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/2023/11.
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:
Best practices for large UAV (drone) image sets? And a couple of ideas to spitball on that front.[edit]
Hi all,
I searched Commons policies and the Village Pump about uploading large aerial image sets, particularly related to drone mapping, but haven't found clear guidelines. I'm wondering if there are general principles for handling such image sets, often consisting of hundreds or thousands of photos.
Many users generate and upload orthomosaics to OpenAerialMap, which is excellent. But making the original aerial photos available on Commons could bring additional benefits:
Users could download and process them for specific uses (generate meshes, height maps, point clouds, customize resolution, etc.).
Different batches could be combined for improved reconstructions. Images from different periods could generate separate point clouds to compare changes over time (e.g. Notre Dame before/after fire, environmental shifts).
WMF (or a third party) could develop a tool to continually generate updated orthomosaics or 3D models from uploaded aerial photos. This would enable 4D comparison of any surveyed area over time.
The tool could tag uploaded photos by location, allowing all photos of a spot to be easily found without the need of manual tagging.
It could add geolocation to non-geotagged photos, or improve precision on poorly geolocated photos.
It could highlight under-mapped areas needing more coverage.
After decades of accumulating aerial photos, we'd have an open, historical dataset showing morphological changes across the planet - extremely valuable culturally and environmentally.
All of this could encourage uploading aerial mapping photosets, and I have tens of thousands that I could contribute immediately. But some challenges exist:
Uploading thousands of images from one area is non-trivial. No one could reasonably title and tag each individually. If issues arose requiring deletion of some photos, it would be unrealistic to review hundreds of thousands of them manually each time.
Many similarly-titled photos could clutter Commons searches.
It could be useful to allow image set uploads as a single item, or grouping them to enable efficient processing. Maybe something along those lines already exists that I'm unaware of.
Copyright is complex and evolving for aerial imagery, even in FOP countries, so a cautious approach is warranted and the topic deserves an entire thread about it. But we can still explore processes and allow non-controversial uploads. In some cases, aerial photos could be accepted but not made publicly available while still being used for improving orthomosaics and geolocation. I welcome perspectives on these possibilities.
In summary, I'm very interested in approaches for handling large aerial photo sets on Commons. Please share any insights you may have on this topic.
I think uploading them and having the needed georeferencing data as structured data is no problem with the current tools. The problem is that we do not have tools to really use them. To make them available in a usable way it would need WMS and tile services. To set this up as a Wiki (in terms of a project where everyone can contribute) would basically mean to create a new Wikimedia project. I do not think this realistic in the near future as between the last two project Wikidata and Wikifunction there where ten years. But in the long term this might be possible also as cooperation with OSM. GPSLeo (talk) 15:58, 6 November 2023 (UTC)Reply[reply]
@GPSLeo: completely agree that ultimately this sounds like a new project (let's call it WikiEarth because... why not). Integration with OSM would definitely be a must. But while we look forward to that, maybe a few minor adjustments could be made here to allow for that kind of material to be uploaded without generating a mess, even though there's nothing stopping us from doing it now. I alone would bump up the total amount of images on the Commons by about 0.1%, and I'm not even a serious mapper. So with that in mind I'd probably think about implementing a new type of "object" called "aerial mapping set" or something similar. Images inside it wouldn't be found on searches unless taken out of the set manually, but they could still be found by location. Does that make sense? With that done, one could easily download sets, process them and upload them to OpenAerialMap with huge gains and little effort on the short term. Rkieferbaum (talk) 16:22, 6 November 2023 (UTC)Reply[reply]
I do not think that we could get a new kind of file page type for this. Obviously this is nothing where edits or uploads are made manually this is a field for bots and scripts. Maybe store the information as a template that is added to all files of the set. The image sets should have a category for each and the files should not be categorized in other categories. To reduce the amount of files there should be a guideline on tile sizes where the file size is below the limit but the amount of tiles for an area is not to large. If you are interested in defining these guidelines I would propose to set up a page at Commons:Wikiproject Maps. GPSLeo (talk) 16:37, 6 November 2023 (UTC)Reply[reply]
"the files should not be categorized in other categories" Why on earth not? If one of the images shows, say, a football stadium, then it should be added to the category for that stadium, where it may well be the only such image, and thus very useful. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:52, 11 November 2023 (UTC)Reply[reply]
@Pigsonthewing: I think he means, and I agree, all the images in the set as a batch process. Any images that are individually useful should definitely be adequately categorized.
I haven't had the time these days, but I'll draft a WikiProject and ping you all to weigh in in the coming days. Maybe post it in Proposals as well. Rkieferbaum (talk) 14:05, 11 November 2023 (UTC)Reply[reply]
Yes I want to avoid that categories of for example municipalities become filled up with thousands of files for different aerial photos and maps. GPSLeo (talk) 14:28, 11 November 2023 (UTC)Reply[reply]
and that's all you have from wikimedia foundation. a lot of even the most basic functionalities are actually realised by user-contributed scripts, e.g. Help:Gadget-GallerySlideshow.
all the uses of these photos you described are beautiful, but to make them happen, you can only hope that you yourself or other users (aka volunteers) code up the whole project, i'm afraid. RZuo (talk) 18:47, 6 November 2023 (UTC)Reply[reply]
I think GPSLeo's pointers above are a very good short term solution. A workflow where sets are uploaded and kept with no categories except one specific to the set and all relevant data is populated by set through the use of scripts. That settles it for now, whatever else might come can be built around that. I'm not sure, technically, how to get that done, but hopefully people familiar with scripts and templates will come along and weigh in. I'll invite tech-savvy people over to see if they can help. Rkieferbaum (talk) 19:00, 6 November 2023 (UTC)Reply[reply]
be cautious about setting the filename format. because changing thousands of filenames afterwards (if you change your mind or have made mistakes) is troublesome and will create a lot of com:move logs.
