This page is used for discussions of the operations, technical issues, and policies of Wikimedia Commons. Recent sections with no replies for 7 days and sections tagged with {{Section resolved|1=--~~~~}} may be archived; for old discussions, see the archives; the latest archive is Commons:Village pump/Archive/2024/12.
Please note:
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.
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).
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:
SpBotarchives all sections tagged with {{Section resolved|1=~~~~}} after 1 day and sections whose most recent comment is older than 7 days.
November 13
Long-term disputes on various wikis involving a cross-wiki IP author
There are numerous disputes involving an IP user indulging in cross-wiki spam, particularly articles on West Germanic varieties. I am hounded for a while.
From what I can see, he's a user who upload much about mapping and cartography, for which is great, but to engage in further conversation with German Low, or etztes with or without a s, Low Saxon with Low German to Low German only, for me all this and the on-going conversation, does not contribute to anything positive. Germanic lang have much variation, as well as French or Latin, especially from those area. You could simply add a variant, or suggest that it might be spelled with a different phonetic sounds. I did review quite a few contribution he made, and this could be solve quickly. In my opinion he is contributing for which is great, if you are trying to bring post back from 2003... It seems like he is using the same account, and he will keep using it since he's in love with the appreciation of contribution... I suggest to close this topic for now and simply add a watch alert.
@Sarcelles: Is this some sort of request for administrative action? If so, it belongs on the appropriate Administrators' noticeboard, not on the Village pump. Conversely, if it is something you are just bringing up for general discussion, I don't know what you want discussed. - Jmabel ! talk18:37, 13 November 2024 (UTC)[reply]
None of these accounts have edited in recent weeks, some not in as long as half a year, so it is hard to imagine what anyone can do about this at this point. - Jmabel ! talk18:40, 13 November 2024 (UTC)[reply]
It can be anticipated, that this author continues to be active on several wikis including Commons. I think this is a good place to discuss this cross-wiki spam. On en.wiktionary I have been removing numerous typical edits by this user. Sarcelles (talk) 14:29, 16 November 2024 (UTC)[reply]
File:Niederfränkisch.png is a file of this kind. It attempts to picture Low Franconian varieties in Europe. It has the following threefold-division:
A minor transitional area to Low Saxon, in the Netherlands
East Bergish running from near the city centre of Essen to Westphalia, also quite small#
A somehow larger area cutting through all of the following: an arrondissement bordering to Brussels, Antwerp province, Dutch Limburg, Belgian Limburg, Duisburg, Düsseldorf, Wuppertal, German-speaking Belgium and French-speaking Belgium.
This is called Nordniederfränkisch (North Low Franconian) and running from France to Holland, Friesland province, Brussels and Westphalia. Sarcelles (talk) 11:45, 22 November 2024 (UTC)[reply]
Dialects of Dutch and German This is a typical German map of some of the dialects from Italy to Denmark. The author is MicBy67, User:Postmann Michael (The discussion creator's blatant and deliberate lie has been corrected! The map creator was not the Commons accounts mentioned, but the account User:Et Mikkel~commonswiki! Let's stick to the truth for once! --MicBy67 (talk) 02:34, 7 December 2024 (UTC)). There should be a further discussion of this issue. Sarcelles (talk) 20:07, 5 December 2024 (UTC)[reply]
Yawn!
Cook up a fresh idea! ;)
Hey, did you happen to catch the latest post on the discussion page? Just checking!
@MicBy67: consider yourself formally warned that the wording of your posts here has been unnecessarily uncivil, and continuing in this vein will probably result in me (or someone else) blocking you. - Jmabel ! talk17:53, 6 December 2024 (UTC)[reply]
@Jmabel: I take note of that. And I'm aware that the discussion creator (or his mentor) will bring up the old story again with the original account (Postmann Michael) was blocked on the German-speaking Wikipedia because of POV from dubious sources, trivialization of National Socialism. Harmful to Wikipedia on the one hand, and with the successor account (Et Mikkel) was permanently blocked on the German-speaking Wikipedia as a way to circumvent the blocking on the other hand.
Nobody really cares about the past two decades anymore!
What is striking, however, is the fact that the discussion creator is trying to construct a connection between the IP's and me. And is cross-wiki hounding actually allowed on Wikimedia Commons? I am asking now for an interested friend … --MicBy67 (talk) 02:55, 7 December 2024 (UTC)[reply]
Are the accounts I allegedly attacked just one that launched a smear campaign against me, to which I responded sarcastically? By the way, this is irrelevant to Wikimedia Commons!
I doubt the identity of the account stated. And that is my right.
You were for yourself banned from the German-speaking Wikipedia for years because of “pointless article work,” right?
Do you haven't anything better to do than try to link my account to the Paderborn IPs in all Wikimedia/Wikipedia projects?
Sarcelles has a history at Dutch Wikipedia of dropping questions about dialects and languages and their boundaries, aimed at construing one opinion as being the truth and falling silent when objections arise. Even so one day before arrival of the archive bot he used to add a random remark to avoid archiving so I took to manually archiving his messages or scrapping them altogether. Also I often alert unsuspecting users to this behavior implying that answering is pointless.
Although Sarcelles poses no acute threat to the wikis, I would be relieved if he could be banned for good from all projects. These tedious and time-consuming discussions lead nowhere, least of all to our prime objective. His minor contributions in the main space do not in any way compensate for the inconvenience.
Not. They will be resolved when the linguists reach consensus i.e. not before 2080. Our task is to describe the debates, not to resolve them → bertux16:14, 19 December 2024 (UTC)[reply]
I reject this type of “apology.” It is not sincere and is merely intended to fool!
First and foremost, the person who initiated the discussion is a die-hard advocate of a “missionary” mission. That's always been the person behind this account!
I'd like to remind you that this account initially solely campaigned on Wikipedia for “human rights violations” outside German-speaking countries and indicated in its blocking procedure on German-speaking Wikipedia that it had created these “reports” or “articles” at Wikipedia “at the risk of its life.”
We all realize that this statement is complete garbage!
The owner of this account has primarily focused on languages and dialects. He now describes himself as one of the “most experienced linguists,” most likely due to his participation in dubious accounts.
In addition, there is a tendency to conduct monologues that usually lead nowhere!
Based on my history with this account, I expect this discussion will go nowhere and reopen in a month or two with a different lemma—as is typical with the person behind this account! --MicBy67 (talk) 01:06, 22 December 2024 (UTC)[reply]
December 10
"prove a license"
Hi, swetrails.com has many images uploaded by a particular user with CC-BY 4.0 (example)- so far so nice and good. On the page, the license is shown as a text overlay on each image but not embedded in the image. I uploaded several of those images with appropriate license information etc. to wikimedia commons. Per chance, over last weekend, the swetrails.com server was not accessible, which made me think, what actually would happend if the server were taken down for good and the copyright owner then denied they ever granted the CC-BY 4.0 license? There is also no contact information anywhere on the site, neither as Imprint nor for the particular contributor. Since nothing is embedded in the original image, it would not be possible to prove that the license was indeed granted at some point in time. How would Wikimedia handle such a case?
--Uli@wiki (talk) 12:10, 10 December 2024 (UTC)[reply]
It's not really clear to me how you could do it on swetrails.com, but sometimes people will save a copy of the page to the Wayback Machine and use it as the source. Otherwise you can just ignore it since they don't really care about dead links on here to begin with. We aren't as dependent on external sources as a project like Wikipedia would be and their pretty lax about them to. So it's probably a none issue as long as you provide a source to begin with. --Adamant1 (talk) 12:19, 10 December 2024 (UTC)[reply]
The VRT is a voluntary team that handles emails with copyright issues but also permissions. If the team gets permission for the use of Roblox screenshots here, then it would be possible --PantheraLeo1359531 😺 (talk) 08:50, 13 December 2024 (UTC)[reply]
Someone contacted VRT a month ago to warn us about this user uploading questionable AI upscales of paintings and other artworks.
One was nominated for deletion, but there are hundreds others and the user nominated them as valued images (and of course replaced the original image on all Wikipedias).
