Commons talk:Project scope
Outdated (does not reflect current admin practices): policy amendment for in-scope exceptions
[edit]I suggest adding something like the below to the section "Must be realistically useful for an educational purpose":
Moreover, administrators can decide that an image is to be deleted despite being legitimately in use on another Wikimedia project site. Elaborating multiple specific realistic use-cases is also not sufficient for an image to be in scope if participants of a deletion discussion vote otherwise and a closing admin agrees. Subsequent undeletion requests can be rejected if the majority by headcount of participants vote for keeping the file deleted.
This is to reflect actions such as the deletion of these files and the rejection of their undeletion:
- Commons:Undeletion requests#File:Female barbarian from space vs giant ant.jpg
- The image was for example useful for illustrating how contents (not barely decipherable low-quality covers) of vintage pulp sci-fi magazines looked like and can help illustrate a subject of a genre. Like some other deleted images there were very few other image in its category and it could be used in addition to these (all low-quality very old) ones.
- Commons:Undeletion requests#Files in Category:Giovanna IV di Napoli by Bing Image Creator
- The rationale was Unused AI-generated fan art, out of scope but they were neither unused nor fan art. Multiple use-cases have been elaborated in addition. They have been clearly marked as AI art and depicted high-quality medieval scenes including a character that looks similar to a historical person.
This seems to happening increasingly often with AI art against which there seems to be a bias. For example, a near-exact look-alike of this image was deleted despite the elaborated use-cases which may not be obvious when just briefly thinking about it (assuming we can and must always anticipate educational usefulness cases rather than realistic broad potential) as well as the most high-quality image for illustrating Category:Anachronism in art and fiction at the time. However, at the time it was not in use and no undeletion request was made. At the same time there seems to be a bias pro photographs where nearly everything is kept/not discouraged if at high-resolution even if no specific use-case is clear, there are countless photographs of the same, and no use-case gets elaborated in DRs. However, let's just focus on deletions that aren't compatible with current WMC policy here. Prototyperspective (talk) 14:13, 28 December 2023 (UTC)
- The first example here seems rather tendentious. You added it as a second image to a Wikidata item that already had a longstanding image. Normal practice on Wikidata is to have only one image per item, and the [:File:Amazing stories 193702.jpg|existing image there] was certainly an appropriate one. - Jmabel ! talk 19:13, 28 December 2023 (UTC)
- That image is low quality – low resolution and you can't even decipher what it's supposed to depict if you didn't know it beforehand. Also it's very old, doesn't necessarily look like insects, and is only a cover with text where a high-resolution image of such contents was previously missing and nicely complements it. It could probably also be useful in some Wikipedia article but why would Wikidata not be enough? It's educationally valuable, in scope, and rather unique here. Again if both neither realistical educational value (multiple have been specified) nor being in use does not matter, then the policy needs a change. Prototyperspective (talk) 19:33, 28 December 2023 (UTC)
- The second example is beyond tendentious, into the realm of the laughable. No, these were not "high-quality medieval scenes". Leaving aside the six-fingered monarch and the based-on-roughly-nothing representation of Giovanna herself, they were at about a level suitable for illustrating a children's book, and to the best of my knowledge they were not in use by any other project (though I'm ready to stand corrected on that last point; am I wrong?). There was a strong (about 80%) consensus to delete. - Jmabel ! talk 19:22, 28 December 2023 (UTC)
- Yes some of these had, probably easily fixable, misgeneration issues. However, most didn't have and they weren't low quality when you compare it with other art that depicts such scenes in terms of various aspects such as within high-resolution modern/digital art. These images are not (only) about any person depicted in them despite the file/cat title. For example it can illustrate how AI art can be used to depict scenes of the past or long gone people where we only have very few (in this case one) image made during their lifetime depicting them.
- So you are saying consensus overrides being in use. Again, this is not what this policy claims so it needs an update so that it means that votecounts are more important than that. However, I wouldn't call three votes (+admin) with refuted arguments "strong" consensus at all. And yes, like I said you are wrong on that. Prototyperspective (talk) 19:38, 28 December 2023 (UTC)
- I hope we're not using user-generated AI art to illustrate historical scenes. There are serious accuracy problems with that – here is one example.
