Eye of the beholder

After seeing this discussion in far too many other threads I decided to Google some of my images that may be considered as out of scope. To my amazement I actually found three that are used out of context to bring them within scope. This one is shown at the end of an essay to point out 'eye candy', I think. This one as an example of an image that violates site policy. This one I can't understand the context after translation but others may be able to. Ours is not to judge whether or not others may find images useful. The arguments about 'server space' are very common and should not be used as commons doesn't use all that much in the big scheme of things; it only uses about 15 to 20 tera-bytes. I see far too many DR that are created for personal views of images and tagged as 'out of scope' as an excuse to list them. I can see this being valid for low resolution images that can be easily replaced with better images like most of those in Category:License plates of Alberta, its sub cats and parent cats. Images that aren't in cats may not stay that way long if someone does a search and adds ones they use to more cats. When images and categories are taken to DR then 'out of scope' should not be the only reason to discuss them. This DR is an example of that. As soon as we see 'out of scope' on a controversial image then we should question the true agenda of the nominator.--Canoe1967 (talk) 17:14, 14 March 2013 (UTC)

Sorry, but "server space" was never a valid argument and can never seriously be used. From the moment a file is uploaded to Commons, it will remain on the server, even if its deleted. So, a deletion doesn't free any server space. --Túrelio (talk) 17:21, 14 March 2013 (UTC)

I would disagree with server space never being a valid argument: The goal isn't really that uploads against policy are be deleted and left stored on the server; the goal is that files that are against guidelines aren't uploaded and stored and taking up space in the first place. --Closeapple (talk) 23:27, 14 March 2013 (UTC)

Indeed. If space is never freed from the server, that's all the more reason to conspicuously publish, and where necessary preventatively enforce, policies setting out what is and isn't an appropriate upload. —Psychonaut (talk) 08:49, 15 March 2013 (UTC)

Nice thoughts, but rather meaningless (sorry) with our current upload procedure. Around 20% of new uploads (8,000/day) have problems (missing source, author, license, wrong license) or are plain copyvios. Uploaders are lying in your face about being the creator of stuff they copied from the web. Do you really think any BOLD warning at upload will change that? The same with out-of-scope uploads.

So, if you want to take server space into account, the upload procedure of Commons needs to be cut in 2 separate steps.

• 1) upload to some pre-Commons server, from which the files are really removed, not just hidden. Images from this server should not be able to be linked into any other project.
• 2) acceptance-check for copyright status and scope, which leds either to true deletion from pre-Commons server or transfer to Commons server.

As a collateral benefit, this might also deter the upload of attack images, as they cannot immediately be posted into articles/pages.

--Túrelio (talk) 09:07, 15 March 2013 (UTC)

That's a fascinating idea! I rather like it in principle, despite the obvious workload issues. I would say that, as with Wikipedia, the bigger Commons' collection gets, the less likely that any new contribution is really necessary. Therefore, over time we should become less focussed on quantity of new contributions, and more on quality of new contributions. At some point in the crossover from quantity to quality, the approach you suggest becomes rather logical. Whether we've already reached that point is of course a matter of opinion. Rd232 (talk) 12:17, 15 March 2013 (UTC)

No, it is not completely meaningless: just because a remedy is not 100% effective does not mean that it is useless. There are already conspicuous warnings and instructions on our upload page, and I would wager that were they to be removed completely the percentage of copyright violations and other problems would be a lot higher than 20%. —Psychonaut (talk) 10:02, 15 March 2013 (UTC)

 Comment Interesting thing put forth here. I had never wondered that the image is not actually deleted. I knew that admins can see it. But it hadn't struck me till now. We can do one more thing about this. Just as for procedure, we can somehow permanently delete all files which have been deleted and stayed deleted for say 5 years or so. (You see, just how offices maintain a hard copy of a file for some years and then they trash it.) but is permanently deleting any file possible? §§Dharmadhyaksha§§ {T/C} 11:57, 22 March 2013 (UTC)

I think that oversighters and MW-developers can really delete things. --Túrelio (talk) 12:55, 22 March 2013 (UTC)

Oversighters can't delete things - only hide them from everyone who doesn't have "oversight" permission. Developers can do anything. Rd232 (talk) 16:02, 22 March 2013 (UTC)

