Commons:Village pump/Archive/2024/07

Urgent: As Media of the Day predestined video with all white preview image

Hello, this is kind of urgent: The video File:Drone video of Keila waterfall and manor in Keila-Joa, Estonia.webm which is selected as Media of the Day for 8th of July, so in just 3 days (without the recent one I am writing this), does only display an all white preview image. Can do someone a fast repair? — Speravir – 23:07, 3 July 2024 (UTC)

Thanks for bringing this up, but there is a simple solution: we can choose a certain time as the frame that is the thumbnail. Input thumbtime= to make a different thumbnail. E.g. [[File:Drone video of Keila waterfall and manor in Keila-Joa, Estonia.webm|thumb|thumbtime=10]] looks like:

So we should be okay as long as someone chooses a good time for a thumbnail frame. —Justin (koavf)❤T☮C☺M☯ 23:34, 3 July 2024 (UTC)

Justin, I know, but this does not work for Media of the Day. I have forgotten to link to it above, but added this just now, or take a look from here: {{Motd/2024-07-08}} – yes, it is a template. — Speravir – 23:44, 3 July 2024 (UTC)

What do you think of this: https://commons.wikimedia.org/w/index.php?title=Template%3AMotd%2F2024-07-08&diff=891850543&oldid=890322763 —Justin (koavf)❤T☮C☺M☯ 23:49, 3 July 2024 (UTC)

Oh, and ! I think, this should be documented somewhere, {{Motd}} or {{Motd filename}} or both. By the way: Strange enough, even thumbtime=0 seems to work. — Speravir – 00:07, 4 July 2024 (UTC)

So, I consider this fast resolved. — Speravir – 00:07, 4 July 2024 (UTC)

@Koavf: Unfortunately your change broke the template, it seems. Wouldn’t Template:Motd/2024-07-08_thumbtime be the right way to handle this? --Geohakkeri (talk) 07:32, 4 July 2024 (UTC)

It's only broken on the language-agnostic version, not the translations. Template:Motd/2024-07-08 thumbtime does not display the actual media. Whatever solution others think works is fine by me. —Justin (koavf)❤T☮C☺M☯ 07:35, 4 July 2024 (UTC)

At least now it displays a reasonable title, even if it has an extraneous bit. —Justin (koavf)❤T☮C☺M☯ 07:37, 4 July 2024 (UTC)

And now it just looks correct on the template. Great work. —Justin (koavf)❤T☮C☺M☯ 08:07, 4 July 2024 (UTC)

Thanks. Templates can be quite tricky sometimes. --Geohakkeri (talk) 08:21, 4 July 2024 (UTC)

Thank you, Geohakkeri for the custom fix. It should in general, though, be available as template parameter. I do not have just gone into the deep details of it yet. — Speravir – 00:03, 5 July 2024 (UTC)

@Speravir: No, it was Justin (koavf) who made the custom fix. I fixed it in general. --Geohakkeri (talk) 06:38, 5 July 2024 (UTC)

@Geohakkeri: I referred to your Special:Diff/891951137 and did not like this approach – I thought of a parameter {{{thumbtime}}} instead. But I’ve also said I hadn’t looked into details. I’ve meanwhile noticed that my idea could not work, and you’ve essentially just adapted some code from {{Motd}}. — Speravir – 23:49, 5 July 2024 (UTC)

This section was archived on a request by: --Speravir 23:49, 5 July 2024 (UTC)

Disease-related deaths in Beijing

Everything, I sampled in Category:Disease-related deaths in Beijing, is complete bollocks.

Executions, airline crashes, and old age are not Disease-related. Looks like vandalism to me. What to do? Broichmore (talk) 18:14, 1 July 2024 (UTC)

How do I request the speedy closure of a deletion nomination?

Hi. I have commented on the deletion nomination for an image of an Indonesian Government official. The original reason for the nomination was (in Spanish) "Who is it?". As far as I can figure out, simply not knowing who a person is, isn't a reason for a deletion nomination. So, I believe the nomination to be invalid, but can't find out how to get it reviewed and potentially shut down. In my opinion, it was possibly a bad faith nomination, because as soon as I categorised it (as it hadn't got any categories when uploaded), the nominator responded to say that he was "pleased to have drawn attention to the lack of categories" and would "thank me again when I added the correct ones". I'm not sure what to do with it. DaneGeld (talk) 12:38, 1 July 2024 (UTC)

@DaneGeld: Why haven't you linked it in the discussion here? If everyone is now in agreement that it should be kept, including the original nominator, then any experienced user (not just an admin) can close it as a speedy keep by clear consensus. - Jmabel ! talk 19:25, 1 July 2024 (UTC)

And ideally shouldn’t be closed by an involved user with the DR (other than the nominator). Bidgee (talk) 19:38, 1 July 2024 (UTC)

@Jmabel - I didn't link the discussion because I have fallen foul of being misconstrued as fishing for comments on discussions in the past, which is apparently frowned upon. I simply avoided it as I was asking for advice, rather than seeking others to comment on it. DaneGeld (talk) 21:48, 1 July 2024 (UTC)

@DaneGeld: in the future: state what you are wanting as neutrally as possible, and link the mention on VP or similar page from the discussion in question, so that you are clearly not doing anything surreptitious. - Jmabel ! talk 21:53, 1 July 2024 (UTC)

I've closed this nomination. Omphalographer (talk) 20:18, 1 July 2024 (UTC)

Thank you. DaneGeld (talk) 21:52, 1 July 2024 (UTC)

Commons Gazette 2024-07

Currently, there are 184 sysops.