In this revision, After categories, I see that wiki code has been added, which is not working. Can anyone take a look at this? --Saroj (talk) 17:50, 7 November 2023 (UTC)Reply[reply]
It looks like he's using some smartphone app and trying to add structured data, which is not at all working. —Justin (koavf)❤T☮C☺M☯ 18:01, 7 November 2023 (UTC)Reply[reply]
I investigated this further. This is a bug in the VisualEditor I created a ticket on this
I just noticed nothing on this page has been approved since August. Any admins feel up to reviewing it? Adam Cuerden (talk) 20:21, 8 November 2023 (UTC)Reply[reply]
Last I looked, the bot that moves categories (and takes care of the various aspects of that) was not running. Unless it's been fixed, {{Move cat}} is moot. I've done some of this by hand, but there is way too much for any small number of people to handle that way. - Jmabel ! talk 02:58, 9 November 2023 (UTC)Reply[reply]
@Jmabel: That is a problem, though there's also the image replacement requests section, where the bot does work, but hasn't been updated. I can try to help manually move a couple categories, though: There's some that are quite small. Adam Cuerden (talk) 10:44, 9 November 2023 (UTC)Reply[reply]
Well, I did Amazon Studios to Amazon MGM Studios. That's one. Adam Cuerden (talk) 13:50, 9 November 2023 (UTC)Reply[reply]
Thanks to @Jmabel for flagging this to me, I found the bot on Toolforge and have re-enabled it. Me doing this is not sustainable, so I've filed T350953 for finding a new maintainer for these bots. Legoktm (talk) 16:20, 10 November 2023 (UTC)Reply[reply]
@Legoktm: thank you so much! (yeah, I know I already thanked you in person, but I wanted to do so publicly.) - Jmabel ! talk 18:41, 10 November 2023 (UTC)Reply[reply]
Here's hoping someone will take this over. Legoktm tells me he figures what he did will only hold us a few months. - Jmabel ! talk 18:41, 10 November 2023 (UTC)Reply[reply]
On November 8, 2023, a user started change the categories around Category:Roads and streets. The discussion Commons:Categories for discussion/2022/06/Category:Streets named after Kyiv was cited as a reference. Thematically, the two terms were previously separated (Roads for roads outside built-up areas, Streets for built-up areas - inside cities). However, this separation has far-reaching consequences and does not correspond to current practice. Wouldn't it make more sense to discuss such a separation beforehand? --XRay💬 09:53, 9 November 2023 (UTC)Reply[reply]
Admittedly, I had started to restore the original layout. However, I stopped the actions because I became aware of the somewhat hidden discussion. --XRay💬 10:00, 9 November 2023 (UTC)Reply[reply]
I had changed the categories around Roads and streets per the consensus I had evaluated in the CFD I have cited. Otherwise, my preference is to keep roads and streets separated, like roads for intercity road transport and streets for intracity road transport. In fact, terms like Stroads would not appear if streets were considered a subset of roads. Sbb1413 (he) (talk • contribs) 10:05, 9 November 2023 (UTC)Reply[reply]
That's when you realize how limited your own English is. I've never heard or read the term stroads. --XRay💬 10:09, 9 November 2023 (UTC)Reply[reply]
I had already suspected that. In the German language, there is no such separation in this form. I also had to read and understand the reference to the categories here. The linguistic diversity does cause some problems here. It reminds me of the term kindergarten, a foreign word in English, and the use of singular and plural. But not only is the word different, a kindergarten in Germany is different from a kindergarten in the USA.--XRay💬 10:36, 9 November 2023 (UTC)Reply[reply]
Good to read: en:RoadUnlike streets, whose primary function is to serve as public spaces, the main function of roads is transportation. --XRay💬 10:50, 9 November 2023 (UTC)Reply[reply]
IMHO roads and streets as category name is good for the summary category - as today. Roads and streets should be a subcategory - as today. Stroads should be a subcategory of roads and (!) streets. This would also correspond to the usual division of categories here. Streets as a subcategory of roads and vice versa should be re-sorted appropriately. --XRay💬 10:29, 9 November 2023 (UTC)Reply[reply]
If it were me I'd get rid of Stroads altogether since the term is just a pejorative slang word for a street with extra lanes (really, "stroads" are just boulevards without the trees, but that's beside the point). Plus it goes against the rule that "There should be one category per topic; multi-subject categories should be avoided. The category name should be unambiguous and not homonymous." What I'd do instead is create categories for roads and streets based on the number of lanes. Or really just move everything currently contained in Stroads to Boulevards since there's really difference except for the aforementioned trees or lack of them, but I don't think it matters as there's plenty of treeless boulevards out there. "Roads" and "streets" should be seperate categories though. --Adamant1 (talk) 12:03, 9 November 2023 (UTC)Reply[reply]
"Stroads" is an obscure neologism that we should not use. "Streets and roads" is fine.
At least some of these changes have been rather strange, e.g. [1]. The streets in NY's Central Park are simply city streets that run through the park. Why would they be in a "roads" category rather than a "streets" category? - Jmabel ! talk 13:15, 9 November 2023 (UTC)Reply[reply]
The closed CFD discussion had a consensus that Streets is a subset of Roads. Especially Auntof6 said, "Why are these categories named "roads and streets"? Isn't a street a type of road? Isn't this like the argument we had a while back to change "Buildings and structures" to just one or the other? The argument was that saying "buildings and structures" (and in this case, "roads and streets") is like saying "carrots and vegetables"." All other CFD participants agreed with them. Sbb1413 (he) (talk • contribs) 13:31, 9 November 2023 (UTC)Reply[reply]
i guess the length of this discussion may mean a cfd for "roads" or "streets" may be necessary.
having lived in 3 different countries with their own languages, i understand why the definitions can be so murky.
my suggestion: keep roads and streets separate, keep streets as a subcat of roads. some countries might not have the streets cat because of linguistic preferences, so let them be.