You should also ping User:Archaeodontosaurus. He is a longtime and extremely productive and active user, and it's not appropriate to talk behind his back. We're going to have to discuss AI images going forward, and as they get better, the issues around them will get trickier. -- Ikan Kekek (talk) 08:42, 15 December 2024 (UTC)[reply]
Thanks for pinging him. It looks like we're working toward a consensus. I agree with the concerns expressed below, but I have to wonder whether we'll even recognize AI in a few years, and that's when problems of accuracy and authorship will become critical and possibly irresolvable. -- Ikan Kekek (talk) 12:52, 15 December 2024 (UTC)[reply]
Part of what bothers me about these images is that they're all being uploaded without any disclosure that they've been AI upscaled, and without making the original image available. The Fra'Mauro map is a great example - the original would have been a great image; the upscaled version is a mess. And I'm worried that a lot of the other upscaled images may have hidden issues as well. Omphalographer (talk) 08:50, 15 December 2024 (UTC)[reply]
I wondered if the upscaling might be occurring in-camera without the uploader being aware that this was happening, but from the EXIF they're using a Nikon D850 from 2017, which I guess rules that out. Belbury (talk) 10:14, 15 December 2024 (UTC)[reply]
Comment I am very open to discussion. Making AI takes me a lot of time. It is difficult and the results are questionable. We must explore this path, it is a tool, we must learn.
I understand the reluctance.
We could find a simple solution: I propose that there be a label to be placed on the images saying whether there was an AI intervention or not.
Could one tell me a reasonable use case for AI upscaled raster images that falls in our scope (beyond demonstrating the technique)? {{AI upscaled}} words it this way: "This process may have introduced inaccurate, speculative details not present in the original picture." I think that any media that relies on details to be understood (portraits, maps, biological organisms, depictions of machines and mechanics...) should not be processed in a way that introduces "inaccurate, speculative details". That would be kind of "alternative facts" that I would not welcome. Regards, Grand-Duc (talk) 09:36, 15 December 2024 (UTC)[reply]
Putting aside the scope and accuracy issues, why even bother upscaling a portrait to begin with? Because I really don't see the benefit of it regardless of if AI upscaled images are accurate or in scope anyway. --Adamant1 (talk) 10:14, 15 December 2024 (UTC)[reply]
Archaeodontosaurus, would you be able to add the {{AI upscaled}} template to all of the files you have uploaded which were upscaled by AI? Another editor could also automate this templating process for you if you provided a simple list of the files. --Belbury (talk) 10:57, 15 December 2024 (UTC)[reply]
Comment As far as overwriting files with upscaled versions goes, there is already a policy in place on Commons, which states not to do that. In case of overwritten works, those edits should be reverted. The upscaled versions could still have a place here on Commons as separate uploads, though I agree with the others that their status as upscaled with AI should still be disclosed. ReneeWrites (talk) 14:33, 15 December 2024 (UTC)[reply]
There's a bit of a hole in this policy in that it doesn't say anything about uploading upscaled images as the only version of an image. We may need better guidance to users about this, probably along the lines of "if you're going to upload an upscaled photo, upload the original first". Omphalographer (talk) 19:26, 15 December 2024 (UTC)[reply]
The DR has been closed as keep since it is COM:INUSE. I don't think any AI upscaled file should be a Valued image, and this is definitely a cross-wiki issue that needs resolving because AI upscaled images should not replace originals. Abzeronow (talk) 19:37, 15 December 2024 (UTC)[reply]
@Abzeronow: Just an FYI, but I started a discussion about AI artwork on Wikipedia a while ago here. The discussion hasn't gotten many comments yet, but it sounds like AI generated or manipulated images clearly go against the guidelines on Wikipedia's end. Especially in cases where it isn't explicitly clear that's what they are. I'm not going to claim to know how that meshes with the guidelines on Commons, but I really don't think someone can claim AI manipulated or generated images are "legitimately in use" on Wikipedia if there's multiple users and guidelines saying they don't belong in articles. At that point the images clearly aren't being used in good faith. I'm not going to suggest it myself, but maybe it would be worth editing Commons:Project scope to say as much. --Adamant1 (talk) 19:52, 15 December 2024 (UTC)[reply]
I think AI advocates have good faith views. As someone who is somewhat resistant to technological change, I can see their viewpoint while disagreeing with it. And I think Archaeodontosaurus has good intentions. Abzeronow (talk) 20:03, 15 December 2024 (UTC)[reply]
Just to be clear, I meant it purely pragmatically. I don't really know or care what Archaeodontosaurus or anyone else's intentions are but there's no reason images should be kept on our end if the usage violates Wikipedia's guidelines to begin with. I assume they wouldn't consider it good faithed on their end to put AI generated artwork in articles anyway though. But hey, whatever. It's all about the vibes right? Who cares about policies, standards, or what Wikipedia users want. The more important thing is not offending anyone who advocates for AI artwork by saying that people on Wikipedia don't think it belongs in articles. --Adamant1 (talk) 20:14, 15 December 2024 (UTC)[reply]
I think the English Wikipedia might be something of an outlier for actually ruling out AI-upscaled historical content in its MOS. Most upscaled photographs I see going through Commons are for the French and Russian projects. Belbury (talk) 20:30, 15 December 2024 (UTC)[reply]
Oh yeah for sure. I don't know about the French language Wikipedia, but the Russian Wikipedia has absolutely no standards what-so-ever. At least from what I've seen they are totally fine with clearly inaccurate historical content. Everyone is different though. So I wouldn't go as far as saying every project doesn't allow for AI generated or altered images, but English Wikipedia clearly has a problem with both. --Adamant1 (talk) 20:35, 15 December 2024 (UTC)[reply]
There's also quite a few projects which don't have a position on AI content simply because the local community is small and hasn't made it a priority to develop policies. AI content ends up getting used on these projects not so much because it's permitted, but because no one's telling editors to not use it. Omphalographer (talk) 21:07, 15 December 2024 (UTC)[reply]
Then some Russian-speakers should do a RFC or some other policy discussion on Russian Wikipedia to see if the community thinks AI content violates rules. It is not in Commons purview to usurp that decision making. Abzeronow (talk) 21:33, 15 December 2024 (UTC)[reply]
Well, only when the English Wikipedia doesn't want it on their project right? Then it's totally cool to usurp their decision. Otherwise we have to respect the projects decision making. --Adamant1 (talk) 21:55, 15 December 2024 (UTC)[reply]
The only usage on Enwiki is a talk page, and the AI upscaled nature of it is discussed. It's used on other Wikipedias though and we don't privilege Enwiki above other Wikipedias. Abzeronow (talk) 21:19, 16 December 2024 (UTC)[reply]
I don't think we do either which is why I didn't claim we did, but I don't think we have a policy on Commons saying Commons:Project scope doesn't apply when AI is involved either. Or are you going to argue that Commons:Project_scope#File_not_legitimately_in_use shouldn't matter to AI generated images just because some random people who have absolutely nothing to do with this what-so-ever are generally good faithed about it? --Adamant1 (talk) 21:52, 15 December 2024 (UTC)[reply]
Not sure whether that close was the right call. As a counterpoint, Commons:Deletion requests/Files in Category:AI-upscaled paintings was closed last month with all the images being either deleted or reverted/redirected to non-upscaled versions, despite some being in use at the time.