---

Edited by RZuo (talk).

---

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

--RZuo (talk) 21:23, 1 July 2024 (UTC)

Discussion on UAE copyright law before 1992

Hello. Kindly visit Commons talk:Copyright rules by territory/United Arab Emirates#‎UAE Copyright law before 1992 for the discussion. Regards, JWilz12345 (Talk|Contrib's.) 02:49, 2 July 2024 (UTC)

What's the best title for the category

Video game videos by name or Videos of video games by name--Trade (talk) 12:53, 1 July 2024 (UTC)

It seems like similar categories start with "videos", but "videos of video games" just sounds clunky. So who knows. --Adamant1 (talk) 12:58, 1 July 2024 (UTC)

I don't think you can get around repeating "video".

Also there is Commons:Categories for discussion/2024/04/Category:Video game videos. Enhancing999 (talk) 20:09, 1 July 2024 (UTC)

Another existing option is "Videos related to video games", which seems to be what the main category actually contains. "Video game videos" sounds like cutscenes and "Videos of video games" sounds like gameplay. Why "by name" and not "by game"? Sinigh (talk) 10:36, 2 July 2024 (UTC)

We already use by name for "Indie video games by name‎" and similar others so i assumed we needed to be consistent? Trade (talk) 15:01, 2 July 2024 (UTC)

Yes, but what I reckon is that "Indie video games by name‎" refers to the names of the indie games and, following the same logic, "Videos of video games by name" would then refer to the names of the videos rather than the names of the associated games. But I'm not sure, so take this with a grain of salt! Sinigh (talk) 16:24, 2 July 2024 (UTC)

Acceptableness of having a source template also provide author information

Hello, {{Web source}} is a template for providing links to the sources of files uploaded from other places on the web. It currently has two (undocumented) parameters which are set up to provide author information, |author= and |photographer=. This seems a little odd to me because usually in templates like {{Information}} author and source information is listed separately instead of on the same line. For this reason, I am wondering if the current set up with {{Web source}} is inline with general Commons best practices and if it would be worthwhile to deprecate the parameters. For reference, the |photographer= parameter is currently only used about a dozen times and I haven't been able to find any uses of |author=. Thanks for any feedback and take care. —The Editor's Apprentice (Talk) 00:49, 2 July 2024 (UTC)

That's probably why they aren't documented. - Jmabel ! talk 03:25, 2 July 2024 (UTC)

The template was apparently thought primarily for artworks and to be used within template:artwork and similar templates and situations (although it can certainly be used in other contexts also). The template:artwork has a line "source/photographer", on the same line (which is somewhat strange and impractical, but it's been like that forever and unlikely to be changed easily). Therefore, it makes sense that template:web_source is compatible and includes relevant related parameters. I suppose that, in the line source/photographer of template:artwork, users may find practical to place the photographer inside the same template:web_source rather than outside it. Also, it must be noted that template:web_source is embedded in other templates, such as template:From_Google_Art_Project. Not sure about the parameter "author", but it might have a potential use in some such other circumstances. I would hesitate to just remove stuff from templates like that. Did you think of consulting the creator of the template (who is still active on Commons)? -- Asclepias (talk) 05:40, 2 July 2024 (UTC)

You're right, the |photographer= parameter does indeed mirror the set up of {{Artwork}}, so its not a complete anomaly and it would make sense for them to be compatible since they are often used together. I indeed share your hesitation to outright remove functionality from templates which is in part why I started this discussion. You're also right to ask about my consulting Zolo as the creator of the template. Usually in similar circumstances that is something I do, but since I was thinking about general best practices/acceptability, in this case I ended up going to a public forum first. That said, I would be interested in hearing about @Zolo's intentions and thought process, both for {{Web source}} as a whole and the |author= and |photographer= parameters in particular. Thanks to you both and please take care. —The Editor's Apprentice (Talk) 20:06, 2 July 2024 (UTC)

The template:artwork displays at the front end, "source/photographer", but not within the code, presumably photographer intended there is the scanner, and not the original photographer. Scanners in that sense exist for any and all GLAM products we have. These days we enter the url there, only a few years ago that might have, had to be the GLAM or an individual by name.

Interestingly |photographer= can be added into the artwork template, so I'll use it from here. Before I was putting that name under artist.

I'm probably doing this the wrong way, but for author I sometimes put in the author of the book the photo came from, or the full name of the artist/ photographer, whereas in artist I might put in the shortened signed name from the piece in question. Then sometimes I want to use the author field to put in the full unabbreviated name.

In any event, if the photo came, from say, en:Life (magazine), I would have; Life under the author, after all, they probably commissioned it, the photographer being a staffer.

I'm sorry, I wouldn’t use this template for anything. It's far too glib. Most of the photographs I upload come from GLAM's, and I use the {{Artwork}} template for them. The GLAM collection information being important here, especially the accession number. I really don't see why we would want to encourage glib uploading of any item, ignoring attribution, and this template, by its brevity, does exactly that.

