Wikipedia talk:Non-free content/Archive 53

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 50

Archive 51

Archive 52

→

So what are we supposed to do?

After months of kerfluffle regarding numismatics articles, we're now effectively back to where we started. I know the supporters of the use of these images would say we're right back to where we should be. Please, bear with me. I'm not trying to start another war.

Allow me to use a case example. As of now, Commemorative coins of Poland: 2008 is the #1 user of non-free content by article on the entire project. See this report (note: #1 is inaccurate, as those files have recently been deemed free license for our purposes). This article uses 44 non-free images. I note several problems with this article and the images it uses;

There is 44kb of text in the article. Do we really need 44 images to support 44kb of text? There's no policy or guideline on this of course, but there is no featured article on the project (based on sampling of WP:FA) with anywhere near this 1:1 ratio of non-free content usage.
22 of the 44 images have dimensions exceeded 1000 pixels. Per this template noting NFCC policy, the sizes need to be reduced.
17 of 18 of the external references used link to the Polish Mint (3) or the National Bank of Poland (14). These are primary sources. Per that policy, "Do not base articles entirely on primary sources". Sure, 1 of the 18 is not a primary source, but the article is effectively based on primary sources.
The only non-primary source is the Polish Numismatics Society. The only place that is used is to support assertions of the market value of five coins. Further, a numismatics society is a tertiary source. There's no secondary sourcing for this article at all.
All of the uses of images violate WP:NFTABLE, which discourages the use of non-free content in tabular format. The problem here is that there is no critical commentary of the designs in any respect. The entire meat of the article is comprised of descriptive summations of the 44 designs.
All of the images have a fair use rationale. To some, that would appear that it meets WP:NFCC #10c. But, on closer inspection every single "purpose of use" stated for the images is "Currency picture". A few have additional text (below the rationale box) of "Demonstrative of the layout of Polish currency". If that is a valid reason for use, then there would be absolutely no reason why we can not use non-free images for every single currency item in the world. Sure, you could update the rationale to something lengthy and sounding good, but in practice the only thing these images are being used for is displaying the coins. That's it.
Per WP:NOTGALLERY, Wikipedia is not a mere collection of photographs. But, that's what we have here, plus technical details. There's no encyclopedic content.
Per WP:NOTEVERYTHING, Wikipedia is not a complete exposition of all possible details. Yet, that's what we have here. There's no summary of accepted knowledge regarding these coins. It's just a reposting of technical details from primary sources. If I were to create an article on a company listed on a stock exchange that had the name, logo, and technical details of the company and nothing else it would likely be deleted. But, this doesn't?
Per WP:INDISCRIMINATE, Wikipedia is not an indiscriminate collection of information. But, that's what we have here; an indiscriminate collection of every technical detail of each of these coins.

As it stands, I could take a number of actions. I could;

Place the article for AfD. It would be kept, even though it's based entirely on primary sources which is prohibited by policy (the one tertiary source not withstanding).
Remove the images as violating WP:NFCC #8 and #3a. They would be restored by several people who insist it is appropriate use. This already happened.
Tag all the images as disputed fair use because the rationales are woefully weak, and without changing the way in which they are used in the article they will remain woefully weak. All those taggings would of course be undone, and I'd probably be accused of a WP:POINT violation. There might be an effort to lengthen the purposes of use, and almost certainly without changing the nature of use, leaving the rationales still woefully weak.
Rewrite the article to be more of an article rather than a reposting of technical details of these coins. It would be reverted for not following Wikipedia:WikiProject_Numismatics/Style/Currency_article#Coins.

Now, please understand, I'm not assuming bad faith here. I am well cognizant of what typically happens and this is it. So, regardless of policy we are stuck with an article that fails our requirements in a large number of ways. In this case, we've carved out a specialist class of articles that comprises some of the heaviest users of non-free content on the project (60% of the non-free content in the top 20 alone) that are untouchable. Numismatics now effectively have a special exclusion that makes them immune to actions based on the violations I've noted above. No, it's not written down anywhere, but that's effectively what we have. Is this really what we want? --Hammersoft (talk) 20:30, 19 July 2011 (UTC)

While I think there would be more trouble if these were the common currency of Poland, the fact they are "commemorative" coinage means they shouldn't get anywhere near the same type of pass (what exists of that) for regular country currency. Add to that that each table explains exactly what the two sides of the coin bears, and the visual representation becomes unnecessary.

I wouldn't "fight" this on the notability aspect. I'm sure with a bit of prodding, there could be sources from newspapers explaining the release of these coins, so that's not the way to do it.