It seems reasonable to expect that where this map image has been added to an article by someone other than the uploader, that person did not realise that it was an AI-upscaled version with smudged inaccuracies, and that they would not have used it if they'd been made aware of that. It's been the first search result on Commons for a year, presumably because it's a Valued Image, and is of a significantly higher resolution than all the other copies we have. Belbury (talk) 20:23, 15 December 2024 (UTC)[reply]
I just had a train of thought... Does somebody know about a Wikimedia project, be it a Wikipedia, -news, -voyage, -species, etc., who doesn't have criteria for verifiability and a need to name sources for its contents? If there's none as I tend to assume, then the whole point of whether AI generated or AI upscaled images are in scope (also per COM:INUSE) is moot IMHO and every addition to any Wikimedia project, besides a few usages to illustrate AI technology, from the outset against the basic requirement of verifiability, even if any project does not have a explicit policy about AI media. AI modified graphics are never verifiable and traceable back to their sources, inherently so due to the underlying software techniques of the AI models. They are always a condensation of most statistically probable colour and luminosity values for each pixel for any given prompt. Regards, Grand-Duc (talk) 01:04, 16 December 2024 (UTC)[reply]
As a past contributor to enwikiversity: those guidelines have never been enforced. The "Review Board" proposed in those guidelines is entirely hypothetical, and the project has been a magnet for crank science for years. This is part of why I'm no longer involved with that project. Omphalographer (talk) 03:20, 16 December 2024 (UTC)[reply]
Wikisource doesn't really care where the images on the author pages come from; they're not part of our primary content. And OCR is at least AI-adjacent, and it's heavily used on Wikisource. Honestly, any digital picture is full of "statistically probable colour and luminosity values"; converting the raw to JPEG involves approximating the data from the camera with cosines that hopefully don't change any important details as well as storing only one color sample for every four luminosity values (often after extrapolating the color sensors of the camera, which has RGB filters on each pixel, so they were a bit questionable to start with), plus in-camera corrections for lenses, plus a pile of numbers set by the camera or photographer (ISO, f-stops, camera speed, plane of focus) that may or may not be preserved by the EXIF data. It's always better to have an actual picture of a peach; but if I'm comparing a human drawn T-rex with an AI drawn T-rex, and the paleontologists say the latter is more accurate, I don't think verifiability changes anything.--Prosfilaes (talk) 03:42, 16 December 2024 (UTC)[reply]
Comment More on topic, I was comparing some of the upscaled images with the originals. The text on the map in File:(Venice) Fra'Mauro's World Map - Biblioteca Nazionale Marciana.jpg is clearly smeared to the point of being unreadable. Whereas the text in the original is perfectly fine. People can debate all day if someone intentionally uploading a clearly blurry, smeared file is good faithed or if a map that no one can even read because it's smeared is educational. It's pretty clear that these images shouldn't qualify as "valued images" though. Per Commons:Valued_image_value "valued images are less about perfect technical quality...and more about the useability of the information on the image page. There's nothing usable about a smeared, unreadable image of a map. None of the other files seem to be any better either. So all of these AI upscaled images should have their "valued images" status revoked since they clearly don't meet the standard. --Adamant1 (talk) 07:39, 16 December 2024 (UTC)[reply]
Artifacts from the upscaling are all over the drawing if you view it at full resolution. If you put them side by side you can see that the cardboard of the original has a texture of its own, and that texture returns in the other version, but here they are morphed into paint blots. As for the claim that no AI was involved, maybe by that he means he used a different post-processing method, but whatever you want to call it has severely impacted the integrity of these images. I would honestly like to see the originals uploaded to Commons as well, even if Archaeodontosaurus feels they are of inferior quality, but other people may feel differently about that. ReneeWrites (talk) 22:01, 16 December 2024 (UTC)[reply]
Ooof! yeah, that totally blows out a bunch of legitimate texture, to the point where if it is in any sense a photograph, then the "camera" in question is doing a bunch of dubious post-processing.
Given that the VI system has no described process for removing the status from an image, and that nobody has answered a project talk page thread asking about this from six months ago, I've been bold and removed the {{VI}} template from this map, to see if anybody connected to the project reacts to that. Commons shouldn't be boosting this to the top of search results as its "most valued image" of Fra Marco's map, when it is of significantly lower quality than other photographs in the same category. Belbury (talk) 11:34, 18 December 2024 (UTC)[reply]
I asked about it myself two days ago and haven't gotten an answer yet. From what I remember with prior incidents the images were just deleted, which isn't possible here even though it clearly should be. So just removing the templates seems like the only other option outside of waiting for months until someone responds with an actual solution. There really should be a more formal process for revoking valued image status from a file though. I'm actually kind of surprised there isn't one. --Adamant1 (talk) 11:42, 18 December 2024 (UTC)[reply]
Comment I confirm what I have already said, and I urge you to go and see the works of Suzanne Valadon, they have the particularity of being very colorful... See it for yourself and do not trust images from the Web which are of poor quality. --Archaeodontosaurus (talk) 06:24, 17 December 2024 (UTC)[reply]
@Archaeodontosaurus: No offense, but the images are clearly AI upscaled and there isn't any reason you wouldn't know about it. Especially since multiple people have confirmed it at this point. Can you at least stop nominating your uploads for "valued image" status until this is sorted out? Otherwise it seems like your just intentionally ignoring the issue, which isn't a super great way to deal with it. --Adamant1 (talk) 08:34, 17 December 2024 (UTC)[reply]
Would people support replacing this image with one that's maybe color-tuned to be a but less yellow but not AI upscaled? I'd be happy to do that. I don't think AI-upscaled images can ever be acceptable. AI upscaling introduces inaccuracies, it's a guess about what extra details are present. We should be about facts-we should not use a fake image over a real one over it just because it looks nicer. Otherwise I'd support deletion, removal of valued image status, etc. Blythwood (talk) 03:33, 18 December 2024 (UTC)[reply]
"Ever be acceptable" as valued images in particular? Or in general for Commons? Because I can agree with the former, but as to the latter, I think AI upscaled images can be very instructive here at Commons. —Justin (koavf)❤T☮C☺M☯03:39, 18 December 2024 (UTC)[reply]
Hola @AbchyZa22: no soy un administrador, pero las imágenes generadas mediante AI no están sujetas a copyright (más información), y quedan, por tanto, automáticamente en dominio público, luego, hasta donde yo sé, cualquier imagen generada mediante AI por un usuario puede ser subida sin problema a Wikimedia Commons, en cuanto a copyright (siempre que además sea considerada como apta para una finalidad educativa). MGeog2022 (talk) 15:33, 16 December 2024 (UTC)[reply]
AbchyZa22 wrote here to warn administrators that images generated by Grok AI (created by Elon Musk) include a non-freely licensed watermark, so when images generated by this AI tool are uploaded to Wikimedia Commons, and they contain that watermark, it may be necessary to delete them (or maybe removing the watermark is enough, since AI generated images are public domain in any case). MGeog2022 (talk) 19:43, 16 December 2024 (UTC)[reply]
Vale, el logo es dominio público, luego solamente se trata de una cuestión estética. En ese caso, no es algo que sea muy importante (no creo que sea tan necesario que los administradores lo tengan presente).
---
OK, the logo is in the public domain, so it's only a purely aesthetical issue. In that case, this isn't a very important issue (I don't think it's so necessary for administrators to keep this in mind). MGeog2022 (talk) 20:55, 16 December 2024 (UTC)[reply]
@AbchyZa22, bien, por tanto, no hay ningún problema con el copyright, como te decía (sólo podría haberlo si el logo no fuera libre, pero no es el caso, porque ya vimos que es un logo de texto, por lo que aplica PD-textlogo). Las imágenes generadas por AI, como tales, no tienen copyright, incluso en el caso de que el proveedor de la AI afirme que sí lo tienen (puedes ver aquí lo que dice al respecto la política de Wikimedia Commons (en inglés). MGeog2022 (talk) 14:00, 24 December 2024 (UTC)[reply]
A dangerous precedent - DMCA after false relicensing
I wish to make the community aware of a serious issue that seems to be arising as Commons ages. I had a habit of uploading material in 2014-2016 from other sources to Commons, with much of it lost from the internet today, no longer accessible through traditional sources. I did not have the habit of archiving all pages with license information at that time, something I only later began with.
The reason I bring this up here, is twofold: 1) this is the second time I have seen this bait-and-switch with regard to licensing, with me having to spend time to comb through the Internet Archive last time to identify that it was originally licensed under CC-BY, and 2) that I do not find that the message on my talk page abides by community guidelines or takes account the WMF's role on Commons. It makes threats of banning me relating to repeat DMCA violation - which can not be grounds for a ban. Rather the question is if the copyright has been violated, which it hasn't, an issue which as far as I can see has not even been explored here. This has a potential chilling effect, in part because it requires me to disclose my name to counter-claim, which violates community guidelines - but also in part because this legal battle should be precisely for the WMF to take, not me personally. This is about a private entity, uploading to a YouTube channel which they still control, then revoking the CC-license, and issuing a DMCA-takedown request 9 years later.