We face an oncoming rush of AI derived images of epidemic proportions. The provenance and sourcing of where images come from, and their legitimacy is something we need to beef up, and I fear the project is asleep at the wheel on this issue. – Broichmore (talk) 11:25, 2 July 2024 (UTC)

Must admit. I'm getting tired of these author templates that takes up half my screen for no good reason other than to look fancy--Trade (talk) 15:06, 2 July 2024 (UTC)

[Month] [Year] in [Place] - photos taken on, or photos of events that occurred

I have a question regarding categories [Month] [Year] in [Place], e.g. Category:June 2024 in Kraków. If there's a subcategory belonging to it, does it mean:

• Photos are be taken on [Month] [Year], or,
• Photos of the event that occurred on [Month] [Year]?

For example:

• Category:2024 Jagiellonian University pro-Palestinian campus occupation‎

Category contains photos taken on June 2024, but the occupation event started on May 2024. Apart of the Category:June 2024 in Kraków, should it belong to Category:May 2024 in Kraków then?

Dwxn (talk) 11:16, 2 July 2024 (UTC)

If a category like this belongs as a parent category of Category:2024 Jagiellonian University pro-Palestinian campus occupation‎, yes, I'd also include Category:May 2024 in Kraków. But my own feeling is that it should simply use Category:2024 events in Kraków, and the month-specific categories should be on individual photos. - Jmabel ! talk 19:10, 2 July 2024 (UTC)

Suppression d'un visuel

Bonjour,

Il y a plus de 11 ans, une personne a posté une photo de moi prise dans un cadre privé. Aujourd'hui, je souhaite vraiment qu'elle soit supprimée car elle peut me porter préjudice au travail. J'ai proposé sa suppression à plusieurs mais elle n'est jamais supprimée. Je ne sais plus quoi faire. Merci de m'aider. — Preceding unsigned comment added by Sayamelyay5 (talk • contribs) 21:34, 1 July 2024 (UTC)

@Sayamelyay5:

1. Is the picture in use on one of our sister projects, such as one of the Wikipedias?
2. Have you requested deletion as a personal courtesy or on some other basis?
3. You say the picture was the picture was taken in a private place. Was it with or without your consent?
4. Also, could you say what country, because issues about this vary widely from country to country.

Jmabel ! talk 21:56, 1 July 2024 (UTC)

Also, having looked slightly further into this: are you sure the picture in question has not already been deleted? Please check. - Jmabel ! talk 22:06, 1 July 2024 (UTC)

See Commons:Deletion requests/File:Sandrine Lefebvre-Reghay photo.png. I deleted the file in question as F10. Pi.1415926535 (talk) 22:11, 1 July 2024 (UTC)

Mille mercis. Y a t il un délai de suppression dans Google ? Bien à vous Sayamelyay5 (talk) 22:20, 1 July 2024 (UTC)

Bonsoir,

J'ai vérifié. Je la vois toujours sur Google en tapant mon nom.

Bien à vous Sayamelyay5 (talk) 22:19, 1 July 2024 (UTC)

Google takes some time to update its cache.

Anyway, the photo is no longer in Commons. If what Google shows is a problem, it should be dealt with Google. Pere prlpz (talk) 08:17, 3 July 2024 (UTC)

Template:Swiss Government Portrait

Could anyone proficient in German fix the dead link? --Trade (talk) 21:24, 2 July 2024 (UTC)

I am not proficient at all but I guess the links should be https://www.admin.ch/gov/de/start/bundesrat/geschichte-des-bundesrats/bundesraete-und-ihre-wahl/alle-bundesraete-liste.html and https://www.admin.ch/gov/de/start/bundeskanzlei/bundeskanzlerin/alle-bundeskanzler-liste.html. --Geohakkeri (talk) 05:53, 3 July 2024 (UTC)

The Community Wishlist is reopening July 15, 2024

Here’s what to expect, and how to prepare.

Hello everyone, the new Community Wishlist (formerly Community Wishlist Survey) opens on 15 July for piloting. I will jump straight into an FAQ to help with some questions you may have:

Q: How long do I have to submit wishes?

A: As part of the changes, Wishlist will remain open. There is no deadline for wish submission.

Q: What is this ‘Focus Area’ thing?

A: The Foundation will identify patterns with wishes that share a collective problem and group them into areas known as ‘Focus Areas’. The grouping of wishes will begin in August 2024.

Q: At what point do we vote? Are we even still voting?

A: Contributors are encouraged to discuss and vote on Focus Areas to highlight the areas.

Q: How will this new system move wishes forward for addressing?

A: The Foundation, affiliates, and volunteer developers can adopt Focus Areas. The Wikimedia Foundation is committed to integrating Focus Areas into our Annual Planning for 2025-26.

Focus Areas align to hypotheses (specific projects, typically taking up to one quarter) and/or Key Results (broader projects taking up to one year).

Q: How do I submit a wish? Has anything changed about submissions?

A: Yes there are some changes. Please have a look at the guide.

I hope the FAQ helped. You can read more about the launch.

You are encouraged to start drafting your wishes at your pace. Please consult the guide as you do so. Also if you have an earlier unfulfilled wish that you want to re-submit, we are happy to assist you draft.