Road: a long, narrow stretch with a smoothed or paved surface, made for traveling by motor vehicle, carriage, etc., between two or more points.
Street: a public thoroughfare, usually paved, in a village, town, or city, including the sidewalk or sidewalks.
Highway: a main road, especially one between towns or cities.
This means that streets might be a subset of roads, although some users may argue otherwise. Sbb1413 (he) (talk • contribs) 16:03, 9 November 2023 (UTC)Reply[reply]
Another possibility would be to use a system like OpenStreetMap there they define "highway=* is the main key used for identifying any kind of road, street or path". We could make the general Category:Ways the root category for all kind of roads, streets, paths, ways, highways or tracks. GPSLeo (talk) 16:45, 9 November 2023 (UTC)Reply[reply]
Yes this might be a good solution. So we might should suggest to name all these categories "Thoroughfares of", "Thoroughfares named after" or similar. The only problem is that thoroughfares would also include rail or water based transport. Maybe highway is better as enwiki states "A highway is any public or private road or other public way on land." The dewiki also states that highway is in the US the legal term for all public traffic areas. Therefore highway would be the root category for all vehicle land not rail transport and would be in the Category:Thoroughfares. Paths, Trails and Tracks would not be highways as they are not included in the definition of public traffic areas. GPSLeo (talk) 21:43, 9 November 2023 (UTC)Reply[reply]
"Thoroughfares" is a pretty obscure word. I bet even a lot of native speakers don't know that word, or would not be quite sure what it means. - Jmabel ! talk 22:22, 9 November 2023 (UTC)Reply[reply]
road is the most generic word. for example, all land transport except railways is known as road transport.
i dont see why cat:roads cannot serve as the parent cat. any country doesnt have roads but something else that people in other countries would call roads? RZuo (talk) 07:17, 10 November 2023 (UTC)Reply[reply]
But I think roads only refers to vehicle based transport. Therefore transport by foot without a vehicle would not be covered by this term. GPSLeo (talk) 09:28, 10 November 2023 (UTC)Reply[reply]
We'd still call it a "road" or "street" if it is on that scale, but closed to vehicular traffic.
The only vocabulary trickiness here I can think of is that alleys (not to be confused with the German Allee, quite different) are often not considered streets and possibly not roads. You'll definitely often here reference to the "streets and alleys" of a city. - Jmabel ! talk 12:04, 10 November 2023 (UTC)Reply[reply]
It is interesting to see the effects of linguistic differences and national circumstances. My impression is that any interpretation of "road" or "street" ultimately does not create a universally valid solution. Ultimately, we can only decide on one solution, which will then (hopefully) be accepted, but will also contain errors. I was already confused in the USA with interstate, highway and freeway, I am not very surprised about the explanations for the situation in India and here in Germany I see things just as inconsistently. In Germany, we mainly have street names in urban areas. "Mainly" means that there are also such roads outside towns, there are entire settlements that have a street-like name, etc. Out-of-town roads (district roads, state roads, federal roads, highways, motorways) also run through villages. It would at least be great if we could find one (!) common generic term. --XRay💬 12:47, 10 November 2023 (UTC)Reply[reply]
I also like the "thoroughfares" idea. I'm sure you can find many different definitions for streets and roads, and you can find as many different examples of things called "street" or "road" that don't match those definitions. That's true in any one country, let alone around the whole world. Moreover, not too many things are named "X thoroughfare" so we'll avoid much confusion on that level. - Themightyquill (talk) 08:23, 11 November 2023 (UTC)Reply[reply]
I was just about to make the same suggestion: thoroughfare (Q83620). This includes walkways, bikeways, paths, roundabouts, streets, roads, interchanges, and so on. --XRay💬 08:46, 11 November 2023 (UTC)Reply[reply]
In US areas I have been in, I have also seen ways, "Boulevard" names, boulevards, avenues, places, squares, an "Esplanade" name, esplanades, circles, drives, courts, walks, rights of way, ramps, entrances, exits, piers, wharves, crossings, landings, expressways, extensions, centers, plazas, skyways, terraces, tunnels, overpasses, underpasses, and even an "Ocean" name (One Atlantic Ocean is a pier in Atlantic City, NJ). These are overlaid with designations and alignments such as county roads, highways, parkways, pikes, turnpikes, thruways or throughways, state routes, US routes (part of the en:United States Numbered Highway System), Interstates (part of the en:Interstate Highway System), bus routes, train routes. The United States Postal Service maintains a list of their abbreviations here. Then there are the rues in Montreal, Quebec, Canada (and probably other places in Quebec). Many of them descended from paths made by horse-drawn buggies, and before them by livestock. — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 13:22, 11 November 2023 (UTC)Reply[reply]
Thoroughfares already exists, but it's not complete. (See above, RZuo.) There is a lot of works to rename/remove all the "roads and streets" categories, but it's a good solution. --XRay💬 08:52, 11 November 2023 (UTC)Reply[reply]
I found 71 categories with "roads and streets" (Search: Category: intitle:/roads and streets/). --XRay💬 08:54, 11 November 2023 (UTC)Reply[reply]
I think most categories are just named streets but also contain images of roads or other transportation infrastructure. GPSLeo (talk) 09:00, 11 November 2023 (UTC)Reply[reply]
Excluding category redirects, I found over 1,000 categories with "roads and streets" (search query: Category: intitle:/Roads and streets/ -hastemplate:"Category redirect"). --Sbb1413 (he) (talk • contribs) 18:00, 11 November 2023 (UTC)Reply[reply]