But it is non-free, it is a list-style article with what otherwise seem to be non-notable individual coins, and there's no discussion of the artwork on the coins beyond the description. Clearly this needs to be reduced. The numismatics projects has not provided any strong reason for keeping them beyond that being their status quo, and seem unable to accept the use of a few example images as demonstration. (It would be more difficult with the commemorative coins as they have no fixes series, but again, the visual appearance of the coins is not necessary to understanding). --MASEM (t) 20:49, 19 July 2011 (UTC)

One might ask how removing any of the coin images from that article about coins actually furthers the non-free content mission. They are not replaceable by free content, there is no legitimate question about the legality of such an informational use, and there is no concrete benefit given to downstream users by reducing an article from 15 non-free images of coins to 10 or some other arbitrary lower number. The only thing that might be accomplished would be rote compliance with stated guidelines and nothing more. So I fail to see how there is any 3rd Pillar interest here that could possibly outweigh the 1st Pillar interest in providing information. This simply isn't a battle worth fighting. postdlf (talk) 21:10, 19 July 2011 (UTC)

That's a slippery slope argument that can be applied to a huge list of articles, not just coins. The thing that WP can take advantage of is that we are not the end-all, be-all of information, as some tend to think we are. We are a tertiary source, summarizing primary and secondary sources. We want people that are researching a subject to use those resources for better authority on the topic. To that end, there is no need to display every coin minted by a country if it conflicts with our free content mission. A representative sample helps to identify the topic but beyond that it only hurts the mission. --MASEM (t) 21:22, 19 July 2011 (UTC)

The opinion that it "hurts the mission" is exactly what I was contesting. postdlf (talk) 23:15, 19 July 2011 (UTC)

The mission aspects are two-fold: we want to educate at a tertiary level and maintain a free content encyclopedia that uses non-free content under exceptional circumstances. To education a reader on the commemorative coins from a country, I argue that we maintain the same educational mission by showing a representative sample alongside descriptions of what coins not illustrated contained, and links to where non-free content is not an issue where these coins can all be displayed, and at the same time, we improve our goal to the free content mission. --MASEM (t) 23:28, 19 July 2011 (UTC)

Are vector files of copyrighted/trademarked images allowed?

I'm not sure if this is the proper place to ask, but it seems as good as any since my question deals with non-free content. I sometimes help improve images over at the Graphic Lab, and once in a while there will be a request to create an SVG version of a raster image (usually a logo) that is copyrighted and requires the addition of a fair use rationale for every article in which it appears. And usually someone will oblige and create a vector version, which makes the image infinitely scalable. My question is, how does this not run afoul of the image use policy and non-free content guidelines? (Or maybe it does?) Since in more than one place I have read that part of the fair use rationale is that the image being used is sufficiently low resolution, it seems like making a version with an effectively infinite resolution would be unacceptable. Could anyone clarify or point me to any prior discussion on the topic, if any exists? -MissMJ (talk) 04:16, 17 July 2011 (UTC)

One has to make a call if the logo fails the Threshold of Originality - that it, if the logo is simple enough - containing just text, basic shapes, and the like - there is no "sweat of the brow" in creating the work and it becomes ineligible for copyright. Such logos that fall within this are completely appropriate to put into SVG and allow for infinite scaling. On the other hand, more complex logos that do pass the Threshold and is copyrightable, must follow NFCC, and an SVG image of that logo would be inappropriate due to the infinite scaling (the recreation of the logo is a derivative work, the copyright still to the original creator). --MASEM (t) 04:23, 17 July 2011 (UTC)

So if a logo is already tagged with {{PD-textlogo}} (or should be), like File:Boeing wordmark.svg, then it's fine to create a vector version of it, but if it's tagged with {{non-free logo}}, like File:Whirlpoolcorp2010logo.jpg, a vector version shouldn't be created? But then you run into something like File:Boeing-Logo.svg, which is a non-free logo, and has all the requisite qualifiers of needing to be sufficiently low res, but an SVG of it exists anyway. Seems kinda shady to provide it, even if it's followed by "this image should not be rendered any larger than is required," because why not just have a small enough raster file and avoid the issue altogether?... -MissMJ (talk) 04:55, 17 July 2011 (UTC)