You can start your draft (see an example) and don't hesitate to ask for support when drafting by sending me a link to your draft/sandbox via Meta email to help/review it. Alternatively you can leave the link in the Drafts List. –– STei (WMF) (talk) 14:36, 4 July 2024 (UTC)

Script for deletion sorting?

Is there a script/tool I can use to assist deletion sorting? I regularly come across DRs that should be categorized, or which were never updated with the result. There are a few such scripts on enwp, but deletion sorting works differently there. — Rhododendrites ^talk |  14:17, 2 July 2024 (UTC)

Apparently not, so now there's User:Rhododendrites/drsort.js. Still a work in progress. Please leave feedback/advice on the talk page. — Rhododendrites ^talk |  13:26, 6 July 2024 (UTC)

IA Upload issues with DJVU files

I've uploaded the below three DJVU files from the Internet Archive using the IA Upload tool and there are issues. There is a warning in the files saying "This file type may contain malicious code. By executing it, your system may be compromised." and the files' dimensions are 0 x 0. Any idea what is going on and how to fix this?

• File:Hannibal and the Crisis of the Struggle between Carthage and Rome.djvu
• File:Theodoric the Goth, the Barbarian Champion of Civilisation.djvu
• File:Oliver Cromwell and the Rule of the Puritans in England.djvu

Ciridae (talk) 06:28, 12 July 2024 (UTC)

I purged the pages and the details came right up. This is something that has been happening for a while with PDFs and DJVUs. Likely the file page is sometimes rendered before the metadata is available (race condition). It then gets cached until you edit or purge it. —TheDJ (talk • contribs) 08:33, 12 July 2024 (UTC)

@TheDJ: Thank you! What are the steps to purge a page? There is a similar issue with File:The National Intimate Partner and Sexual Violence Survey, 2010-2012 State Report.pdf. Ciridae (talk) 17:14, 12 July 2024 (UTC)

Help:Purge is the solution. Ciridae (talk) 17:57, 12 July 2024 (UTC)

This section is resolved and can be archived. If you disagree, replace this template with your comment. Ciridae (talk) 17:57, 12 July 2024 (UTC)

Categories by day: date format?

Dates that clarify or disambiguate categories, such as categories for individual sports matches, should they be written as 6 October 2021 or 2021-10-06?

I would assume that the all-numeric format is the better option, since it automatically sorts similarly named categories by date, and since categories for specific days use that format (e.g. Category:2021-10-06). But both formats are in use and I haven't been able to find any instructions as to which one should be used in these cases.

Sinigh (talk) 11:38, 2 July 2024 (UTC)

In the HTML script stick with 2021-10-06 format. Broichmore (talk) 18:24, 2 July 2024 (UTC)

Yes, definitely! But what about the name of the category? Sinigh (talk) 18:33, 2 July 2024 (UTC)

If it's down to an individual day, I'd strongly favor ISO notation (2021-10-06 in this case). If it was just the month, then June 2021. - Jmabel ! talk 19:12, 2 July 2024 (UTC)

 Agree. I think both of those choices make the most sense for category trees. Sinigh (talk) 20:28, 2 July 2024 (UTC)

I've been using the format '2023-01 text' for categories where only the month matters (if not only the year). I think that is the most reasonable standard so that the cats are sorted chronologically on category pages. Prototyperspective (talk) 21:48, 2 July 2024 (UTC)

You can also solve that with a DEFAULTSORT. - Jmabel ! talk 22:45, 2 July 2024 (UTC)

I know but they are (more) unreliable and why not simply use this format putting the verbatim month into the title isn't useful and just causes problems due to people not adding a sortkey and the title longer. Prototyperspective (talk) 11:40, 4 July 2024 (UTC)

Yes, there are defaultsorting navigational templates for month categories. Sinigh (talk) 13:19, 3 July 2024 (UTC)

"both formats are in use"

where is "6 October 2021" systematically in use?

in any case, prefer yyyy-mm-dd over other formats. RZuo (talk) 16:33, 4 July 2024 (UTC)

Good question, I should have provided an example in my first post. I'm referring to categories like those you'll find in this one: Category:Women's association football matches in Sweden. In this particular case, there seems to be a preference for a comma followed by "M Month YYYY", but dates in brackets are not uncommon elsewhere. (This category also showcases the typically arbitrary mix of "vs", "v", and hyphens both with and without spaces, and not a single unspaced en dash to be seen.) Sinigh (talk) 18:39, 4 July 2024 (UTC)

Tyresö FF-Umeå IK, 16 April 2013‎ should be the agreed format IMO. This is the date format used by Wikipedia too, in the main. The numerical date in reverse, can be hidden from view to force sorting. Broichmore (talk) 09:35, 7 July 2024 (UTC)

If so, please note that the first punctuation mark should be an en dash: Tyresö FF–Umeå IK. As for the date format, that soccer game is actually a very good example for this thread. The category for the date itself is numeric: Category:2013-04-16, which contains two sports games with Day-Month name-Year dates, but also 60+ categories, in and including Category:Photographs taken on 2013-04-16‎, that instead use the all-numeric format. At the same time, there is indeed a general preference for the Day-Mn-Year format across Wikimedia. So you understand why I saw the need to ask my initial question.