Lets get these developers to get all old junk deleted. I mean, what's the point of pretending to delete copyrighted images when they are actually in use at least by a smaller group of people. Its like, i am all alone, so laws don't apply for me, i am allowed to murder. Also.... One of the editors is giving me this as a reason to not delete stuff, as it doesn't really get deleted. §§Dharmadhyaksha§§ {T/C} 16:37, 22 March 2013 (UTC)

One good thing with not deleting things permanently is that the files can be placed in Category:Undeletion requests for undeletion when the copyright expires. --Stefan4 (talk) 16:50, 22 March 2013 (UTC)

We don't do that for all images. We should at least do that. Or else there is no point in keeping them invisible. But what about the out of scope images, the facebook type personal images? §§Dharmadhyaksha§§ {T/C} 06:23, 23 March 2013 (UTC)

I somehow like the idea of a pre-Commons were things can be realy deleted also this also gives some Potenzial for abuse. I think however that you would need to be able to link imediately to the picture because there will be a huge backlog immediately.--Saehrimnir (talk) 16:14, 8 July 2013 (UTC)

A thought experiment

I'd like to propose a thought experiment which might help identify a mutually agreeable lower bound on "realistically useful for an educational purpose". It has been argued by some contributors here that if a particular well-defined topic is in-scope on some Wikimedia project, then media which illustrates this topic, and which does not violate any other non-scope related policies such as copyright, should be considered in-scope for Commons. The argument further goes that such images should never be deleted, even if they are not currently in use, because there is always some potential educational use which might be exploited at some point in the future; such future contributors, whether they're external or on a Wikimedia project, should have the benefit of choosing from a rich variety of freely licensed media when looking to illustrate their works. Now for the thought experiment, I propose applying this argument to an admittedly ridiculous extreme. By doing so I'm not trying to invalidate the whole argument, but rather to test its limits, and if they are found wanting, to solicit opinions on what limits need to be added to make it useful as a policy.

But what if I were to upload several more images, each showing another uniform random sample? Say, ten more? Again, these still might be useful for showing the consistency in density when presented together. Or perhaps if I left these ten images on Commons long enough, contributors to Wikipedia might pick and choose different ones to illustrate different articles on randomness-related topics. Perhaps eight of the images might end up being used, but never more than three in the same article. Does that still prove that all eight are in-scope? Or would anyone now start to argue that the images are redundant, and that all but three should be deleted, since the Wikipedia articles which use the deleted ones can be changed, without loss of information, to use the three remaining images? If not, what of the two images which were never used on a Wikipedia article? Are they in-scope? After all, eight of their brothers were, so there's certainly the potential that the other two will eventually find use.

Taking the experiment even further, what if I were to upload a hundred more distinct random samples? Or a thousand? Or a million? Is there any point at which the images are no longer "realistically useful for an educational purpose"? If so, what is this point, and how do you know when you've arrived at it? Can you formulate some sort of test which would apply in the general case—that is, not just to images of random data, but to images on any other topic which is generally considered to be in-scope on Wikimedia projects? —Psychonaut (talk) 09:55, 15 March 2013 (UTC)

I would answer it this way: all of the images are in scope no matter how many millions there are. They would be deleted for reasons of redundancy (2.3.5 not 2.2). To try and take a real-world example: a category for a very widely photographed building has shots of the whole structure from almost every angle you can take a photo from. If there is a mediocre-quality shot then it is less useful, if it's terrible then it's in danger of getting chucked. If someone treks into the middle of the jungle and takes a rubbish photo of an ancient temple, but they're the only person to do it, then the photo is safe because it's the only one we have. See also: rare vs common species categories. In-scope media is regularly deleted for other reasons --moogsi (blah) 11:51, 15 March 2013 (UTC)

The policy you cite is worded in such a way that it covers only duplicates which are inferior in quality: "At deletion requests you will need to provide reasons why a particular file is inferior to the alternative version." (All the deletion templates it lists in that section make similar reference to "better" versions of the images.) In my hypothetical example all the images are of the same quality, and they are all unique, so I don't think that 2.3.5 of the deletion policy applies. On the other hand, Commons:Project scope also addresses redundancy, but in a way that is more specifically applicable to this thought experiment: it says that "files that add nothing educationally distinct to the collection of images we already hold covering the same subject" are not in scope. The question is therefore at what point, if any, does adding more unique, good quality images on a single topic violate this clause? —Psychonaut (talk) 12:05, 15 March 2013 (UTC)

