 1800 of these, I'm guessing this should be a bot task. It would be pretty tedious to hand-create 1800+ category redirects. - Jmabel ! talk 22:43, 22 October 2022 (UTC)Reply[reply]

I went to see where I could request such a task, but commons:bots/Requests seems to be geared towards botmasters. Arlo James Barnes 02:35, 24 October 2022 (UTC)Reply[reply]

@Arlo Barnes: I think you want Commons:Bots/Work requests. - Jmabel ! talk 14:45, 24 October 2022 (UTC)Reply[reply]

Videos whose components are hard to verify[edit]

i run into a problem with File:Soong May-ling gives a speech to the Congress of the United States of America.ogv. the video part is slideshow of old photos whose sources are hard to verify. the audio part is a recording of a speech delivered to the US congress. its youtube source is not any US gov agency .

so it's definitely not a "video of 1943". for these videos that have parts from different periods and are stitched together, how should they be properly categorised? RZuo (talk) 18:45, 21 October 2022 (UTC)Reply[reply]

Certainly not as "Videos of 1943". Ruslik (talk) 19:03, 22 October 2022 (UTC)Reply[reply]

October 23[edit]

Can someone download a hires version of this image?[edit] When you use the download function you get a highly compressed version. See File:Maurice Rumford Pierce (1887-1968) in 1917.jpg where I can't make out the signature of the creator, yet the website contains a hires version. Is there a way I am not aware of? --RAN (talk) 16:22, 23 October 2022 (UTC)Reply[reply]

There is no simple way to do this. Ruslik (talk) 17:25, 23 October 2022 (UTC)Reply[reply]
The IIIF manifest is at (in our own Mirador viewer).
Nice people also provide the original in the manifest, unfortunately this website doesn't seem to be run by nice people. You can use a IIIF downloader to get all the tiles and make a larger image out of these. Given that IIIF is getting used quite a lot these days, maybe we should provide a downloader on Toolforge. Multichill (talk) 19:16, 23 October 2022 (UTC)Reply[reply]
  • I hate archives that hoard their material. I respect that they have preserved the material, but why not make it widely available at a high quality, without destructive watermarks? --RAN (talk) 21:28, 24 October 2022 (UTC)Reply[reply]
I got it with the Dezoomify tool. Yann (talk) 19:27, 23 October 2022 (UTC)Reply[reply]

October 24[edit]

dark mode?[edit]

Is there a way to turn on dark mode on commons? thx. limitless peace Michael Ten (talk) 00:56, 24 October 2022 (UTC)Reply[reply]

There is not currently a built-in dark mode, but you can edit special:mypage/vector.css (substitute the name of another skin if you're not using vector, for example if you edit on mobile it would be `minerva`) to add in themes others have designed (some of which may be better than others). Currently I have Commons on light mode, but I've used d:user:Arlo Barnes/vector.css on Wikidata, it should largely be the same. Arlo James Barnes 02:30, 24 October 2022 (UTC)Reply[reply]
@Michael Ten: Hi, and welcome. No, sorry, there is not. I tried using the Dark Theme Tab in Chrome, but it has no way to distinguish on my watchlist between unread (bold, with a filled marker) and read (normal, with an unfilled marker). Arlo's solution, which I implemented in this edit for my monobook skin, leaves lots of light stuff (it's better in vector, but not updated for vector 2022).   — 🇺🇦Jeff G. please ping or talk to me🇺🇦 05:41, 24 October 2022 (UTC)Reply[reply]

October 25[edit]



I ran several times into the fact that there are several places where the guidelines regarding Depicts are described or discussed. And the ONE place where is actually not available is in the guidelines space themselves. This makes it very complicated to explain new comers how to add depicts in the most useful and in the least inconvenient way (enough depicts... but not too many depicts etc.). It also make it difficult for independent new comers to actually find out which guidelines they should follow.

So... I have attempted to merge and summarize what I could find as guidelines, discussions in talk pages and rationale, to keep the nutshell (and link to the other pages for more details). There is nothing really new... it is mostly a summary of the most important points. I am aware that there is no full consensus on how tagging should operate and I understand why. But it feels like the absence of visible and findable guidelines is making the situation worse.

Hence... what do you think of the simple Commons:Depiction guidelines ? Good move ? Decent ? Terrible ? Anthere (talk) 00:37, 25 October 2022 (UTC)Reply[reply]