There was never a source linked on these file pages and the license was never proven. When noticed by someone these files would already have been deleted. GPSLeo (talk) 19:17, 16 December 2024 (UTC)[reply]
Of course there was, are you telling me that it was unlicensed for 9 years? And what do you mean by proven? The page included a link to a page that upon being uploaded in 2015 included correct license information, that subsequently was changed. As I said, this is not about "proof". The matter at hand is about pages changing their licensing. I don't care one bit about the file itself, it was of marginal use. CFCF (talk) 19:19, 16 December 2024 (UTC)[reply]
Your license information was Anatomy video from youtube by Kenhub. Licenced as CC-BY as of download date 3/1/15. This is quite plausible, though you didn't provide a link to the YouTube video directly. I'm absolutely assuming your good faith and willing to believe that this was the case, but we would need proof (see my comment below). Gestumblindi (talk) 19:23, 16 December 2024 (UTC)[reply]
(edit conflict) Just as a service, if you intended to ping Joe Sutherland, your link didn't work, because his user account is User:JSutherland (WMF), but this link should generate the ping now. - Without deeply delving into the specific case, I agree that, if the videos ware originally uploaded to YouTube under CC-BY and the license was changed later (after you transferred the videos to Commons in 2015), you are of course not at fault and could have done nothing to prevent this. There was no License review as far as I can see (as an admin), so we have no proof, but maybe the original license can still be found at the Internet Archive or the like. As CC-BY licenses are irrevocable, maybe the WMF itself should file a counter-notice in case it can be shown that the content was - and therefore would still be -, in fact, freely licensed? Gestumblindi (talk) 19:21, 16 December 2024 (UTC)[reply]
As response to both comments, that seems possible I could have for whatever reason have omitted the link in lieu of the name of the video. I don't think it's likely that the correct license can be found, as YouTube is generally not covered by the Internet Archive, and I have no idea at what date they relicensed. Also, frankly - I don't think it's worth it for these files (also why I only uploaded 2 of maybe something like 50 files at the time). I never did find any use for them in any project, and restoring them would likely leave them as orphans. However, I think the issue itself is the dangerous thing. We have amassed a huge trove of material that could be relicensed incorrectly, and even if material has been "proven" - does that imply that a permanent record of the licensing has been kept? CFCF (talk) 19:43, 16 December 2024 (UTC)[reply]
YouTube video pages which are linked from Wikimedia web pages are generally pretty well covered by the Internet Archive. Unfortunately, since none of the source videos were linked on these files, it's quite possible that the original videos weren't archived. Even if they were, we have no easy way to find them. (Kenhub appears to have refreshed their YouTube content at some point between 2015 and now; the original videos don't seem to be online anymore.) Omphalographer (talk) 20:59, 16 December 2024 (UTC)[reply]
Asking for license review before it's too late, as I just did here, is a good option, especially for sites like YouTube and Flickr (this fact should be better known). Sadly, it seems that for this case, it was too late. Linking to the exact video, if a Commons bot automatically archived the page in Wayback Machine, would be another way to prevent this. About a year ago, a survey was carried out, and I created this proposal to try to avoid this kind of problem. In fact, I wasn't aware of the existence of license review then: that's basically what I was asking for, and nobody said in the discussion page that it already existed (perhaps what I exactly meant was not well understood). MGeog2022 (talk) 21:17, 16 December 2024 (UTC)[reply]
I don't think anyone is suggesting to tackle this one file at a time. I don't care about this file, and came here to make this general point. CFCF (talk) 20:15, 16 December 2024 (UTC)[reply]
The "Show more" link on the Wayback Machine archive doesn't work properly for me, but looking at the HTML source isn't hopeful (whitespace elided): <h4 class="title">License</h4><ul class="content watch-info-tag-list"><li>Standard YouTube License</li></ul> --bjh21 (talk) 22:38, 16 December 2024 (UTC)[reply]
That's what I was looking for. IMO this doesn't look good. I support deleting these videos unless we get some evidence of a free license. Yann (talk) 18:07, 17 December 2024 (UTC)[reply]
@Yann: I've found some evidence of a free licence. Since this is complicated and not really related to the original subject of this thread, I plan to start a discussion about it over on COM:VPC. --bjh21 (talk) 14:08, 18 December 2024 (UTC)[reply]
Unfortunately, one of the details of the DMCA is that a web publisher relying on the DMCA, like Wikimedia does, must block users that get too many DMCA strikes. You do have the right to counter the DMCA, at which point the company would have to drop it or initiate legal action against you. I don't think it's an issue here; Wikimedia just doesn't get enough DMCA strikes to ever block someone over them.--Prosfilaes (talk) 21:58, 16 December 2024 (UTC)[reply]
I agree with attempting to counter the DMCA takedown notice if you believe that it is wrong. Sometimes DMCA notices are overzealous, specious or downright fraudulent. During my time at Wikimedia, there were several successful counters, and I have myself countered takedown notices from YouTube successfully as well (for public domain music, for example. Bastique☎ let's talk!17:11, 17 December 2024 (UTC)[reply]
Typically, the WMF refuses to act on DMCA complaints that are clearly bogus. In this case, however, I think the WMF took the correct action. There is no evidence to point to in order to back up the free license claim, other than CFCF's statement. Thus it is up to CFCF to dispute the DMCA take-down. It would be dangerous folly for the WMF to risk its DMCA safe-harbor status by blindly defending every challenged upload. The best way to prevent this from occurring in the future is to simply link to the YouTube video from the file page, as this will cause the YouTube page to be archived in the Internet Archive by IABot. Nosferattus (talk) 18:49, 17 December 2024 (UTC)[reply]
Nosferattus - Please take a look at the discussion started by bjh21 on the copyright VP - people have managed to dig up some evidence of CC-licence.
What I think is important to note here, is that this is a DMCA complaint that is bogus, but not clearly so - because the ability to dig up evidence is limited because it was so long since the files were uploaded. I think this is an important issue, as it is not isolated to this case. The fact that the DMCA led to a takedown, and a legal threat is indicative of a broader problem - something that we as as community should strive to solve sooner rather than later. CFCF (talk) 08:03, 23 December 2024 (UTC)[reply]
I doubt it will ever happen, but I've been arguing for a while now that they should allow for fair use on here. That seems to be the only way to deal with these types of things to any meaningful degree outside of just taking down the file regardless of if the license was changed at some point. Expecting people to create mirrors of licenses on archive.org as fall back doesn't seem like a sustainable solution though. --Adamant1 (talk) 20:27, 23 December 2024 (UTC)[reply]
How would that help here? Fair use is pretty limited and wouldn't apply to videos like these. Someone's photo of a castle or a mountain is unlikely to be fair use. The whole of 1984 (deleted last year) couldn't possibly be fair use.--Prosfilaes (talk) 01:24, 24 December 2024 (UTC)[reply]
As far as I know fair use still applies to movies and other videos. With something like this it would be more about giving the WMF a reason to not take down the video so they can see if the people who gave them the DMCA notice to will actually challenge it. 99% of the time no one wants to spend the time or energy actually taking someone to court. Especially with something that was already PD to begin with. There's no reason the WMF couldn't use fair use as a reason to keep the video up in the short term and then say the license was changed "but fair use anyway Your Honor!" if it ever actually goes to court though. But I certainly wouldn't advocate for something like that with a full movie being hosted on here. Fair use obviously isn't a valid reason to host the video in that case. --Adamant1 (talk) 01:31, 24 December 2024 (UTC)[reply]
No, "fair use" wouldn't work here. English-language Wikipedia is quite strict in its application of fair use, as you certainly know: There needs to be a specific rationale for each file kept, it needs to be in use; and for images, the size is limited to what is necessary for display in the article (so, no full resolution). That's because for "fair use", there needs to be a particular "use" indeed for it to apply, it's not applicable for just hosting files that might be educationally useful, whether actually in use or not (which is what Comnmos does). Gestumblindi (talk) 12:06, 24 December 2024 (UTC)[reply]
I tried to improve some of the descriptions of the diagrams and missing categorization.