IMO it is not necessary to change the categories with "named after". Next step of search: Category: intitle:/roads and streets/i -intitle:/Roads and streets (in .*? )?named after/ -hastemplate:"Category redirect". --XRay💬 19:15, 11 November 2023 (UTC)Reply[reply]
I'm not sure "thoroughfare" works. At least in the United States it insinuates a main road going through a town that forms a route between two places. Well there are many back (or country) roads that don't fit that. Including agricultural and forest service roads. I don't think it's necessarily workable to sort categories for roads based on if they are "main" roads, back roads, country roads, serve agricultural or forest service purposes, Etc. Etc. either. OpenStreetMap's highway tag barely manages to organize most types of roads into the same schema and it only really works in a small a portion of Europe. That's even to account for how roads are classified outside of the United States and Europe either. Really everything should be a sub-category of "roads" and just call it good there. --Adamant1 (talk) 19:16, 11 November 2023 (UTC)Reply[reply]
So you would be fine if for example this would be categorized as "road"? I would also not categorize this as "thoroughfare" as the dewiki article linked on Wikidata refers to "Verkehrsweg" as the term for official public transportation infrastructure what this trail is not. It is an official trail but not in terms of the road law. GPSLeo (talk) 20:59, 11 November 2023 (UTC)Reply[reply]
I'd consider that a path and images of paths would probably best in their own category scheme separate from roads. Roads usually have an "improved surface" though. Whatever that means, but it wouldn't include what's essentially an improvised primitive trail. You might say whatever this is would qualify as a road. But then I don't think places with wood slats above a surface to make it more easily crossable counts as an "improved surface" or therefore a "road." Anymore then you'd call some branches laid across a creek so it can be crossed more easily a road or even call some stepping stones one. Again maybe a path, probably a trail, but a road? No. --Adamant1 (talk) 21:30, 11 November 2023 (UTC)Reply[reply]
Okay, paths are no subclass of roads. But how to name the common category of roads and paths? GPSLeo (talk) 21:42, 11 November 2023 (UTC)Reply[reply]
Don't. Just keep them separate. Confine roads to "improved surfaces" for vehicle traffic (I.E. compacted surfaces like gravel and paved roads) and anything below that to paths and/trails. There's some more fuzzy classifications there for sure, like compacted dirt track roads, but that's the simplest way to do it IMO regardless. Realistically I don't think any IRL outside of OpenStreetMap would classify that as a road anyway. I know where I live it would just be a path. Although we don't really have "track roads" to begin with, but that's besides the point. --Adamant1 (talk) 21:57, 11 November 2023 (UTC)Reply[reply]
So you would be fine if for example this would be categorized as "road"? In northwestern Montana it would probably have a state highway number. But, more seriously, that looks more like a trail or footpath than anything I'd want to call a road, thoroughfare, or whatever.
I notice someone has started changing things to "thoroughfare". I really don't think that's a good choice, especially because (like "road") it connotes something with a lot of traffic. I personally would be fine with "streets and roads" even though it's not the way we usually do things. The vocabulary is complicated here, including as you move around the English-speaking world. I could also live with just "roads", because technically a street is a type of road, but it "feels weird": at least in the U.S., in any thickly settled area, "road" suggests something major. I can't imagine asking an urban dweller in the U.S., "what road to you live on?" - Jmabel ! talk 04:11, 12 November 2023 (UTC)Reply[reply]
That means, Roads and streets would act better as an umbrella category than Thoroughfares. However, I had subscribed to the "thoroughfare" idea because both roads and streets can be defined as thoroughfares. I had previously cited the Dictionary.com definition of a street ("a public thoroughfare, usually paved, in a village, town, or city, including the sidewalk or sidewalks"). You can exclude walking infrastructure like Paths and Walkways from thoroughfares, as they are not subjects of vehicular traffic. I have included them mainly for convenient purposes. Sbb1413 (he) (talk • contribs) 04:22, 12 November 2023 (UTC)Reply[reply]
The probably with that, I.E. making "thoroughfare" as the top level category for both streets and roads, is the whole "including the sidewalk or sidewalks" part of your definition since not all roads have a sidewalk, obviously. So you can't make Category:Roads as a child of Category:Throughfares. I guess we could just ignore that, but then there's other problems with "throughfare" being the top level category as well. For instance like I think I've already pointed out it insinuates a main road that goes from one place to another. Again as I think I've pointed out already, along the lines boulevard. Yet many roads and streets are not main throughfares. Plus they often have end points. If you want to argue a "throughfare" can end, cool. But I'd point out the word "through" in throughfare, as in "moving in one side and out of the other side." So roads or streets that end inherently can't be throughfares. And yes I'm aware that the word is technically "thorough" but it has the same connotation as "through" lol. --Adamant1 (talk) 07:36, 12 November 2023 (UTC)Reply[reply]
Comment I just added two colons to your contribution because you added two wrong categories to the village pump page by their omission. -- Robert Flogaus-Faust 10:27, 12 November 2023 (UTC)Reply[reply]
"Thoroughfares" is not a solution for everything. But it is a better alternative to many categories with "roads and streets" (such as "Quality images of roads and streets") and a much better alternative than "roads" as an umbrella category. --XRay💬 08:54, 12 November 2023 (UTC)Reply[reply]
At the end of the day I could probably live with it as a parent category for roads, but it really doesn't fit with streets. There's no reason roads and streets have to be in the same parent category either. --Adamant1 (talk) 09:02, 12 November 2023 (UTC)Reply[reply]