You are right that the scope version of redundancy is more relevant. But there is unfortunately no well-defined point. It's an arbitrary decision based on "how many images do you need to illustrate something?" So there is one point of arbitrariness to focus on (the arbitrariness is what upsets people most about the scope policy in general, I think) --moogsi (blah) 12:27, 15 March 2013 (UTC)

Of course there's some degree of arbitrariness, and I'm not advocating that we hard-code a limit on the number of images that can be uploaded for any one topic. But the extremes between specifying such a hard limit and giving users or administrators carte blanche on deciding how much is too much could possibly be narrowed with some guidelines of general applicability. I was hoping that this example (and indeed the entire RfC) could spark some ideas for what these guidelines should be and how these should be formulated and applied. —Psychonaut (talk) 12:45, 15 March 2013 (UTC)

Thanks for this example, I think that's a useful way to stimulate thinking about the nature of "in scope". It prompts me to think that perhaps (as so often) the problem is from thinking in binary terms, "in scope" or "not in scope". Rather, we should think in terms of (a) the benefit of an additional image, which depends on the quality (of which there are various aspects, including resolution, composition, etc etc) and on how different it is from existing images (and "different" could include licensing differences) and (b) the cost of an additional image, in terms of making it harder for users to find and select an image that suits their needs because a larger collection of images is harder to browse and search. The benefit and cost need to be weighed against each other. We do this as part of scope discussions by focussing on (a), because (b) is so difficult to evaluate. But being more conscious of the trade-off may be helpful. Rd232 (talk) 12:34, 15 March 2013 (UTC)

Could we have a bot create a new cat in all cats that has valued, featured, quality, and high resolution images to make searches easier? Two new cats to each: High resolution and then the other 3-4 in another cat. We could also manually add to a third cat that other users can put images in as 'checked as good'?--Canoe1967 (talk) 22:09, 15 March 2013 (UTC)

The problem is that for most areas you actually need the reverse. ie the ability to exclude known weaker images from searches. For example it would be nice to trim the low quality poorly described stuff from Category:Science Museum (London).Genisock2 (talk) 20:31, 14 April 2013 (UTC)

• I think the approximate answer to this thought experiment is that when an individual user using the normal interface thinks it is worthwhile to upload and annotate new random points files, they are at least within a few orders of magnitude of being worthwhile. A user isn't actually going to take his own time to upload millions and millions of random points files. Now, it may be that you still decide a user is being a bit nuts, but there you're bickering over maybe a factor of 10 between what most people think is too much and he thinks is too much. At which point the gap can be bridged by having him try to explain his thought process to others or vice versa. Odds are, he has some legitimate reason to upload random points files (perhaps they're the basis to calculate mathematical patterns of static, or random profiles to sue for some kind of program that automatically draws art based on random points files). The key "biological difference" between the inclusionist and the deletionist is imagination - they latter group simply doesn't have much. But at some point some people out of the consensus should hazily see the possible use. Wnt (talk) 00:44, 1 May 2013 (UTC)

Archive

I'd like to raise the concept of Archived files here. Archived files would have their own Archive: namespace (that's easy to do) and would be excluded from category views unless the user ticks a box on the page (harder to do, but this is a long-term idea, so let's not worry about that). The effect of this would be that Archived files would be excluded from searches and category browsing by default, but could be easily included by the user as needed.

This would be helpful because it would allow us to tidy away files that are clearly superseded by better ones, but which we don't want to delete (eg because they've been around a while, and may be in external use). I think the idea is relevant here because a big reason we worry about Scope issues is that having more and more files makes it harder for users to find what they need - that's why we want to delete things that don't seem to be much use. So, what do people think? Rd232 (talk) 12:34, 15 March 2013 (UTC)