Reviewing it a bit closer, I'm not entirely convinced by the title of the work being used and the person mentioned as author. The PDF includes a one page memorandum and a revised version of a circular. All diagrams are from the circular.
How can we prevent the same from happening again in the future? For this to determine, we would need to know in more detail how the automated process worked. In the meantime, I'd suggest doing it manually and reviewing each file. ∞∞ Enhancing999 (talk) 08:35, 19 December 2024 (UTC)[reply]
The one suggested by the symptoms mentioned above, possibly leading to the upload of 260 files including nine for "a)".
If you did it manually, I think you should more careful review what you are uploading.
Also, please include the subjects of memoranda in descriptions, just mentioning its date, sender, number and addressees isn't really helpful. More importantly, don't delete the information when complete by others. ∞∞ Enhancing999 (talk) 10:25, 19 December 2024 (UTC)[reply]
The suggestion about the missing subject is not really off topic. The diagrams were extracted from there and, as mentioned above, none of the descriptions mentioned it. In the meantime, I added it to several of them. It's is or was also missing in the pdf, Wikidata and the category.
The don't delete is about your deletion at [3]. If the subject of the memorandum is cryptic to you, it really makes me worry about your editing in general. Your replacement description is somewhat pointless as it repeats what's in the infobox and parent categories.
No need to worry about my editing thanks, nor will I reasonably remember how many files I've uploaded among the thousands and thousands I've added in the past two decades. Bizarrely formatted descriptions that are sentence fragments about parts of the contents of a document are not as helpful as the description I left. If you were giving a description of what the United States Constitution was, would "American founding legal document" be better or "Includes Bill of Rights"? Clearly, the former is better. Including obscure and unhelpful descriptions does not help anyone. Finally, I never wrote or implied that the subject of the memorandum is cryptic. Please re-read this exchange and please post more carefully. —Justin (koavf)❤T☮C☺M☯11:03, 19 December 2024 (UTC)[reply]
Maybe you missed that the 1-page memorandum was about transmitting the 78 pages of the appendix. So "Bill of Rights, including Constitution" would have been the better comparison.
You seem to be willfully missing the point about descriptions. I didn't make the original document, so I can't tell you why so many images were included nor why the choice to make certain strings of characters into images was made at all. —Justin (koavf)❤T☮C☺M☯11:11, 19 December 2024 (UTC)[reply]
It was four years ago and I've made hundreds of thousands of edits to WMF wikis since then. What is it you are trying to get at with continuing this thread? What do you want from me? —Justin (koavf)❤T☮C☺M☯11:29, 19 December 2024 (UTC)[reply]
Hopefully, you rarely upload hundreds of files instead of 17.
@Bakhos2010: Without having a specific death date for an author, if it is from 1915 then it's got about another decade until we could use {{PD-old-assumed}} 120 years after creation. Being 120 years old is not a guarantee that it is public domain—for example, it could be discovered that it was by someone young who lived into the 1970s or even longer—but that is our policy on how old something has to be before we assume PD rather than assuming copyrighted.
It is out of copyright in the U.S. (95 years from publication, rather than based on death of author). I don't know much about Kazakh copyright law other than in it being "pretty normal Berned-Convention copyright. If the piece is truly anonymous, some countries would have it enter the public domain 70 years from creation, but for that "we don't know whose work it is" isn't enough, we'd really have to show that it was published anonymously, and I don't even know whether Kazakhstan allows that. You might find someone who knows more at COM:Village pump/Copyright. - Jmabel ! talk06:52, 20 December 2024 (UTC)[reply]
Summary of November Commons Community Call published + next steps
In summary, the main question explored in these calls was whether the Wikimedia Foundation should continue to develop structured data, or make the category system multilingual and easier to use. In both calls, there was broad agreement that we shouldn’t maintain two essentially separate systems. There was a preference by the people who intervened for developing structured data, tempered by an awareness that the community has been using categories for so long that any switch would require a lot of technical and social commitment.
To read the full summary of the call, please refer to the appropriate subpage.
Next steps and upcoming conversations
We are working on the summary of our December conversation about new media and new contributors, and we hope to publish that in early January. You will be alerted in time when this happens.
Also, there will be another two rounds of conversations on January 15 and on February 5, at 08:00 UTC and 16:00 UTC. You can subscribe to the events by going to the Commons Community Calls project page and clicking on the relevant link (you'll be redirected to the event on Meta).
Abzeronow originally proposed one solution for Commons, but Rudolph Buch has suggested several alternatives, and I have a different idea of my own, and I want to make sure we have at least a strong consensus before moving files. Proposals C, D, and E all come from Rudolph Buch; I've done my best not to alter any of his meaning but you can check [5] to verify that. Keep in mind that due to templating, there are many templates on various wikis that will necessarily use File:Flag of Syria.svg.
C) Do nothing and to trust the wiki editors in updating their pages.
D) Rename File:Flag of Syria.svg to File:Flag of Syria (1980).svg without leaving a redirect. This will lead to a huge amount of broken image links (which is bad) but prompt the editors to check what flag is right for the respective page (which is good).
E) let a bot replace File:Flag of Syria.svg by File:Flag of the Syrian revolution.svg at all wiki pages. [If I understand correctly, this means to bot-edit all of the sister projects, rather than change anything at Commons. @Rudolph Buch, please let me know if that is not what you meant.]
I believe the following remark from Rudolph Buch sums up his objection to proposal A (and presumably to proposal B): "Would that automatically feed the new flag into ~500 Wikipedia pages regardless of context and caption? Like when File:Flag of Syria.svg is now used to illustrate that this is the flag that was adopted in 1980 and after the move it shows the 2024 flag without hint in the page history or any other warning to the Wikipedia editors?" User:The Squirrel Conspiracy replied to that (in part), "Correct. However the projects have backed themselves into a weird corner because there's also templates that - instead of asking for an image - automatically pull the file with the name "Flag of [country name].svg" - and those would have the wrong image if we don't move it."
Further thought: in both proposal A and proposal B, if we allow "Move and replace" to take place when we move File:Flag of Syria.svg to Flag of the United Arab Republic (1958–1971), Flag of Syria (1980–2024).svg, that will change all explicit uses of File:Flag of Syria.svg in sister projects to use the new name, which will show up in the relevant page histories, watchlists, etc. It will not affect those pages where a template generates "[[:File:Flag of Syria.svg]]. In contrast, proposal E is likely to change exactly the uses that specifically meant this particular flag, while not solving the issue for template uses, and proposal D will break all the template usages. So 'my own "ranked choices" would be B, A, C, while definitely opposing D or E. - Jmabel ! talk01:28, 19 December 2024 (UTC)[reply]
I'd be fine with B if we know it won't break any templates so for me I'd favor A or B* (*providing it doesn't break templates) over C. I also would oppose D because it breaks pages and would be out of harmony with variants (which is why I proposed the name I did, it stays in harmony with variants). I also would oppose E since it could break templates. Abzeronow (talk) 17:15, 19 December 2024 (UTC)[reply]
I am going with "B". Abzeronow, the original proposer says he is fine with it. I think it works best. No one else seems to be saying Rudolph Buch's approaches are better. - Jmabel ! talk18:23, 22 December 2024 (UTC)[reply]
Made the moves, but the "replaces" apparently did not work as I wished. It looks like there were a lot of uses of things like {{Flag entry|Width=200|Image=Flag of Syria.svg|Caption=Syria}} even for things that were about a specific year. Not a great choice. I think there is a ton of hand work to do, no matter what.