I would argue that that logo should not be present on WP as an SVG, even if it is possible, presuming that the graphics are sufficiently unique for copyright. (I would further argue that because we have the free wordmark, the non-free logo image is unnecessary, but that's assuming that there's no further discussion on the graphic itself, but that's not a discussion here). SVGs of non-free images should not be allowed due to the nature of infinite resolution. --MASEM (t) 06:31, 17 July 2011 (UTC)

Makes sense. Thank you! -MissMJ (talk) 07:07, 17 July 2011 (UTC)

It's been my experience that non-free SVG logos have been allowed for years (provided they have a fair-use rationale and meet all other other requirements for non-free works). In general, they reproduce the logo/trademark more accurately. I know it was discussed here; thought that they used to be mentioned on Wikipedia:Logos but I don't see it now. Use the {{SVG-Logo}} tag. Making our own vector versions of trademarked logos, as opposed to converting ones directly made by the company in questions, is slightly dicier from a trademark perspective though -- they need to be very accurate, as if they are inaccurate in some detail they could be seen as diluting the trademark (even when used in an educational context). It would usually be preferable to find a PDF or other vector version directly from the company in question, to eliminate those trademark questions. Carl Lindberg (talk) 15:29, 17 July 2011 (UTC)

People have apparently tolerated or even supported non-free SVGs for years as Carl above points out, but it doesn't make a lot of sense to me--they are far higher resolution than Wikipedia will ever need from a fair-use standpoint, and raise issues of bad tracing/redrawing of the images: in short I think they violate a whole mess of WP:NFCC, but I've had lots of people tell me that SVGs that can be drawn at X resolution are not "high-resolution", which I simply don't understand. Masem brings up a good point in that trademarked images might be less problematic. Der Wohltemperierte Fuchs^(talk) 15:47, 17 July 2011 (UTC)

There already are too many vector images in Wikipedia that are copyrighted! Evidently, Wikimedia foundation can be sued for distributing copyrighted images in the highest quality possible:∞*∞ resolution. Perhaps no one does that because Wikipedia serves a noble cause. I personally want Wikipedia to be in possession of these copyrighted images which greatly add to the quality of the encyclopedia. We should take up this issue with the legal committee of Wikimedia foundation (in the hope that we will be able to find a justification for continuing our vectorization drive). In the worst case we will have to rasterize all the vector non-free images in Wikipedia to a suitable resolution! We will have to, either start to Imagine a world in which every single person on the planet is given free access to the sum of all free human knowledge!? or amend the law! – ∃ Aditya 7 ¦ 16:55, 17 July 2011 (UTC)

If we allow non-free vector images, isn't it illogical to not allow high resolution non-free raster images; given that high resolution raster images can very easily be embedded in SVG images! – ∃ Aditya 7 ¦ 18:38, 17 July 2011 (UTC)

This comes up every couple of years or so. The standing advice is that the SVG is fine, so long as it doesn't contain a more detailed vector description than that needed to render the logo at the resolution we render non-free bitmap logos at. Yes, the vector description can be cleanly re-rendered at abritrary scales, but so long as it doesn't contain any additional detail intentionally included that would be invisible at the resolution we need, that has been seen to be acceptable.
Ideally, the vector description should come from a pdf issued by the company itself, rather than a user tracing or user-recreation. An important thing is to present the mark to fairly reflect the trademark holder's design, to avoid tarnishing the brand. From a legal perspective, that is probably a much more significant consideration than the absolute size of the image, which, since the basic case for fair use in the way we use the graphic here is so strong, probably would not be held to much affect the degree of the copyright taking. Jheald (talk) 19:25, 17 July 2011 (UTC)

I can argue fairly that an SVG logo that passes the ToO, pulled from a PDF or equivalent created by the company represented by the logo, is fine for these reasons - we aren't recreating anything ourselves. I would question a non-free SVG logo created by a WPian in the same fashion. --MASEM (t) 19:38, 17 July 2011 (UTC)

J, then the question is why use a vector when we can use a raster image that has the same amount of content at a specified resolution, no larger than would ever be necessary for our purposes? Vector non-free images are completely unnecessary and in violation of WP:NFCC, unless you can explain how an image that can be rendered without loss of quality at X number of megapixels qualifies as "low resolution" no matter what. Der Wohltemperierte Fuchs^(talk) 19:46, 17 July 2011 (UTC)