- Still, I oppose the proposed policy change, because I agree with the current text of COM:INUSE: "It should be stressed that Commons does not overrule other projects about what is in scope. If an image is in use on another project (aside from use on talk pages or user pages), that is enough for it to be within scope." I think images should not be deleted as out-of-scope if they are in use on a content page of another project (except maybe some exceptional cases like obvious vandalism). —Mx. Granger (talk · contribs) 20:47, 28 December 2023 (UTC)
- No, certainly not in this case where it's not about specific historical scenes anyway but about artistic depictions of the past. Of course can there be accuracy problems. There can also be accuracy problems with paintings and images made with Photoshop. There clearly are accuracy problems here. However, the images are still useful. It's that people coming to this have one specific potential use-case in mind and then seeing how these images do not match their high standards for that particular usecase narrowly want it deleted. But it's not about that particular use-case(s) anyway / various assumptions are flawed. Secondarily, images that do not meet high standards can still be useful. For example: to illustrate how hamburgers are depicted in contemporary advertisements at high resolution, inaccuracy is not just common or an early fixable disclaimable issue but even required since in that case it's basically the whole point (and there was no such image before – it's different for subjects for which there are multiple equivalents). It's one thing to delete a few images out of a large set but another to delete few in-use ones so I don't see how that is possible with current policy. Prototyperspective (talk) 21:06, 28 December 2023 (UTC)
- So you are saying consensus overrides being in use. Again, this is not what this policy claims so it needs an update so that it means that votecounts are more important than that. However, I wouldn't call three votes (+admin) with refuted arguments "strong" consensus at all. And yes, like I said you are wrong on that. Prototyperspective (talk) 19:38, 28 December 2023 (UTC)
The problem was/is that the files were in use because Prototyperspective cynically decided to add them to other projects once it was clear no one agreed with that the files were eductional. Per the guidelines "The emphasis here is on realistic utility, either for one of the Wikimedia projects or for some other educational use. Not all images for example are realistically useful for an educational purpose. An image does not magically become useful by virtue of the argument that it could be used to illustrate a Wikipedia article on X, merely because X happens to be the subject of the photograph." Prototyperspective seems to under the false impression that them simply saying the files are eductional for "X purpose" or adding them to a sister project automatically makes them in scope. That's not how it works though. There needs to be a "realistic utility" and most people agreed there wasn't one. Partially due to the historical inaccuracy, but also because Commons just don't allow for amateur artwork by non-notable people to begin with. Again per the guidelines, "any use that is not made in good faith does not count."
That said, maybe you could argue there's a double standard when it comes to allowing for amateur photography but not artwork. That's a different discussion though and negate the fact that this whole is thing tendentious and bad faithed on its face. Especially since Prototyperspective knows COM:INUSE isn't a free pass or valid excuse to ignore consensus. Although I'd be open to discussing there being stricter rules around amateur photography, but I doubt Prototyperspective wants to do that and it probably wouldn't go anywhere anyway. That seems like the best way to balance things out if this is really such an issue though. --Adamant1 (talk) 21:54, 28 December 2023 (UTC)
- As an admin, I usually strictly apply COM:INUSE, but of course I wouldn't necessarily see COM:INUSE as applicable if an image was added to projects in the course of a deletion discussion by a participant to make a point. Seems pretty obvious to me and I don't think that any policy changes / additions are necessary. Gestumblindi (talk) 23:18, 28 December 2023 (UTC)
- No, it was not added "to make a point". When specific realistic use-cases are elaborated and not addressed or even refuted afterwards as well as at least one of them implemented via the file being in-use, but the file is deleted regardless how would this be compatible with realistically useful for an educational purpose, nor legitimately in use as discussed above scope delineations? It fails the scope definition twice. The files in use were few and the ant one was unique and the same deletion previously occurred for the only image sufficiently illustrating the modern concept of the category linked above and so on despite being a less notable subject). People just increasingly vote out by headcount various useful relatively unique images made using AI tool indiscriminately and without sufficient basis in policy and kind of ignore all realistic use-cases even when files are in use or when their later equivalents are in use. Prototyperspective (talk) 23:39, 28 December 2023 (UTC)
- Every AI-generated image is "unique" given a sufficiently specific description of its contents. But that doesn't mean they're all useful. Omphalographer (talk) 01:38, 30 December 2023 (UTC)