I'll do my best to take this on, starting with Commons itself, then en-wiki. If someone wants to help on some other wiki, please say so here and have at it. - Jmabel ! talk18:39, 22 December 2024 (UTC)[reply]
@Abzeronow: are you sure about that Commons Delinker command you just gave? I'm going through these by hand, and seeing some I don't think should be changed, although admittedly the bulk of them should. - Jmabel ! talk20:03, 22 December 2024 (UTC)[reply]
I think it probably should be removed. I'm finding it runs about 60% should change, 20% certainly shouldn't, and an awful lot of tricky judgment calls where I am trying to leave messages for more appropriate people to decide. - Jmabel ! talk20:19, 22 December 2024 (UTC)[reply]
@Abzeronow: I'm finding more and more that should not change. Yes, we should definitely remove the command. In fact, since you said you are OK with that, I'll do it. - Jmabel ! talk20:23, 22 December 2024 (UTC)[reply]
Now that I have a larger sample: at the early time of my remark above, I just happened to hit a bunch that should change. I've looked at maybe 1500 pages now, and less than a hundred specifically wanted the Assad-era flag. So (1) this is overwhelmingly correct as it is and (2) there is still going to be a lot of hand-correction in a lot of wikis, way more than I personally can do. - Jmabel ! talk22:37, 22 December 2024 (UTC)[reply]
I'm not opposed to the proposal itself (in fact I do support it!), but the point is we should first clean up old usage of the flag, and then change the redirects, since this is a national flag widely used on all wikis. The issue was brought to me by members of the local community finding lots of articles showing historically erroneous Syria flags, which could not be instantly changed at once, and need time or outside assistance (e.g. global replace tool) for doing so. —— Eric Liu（Talk）05:07, 24 December 2024 (UTC)[reply]
@Ericliu1912: Based on my experience so far in cleaning up several hundred of these in en-wiki, it is going to be very hard to identify what needs to be cleaned up if we don't make the change first. How do you propose to identify the places that are affected? - Jmabel ! talk05:11, 24 December 2024 (UTC)[reply]
And is it possible to have a one-time global replace done, replacing all non-Country data usage of "File:Flag_of_Syria.svg" with "File:Flag_of_Syria_(1980–2024).svg"? I guess that would ultimately do the job. —— Eric Liu（Talk）05:18, 24 December 2024 (UTC)[reply]
@Ericliu1912: No, that would definitely not do the job. It's a long story, some of which is above. I want to give you this quick reply now, because explaining in detail will take 15+ minutes. I'll write out the more complicated picture and then post that. - Jmabel ! talk05:27, 24 December 2024 (UTC)[reply]
I guess it is appropriate that we leave notes to the communities using Country data Syria templates on their Village Pump respectly. —— Eric Liu（Talk）05:39, 24 December 2024 (UTC)[reply]
But I wonder why it'd not work? As all direct (non-Country data) global usage of "File:Flag_of_Syria.svg" currently were indeed just "File:Flag_of_Syria_(1980–2024).svg", the replacement should mostly be smooth and sufficient. Even is it not enough in some cases like certain template wrap usage, we could still go ahead and replace most of the current links first, that should also decrease the burden for the remaining manual changes. —— Eric Liu（Talk）05:41, 24 December 2024 (UTC)[reply]
@Ericliu1912: Why a global search-and-replace is a bad idea here (and also almost impossible to do in an effective manner):
Many—I strongly suspect most—of the places where the Syrian flag is used should switch to the new flag. The following is a representative set of examples, though certainly not exhaustive.
All geographical articles should be using the current flag, not the flag of a prior regime.
There are presumably a lot of templates in Category:Templates related to Syria that use a flag. Those should all use the current flag, not the flag of a prior regime.
Any infoboxes related to geography that contain a flag should all use the current flag, not the flag of a prior regime.
As far as I'm aware, the new government of Syria, presuming it is widely recognized, which appears to be happening, will inherit (or already has inherited) all of Syria's positions in international organizations, e.g. the UN and its various affliates, the organization of non-aligned states, status as a signatory of various treaties unless the new government were to renounce those. All of those should update to the current flag.
If you think about how flag files are used, search-and-replace is very tricky. They are almost never used in a syntax that writes out File:Flag of Syria.svg in the wikitext. For example, there are templates that effectively do File:Flag of [COUNTRY].svg, or that get at these other ways. If that's not clear, I'll elaborate; I'm trying to get you a response quickly, so I'm not approaching this at essay length.
Also: when the template is updated, if there is anything that should permanently use the current revolutionary flag regardless of future regime changes, there should be a way to specify that, too. Let's not get caught in the same thing again! (en-wiki has already done this.) - Jmabel ! talk05:50, 24 December 2024 (UTC)[reply]
I understand the difficulties, so I suggest that we at least (1) replace direct file links and update about ~110 Country data Syria templates (which is the most obvious and widely-used template), adding a "1980" alias for them (and maybe an "opposition/revolution" alias too, just in cases which do "permanently use the current revolutionary flag"); (2) list the rest of the templates that indeed embed File:Flag of Syria.svg (in a historical context) and try to do the replacement; (3) regretfully ignore the rest like File:Flag of [COUNTRY].svg you mentioned above and change the redirect to the opposition flag, at the same time also notifying the communites, reminding them to finish rest of the work. —— Eric Liu（Talk）06:05, 24 December 2024 (UTC)[reply]
@Ericliu1912: In my experience (mainly Commons and en-wiki) there is very little correlation between how the file was used (in a technical sense) and why it was used (to refer to a regime or a country). I think each wiki is going to have to work out for itself what is right for how usage is shaped on that wiki. No matter how we do this, there is going to be a LOT of hand-work, because neither case ("it meant the country" or "it meant the regime") is clearly dominant. This isn't going to be an 80-20 case, it's more in the range of 60-40. My personal guess is that more cases mean country than regime, but (on en-wiki, at least, which I'm guessing is typical of the Wikipedias in this respect) it's not dramatically more.
The more a given Wikipedia covers events relative to how much it covers geography, the more often it will mean a regime. But right now it is totally jumbled together.
This suggests a large area in which we have not at all future-proofed (for the hundreds of other countries in the world). Wikipedia wasn't around in 1989-1992, or we would have recognized this as a potentially major issue up front when we designed flag-related templates. - Jmabel ! talk06:16, 24 December 2024 (UTC)[reply]
I hope that you are in good spirits. My name is Chinmayee Mishra and I am a part of the Let’s Connect working group - a team of movement contributors/organizers and liaisons for 7 regions : MENA | South Asia | East, South East Asia, Pacific | Sub-Saharan Africa | Central & Eastern Europe | Northern & Western | Latina America.
Why are we outreaching to you?
Wikimedia has 18 projects, and 17 that are solely run by the community, other than the Wikimedia Foundation. We want to hear from sister projects that some of us in the movement are not too familiar with and would like to know more about. We always want to hear from Wikipedia, but we also want to meet and hear from the community members in other sister projects too. We would like to hear your story and learn about the work you and your community do. You can review our past learning clinics here.
We want to invite community members who are:
Part of an organized group, official or not
A formally recognized affiliate or not
An individual who will bring their knowledge back to their community
An individual who wants to train others in their community on the learnings they received from the learning clinics.
To participate as a sharer and become a member of the Let’s Connect community you can sign up through this registration form.
Once you have registered, if you are interested, you can get to know the team via google meets or zoom to brainstorm an idea for a potential learning clinic about Wikimedia Commons or just say hello and meet the team. Please email us at letsconnectteam@wikimedia.org .
We look forward to hearing from you :)
Are there any templates, or any other standard way to notify users about possible pseudoscientific images etc.? I am talking about cases where deletion is not an option (for example, it may be legitimately used in a wikipedia article describing some pseudoscientific view, or when there are different views about it on local project). --Hwem (talk) 16:11, 19 December 2024 (UTC)[reply]
This sounds like a good idea, but it could be used to edit war over controversial concepts are disputed within the scientific discourse. Social sciences are also a form of science and more often than not different countries have different official narratives which they deem as "scientific history" that would quickly turn this template into a nightmare. Don't even get me started on medicine, for years now the Chinese Communist Party (CCP) has been arguing that "Western medicine" (their word for scientific medicine) isn't definitive and that "Chinese medicine" (Traditional Chinese Medicine (TCM)) is "just as valid" (this is despite of decades of the same Communist Party trying to eradicate TCM and promoting some forms of scientific medicine within the People's Republic, likewise, the Republic of China (Taiwan) also grants special privileges to TCM). A lot of pseudoscience gets "legitimised" when going over national borders.