Maybe it's intentional that date formats vary depending on the subject on the category? Sinigh (talk) 13:31, 7 July 2024 (UTC)

sports matches is a topic that needs a standard the most, but it seems commons never had a standard?

i just made some considerations:

1. there're two kinds of matches: two sided matches and "first past the post" matches (like track and field, borrowing an election jargon).
2. i think for two sided matches it's easy. the category title should include both teams' names and the date, that's all.
3. in the rare case when on the same day there are more than one match between teams with the same names (e.g. both a football match and a handball match between both france and japan), then we can add the actual sport as the disambiguation, e.g. France vs Japan (2022-11-11, U17 women's football), France vs Japan (2022-11-11, women's football). in the rare case when two sides face each other more than once on the same day, titles can be John Doe vs John Smith (2022-11-11, 1st), John Doe vs John Smith (2022-11-11, 2nd)
4. vs, v., -, or something else, which symbol to use, is up to community to decide.
5. for matches that're not between exactly two sides, i guess titles can just be the name of the event, from big to small, e.g. 2024 summer olympics - women's 100m freestyle - qualifier - group b (2024-11-11).
6. absolutely no reason not to use yyyy-mm-dd. other formats are messy for everything -- appeareance, sorting, intelligibility for non-english speakers, problem of preference between en-gb and en-us...

RZuo (talk) 21:26, 7 July 2024 (UTC)

We need someone to maintain CropTool

It would appear that CropTool is once again broken to the point of unusability. The person who got it working again in February 2024 did not take on responsibility to maintain it over time, which is of course their prerogative. We really need someone to maintain this quite valuable tool. - Jmabel ! talk 17:58, 3 July 2024 (UTC)

I hope this comes across without any challenging or derisive tone, Jmabel, but I have no idea what you're talking about. I've used CropTool hundreds of times and very rarely had any problems. I've used it in the past 48 hours. What kinds of issues are you experiencing? Is this a browser thing? —Justin (koavf)❤T☮C☺M☯ 18:10, 3 July 2024 (UTC)

@Koavf: A variety of issues, ranging from not getting it to load the image at all to going all through the process and not having it saved. I literally cannot remember the last time it worked for a rotate-and-crop (not counting a multiple of 90°), but it's been months. After half a dozen times in a row that it didn't work for me a couple of weeks ago, I've just switched to download, use GIMP, upload. We keep getting reports here and the VP (etc.) that it isn't working, and you are actually the first I've heard from in a month saying that is not a 100% experience; I'm actually a bit surprised to hear it is sometimes working. - Jmabel ! talk 20:12, 3 July 2024 (UTC)

I don't use it very often, but I've used CropTool a few hundred times in the last year and never noticed any issue. However, most times I don't rotate the image. Pere prlpz (talk) 20:34, 3 July 2024 (UTC)

For rotations, I just use a little tool to request User:Rotatebot make the change. —Justin (koavf)❤T☮C☺M☯ 21:13, 3 July 2024 (UTC)

Unless I'm mistaken, Rotatebot will only do multiples of 90°; it will not do (for example) a 1.27° rotation and corresponding crop, nor do I readily see how anyone could know that is exactly what they want without a visual tool. - Jmabel ! talk 22:25, 3 July 2024 (UTC)

It can do rotations down to 1 degree. From the tool I have:

If you request a rotation by 90, 180 or 270° Rotatebot will do this in a few hours. If you request a rotation by any other angle it will probably take longer.

So as long as you can translate whatever you want into 360 degrees, then you should be good. (I should also point out that I've never asked for a rotation other than 90/180/270.) Again, I don't want to downplay if you or anyone else is having problems, but between CropTool cropping and the RotateLink gadget, I don't see the immediate issues. Agreed that long-term maintenance is certainly highly important tho, as these tools will break somewhere down the line. —Justin (koavf)❤T☮C☺M☯ 22:36, 3 July 2024 (UTC)

For the record, I tried the CropTool again, since several people above said it was working. I started from a 4000 x 6000 px image already on Commons. The tool spent somewhere upwards of 30 seconds failing to fetch the image, then prompted me again for a URL. Repeated twice, at which point I'm comfortable in saying it is still broken for me. - Jmabel ! talk 03:34, 5 July 2024 (UTC)

I just tried to crop and rotate and image and it worked: File:Edifici Catalana Occidente (cropped).jpg. Maybe I was just lucky or maybe there is a problem related with the image you tried, your configuration or your equipment. What file did you try to crop?--Pere prlpz (talk) 08:43, 5 July 2024 (UTC)

I think mere corp is more likely to fail for larger file sizes and during busier hours. Other features are completely broken. Enhancing999 (talk) 09:13, 5 July 2024 (UTC)

Recently I've been unable to overwrite using CropTool, getting the error "The overwrite option is disabled because this is a multipage file", for example when trying to crop File:Galerius Arch (Thessaloniki).jpg. I've never seen this error until recently. It seems to work correctly when saving a new file. Consigned (talk) 18:03, 5 July 2024 (UTC)

Occasionally, I get that message (for single page files or files with a blank second page), but it still overwrites. Enhancing999 (talk) 18:06, 5 July 2024 (UTC)