Yeah, "thoroughfares" may not fit well with streets, but both Dictionary.com and Merriam-Webster define a street as a thoroughfare. Sbb1413 (he) (talk • contribs) 09:09, 12 November 2023 (UTC)Reply[reply]
I could easily find other places that don't define streets using the word thoroughfare and I've already pointed out the word insinuates a through road. So I don't think just defaulting to dictionary.com's definition is a good or valid way to do this. Otherwise what makes it more legitimate then other sources except that the definition fits your personal opinion? I don't think that's how we usually choose category names either BTW. There's no reason we can't come up with something that is widely used, fits for both roads and streets, and satisfies most (if not all) people in the conversation though. --Adamant1 (talk) 09:26, 12 November 2023 (UTC)Reply[reply]
Please have a look to en:Thoroughfare. IMO it fits well for roads, streets and a lot of more. But you're right, there are differences. It's more road than street. (In German it's easier, it's all "Straße". Or more common "Weg". ;-) ) --XRay💬 10:24, 12 November 2023 (UTC)Reply[reply]
The more I look into the subject, the more difficult it becomes. There are so many differences between the linguistic and national situation in German-speaking and English-speaking countries alone that a sensible, uniform solution for commons seems impossible. The best uniform solution currently seems to me to be "transportation network" - which will hopefully also include the dead ends. --XRay💬 10:34, 12 November 2023 (UTC)Reply[reply]
... or "transport infrastructure". --XRay💬 10:41, 12 November 2023 (UTC)Reply[reply]
Because of the problems with the use of the term Thoroughfares, I am no longer subscribing to its idea. Instead, I would like to restore the Roads and streets category, which would include all the non-rail land structures used by land vehicles or pedestrians (roads, streets, stroads, alleys, paths, walkways etc.). Sbb1413 (he) (talk • contribs) 15:14, 12 November 2023 (UTC)Reply[reply]
One other thing I want to remind everyone of, and not just for this particular set of categories: Categories are about helping people find stuff, using terms they are likely to use and understand. It is not intended to be ontologically pure, though it's nice when it can be. - Jmabel ! talk 22:00, 12 November 2023 (UTC)Reply[reply]
Blockers to automated import of structured data[edit]
I recently came across the AC/DC gadget (maintained by @Lucas Werkmeister, I think?), which helps with adding structured data. A lot of the work it does seems like it could almost be automated. For instance, I could go to Category:The Wharf (Washington, D.C.) and add P180 (depicts) > The Wharf (Q99344305) for all the images, but that information is basically already established, because the category is already linked to the Wikidata item. It'd be neat if the gadget suggested the linked item as the entity depicted. It'd be even neater if we could find a way to import depict statements for all categories linked to a Wikidata item.
Naturally, an action at that massive scale would require careful consideration. It also seems like, ultimately, the way we'd get more comprehensive structured data for our collection. I'm wondering if it was considered when structured data was first implemented, or if there has been any other discussion about it.
I'm also interested in discussing potential blockers (i.e. situations we'd have to account for in which automatically adding the statement would not be appropriate. One big one that occurs is set categories vs. topic categories (I linked to the Wikipedia guideline, which explains it more clearly, but the Commons analogue is here). Set categories are straightforward, but for topic categories, we might have e.g. an image of a document written by Jane Smith in Smith's category, and we wouldn't really say that it depicts Smith.
Are there other potential issues that I'm not thinking of? And do folks have any other general thoughts about importing structured data from Wikipedia? Cheers, {{u|Sdkb}}talk 21:43, 9 November 2023 (UTC)Reply[reply]
FWIW for many categories few (if any) of the images in the category could usefully be said to "depict" the item for the category. We are constantly having problems with people adding things like "depicts Seattle" or "depicts Chiquita Brands International". - Jmabel ! talk 22:27, 9 November 2023 (UTC)Reply[reply]
Agreed with Jmabel here. At the mass scale you're suggesting, error rate would need to be extremely low to allow them to be reliably fixed by humans. Even with careful design of the tool, it won't be able to account for all the subtle complications in categories. Pi.1415926535 (talk) 23:35, 9 November 2023 (UTC)Reply[reply]
Yeah. It's somewhat w:garbage in, garbage out, so we wouldn't need to get the error rate to zero. But certainly lower than what it'd be if we went ahead at the moment. Maybe better establishing/tagging which categories are set categories vs. topic categories would be an interim step here. {{u|Sdkb}}talk 06:58, 10 November 2023 (UTC)Reply[reply]
Every time I use Depictor I have the same idea: that so many of these associations are already established by the fact that the file is in the category! But the thing is that definitely not all of them are, and for some categories I'm surprised at the large number that aren't. So I'd be wary of automating anything. I do think a quicker way of selecting for depicts (e.g. something like Cat-a-lot) would be great though. SamWilson 01:41, 10 November 2023 (UTC)Reply[reply]
@Samwilson, @Jmabel: FWIW, if you open cat-a-lot and select some images and then open AC/DC, it’ll prefill the files to edit with your selection. (I also created PagePile Visual Filter to similarly filter other sets of pages in a similar way, e.g. categories that are too big to fit in one screen; you can use the resulting PagePile with AC/DC.) Lucas Werkmeister (talk) 14:45, 10 November 2023 (UTC)Reply[reply]
@Lucas Werkmeister: so if you already know what tool to use, this lets you know that the tool does this.