I like the idea of warning ⚠️ people of pseudoscience, as a free educational document that explains a concept in pseudo-medicine can have real life negative consequences and even kill people, the definition itself of something that is or isn't considered to be a "pseudoscience" is more difficult to state in an authoritative voice. Of course, nobody is going to dispute that the Miasma Hypothesis is still considered to be scientifically valid, but I've seen Dutch medical journals pussyfoot around outright calling commonly used forms of pseudo-medicine as pseudoscientific often using terms like "currently unproven" and "alternative medicine" rather than calling it fake medicine (pseudoscience).
"From" is probably best. "By" is OK, but could be misunderstood as referring to who is doing the quoting rather than who is quoted. "Of" is pretty much wrong for a category name, though I can see where it could come up in a sentence.
"Here are a set of quotations from Rousseau." - good
"Here are a set of quotations by Rousseau." - confusing, could mean Rousseau is quoting something, rather than being quoted
"Here are a set of quotations of Rousseau." - clear, but a little odd.
"Here are a set of quotations of Rousseau by Will Durant." - Will Durant is quoting Rousseau, and probably the clearest succinct way to say that.
"Here are a set of quotations from Rousseau by Will Durant." - Will Durant is quoting Rousseau. Probably clear, but a little awkward.
Printed images being put in categories for photographs
(@Jmabel: since I thought he started a similar discussion although I can't find it). Recently I've been going through and trying to find improperly categorized images of postcards. It seems as though a lot of people have been putting images of printed postcards in categorize for photographs. Simply because they depict a specific place, person, or object. This seems wrong because a printed postcard clearly isn't "an image created by light falling on a light-sensitive surface." At least not in any way that matters. After doing some research, it seems to be a much larger problem outside of postcards. For instance, this image is in Category:Photographs of Rome in the Rijksmuseum Amsterdam even though the image in the book clearly isn't a photograph or based one. So is there any consensus about what does or doesn't belong in a category for photographs? I'm sure there's a line there, but it seems weird to put every image of a place, person, or object in a category for photographs simply because the postcard or whatever it is might have been based on one. --Adamant1 (talk) 15:01, 20 December 2024 (UTC)[reply]
@Adamant1: I would definitely take the view that your example is not a "photograph of Rome". It's a photograph of a drawing of Rome, but not directly a photograph of Rome. That category also contains a number of photographs of paintings that are (presumably) in Rome. I'd say they shouldn't be there either (a photo of an object in Rome not necessarily being a photo of Rome itself), but that's at least somewhat arguable. --bjh21 (talk) 15:55, 20 December 2024 (UTC)[reply]
I'd agree that if it doesn't even derive from a photograph it absolutely should not be in a "photograph" category.
If they do derive from a photo… I don't really know how to best handle that in category terms (and I guess I wouldn't worry a lot about it). In looking into this, I came across an interesting example.
Clearly both retouching and colorizing the same photo; clearly at least one of the two must be unfaithful to the photo in ways we would never accept from our own contributors of their own work. - Jmabel ! talk18:03, 20 December 2024 (UTC)[reply]
That's unfortunately very common with early postcards - very often, the colorists were working off a B&W image without ever seeing the scene in person. They would also often alter the sky, as seen in your example. Pi.1415926535 (talk) 18:35, 20 December 2024 (UTC)[reply]
Colour difference
Hello
I recently created a SVG file of a PNG logo with Inkscape, I used the colour picker tool and everything seemed fine. I then save as optimised SVG and uploaded the SVG file to Commons, but the red square seems darker than on the original PNG (the blue and yellow also seem a bit different).
I then uploaded this original PNG, but its colours are the same as the SVG.
I then took a screen capture of the original, compared it to the Commons files, and the colours are definitely not the same (3rd image in the gallery below).
SVG
PNG
Compared screenshots
Is it a known issue, or am I missing something ?
Regards, Şÿℵדαχ₮ɘɼɾ๏ʁ 19:43, 20 December 2024 (UTC)
Hello ReneeWrites. I've done more tests and the problem seems to be linked to the web browser used (see the new version of the screenshots compared above, CTRL+F5 to clear the page cache).
So Firefox seems to display different colours on linguistlist.org and Commons whereas Chrome and Edge don't, but all three browsers seem to display slightly different colours to the PNG/SVG files.
I'm a bit confused and can't figure out where the problem really lies...
Regards, Şÿℵדαχ₮ɘɼɾ๏ʁ 15:56, 21 December 2024 (UTC)
I don't think it's a browser issue necessarily. I notice that the color picker is sometimes slightly off when I'm working on a file in Illustrator, at which point I eyeball the process. ReneeWrites (talk) 23:54, 21 December 2024 (UTC)[reply]
As someone who is not an expert, that doesn't track to me: the SVG defines these colors with hex codes, which are at a re-encoding of RGB values. My eye doesn't see a difference between the SVG and PNG logos and an online color picker calls the PNG's red #800000, the PNG thumbnail of the SVG #810000 (!!!), and the code of the actual SVG itself defines it as #800000. The codes fro the blue and yellow squares have the same values in all three files.@SyntaxTerror: , this warrants a ticket at phab:. Are you willing to make one? If not, I will. —Justin (koavf)❤T☮C☺M☯19:51, 22 December 2024 (UTC)[reply]
Hello Koavf and thank you for your reply. I was also thinking of making a ticket on Phabricator, but I still don't know if this problem is related to MediaWiki or browsers (or maybe my computer?).
If you'd like to make a ticket, it's no problem with me as I'm not a native English speaker, and technical terms can sometimes be difficult for me. Mention me anyway so I can answer any questions. Regards, Şÿℵדαχ₮ɘɼɾ๏ʁ 00:43, 23 December 2024 (UTC)
SVG uses the sRGB colorspace. librsvg turns SVG into a PNG bitmap. By default, PNG does not have a default sRGB colorspace, but its colorspace can be set to sRBG. Last time I checked, librsvg or the libraries it uses do not set the PNG colorspace to sRGB. Consequently, we should not expect the colors to match. There are also questions about screen grabs and color pickers. Are the colors before or after the system color correction? Glrx (talk) 01:17, 23 December 2024 (UTC)[reply]
Photography as a verb
Normaly photografers respect each other and move away so you can take a picture. The probably professional photografer did not care and was buzy with his two subjects. The lighting created a surprising effect with shadows. Is there a category for a working photografer. I wanted to use photoshoot, but this category does not seem to exist.Smiley.toerist (talk) 23:08, 20 December 2024 (UTC)[reply]
At Commons:Questionable Flickr images, there is a list of blacklisted users. Sometimes I think, "Hmmm. This image from that account looks good. I wonder why it is blacklisted?"
Looking at the list, the reason in most cases is "Flickrwashing." Personally, I do not think that it is very helpful. So, I wonder if we could agree that for new requests, it would be good if a link to a discussion is included?
Next is when do we blacklist? What if a user has 99 good images and 1 bad? Or 50 good and 50 bad? Or 1 good and 99 bad? I think in some cases the reason to blacklist has been derivative works and the lack of Freedom of panorama. Personally, I do not think we can set up a general rule, but if a Flickr user is in bad faith and tries to push files to Commons, we should always blacklist. If most of the uploads are bad, we should also blacklist. But if it is only a smaller part of the photos or if the reason is derivative works, we should not blacklist.
You could say, "If you think it's a problem, then just watch the damn page and fix the requests." You would be right. But I also think that it would be good if we had some guidelines about what to add as a blacklist reason and when to add or remove. For example, I have sometimes thought about going rogue and just removing all requests without a link to a discussion. But the result would probably be a lot of bad images and angry users.
So, even if we can't set up a rule saying that if the bad ratio is > 40%, then blacklist, perhaps we can set up a few "rules of thumb"?