Now I got what I think may be the same error as you. When trying to rotate 270º (without cropping) file:Diagonal 441 - Muntaner 223-225 - 20240618 173626.jpg, I got "[Error] undefined". Cropping the same image without rotating seems to work fine.--Pere prlpz (talk) 22:06, 7 July 2024 (UTC)

Distributed_by

See: File:Onorio Moretti (1881-1939) obituary in The Boston Globe of Boston, Massachusetts on October 24, 1939.jpg in the structured data. I want to allow Distributed_by to be used in structured_data. AP UPI news articles are distributed to news outlets. Recordings and movies have distribution companies involved. I tried to make a change at Distributed_by but it did not work. RAN (talk) 15:05, 7 July 2024 (UTC)

• What I am looking for is to allow distributed_by in structured data. --RAN (talk) 20:00, 8 July 2024 (UTC)

Voting to ratify the Wikimedia Movement Charter is ending soon

You can find this message translated into additional languages on Meta-wiki. Please help translate to your language

Hello everyone,

This is a kind reminder that the voting period to ratify the Wikimedia Movement Charter will be closed on July 9, 2024, at 23:59 UTC.

If you have not voted yet, please vote on SecurePoll.

On behalf of the Charter Electoral Commission,

RamzyM (WMF) 03:45, 8 July 2024 (UTC)

Identifying and categorising special building in Japan

This building is just outside de Category:Hakata Station main station. It has green plants and even a waterfall.

Smiley.toerist (talk) 11:37, 8 July 2024 (UTC)

It’s the Miyako Hotel, Hakata, at (33.58977,130.42283). Dogfennydd (talk) 17:22, 8 July 2024 (UTC)

@Dogfennydd@Smiley.toerist done categorizing! It is a very good thing that Japanese copyright law does allow commercial uses of images of their architecture. JWilz12345 (Talk|Contrib's.) 00:42, 9 July 2024 (UTC)

Long term preservation of media files

In relation to the point Long-term structured digital preservation of humanity's media that has been included here, I'd like to comment on a general idea. I know that it isn't feasible at the moment, but I think about it as a wish (or dream) for the future. There are new very promising storage technologies in development, that will be great for very long term data storage. Once they are available at an affordable cost, many people or institutions will use them to ensure their data is never lost. But what data deserves that more than... human knowledge put together? This goes beyond Commons: all Wikimedia text dumps should also be included there. Text dumps are currently "only" about 25 TB in size (even if including last 5 versions; each one of them covers full history for all Wiki pages). All media files in Commons are currently 543.48 TB in all. 5D optical data storage disks could store up to 360 TB of data, so 2 of them could store all Commons media files, plus full text history of all wiki pages from all Wikimedia sites (including Commons itself). I don't know when it will become affordable, but, once it is, I think is absolutely desirable that WMF produces, each year or each few years, those 2 (or 3 or 4, no more should be needed, even if total size grows a lot) disks of human knowledge. Each pair/group of disks could be stored in different places, so very long term preservation of all content is ensured. If cost ends up being really low, its use can also be very promising for Internet Archive, so the earthquake risk for its contents (that I have expressed several times, both here and in Wikipedia village pump, and even at their own forum) would be virtually eliminated. MGeog2022 (talk) 13:04, 6 July 2024 (UTC)

It is very important to think about how human knowledge and documents of it can outlast many Millennia. Right now, there are several researches going on, which media might be the best for storing many hundreds of terabytes of data like "Project Silica" or "Cerabyte". Researching takes a lot time until respective storage media is available for the market. For writing and using the said 5D medium, we need a machine with a femtosecond laser, which is very uncommon. We need storage media where writing and reading will also be possible in the far future. Preserving data amounts under 1 PB isn't that expensive and can be assured easily. Next to classic HDDs, files could also be stored on LTO tapes, which are rather cheap and are good for archiving. Apart from this, we need distributed archiving and redundancies, and storage media that is not exposed to (human) failures and errors, or malware, or other threats. But as storage medisa gets better from year to year, we already are to store many many Petabytes on small space (the Internet Archive has a high storage density with its petabox). And in theory, you would need only fourty 26 TB HDDs for one petabyte, which fit in any (smaller) room. :) --PantheraLeo1359531 😺 (talk) 14:21, 7 July 2024 (UTC)

But the biggest problems are errors made by human users, disasters, wars, and people who don't think about or don't care about archiving their works (photos or videos of events etc.) they created. --PantheraLeo1359531 😺 (talk) 14:22, 7 July 2024 (UTC)

Another problem is how organisations like Internet Archive or Google, or Meta or whoever cares about preservation. The Internet Archive stores very many petabytes that have to be managed and to be maintained. Companies like Meta or Google hold high relevant files (look at images contributed to Google Maps; many people use Instagram to post relevant scenes) alive. What happens after their liquidation? How are their files treated and how to handle issues like copyright and privacy? What about archives of administrations, of towns, of newspapers, etc. pp... Many questions, but unfortunately too few answers. --PantheraLeo1359531 😺 (talk) 14:27, 7 July 2024 (UTC)