I'd really like to see, for this and other tasks, some things where a user has a fair chance of starting from what they want to do and find a tool that does it. - Jmabel ! talk 12:31, 11 November 2023 (UTC)Reply[reply]
Aside: It irks me that so many of our tools do not share a common user interface, or common behaviours. This is barrier to entry for new users. A well as the difference between selection in ACDC and Cat-a-lot; category selection works differently in Cat-a-lot and HotCat, for example. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:41, 11 November 2023 (UTC)Reply[reply]
Is there any way to load news articles under copyright, with a restoration date, without going through a speedy deletion process, so that they never actually appear and are restored at the appropriate date? Can we preload works expiring on January 1, 2024. RAN (talk) 04:19, 10 November 2023 (UTC)Reply[reply]
For that particular case, why bother? Wait 8 weeks. - Jmabel ! talk 12:05, 10 November 2023 (UTC)Reply[reply]
I have some 2025 material too. --RAN (talk) 13:01, 10 November 2023 (UTC)Reply[reply]
Off the top of my head: slap a maintenance category on them, and we can delete the whole category with VFC. But unless you make a list of them in Category:Undelete in 2025, it may be difficult to find when the time comes. - Jmabel ! talk 18:47, 10 November 2023 (UTC)Reply[reply]
We had a a pretty long discussion here about the categorization of village names in the area once known as Nagorno-Karabakh Autonomous Oblast (and later known interchangeably as the Republic of Karabakh/Artsakh) last month. Since Azerbaijan ethnically cleansed the area, there has been a lot of activity here in the sub to delete all of the Armenian category names of the settlements and replace them with Azeri names only. We have other areas where multiple names are used, and I think in this region it will be important to keep both names as well. In the case of many of the villages, they have (and continue) to be written about mainly using the Armenian names. We have examples of Catalon/Spanish names coexisting such as Category:Donostia-San Sebastián, old and new names coexisting, such as Constantinople and Istanbul (as totally separate categories), and I am okay with either solution, or with just having both an Armenian settlement name category and an Azeri one coexisting side by side for all the settlements of the former Nagorno-Karabakh region, and have them categorized into their regional categories as well. I know most of the users/world do not care so much about this region, but simply for the practical value of storing and finding information by users, both readers and uploaders of content, this solution is important, and considering there are alternate arrangements whenever it seems helpful, I see no reason why this solution wouldn't be quite helpful here. --RaffiKojian (talk) 04:44, 10 November 2023 (UTC)Reply[reply]
You're rehashing the same argument. In the previous thread, several users, including an admin, explained why it's impractical to have two or more separate names in a category title. We were close to reaching a reasonable agreement until you went back to your old stance of using multiple names in a single category title. Please stop wasting the community's time. — Goldentalk 05:40, 10 November 2023 (UTC)Reply[reply]
That's nice you consider this a waste of time, while spending so much of your time erasing all of the Armenian village names. But the fact is that here we have other solutions for special cases which accommodate multiple names, and others chimed into the conversation stating as much and sharing the opinion that they agree to keep the Armenian names, and you conveniently ignored that and then just dove back into the erasures. I believe there is a solution to be found that would help the majority of users working in these categories find things, which is the very reason for categorization, and I think it's to double-up on settlement name categories - one in Azerbaijani and one in Armenian transliteration. A simple, elegant solution that would serve users regardless which language they know the name of the settlement in. RaffiKojian (talk) 19:24, 10 November 2023 (UTC)Reply[reply]
Hello Golden, you are a native speaker of Azerbaijiani, the country who recently took the area by military force. As a Wikimedia user with no consensus backing your wished migration you may want to sit this one out, delegate that decision to the Commons community and freeze your categories moves. Seeing Azerbaijiani users press for renaming this early will be perceive as brutal, possibly nationalism bigoterie and online harassment. We have nothing to gain from such rush.
Same for RaffiKojian, I encourage you to take some distances with this topic. Hugo en résidence (talk) 17:08, 13 November 2023 (UTC)Reply[reply]
I have not moved any category with an Armenian name to an Azerbaijani one since the first discussion where it was first contested. However, I still find your statement problematic. Being Azerbaijani or Armenian does not inherently make us unable to contribute constructively to topics about our own region. Asking Azerbaijanis and Armenians to distance themselves from a topic that is literally about their countries doesn't sit right with me. — Goldentalk 18:08, 13 November 2023 (UTC)Reply[reply]
I'd say that is above the U.S. TOO. Remember that TOO issues are all based on national law, so an example from a different country isn't worth much. - Jmabel ! talk 12:08, 10 November 2023 (UTC)Reply[reply]
i just noticed that if you are not logged in and want to create a new page, you will see a message, which i think is produced by MediaWiki:Newarticletextanon. it seems the box (produced by Template:Anon-warning-no-edit?) only has english and a handful of other languages. we need to translate it, but i dont know how. RZuo (talk) 16:16, 11 November 2023 (UTC)Reply[reply]
@RZuo: Editing in the MediaWiki namespace is restricted to Interface Admins, but they don't have their own noticeboard yet, so use COM:AN. — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 04:25, 12 November 2023 (UTC)Reply[reply]
File:Seal of New Hampshire.svg The NH state seal is licensed on Wikimedia Commons as "public domain". However, there is a NH state law which states that No person shall manufacture, sell, expose for sale, or have in possession for sale any article or substance, being an article of merchandise or receptacle of merchandise or article or thing for carrying or transporting merchandise, or sell, expose for sale, give away, or have in possession for sale or to give away or for any purpose any article or thing to advertise or promote services, upon which shall have been printed, painted, attached, or otherwise placed a representation or likeness of the state seal
This is concerning because "Whoever violates the provisions of RSA 3:9-a shall be guilty of a misdemeanor if a natural person, or guilty of a felony if any other person" — Preceding unsigned comment added by A.FLOCK (talk • contribs) 06:35, 12 November 2023 (UTC)Reply[reply]
Same with images of currency, non-copyright restrictions. You must follow the law, but that has nothing to do with copyright. --RAN (talk) 00:07, 15 November 2023 (UTC)Reply[reply]
Hello, I'm looking for the source (exact page) of this image of a St George knight. The image description says "Vies de Saints (BNF Richelieu Manuscrits Français 185)." I checked this manuscript« La Legende des Sains » [de JACQUES DE VORAGINE], traduction de « JEHAN BELET », but I can't find it. Does anyone know the exact source (page)? -Artanisen (talk) 01:35, 13 November 2023 (UTC)Reply[reply]
Thanks for your help HyperGaruda. So the information was incorrect since 21 August 2016. Dbachmann should be notified that he or she shouldn't add incorrect source data. -Artanisen (talk) 12:03, 14 November 2023 (UTC)Reply[reply]
The fact that the package is transparent may contribute to a determination of COM:DM, since you could argue that this is a photograph of the mashed yam, while the graphics are just part of the package. -- King of ♥ ♦ ♣ ♠ 10:10, 13 November 2023 (UTC)Reply[reply]
I see your point. Thank you for the tip. --トトト (talk) 14:44, 13 November 2023 (UTC)Reply[reply]
Most of it being simple text probably helps. It's really only the small diagram of the box - which is so simple that there's at least a question as to if it passes Threshhold of Originality - and the picture of the yam (Which probably does pass Threhold of Originality, but is a very small part of te image) that are at all problematic. Adam Cuerden (talk) 22:59, 13 November 2023 (UTC)Reply[reply]
I think what little here is copyrightable is de minimis. I'd say "keep". - Jmabel ! talk 01:51, 14 November 2023 (UTC)Reply[reply]
Threshold of originality : does creator or uploader localisation matters ?[edit]
Hello all,
I, an user based in France, recently uploaded a few French institutions logos based on their lack of originality. The visuals were created in France. The copyright holders are French public institutions.