- No, it was not added "to make a point". When specific realistic use-cases are elaborated and not addressed or even refuted afterwards as well as at least one of them implemented via the file being in-use, but the file is deleted regardless how would this be compatible with realistically useful for an educational purpose, nor legitimately in use as discussed above scope delineations? It fails the scope definition twice. The files in use were few and the ant one was unique and the same deletion previously occurred for the only image sufficiently illustrating the modern concept of the category linked above and so on despite being a less notable subject). People just increasingly vote out by headcount various useful relatively unique images made using AI tool indiscriminately and without sufficient basis in policy and kind of ignore all realistic use-cases even when files are in use or when their later equivalents are in use. Prototyperspective (talk) 23:39, 28 December 2023 (UTC)
- Support adding the first sentence of the proposal. The other sentences are unnecessary and Commons doesn't work by voting anyway. Nosferattus (talk) 01:27, 29 December 2023 (UTC)
- I absolutely oppose that sentence "Moreover, administrators can decide that an image is to be deleted despite being legitimately in use on another Wikimedia project site". That would open a can of worms. There is one exception to COM:INUSE, that is copyright concerns, and this is already clear enough. Gestumblindi (talk) 10:54, 29 December 2023 (UTC)
- Oppose. As someone who supported the undeletion requests, this is just disrupting the Commons to make a point. COM:INUSE only applies to good-faith uses. At the time, I made a mistake and came to the conclusion that, despite your bludgeoning, your additions were good-faith uses, and supported the undeletion request on that basis. Based on your subsequent conduct, and reconsidering my view of your prior conduct, I am forced to conclude that these were not, in fact, placed in good faith, and rather that they were placed either to 'save' an image from deletion without considering the appropriateness of the placement or make a larger point about AI images. Prototyperspective: you need to drop the stick. —Mdaniels5757 (talk • contribs) 22:00, 29 December 2023 (UTC)
- Oppose - I see no evidence that current practise is out-of-alignment with policy, nor do I see evidence of a genuine issue requiring a policy change. Current policy requires a file be "legitimately in use." An advocate of retention slapping an unused image into a sister project during a scope DR is not legitimate--there are numerous DRs with this finding and, as a symmetrical corollary, the inverse of removing an image to influence deletion is equally illegitimate. Alternatively stated: changing an image's use to influence the outcome of a scope-related DR is not legitimate vis-a-vis INUSE. Such gaming is nothing but a disingenuous ploy, a failure to appreciate the spirit of assessing dispassionate measures of educational use. Discussions here and elsewhere suggest that the flaw lies not with policy or practise, but with Prototyperspective's (mis)understanding of "legitimate" and (in)ability to listen to the many contrary opinions. Эlcobbola talk 23:28, 29 December 2023 (UTC)
- All fine but one thing is not true at all: I did listen to the contrary opinions and directly addressed them. Besides elaborating unaddressed realistic usefulnesses, one thing I asked is simply why it would be fan art which simply wasn't answered.
- While changing an image's use to influence the outcome is not clear from the current policy phrasing, that's not what has been done; nevertheless your earlier elaborations make sense, contain good points that refer to policies, and I hope we can agree to disagree on how educational use is assessed in regards to the spirit of this policy as it's currently written. Probably last comment of mine here. Prototyperspective (talk) 23:41, 29 December 2023 (UTC)
COM:INUSE and non-content pages
[edit]On Commons:Deletion requests/Files in Category:Hackers, User:Matrix raised some concerns that the wording of this passage leaves some doubt as to whether files used in userspace are considered "in use":
A media file that is in use on one of the other projects of the Wikimedia Foundation is considered automatically to be useful for an educational purpose, as is a file in use for some operational reason such as within a template or the like. Such a file is not liable to deletion simply because it may be of poor quality: if it is in use, that is enough.
[ ... ]
It should be stressed that Commons does not overrule other projects about what is in scope. If an image is in use on another project (aside from use on talk pages or user pages), that is enough for it to be within scope.
Would it be appropriate to add some wording such as:
A media file that is in use in a content page on one of the other projects of the Wikimedia Foundation is considered automatically to be useful for an educational purpose, as is a file in use for some operational reason such as within a project page, a template or the like.
to clarify that images used in non-content namespaces are not considered COM:INUSE? Omphalographer (talk) 20:59, 7 June 2024 (UTC)
- Partly, this seems to follow from the next sentence: "The uploading of small numbers of images (e.g. of yourself) for use on a personal user page of Commons or another project is allowed as long as that user is or was an active participant on that project." So, COM:INUSE already says that images "for use on a personal user page of Commons or another project" are only acceptable in "small numbers". I don't see a pressing need for changes to the existing wording. Gestumblindi (talk) 21:47, 7 June 2024 (UTC)
The definition of "legitimate" use regarding COM:NOTUSED
[edit]The section COM:NOTUSED introduces the concept of a file not "legitimately" being in use, but doesn't adequately define it. There have been several discussions about projectspace/userspace, but I want to address ambiguity where it comes to mainspace (as in Wikipedia articles, Wiktionary word pages, etc.): the role of the ratio of volunteers to content.