According to the Wikipedia article about 5D optical data storage, it was used in 2018 to store a copy of Isaac Asimov's Foundation trilogy. That made me think that it was a matter of time (let's say 15-20 years, for example) for it to be available at an affordable cost to WMF. Of course there are lots of storage technologies available; the point with 5D storage is that the same physical disc could last for (theoretically) billions of years, so, once a backup is created, it needs no further maintenance. Well, up until 2021, there were no true backups of Wikimedia Commons media files (only several production copies); now, there are backups in both 2 main WMF datacenters. Before 2013 or so, there was only 1 main WMF datacenter (no copies or backups outside of it). I've read about plans to also implement offline backups that can provide even bigger security. Being optimistic, if the vast majority of files survived for many years when there weren't even proper backups (and even when all copies were at the same datacenter), they are much more likely to survive now that they are backuped following established standards. When the day arrives that high capacity media with an indefinite lifetime is available, it would be great to also write those backups to such media, but probably it isn't strictly needed, since well-maintained, frequent backups would in fact get almost the same level of security.

By the way, I think it is worth publicly recognizing the work by Jaime Crespo of WMF, in backups, and particularly media backups. They were an absolute need, and it seems that almost all work was made by him; it's really great.

What about archives of administrations, of towns, of newspapers: this is a complex issue. Data that enters public domain can be uploaded to Internet Archive, or, if applicable, to Commons. Governments of all levels should take care of their historical data, and eventually, to have a kind of world central repository, most of it could also end up in Archive or, in some cases, Commons. Archives of important newspapers are usually well-kept, and, if they enter public domain, could also be centralized in Internet Archive or a similar site.

Internet Archive or Google, or Meta: Google and Meta have enough money to store an enormous size of data, but their purpose is to monetize it, not to preserve it indefinitely. On the other hand, Internet Archive's problem is the combination of its small budget with the really big amount of data that it stores, probably along with some wrong decisions they have made. MGeog2022 (talk) 20:27, 7 July 2024 (UTC)

Sorry if I am trying to go too far too soon: thinking more about reality and less about dreams :-), probably the suggestion I should make is to implement offline backups (for example, on tape), if there are no immediate plans for it just now. MGeog2022 (talk) 20:38, 7 July 2024 (UTC)

For me, talking about these topics and possible scenarios is interesting, and worth to discuss :) --PantheraLeo1359531 😺 (talk) 06:44, 9 July 2024 (UTC)

• Watch some of the videos on Iron Mountain, Boyers, Pennsylvania facility storing digital and physical media, and Granite Mountain Records Vault and compare to 2008 Universal Studios fire original master recordings; and w:Double Fold where archives discard/sell original material after microfilming. --RAN (talk) 15:15, 7 July 2024 (UTC)

U4C Special Election - Call for Candidates

You can find this message translated into additional languages on Meta-wiki. Please help translate to your language

Hello all,

A special election has been called to fill additional vacancies on the U4C. The call for candidates phase is open from now through July 19, 2024.

The Universal Code of Conduct Coordinating Committee (U4C) is a global group dedicated to providing an equitable and consistent implementation of the UCoC. Community members are invited to submit their applications in the special election for the U4C. For more information and the responsibilities of the U4C, please review the U4C Charter.

In this special election, according to chapter 2 of the U4C charter, there are 9 seats available on the U4C: four community-at-large seats and five regional seats to ensure the U4C represents the diversity of the movement. No more than two members of the U4C can be elected from the same home wiki. Therefore, candidates must not have English Wikipedia, German Wikipedia, or Italian Wikipedia as their home wiki.

Read more and submit your application on Meta-wiki.

In cooperation with the U4C,

-- Keegan (WMF) (talk) 00:02, 10 July 2024 (UTC)

Negative boosted AI images

Should Template:Negative boosted template be added to Template:AI upscaled and Template:PD-algorithm? You could argue that most people who are using Commons would rather prefer to find non-AI images when using the search function.--Trade (talk) 21:04, 7 July 2024 (UTC)

It depends on the subject, for some subjects (and not only AI-specific ones) some AI images are the most relevant, especially when it comes to recent art genres. It's more the quality of the image that matters than how it was created. There may not be very many cases of such but I oppose it being added to PD-algorithm. I'd support it being added to items in Category:AI misgeneration. Prototyperspective (talk) 21:30, 7 July 2024 (UTC)

For actively bad images it would make sense to also include the supercategory of Category:Poor quality AI-generated images, which can look superficially useful as a thumbnail but have some fundamental error at full size. Belbury (talk) 13:22, 9 July 2024 (UTC)

(Wrt Template:AI upscaled) Overwriting images with upscaled versions (whether with AI or otherwise) is against Commons policy. Should these edits not be reverted? ReneeWrites (talk) 23:00, 7 July 2024 (UTC)

Commons is chock full of improperly overwritten files. The way the regulars talk about it here on VP, you would think something's being done, but not really. RadioKAOS / Talk to me, Billy / Transmissions 00:04, 8 July 2024 (UTC)

In fairness that's why COM:RFR was created. I don't know how effective it is in curbing destructive overwritings but it's not like there have been no attempts. The issue is that judging by what pages the template links to, the vast majority of these upscalings were done by an experienced Commons user. I assume this was a blind spot for them and they're just not aware of this policy. ReneeWrites (talk) 07:10, 8 July 2024 (UTC)

@ReneeWrites: they absolutely should be reverted.