But I'm realizing that Commons:Threshold of originality (TOO) does not states if I —a French user uploading from France— and my uploads —designs created in French juridiction— can be shared using USA-based legal copyright logic. Should my upload logic be based on the design's national juridiction (therein, France copyright laws and interpretations) ? Do we have some clarification on this ? Are there some users with proper comparative law or legal know how to enlighten us on this ? Hugo en résidence (talk) 09:56, 13 November 2023 (UTC)Reply[reply]
We generally do not care about the country of the uploader or of the creator of the specific file (e.g. if a user decides to recreate a logo as an SVG). What matters is the country of the organization that the logo is meant to represent. Here this distinction is not important since it's all France. -- King of ♥ ♦ ♣ ♠ 10:07, 13 November 2023 (UTC)Reply[reply]
My understanding is an upload must be a free image in both the country of origin (e.g., France) AND the US (because the servers are in the US). That can give rise to some anomalies. The foreign country and the US may have different post-mortem copyright expiration dates. The countries may have different thresholds of originality. For example, the UK recognizes sweat of the brow and has a very low TOO, so an image might be below TOO/free in the US but not in the UK; Commons would not allow the image. Countries also have different freedom of panorama laws. For example, The Louvre Pyramid is not a free image on Commons because France does not have freedom of panorama even though the US does. Glrx (talk) 16:20, 13 November 2023 (UTC)Reply[reply]
@King of Hearts and Glrx: thank you. I believe the logos I uploaded are clearlycopyleft below TOO in US, and arguablycopyleft below TOO in France. Moreover, there are from public institutions, so I feel « ok » to take responsibility for this upload under PD licence on the basis of {{PD-textlogo}} and French laws. I also did so in the context of my mission within local public institution University of Toulouse. I hoped USA's servers' more tolerant juridiction would take precedence, but this is ok as well. Hugo en résidence (talk) 16:40, 13 November 2023 (UTC)Reply[reply]
@Hugo en résidence: they are pretty clearly not "copyleft" (rights deliberately released under a free license) unless there is some evidence of that which hasn't been presented. However, they are probably below the threshold of originality for copyright. - Jmabel ! talk 01:56, 14 November 2023 (UTC)Reply[reply]
@Jmabel: , thank for the distinction. I'm editing-correcting it in my former text. Hugo en résidence (talk) 07:52, 14 November 2023 (UTC)Reply[reply]
Hello, following this discussion in my talk page, since I am not a native speaker in English I raise here the topic: when writing the name of a musical act which name begins with "The" (The Beatles, The Rolling Stones), I have always known (since school) that the article must not be capitalised unless it's the first word of a phrase after a full stop. Accordingly, I named the categories, "Members of the Beatles", "Members of the Rolling Stones" and so on. I just want to know whether i am tuned with the rest of the community because the user Bricks&Wood thinks differently. Thanks -- Blackcat 19:23, 13 November 2023 (UTC)Reply[reply]
@Bricks&Wood: Sadly, it's not so. Commons is not reference for itself, and it doesn't honour anything. When the article is the first word of the phrase is ALWAYS capitalised. You made the wrong example. Nowhere in the press "the" before the group's name is capitalised when it's not the first word of a phrase. -- Blackcat 00:14, 14 November 2023 (UTC)Reply[reply]
I already remarked at User talk:Blackcat before discovering this discussion here, but will repeat for broader audience (and more succinctly).
It seems that the consensus in the end is that when "the" is part of the name, it is only capitalized at the beginning of a sentence: en:Wikipedia:Manual of Style/Music#Names (definite article). We should probably follow that, even if it goes against my gut. - Jmabel ! talk 02:19, 14 November 2023 (UTC)Reply[reply]
i agree with Bricks&Wood. i dont think it's a wise practice to force certain nobody's "manual of style" onto real-world people's official names.
that is, if they do it all caps, we do all caps; if all lower case, we lower case; etc. "The" is no exception.