On the English Wikipedia, for example, there is a relatively high ratio of volunteers to content. If an image is added where it is not useful, it is more likely that it will be spotted and reverted. If not, it is relatively easy to recruit participants to a discussion to decide on its use. On projects with a relatively low ratio of volunteers to content like Wikidata or some of the smaller Wikipedias, it is unlikely for a poor image to be reverted and difficult to attract attention to decide the matter. Often, a Commons user with no connection to that language will just go in and remove the image in order to get around INUSE. That works, but it always feels subversive.
How could this page address a scenario like this: a user adds their own very low quality photo to an obscure page of the Latin Wikipedia. If the Latin Wikipedia had more volunteers, it could easily spot and remove this photo, however no volunteer has noticed and none of the people involved with the matter on Commons know Latin. On Commons, someone nominates it for deletion on the basis of it being very low quality but the creator argues it's COM:INUSE on the Latin Wikipedia. What happens next?
The fact is, Commons is a much bigger community than many small Wikimedia projects and has much greater capacity to assess the quality of an image. Is one solution to acknowledge this? — Rhododendrites talk | 17:25, 31 July 2024 (UTC)
- If a photo is added to a small Wiki against the local guideline and there are no local admins who can act against this then this is a case for global admins and stewards not for Commons. We can not make editorial decisions or admin actions on other Wikis. GPSLeo (talk) 18:44, 31 July 2024 (UTC)
- In practice, we often do. All of us have accounts on all wikis, after all. I cannot recall the last time I did so personally, but when I remove something on enwiki, it's in my capacity as an enwiki editor rather than as a Commons editor imposing my will on enwiki. If I happen to see vandalism on a wiki I've never edited before, should I refrain from reverting and go talk to a global admin/steward? Or should I use my account on that wiki to undo an obviously bad edit? The question is, realistically, how confident you are that it clearly contradicts local policies. I don't typically edit other wikis because I don't know, but if it were an obvious case, I'd consider it. Do global admins/stewards regularly make edits that are simple editorial tasks of reverting a subpar (but not vandalism) edit? Regardless, if any mainspace use is automatically "legitimate" as far as commons is concerned, we should be clear about that. If the only way to delete something that's in use is for local users to find consensus against it or to solicit a decision by a global admin/steward, we should probably say that, too. — Rhododendrites talk | 19:10, 31 July 2024 (UTC)
- Seconding this. The role of admins and global stewards is to handle occasional tasks which require admin privileges, like deleting spam pages and blocking disruptive users. They aren't here to review all edits; there aren't enough hours in the day for that. Users of other wikis are expected to use their best judgement when evaluating edits on small language wikis. Omphalographer (talk) 20:17, 31 July 2024 (UTC)
- In practice, we often do. All of us have accounts on all wikis, after all. I cannot recall the last time I did so personally, but when I remove something on enwiki, it's in my capacity as an enwiki editor rather than as a Commons editor imposing my will on enwiki. If I happen to see vandalism on a wiki I've never edited before, should I refrain from reverting and go talk to a global admin/steward? Or should I use my account on that wiki to undo an obviously bad edit? The question is, realistically, how confident you are that it clearly contradicts local policies. I don't typically edit other wikis because I don't know, but if it were an obvious case, I'd consider it. Do global admins/stewards regularly make edits that are simple editorial tasks of reverting a subpar (but not vandalism) edit? Regardless, if any mainspace use is automatically "legitimate" as far as commons is concerned, we should be clear about that. If the only way to delete something that's in use is for local users to find consensus against it or to solicit a decision by a global admin/steward, we should probably say that, too. — Rhododendrites talk | 19:10, 31 July 2024 (UTC)
- First of all it's most hypothetical and has proven to largely be sufficient with it being intentionally broad. Still it may good to clarify this. An image relevant to some Latin-language Wikipedia being included in the article is clearly legitimately in use. If it seems highly illegitimate I think one would should first try to resolve this on the Wikipedia. That is for example pinging authors of the article and asking them. I've done this two times or so after which editors simply removed the identified problematic file after which it was no longer in use. Prototyperspective (talk) 20:14, 31 July 2024 (UTC)