This puts me in mind of the media which is deleted as out-of-scope so so often, the "Facebook photos", i.e. shots of random people and their friends. The (somewhat unfeasible) alternative is to categorize these images in a way that lets people find them and use them. Is there any sensible way we can store them? What about "Fashion in 2013", "People in 2013", for all those future historians? There is of course Category:Unidentified people, populated in great part by DrTrigonBot. Now, I love DrTrigonBot, mostly because it seems to be completely insane, but it also makes me think of something: machine vision will improve the ability to search through media and will make the feasibility of categorizing it seem less important. We are certainly a very long way away from "Computer, show me a picture of...", but my thought of "how is anyone supposed to FIND ANY of these images?" starts to vanish, just a little bit. I like your use of the words "things that don't seem to be much use" - the first section on this page highlighted the impossibility of thinking of all the possible (current and future) uses of a piece of media --moogsi (blah) 13:00, 15 March 2013 (UTC)

Educational scope

The purpose of the wikimedia foundation is to provide educational content, that there are an increasing number of wikimedia projects is an indication that what is educational can take many different forms; though there are of course many overlaps. One would not expect a wiktionary entry on a topic to be identical to the wikipedia entry on that topic nor the wikivoyage one, articles are not even identical across languages. One way for Commons to be educational is as a source of images for the other wikimedia projects, and some believe that this is the only purpose of Commons, hence the oft heard phrase, delete as it is not in use.

However Commons is not wikipedia, nor wikibooks etc, Commons must be educational in and of itself, and must define what is educational in its own terms not that of wikipedia etc. The obvious educational purpose of Commons is as a source of free to use files to others, be that school children looking for pictures for a school project, teachers preparing presentations or a magazine needing images to illustrate a story.

However being a source of images is not the only way Commons can be educational. One bugbear is the upload of personal images, pictures of themselves, family portraits, holiday snaps, pictures of their living rooms etc. However (and apologies to everybody who has seen me make this argument before) Would we delete personal photographs from 1863, 1913, 1963? Such files offer a window into the past, of fashion, dress hairstyles etc. being an archive of files illustrating how people lived/live in a particular time and place is a valid educational exercise; being an archive of internet memes is educational. Being an archive of material that people and researchers can use to conduct OR is educational, because although wikipedia disallows OR enabling research is educational. A book is of course educational but would you argue that a bunsen burner is not. Both are but in different ways.

And yes I am saying pictures of people's living rooms is educational.--KTo288 (talk) 23:34, 18 March 2013 (UTC)

I believe that some old personal photos were deleted in Commons:Deletion requests/Family history.

We do need some photos of people and living rooms and similar things just as we need some photos of Eiffel Towers and Mona Lisa and other things. The problem seems to be the same as suggested above: if there are too many images, then some will be deleted per COM:SCOPE. Since there are fewer photos of 19th century living rooms than photos of 21st century living rooms, a private photo of a 19th century living room is more likely to survive than a private photo of a 21st century living room. --Stefan4 (talk) 00:20, 19 March 2013 (UTC)

Zee problem with that argument is that it ignores the difference in rates photos taken, changes in photo subject and the digitisation barrier. While estimates vary we are probably pretty close to 50% of photos being taken since the year 2000. Of those taken before that point a lot are probably lost (recycling film for its silver content is destroying a lot of material), in the possession of archives without the resources or interest in digitalising or in the possession of families in the same position. Finally subjects have changed. Obviously we all know about the pre 1901 tendency towards formal portraits but even after than point photos tend to be of certain subject families considered to be of significance. Babies, holidays, weddings, parties. As a result photos of everyday life are less common and more valuable than they are in a day and age when everyone instagrams their lunch or I might take 20 photos of my livingroom checking the motordrive on my camera.Genisock2 (talk) 20:58, 14 April 2013 (UTC)

Files as tesserae

Imagine that each file here is individually a tessera. An individual tessera may be plain and boring, or it may be beautiful in its own right, but plain or jewel like a tessera is only part of something greater, i.e. a mosaic. Now imagine that every tile being part of the same Roman floor mosaic and that the subject of the mosaic is knowledge, some tessera belong to at least one pattern in the mosaic, and most in multiple ones. Looking at the patterns they resolve themselves into panels in some areas grainy, in others clearer. The more tessera we have and the better we sort them out into the correct patterns that they belong to the clearer and and better the mosaic.