@RadioKAOS: I certainly revert these on sight, unless it is genuinely own work by the uploader and they make that decision themselves to overwrite. - Jmabel ! talk 02:41, 8 July 2024 (UTC)

The template is to flag that an image may contain speculative AI-hallucinated content, not that the Commons image has been overwritten. Most AI-upscaled images are uploaded as separate files (sometimes, frustratingly, instead of the original, where a user pulling historical photos from old newspapers thinks they look better with the AI treatment). Belbury (talk) 10:57, 8 July 2024 (UTC)

"You could argue that most people who are using Commons would rather prefer to find non-AI images" We can argue lots of things of course, but is this actually the case ... I'd avoid this kind of reasoning, we do not know why people look for the things they look for. Better to make actual judgements on proper arguments. The negative boosted template is a pretty aggressive downrank (similar to that which deletions get) which I would not advise using for anything but literal trash, and thus I would not advise in this case. If it should be downranked with a different value is something that I'd be open to (based on arguments), but we would have to define and add extra levels of derank templates for that to work. —TheDJ (talk • contribs) 14:33, 9 July 2024 (UTC)

Some Commons users won't mind being served AI images. A number of Wikipedia projects seem very relaxed about using AI-upscaled or entirely AI-generated images for biography articles, apparently preferring them to lower-quality historical images. If I myself was looking for a drawing or an icon I wouldn't dismiss an AI-generated image out of hand.

But there is definitely an issue where a user may find it hard to distinguish real photographs from artificial ones, when searching. Right now, a search for "astronaut" is okay for the early results being a mix of photos and cartoony artwork - the photos are all (I think!) genuine, the cartoons are either AI or hand-drawn. There's nothing misleading there. But a search for "london steam train" has a completely fake AI image in the first few results, which only becomes apparent if you click through or hover to read the filenames. Belbury (talk) 13:01, 10 July 2024 (UTC)

Good points, in your second link the problem is mitigated to some extent by the good file naming so when hovering over the AI image one sees a tooltip over it and the URL in the bottom left with "AI-generated" in the title (even when at VP/Proposals many have opposed a policy guideline for descriptive file naming). I think the best approach would be to have a tag in a corner of the thumbnail image that clarifies the image is entirely/largely made using some AI tool, and so even when glancing over the search results. Prototyperspective (talk) 20:09, 10 July 2024 (UTC)

Does Pywikibot still work?

Yesterday I have uploaded File:Brückner Vielflache Fig. 42.jpg using Pywikibot. After one image I have stopped the script, to fix a detail in the description. But since then I was not able to upload the other images. The main error messages are An error occurred for uri https://commons.wikimedia.org/w/api.php and The write operation timed out.

all error messages

Uploading file to commons:commons... Sleeping for 7.3 seconds, 2024-07-10 17:23:06 ERROR: An error occurred for uri https://commons.wikimedia.org/w/api.php ERROR: Traceback (most recent call last): File "/home/PATH/env/lib/python3.12/site-packages/pywikibot/data/api/_requests.py", line 684, in _http_request response = http.request(self.site, uri=uri, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/PATH/env/lib/python3.12/site-packages/pywikibot/comms/http.py", line 283, in request r = fetch(baseuri, headers=headers, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/PATH/env/lib/python3.12/site-packages/pywikibot/comms/http.py", line 457, in fetch callback(response) File "/home/PATH/env/lib/python3.12/site-packages/pywikibot/comms/http.py", line 343, in error_handling_callback raise response from None File "/home/PATH/env/lib/python3.12/site-packages/pywikibot/comms/http.py", line 448, in fetch response = session.request(method, uri, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/PATH/env/lib/python3.12/site-packages/requests/sessions.py", line 589, in request resp = self.send(prep, **send_kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/PATH/env/lib/python3.12/site-packages/requests/sessions.py", line 703, in send r = adapter.send(request, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/PATH/env/lib/python3.12/site-packages/requests/adapters.py", line 501, in send raise ConnectionError(err, request=request) requests.exceptions.ConnectionError: ('Connection aborted.', TimeoutError('The write operation timed out')) WARNING: Waiting 5.0 seconds before retrying.

Was something changed yesterday? My script is the same that has worked for the one uploaded file. --Watchduck (quack) 15:42, 10 July 2024 (UTC)

I have succesfully uploaded photo today. If you dont already do try to use asynchronous=True for upload. If that doesn't help try to set socket_timeout for wikibot. ie.

pywikibot.config.socket_timeout = 120
site = pywikibot.Site("commons", "commons")  # for Wikimedia Commons
site.login()
image_file_path='/tmp/example.jpg'
commons_file_name = "File:example.jpg"
file_page = pywikibot.FilePage(site, commons_file_name)
if file_page.exists():
        print(f"The file {commons_file_name} exists.")
        exit()
file_page.text = wikitext
summary = "Uploading example.jpg"
file_page.upload(image_file_path, comment=comment,asynchronous=True)

--Zache (talk) 16:48, 10 July 2024 (UTC)

pywikibot.config.socket_timeout = 120 worked. Thanks. --Watchduck (quack) 17:02, 10 July 2024 (UTC)