the only exception is mediawiki's technical limit of capitalising the first char, so we cant have iPhone, eBay...--RZuo (talk) 12:25, 14 November 2023 (UTC)Reply[reply]
If the band has been consistent, yes, we should follow them. Have they been? - Jmabel ! talk 16:13, 14 November 2023 (UTC)Reply[reply]
The pictures of historic paintings at the newly-online Philip Mould Picture Archive website (Mould is a well-known commercial art dealer) are not generally available elsewhere. There are some with non-free frames, and some still in copyright. Nor is the resolution high, sadly. Nonetheless, it would be useful to harvest what can from the site, and to add metadata to Wikidata. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:37, 13 November 2023 (UTC)Reply[reply]
We are dangerously close to 100 million files--how should we celebrate?[edit]
I took a look in the archives and didn't see any discussion, so sorry if I missed it. We are very close to 100 million files and for those of us who don't understand how orders of magnitude work, this is the last "nice, round number" we'll hit for a long time. I propose that we should have some special designation on the front page for this and celebrate all of the hard work that everyone has put into making such a huge repository of free media. Thoughts? —Justin (koavf)❤T☮C☺M☯ 23:39, 13 November 2023 (UTC)Reply[reply]
And ignore the persistent "quantity over quality" issues that our limited pool of editors haven't been able to even scratch the surface on correcting, some going back many years? As I pointed out recently, it was all spin when en.wp put all the emphasis on edit count instead of article count during the 20th anniversary, because 6 million articles is dick all to brag about when you look at it realistically.RadioKAOS (talk) 03:15, 14 November 2023 (UTC)Reply[reply]
As the saying goes, one mans junk is another mans treasure. Although there is a lot of low quality stuff on here that should really be dealt with somehow in the long-term. But 100 million files is still a huge accomplishment regardless. It would be cool if nothing else there was at least a special barnstar or something that could be given to whomever uploads the 100 millionth file. That's really my only suggestion though. --Adamant1 (talk) 03:53, 14 November 2023 (UTC)Reply[reply]
I'd worry that the lucky number would end up going to a file that's unsuitable for Commons (like a selfie or copyvio), boring (like a file in a batch import), or something that would be awkward to call out (like a company logo or an image of something unpleasant). Omphalographer (talk) 18:55, 14 November 2023 (UTC)Reply[reply]
The 100 millionth file will only be a guess, when taking into account the thousands of files deleted daily. So we can choose whatever suits the purpose. Yann (talk) 10:35, 16 November 2023 (UTC)Reply[reply]
Yes it is impossible to say what file is the number 100.000.000. The file with the id 100.000.000 exists since February 2021 File:1973 Gedenktag 2020 zum 9. November auf dem Platz Bornplatzsynagoge im Grindelviertel in Hamburg.jpg. I would suggest that we use this number to discuss a redesign of our main page. We should change the links in the "Highlights" section to actual galleries and also replace the "Content" section as the category links are not reasonable usable. GPSLeo (talk) 11:21, 16 November 2023 (UTC)Reply[reply]
I will add a "Woohoo!" and raise a container of my favorite beverage. A banner under the logo would be nice. — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 04:07, 14 November 2023 (UTC)Reply[reply]
Idea : onboard 200 news admins, potentially diving from old admins without recorded blocks. That could help to clean up. Hugo en résidence (talk) 09:52, 14 November 2023 (UTC)Reply[reply]
why dangerous? :D
also if you can, please leave a note at Commons talk:Commons Gazette when the number reaches 100mil. then i can include the news.--RZuo (talk) 12:25, 14 November 2023 (UTC)Reply[reply]
I aborted a batch of uploads of my own photos just now. Since the last time I uploaded anything of mine, Upload Wizard was apparently reconfigured to only allow me to mark them with a CC 4.0 license. I'd rather use the license I've been using and migrate them when I've had the opportunity to familiarize myself with the differences between the two. Any suggestions besides trying to force me to do things this way? I can assure you that I'll take my time and talents elsewhere before I comply with that.RadioKAOS (talk) 03:22, 14 November 2023 (UTC)Reply[reply]
Thank you for sharing your concerns. Could you let me know which license you prefer to use for your uploads? This will help me in understanding your needs better and in finding a suitable solution.- Udehb-WMF (talk) 13:13, 14 November 2023 (UTC)Reply[reply]
@RadioKAOS: You can use VFC promptly after uploading to change the license, with an edit summary indicating that the original license was in error. If you haven't done this to whatever was already uploaded, I strongly recommend doing so. - Jmabel ! talk 16:16, 14 November 2023 (UTC)Reply[reply]
@Udehb-WMF: It looks like they were previously using the CC-BY-SA-3.0 Unported license.[2]Nosferattus (talk) 21:04, 14 November 2023 (UTC)Reply[reply]
@Udehb-WMF: Frankly, I think limiting the UploadWizard to the current CC licenses for own work is a good thing. It makes the process more streamlined and less confusing. Plus the 4.0 CC licenses are a huge improvement over the 3.0 licenses. Nosferattus (talk) 16:29, 15 November 2023 (UTC)Reply[reply]
Expand 400 character limit for strings in Data namespace[edit]
I ran into a 400 character limit for strings in JSON files in the Data namespace. I need room for larger strings. More information at Help:Tabular Data#Data types. Can this character limit be increased? Heyzeuss (talk) 21:29, 14 November 2023 (UTC)Reply[reply]
@Heyzeuss: One problem with having large character limits for strings in the Data namespace is that editors will potentially start including text that is copyrighted. Can you provide more information about your use case and why you need to include strings longer than 400 characters? Nosferattus (talk) 16:38, 15 November 2023 (UTC)Reply[reply]
I'm transcribing the table of contents for a book over at Wikisource. Each chapter in the table of contents has a description that is roughly paragraph-sized, and some of them exceed the 400 character limit. You're right about the copyright issue, but the editors at Wikisource are particular about what authors have been dead for 75-100 years, depending on jurisdiction. There is plenty of public domain text at Wikisource that could be utilized in the form of tabular data, that has strings far exceeding 400 characters. Heyzeuss (talk) 19:59, 15 November 2023 (UTC)Reply[reply]
I would say that a TOC should never include such a large text. Just shorten it and include the original one on the page for the section text itself. The catalogs of libraries even shorten the title of books if they are to long. GPSLeo (talk) 20:35, 15 November 2023 (UTC)Reply[reply]
That's true, but we're transcribing the text from the pages of old books, including the pages with tables of contents. In that situation, we can't truncate the chapter descriptions. Another thing is that they have lists of sub-sections in each chapter. There are other potential uses for long strings, like entries in old dictionaries and encyclopedias. Heyzeuss (talk) 10:11, 16 November 2023 (UTC)Reply[reply]
.svg.png)
Comment I just added two colons to your contribution because you added two wrong categories to the village pump page by their omission. -- Robert Flogaus-Faust 10:27, 12 November 2023 (UTC)
19:23, 13 November 2023 (UTC)