Imagine that the file to the left represents our mosaic, and that each random point in it represents a tessera. If the square is completely black then our mosaic is complete, and if it is white nothing. Despite the complaint that we already have too many files I don't think that we are anywhere close to the level of coverage that this file would represent.--KTo288 (talk) 00:45, 19 March 2013 (UTC)

• I would tend to agree with you in principle, although I think that you overstate the level of development of Commons, I would not say that we have anywhere close to the level of 50% of the media that we need, in fact whole areas are left empty. Let me give you an example, on youtube I often view videos like "top ten reasons why creationists are wrong", I ask myself a question "Are they educational", and the answer is clearly "Of course they are". Now, please show me a single video of that general area, I am not even talking about the specific subject, but show me a video of "Strategies for successful campaign to extend FOP" or "Useful tips on digging a trench". You will find nothing. And the scary thing is that if I were to spend several months right now developing a collection of such videos, somebody would come along while my computer breaks or something, and delete the whole bunch as "not in use on english wikipedia". Sinnamon Girl (talk) 07:32, 13 April 2013 (UTC)

Eh claim it as wikiversity course material. Wikimedia as a whole doesn't really have a howto project at the present time. That said at the present time video content is so uncommon that there is little in the way of attempts to delete any of it.Genisock2 (talk) 21:05, 14 April 2013 (UTC)

But you see, you need to lie, cheat, and try to make it fit into a preexisting category. This shouldn't be needed. Sinnamon Girl (talk) 03:00, 15 April 2013 (UTC)

Well given that your initial claim isn't based on anything it seems strangely appropriate.Genisock2 (talk) 19:02, 15 April 2013 (UTC)

Free files at Flickr and other user upload sites

Stream of thought over this topic-didn't really get that far-but maybe other users may pick up and run with this.--KTo288 (talk) 00:45, 19 March 2013 (UTC)

 QuestionGiven that files are loaded at Flickr and at other sites, e.g. Geograph with Commons compatiable licenses, why copy them to Commons, they would be just as educational and available to users outside of the Wikimedia projects where they are. If we find one that is useful for a particular wikimedia project, why can't we just load that one file to Commons, rather than hoovering up all that we can find by bot.--KTo288 (talk) 00:45, 19 March 2013 (UTC)

 CommentAlthough some Flickr users are no doubt flattered to find their images used in Wikipedia articles, others have been horrified by the way their images have been characterised and of derivative works created from them. This is in no small part due to some users not really understanding the implications of the licenses that they apply to their images. To maintain control of their images, and also perhaps encouraged by Flickr's attempt to monetise content, some Flckr users are revoking the original Commons compatiable licenses and applying less permissive ones to files old and new. It is my subjective experience that Commons compatiable licenses are becoming increasingly rare at Flickr.--KTo288 (talk) 00:45, 19 March 2013 (UTC)

 QuestionIn that case we should grab every free file that we can, while we can then, and keep them all. Tough luck if someone changes their mind licenses cannot be revoked.--KTo288 (talk) 00:45, 19 March 2013 (UTC)

 CommentNo. While legally we would be entitled to do so, such a mindset in my opinion is self defeating. I see Commons as being here for the long term, I hope for it, and the work we have put into it, to survive long after we have gone. The consequences of this is that Commons' relationship with the world at large must also be for the long term. Alienating users at Flickr and elsewhere into believing that we are stealing their images, using them for nefarious purposes and bullying them over copyrights, which they never thought would be used against them, will just turn off the taps of future user created PD and creative commons images. We need to educate, flatter and persuade, not huff and puff about what we are entitled to do.--KTo288 (talk) 00:45, 19 March 2013 (UTC)

Finding the hypothetical user

One argument that I and others have been derided for using is the hypothetical student/educator/researcher who finds the perfect file for his or her work on Commons, and because we cannot guess what that file is we should hold onto everything just in case.

So lets attempt to find this hypothetical user. Wikipedia now has that survey thing, can we add one to our files, along the lines of

Is this file useful to you- 1 to 5

Will you be using this file for an educational purpose-1 to 5

I can't code to save my life so I don't know how hard it is to do, the best way to do it or even if it is possible. We could also conduct a survey advertised on the main banner. To get the best possible coverage and feedback there should be at most a dozen or so questions with maybe yes-no tick boxes. e.g.