I would not eliminate user-made vectors; they probably make sense in some situations (particularly where the trademark may have lapsed but the copyright hasn't, and there are no vector versions available). There is no real difference, copyright-wise, so there is no difference to their "free"-ness. The only issue is one of trademark, really, and that would usually be pretty thin (it would have to be visibly inaccurate I think to be an issue -- we wouldn't want to misrepresent something visibly incorrect as their official logo). There really is no good reason to disallow SVGs in my opinion; they are flexible and would look (much) better in print as well, if that is needed. Wikipedia articles are not always just for websites. Quite frankly companies would often prefer it anyways since they will almost always look better when generated from the vector source. The fair use arguments would be the same; there is no real difference between them and rasters I don't think. The main issue is having more expression than we need; for rasters image size could be argued, for vectors, it's more in the amount of detail present (i.e. if there is lots of tiny details not seen at regular sizes, we could strip that out). The amount of copyrightable expression we host in an SVG is identical no matter the resolution so the image size becomes completely meaningless (just make sure we don't render it any bigger than we really need in actual use I guess). Most logos are designed to look OK at small sizes anyways so there is rarely any of that type of tiny detail. The issue has been brought up several times and I'm pretty sure consensus was to keep them. Carl Lindberg (talk) 01:13, 18 July 2011 (UTC)

It doesn't really matter if there's the exact same visual information or not—resolution is different from detail anyhow. An SVG unequivocably fails WP:NFCC C.2 ; we can use a low-resolution raster, so we should. The whole "companies would prefer it" is at best untestable and at worst plainly false; if it were true companies would be packaging the Illustrator eps files or svgs with their brand packaging. Given that it is easier to use SVGs for other purposes I'm pretty sure most wouldn't. Der Wohltemperierte Fuchs^(talk) 02:15, 18 July 2011 (UTC)

And unneeded detail is the crux of that NFCC criteria, not resolution per se. (And yes, companies often give out vector versions of the logos, though they usually want someone to sign a trademark agreement first -- they may wish the vector versions were not available here, but that is not something that trademark or copyright law would explicitly forbid. There is likely not much they can do about brandsoftheworld.com either.) Carl Lindberg (talk) 03:18, 18 July 2011 (UTC)

Trademark is not the issue here (as long as we're using logos in educational purposes). It is also not a fair-use issue in that such use of logos are ok. It is an issue with non-free content when the logo can be copyrighted. The Foundation Resolution says nothing directly about resolution, but its our NFCC that says "Low- rather than high-resolution/fidelity/bit rate is used". Now, again, I can understand that if the original medium can be grabbed from an official source in an SVG medium, that's great, but really, if we're making the exception for logos, someone is going to complain somewhere down the line that if logos get that freedom, then why not other copyrighted art that *could* be recreated as SVG (there's not much but things like Piet Mondrian's modern art or South Park come to mind). I'd rather see non-free SVG converted to an appropriate sized PNG to avoid raising any exemptions that people will want for other images. I know it's longstanding that it's ok, but this is a very good question raised that we should address.

That said, if there is some way to judge how low fidelity of the details of an SVG image that can be used as a fair metric for judging when a non-free SVG is still representative of the logo without being detailed beyond what non-SVG images at low resolution should be. --MASEM (t) 01:49, 18 July 2011 (UTC)

That is a good example where the NFCC guideline does not work as a one-size-fits-all type of thing. When you are talking about a non-free photograph, resolution is quite important -- there is an awful lot of expression present in a high-resolution photograph that we do not need to illustrate the article; people pay more for larger-resolution photographs for a reason. That part of the criteria is very good when it comes to those. When it comes to simplistic logos, high-resolution versions usually don't add all that much -- the expression present is usually not that different and there are rarely small details which become visible. When it comes to an SVG, there is no resolution, so the criteria is meaningless really. The expression we are hosting and using is almost always the same no matter what resolution we render it at. If an SVG has some intricate details which are only visible at a high resolution, then that is a fair parallel to the photograph-centered NFCC criteria, which would probably be better worded as "more detail than required". They have different parallels for non-free audio or video already; some other wording for vector is likely needed. For example, if they were non-free, File:File:DirigibleR80.svg and File:DouglasDC3Drawing.svg (say the detail around the wheels, or interior of the engine) would have unnecessary detail not visible at the resolution needed to give the user an idea of the image, and we could strip a lot out without really affecting our use. Not as easy as scaling down or downsampling, but something we should think about if possible. Some SVGs may have hidden elements meant for aiding editing which are not visible in the final display -- those would be unnecessary as well. It's certainly possible logos may have stuff like that, though my guess is that is rare, since usually the point is to make logos recognizable at small sizes so there is usually no intricate detail. Carl Lindberg (talk) 03:06, 18 July 2011 (UTC)