"Flickrwashing" is a fairly specific term. It doesn't mean "uploading images with copyright issues". It specifically means "uploading and taking credit for other people's copyrighted photos". Taking a photo of a painting or a non-FOP architectural work isn't Flickrwashing; copying a photo from a stock photo web site or a newspaper and reposting it is.
QFI listings should generally be reserved for instances of true Flickrwashing, not other rights issues. If we blocklisted accounts just because they took a few photos of paintings or toys or whatever, we'd be here all day (and we'd probably be rejecting a lot of perfectly good photos in the process). Commons users importing content from Flickr need to evaluate it for DW/FOP concerns themselves; QFI should be reserved for sneaky copyright violations which aren't apparent from the image itself or its subject matter. Omphalographer (talk) 22:17, 21 December 2024 (UTC)[reply]
I agree 100% with that. So far the accounts I have added to QFI is because they were specifically created for license washing. Yann (talk) 22:56, 21 December 2024 (UTC)[reply]
Some good points above. Two points: 1)Keep in mind that Flickr is in some ways social media, and allows derivative works of copyrighted material (eg photos of posters or book covers, etc) which Commons cannot allow. So Wikimedians copying images from Flickr should keep this in mind as something to consider before uploading a given image to Commons, knowing that while some may be tagged as free licensed on Flickr they are still not properly free for Commons. 2)"Blacklists" are not simply for Flickr accounts that don't live up to Commons standards, nor for accounts that are generally good but the users are sometimes careless with with license claims. It is for accounts that generally or deliberately make false license claims, usually with intent to deceive as to actual license status. -- Infrogmation of New Orleans (talk) 23:06, 21 December 2024 (UTC)[reply]
Hello. I have a running debate with User:Eurohunter on my talkpage about whether we should use the term "of" or "in" with the example of police automobiles. This category should contain files of police automobiles in Germany. So it can be a file of a german police automobile in Berlin, but also a file of a Polish automobile somewhere in Germany. However, this category should contain files of automobiles of the german police no matter whether this was made in Germany or an other country. At least thats my opinion. Is there an other interpretation and should we use both terms simultaneously? Regards Lukas Beck (talk) 11:45, 22 December 2024 (UTC)[reply]
For me in this case it is very clear. All police cars located in Berlin at the moment the photo was taken are "Police cars in Berlin". These cars can also be "Police cars of the Berlin state police" or of other state police of federal police. Of is for the ownership and In is for the location. When categorizing for the manufacturer I would always explicitly state that the of refers the the manufacturer "Policy cars of German manufacturer". GPSLeo (talk) 13:27, 22 December 2024 (UTC)[reply]
+1. I was just in the process of writing a message to the similar effect. There's only 1587 categories with "of" to begin with and most, if not all, of them have to do with the ownership or manufacturer. --Adamant1 (talk) 13:39, 22 December 2024 (UTC)[reply]
Categories combining city and photographer
Hello fellows, we have three categories which (together with their subcategories) sort photos by a combination of the criteria by city and by photographer:
The former category is for categories for photos of specific cities by specific photographers. The latter is for photos of unspecific cities by specific photographers. Ruslik (talk) 19:56, 22 December 2024 (UTC)[reply]
What's the use case for the latter? Surely all of those files would already be categorized as photographs of cities and photographs by that photographer; the added value of a category specifically for photographs of cities which aren't identifiable as any specific city seems minimal. Omphalographer (talk) 23:23, 22 December 2024 (UTC)[reply]
@Aristeas: If it were me I'd up merge whatever subcategories overlap between the two into whichever one makes more sense and then go from there. It be that a lot of the parent categories end up being empty and/or otherwise pointless once the overlap is fixed though. --Adamant1 (talk) 20:25, 22 December 2024 (UTC)[reply]
I doubt there is a useful distinction. Possibly there are techniques that are not "processes" (e.g. the use of a polarizing filter), but every process is a technique. - Jmabel ! talk19:07, 23 December 2024 (UTC)[reply]
I think the intended distinction is that a photographic technique is something that the photographer does when taking a photograph, whereas a photographic process involves film chemistry and the processes used to turn it into a print in a darkroom. This isn't really borne out by the current contents of the categories, though. (And, of course, this distinction completely breaks down when dealing with digital photography.) Omphalographer (talk) 22:35, 23 December 2024 (UTC)[reply]
Maybe it's of some help to look at how other projects deal with this distinction. This is what the category descriptions on enwiki have to say:
Category:Photographic processes - "This category groups together articles that describe procedures by which light-sensitive materials are made to produce an image. The category should not be confused with Category:Photographic techniques, which comprises articles describing procedures related to how a photograph is taken or composed, or is manipulated during or after processing."
Category:Photographic techniques - "This category contains categories and articles relating to the theory and methodology of composing and/or taking photographs, or to their manipulation during or after processing. It should not be confused with Category:Photographic processes, which comprises articles relating to the production of images using light-sensitive materials."
The descriptions are very similar to the ones Omphalographer stated. "Photographic technique" is defined slightly more broadly so it's inclusive of digital photography, while "Photographic process" describes an explicitly analog process. I like these descriptions, they seem clear to me, leave very little room for ambiguity, and solve the issue being discussed above, with maybe one exception: The page for "Post-processing" is a disambiguation page, of which the one relevant to photography is titled "Image editing". This page is categorized as a technique, not a process. ReneeWrites (talk) 14:25, 24 December 2024 (UTC)[reply]
Genre seems to concern itself with subject matter - what a photograph depicts, so that's not really an issue. Style seems to be a bit more problematic. ReneeWrites (talk) 14:58, 24 December 2024 (UTC)[reply]
I got an e-mail ping waking me up, apparently it was this edit, which I found was odd. The page blanking was done on 24 December 2024, but the admin tagged it on 18 November 2022 meaning that more than 2 (two) years went by without anyone finding it problematic that a user page existed unblanked.
I have long noticed that some users treat the unblanked user pages of blocked users as "some sort of backlog" even when the blocking admin didn't decide to perform this form of damnatio memoriæ. What I find interesting is the edit summary "I really can't see any policy-based reason to blank this,m and at the moment the content of this page is relevant to an arbcom case. This is literally an admission by an Enwiki ArbCom member that the blanking of user pages isn't rooted in any policy. In fact, the blanking of a user page would be considered a form of vandalism in any other context.
Historic England have put a number of PD aerial images (example - one of many taken by the RAF prior to 30 June 1957, so now expired Crown Copyright) behind software that makes it difficult, but not impossible to download the high-res versions. They do not assert copyright, but charge for making copies available to download, under what purport to be restrictive terms.
Various methods for downloading high res copies have been described online (see sources at w:SmartFrame).
A claim has reached me that they will launch a new system in the new year, to prevent such downloads.
Comment I am very open to discussion. Making AI takes me a lot of time. It is difficult and the results are questionable. We must explore this path, it is a tool, we must learn.
Question pinging @Yann:any opinion?? (google translator). AbchyZa22 (talk) 21:02, 16 December 2024 (UTC)
Chinmayee Mishra 05:44, 19 December 2024 (UTC)
.jpg)
.jpg)
SVG
PNG
Compared screenshots
_Fra%27Mauro%27s_World_Map_-_Biblioteca_Nazionale_Marciana.jpg){kind=link}
_Self-portrait_by_Suzanne_Valadon_(1894).jpg){kind=link}
_Fra%27Mauro%27s_World_Map_-_Biblioteca_Nazionale_Marciana.jpg){kind=link}
_Apr%C3%A8s_le_bain_(1908)_-_Suzanne_Valadon_-_Petit_Palais,_Gen%C3%A8ve.jpg){kind=link}
.jpg){kind=link}
![[2]](https://commons.wikimedia.org/w/index.php?title=Special:Log&page=File%3AM-21-19+Memorandum+for+Heads+of+Executive+Departments+and+Agencies.pdf-image-004.png){kind=link}
,_Flag_of_Syria_(1980%E2%80%932024).svg){kind=link}
,_Flag_of_Syria_(1980%E2%80%932024).svg&action=edit&redlink=1){kind=link}
,_RP-F-2001-7-164-12.jpg){kind=link}