• Are you a regular user of Commons. Y/N
• Are you using Commons for a Wikimedia project. Y/N
• Do you consider your use of Commons as being for an educational purpose. Y/N
• If yes did you find what you are looking for Y/N
• Are you a student Y/N
• Are you an educator Y/N
• Are you a researcher Y/N etc

We can perhaps work out who our users are and what if anything we are doing right, and perhaps if there is any chance that our hypothetical user exists.--KTo288 (talk) 01:11, 19 March 2013 (UTC)

Political satire and propaganda

The original motivation for this RFC deals with political satire. I believe political satire and propaganda is in scope, as propaganda is by definition "educational". Commons is not Wikipedia, we cannot decide what and how people should be taught to think. Neither does Commons have a WP:NPOV policy.

Take as an example this collection of cartoons on the 2008 South Ossetia war. Is there any cartoon there we would not gladly host if we had the (copy)right to? Then ask this second question: Would the Wikipedia community ever reach a consensus, that including any one them in the Wikipedia article on the conflict would be in line with WP:NPOV? (likely not.)

Political satire is always propaganda. A cartoon that was not biased just would not have the punch line. -- Petri Krohn (talk) 08:28, 2 April 2013 (UTC)

I don't think anyone disputes that political cartoons by well-known people or published in significant media are likely to be in scope. The problem arises with original user-generated content. Rd232 (talk) 22:38, 2 April 2013 (UTC)

I don't think it's a matter of someone who is not 'well-known' or 'user-generated', issues like these always boil down to 'someone you don't like' by someone who doesn't like people. There is a graphics lab, there are valued images, there is a pervasive fundamental objective throughout the project to get people involved in making content. I think read somewhere that Jimbo put his foot in his mouth once suggesting photos should be by notable artists. That went sour and was an instant epic fail, I think he may have made some effort to backtrack I can't recall, because the point is, and the success of the project relies upon, fostering a creative community of content producers. Nothing else would work, after all, anything your bot can copy mine can copy better and so on comes into play. No community, no project. Copying everything into one place just makes it easier for someone to copy the whole project.

Political satire is a proper subject, it's simply less and less welcome as the badly structured project moves towards a police state. Tyrants are no fans of satire. Propaganda however is a different matter, in fact, at the graphics lab, some member of the office I think it was, asked for an image to be worked on, the name of the image and request had the descriptive word/name 'propaganda' in it. Penyulap ☏ 19:27, 9 April 2013 (UTC)

I don't think that the propaganda cartoons of georgia-russia would go in the south osettia war article, maybe one might, if there was a sub article on propaganda and reactions to the war and media coverage and so on, that's where they'd go.

It's pretty hard to actually write those sub articles or good quality articles on parody, memes, propaganda for various topics without pictures to go with them. You can't very well describe the pictures with just text when it would be so much more effective with pics. Hard to have the pics when people with no humour attempt to drive those who have it from the project generally.

There are some notable artists who are well recognised and donate countless pics to the project and they are despised by people conditioned to think themselves god (high self-esteem) when they do nothing all day and never improve themselves (perfect already, just ask their self-esteem).

This effect is even worse when you add 'good at humour' or 'good at parody/satire' to the mix. Then the talent-less and humourless are doubly upset because not only does the other person have talent, but they seem to be enjoying it as well, and the natural way to ease tension is to use humour, which leaves the humourless without retort, unless they have a sysop bit with which to oppress, thereby triggering the parody of wiki management.

I think it's safe to say that this is one encyclopaedia that will never evolve to the stage where it can effectively document parody, satire, memes, comedy-related subjects in an encyclopaedic fashion. Maybe the next one can. Penyulap ☏ 20:02, 9 April 2013 (UTC)

I don't know what you're on about. Category:Political cartoons has a lot of content in it, and there are obviously reams of art produced by notable artists. The issue is simply that whilst photos generally document something and therefore have educational potential if the something is worth documenting, art is from the imagination, and does not necessarily document anything. The distinction (though I've made it clumsily...) isn't that hard to understand. Of course art may be worth documenting in itself, if it's significant. A cartoon in a newspaper is likely worth documenting because of the editorial decision gone into publishing it, regardless of the merits of the cartoon itself. A cartoon I scrawled on the back of my napkin is likely not worth documenting, however stunningly insightful I may think it is. Rd232 (talk) 20:19, 9 April 2013 (UTC)

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.