Our use of logos is not to capture details that can only be seen at high resolution. We use non-free logos to connect the logo to the entity it represents, and should use a size and resolution that is no larger than necessary to assure that connection. Take, for example, the logo for ING File:ING.svg. The definitive element, the lion, is clearly seen at 200px, and it is very doubtful that an average reader, after seeing the lion at that size, wouldn't be able to connect the logo again to ING in the future. There is no need for a 2000px image to show the finer work of the various detail within the lion, unless there is sourced discussion that showing these elements would be necessary for comprehension (checking the ING article, though , that's not the case). --MASEM (t) 03:13, 18 July 2011 (UTC)

[[:File:Pepsi logo 2008.svg|thumb|center|1200px|Nonsense!Prodego ^talk 01:34, 18 July 2011 (UTC)]]

Perhaps you are more of a

fan Masem? Prodego ^talk 02:06, 18 July 2011 (UTC)

No, I'm just following NFCC. The Pepsi logo you included is non-free, and we can't use it on talk pages, period. Free images like the coke one are ok. --MASEM (t) 02:22, 18 July 2011 (UTC)

So, the consensus is to allow non-free images (vector or otherwise) when they contain only the necessary amount of details that are visible at the fair-use resolution (160k pixels). Right? – ∃ Aditya 7 ¦ 05:08, 18 July 2011 (UTC)

Seems to be what may be agreeable here. One aside is that companies will often contribute their own logo, particularly if it has changed as they want to keep it current. Obviously they are granting permissions for its appearance on Wikipedia in the form that they donate, even if it falls in the the non-free category. We should encourage dialog with organizations on the topic of their logos on Wikipedia, and not just rely on mindless application of criteria that may not actually be in the best interests of copyright/trademark holders or Wikipedia. Graeme Bartlett (talk) 08:51, 18 July 2011 (UTC)

Whoops, I seem to have inadvertently poked a hornet's nest here. >_> Some things:

It bugs me that the discussion progressed to only include talk about corporate logos. You will notice I was very careful to talk about images in general and not just corporate logos—even though logos seem to be the majority of these types of cases—since sometimes this applies to things like government emblems, seals, NGO logos (like File:WWF logo.svg), etc. Some of these entities don't publish PDFs with their images in them, so it's not just a matter of "well, we're not recreating it because there's a PDF." Sometimes there is no PDF.

I find the discussion about level of detail to be, quite frankly, utterly inane. Most of the images in question are fairly simple shapes—since these are the types of drawings most appropriate for vectorization. File:Boeing-Logo.svg doesn't have any details that are visible at 1000px that aren't visible at 200px, and neither do File:Pepsi_logo_2008.svg or File:WWF logo.svg. You can't make the argument of "well, we're not including any details in the SVG so it's okay to have it." The level of detail yardstick may work well for raster images, but for vector images it just sounds to me like definitions are being seriously stretched to justify something that otherwise could not be justified.

There was a recent request on Wikipedia:Graphic_Lab/Illustration_workshop to fix File:SouthSudanCoatofArms.svg (request seems to have since been removed for using non-free images in a gallery), because people were using it instead of File:SouthSudanCoatofArms.png. Here's an example of a fairly detailed, copyrighted image. Someone got lazy and instead of doing it properly, just ran it through an auto-trace function. Not only does the resulting SVG look awful, but what is the point of its existence? Using it in its current form dilutes the South Sudan coat of arms because it's not a faithful reproduction of what it looks like, but even if someone spent an ungodly amount of time fixing this SVG, what would be the point? It's not like you could simplify it to fit with the whole "we don't include as many details in our SVGs so it's okay" business, because, once again, it would dilute/misrepresent it. And if you traced it faithfully, well, it would look just like the PNG when rendered at 300px a side!

So it seems to me like creating SVG versions of non-free images is just doing a lot of useless work. Even if it's as easy as getting a vector out of an official PDF and uploading it: why? Why provide an infinite resolution image when you could just use that PDF version to export a raster file at maximum resolution that falls under fair use and upload it. Why do we need a file with infinite resolution if it should/will never be rendered higher than a certain threshold resolution to comply with fair use standards?

Anyway, I just wanted to see what people's thoughts were, and if I was maybe missing something here, in order to formulate a personal policy of whether I would work on requests for vectorizing non-free images. Whatever the consensus ends up being here, I think my policy will be to let someone else handle this kind of work. -MissMJ (talk) 20:22, 18 July 2011 (UTC)

One important point about vector images is that they're not just about higher resolutions. They're also better for lower resolutions, or similar resolutions. For example, we might have a bitmap with a width of 300px, but display it at 200px; or some people may have preferences set to display it at 240px, or 180 px or whatever. Reducing the resolution of a bitmap in addition to the reduced resolution causes blurring -- even with good interpolation software. With a vector image, that can often substantially be avoided.

As for your emphasis on non-profit organisations like the WWF, it seems to me that they have as much interest as anyone in seeing their logo rendered as sharply and clearly as possible, the way it was created to be seen, despite a limited image size; rather than as an unnecessarily blurred and blotchy mess. Jheald (talk) 22:16, 19 July 2011 (UTC)

Most of our logos, etc. are encapsulated in infoboxes where the sizes are specifically decided beforehand, so scaling is rarely an issue for these images (and Wikipedia's PNG scaling seems quite good; JPEG isnt unfortunately.) But that doesn't override the issue of images with higher resolutions than we will ever need. There is no policy issue regarding scaling down. Der Wohltemperierte Fuchs^(talk) 18:56, 21 July 2011 (UTC)

Rational for emailed photoshopped images

There are tons of photoshopped images of no monetary value circulating on the internet. Occasionally one of them will become politically, historically, or culturally notable, such as this one, which was emailed by a government official with a comment tying it to the birther conspiracy. I would expect that such things happen often enough that we might want a specific licensing rational for them. Any thoughts? — kwami (talk) 20:26, 20 July 2011 (UTC)

You want a boilerplate rationale for photoshopped images? Album covers for album articles, yes. For this? Each image would need a different sort of rationale. Unless, we had a boilerplate rationale that said "We believe this photoshopped image qualifies as fair use because it's photoshopped". Hmm. --Hammersoft (talk) 13:39, 21 July 2011 (UTC)

Bibliography different from discography?

I have removed nonfree images at Terence Reese bibliography, explaining that, in my view, the article was similar to a discography and thus the images were not justifiable. The article's primary creator has reverted. Another few sets of eyes/opinions would be welcome. (ESkog)^(Talk) 20:34, 20 July 2011 (UTC)

I agree with your analysis. That said, I believe it would be better to allow images in discographies and bibliographies when, as here. the entries do not have individual articles. Hullaballoo Wolfowitz (talk) 20:57, 20 July 2011 (UTC)

All other points aside, I agree that there's no reason to treat a bibliography different than a discography. postdlf (talk) 22:05, 20 July 2011 (UTC)

The reason we allow covers in articles on individual books and albums is that because 1) the work is sufficiently notable to have an article and ergo it will have commentary on the work itself and 2) the cover then carries implicit marketing and branding of the item in question as to allow it to meet NFCC#8. #1 never exists in the case of a bibliography or discography and thus there's no way NFCC#8 is going to be met. --MASEM (t) 23:48, 20 July 2011 (UTC)

I thought the main rationale for not using non-free images in discographies and the like was that we didn't reuse such images when a link to an article that already uses them is present. That wouldn't apply to entries that lacked separate articles. The identification value is the same in such list entries, and in many instances the entries may be no less lengthy or substantive than a standlone book or album stub/beginning article would be. postdlf (talk) 18:24, 21 July 2011 (UTC)

So the idea is that if something is mentioned it's ok to use a non-free image of it, so long as the one place where it is used is either an article on it or if an article doesn't exist a list that lists the thing? --Hammersoft (talk) 18:36, 21 July 2011 (UTC)

No, that's not what I said nor the idea I communicated. Please take some more time to read and comprehend what I write before responding. postdlf (talk) 18:41, 21 July 2011 (UTC)

I apologize for responding in haste without reading anything you said or taking the opportunity to comprehend. Anyway, my point is I'm trying to elicit from you what your stance in. You've, to date, been evasive about that. It would appear to me from your posts so far that if (a) there is no article on the subject and (b) the subject is referenced in a -ography, that it is ok to include a non-free image of it on the -ography. You're now saying this is incorrect. Ok, so what is your stance? --Hammersoft (talk) 18:45, 21 July 2011 (UTC)

I'm trying to develop one, actually. I made three main points: 1) the prohibition on using nonfree images where an article is linked that uses the image doesn't apply to entries in -ographies that lack separate articles; 2) the identification value of cover art is the same in -ographies as in separate articles for those listed works; 3) -ography entries can be just as substantial as standalone articles on the kind of work listed. My only conclusion, if those points are sound, is that there should at least not be a complete prohibition on nonfree cover art for identification in -ographies of items that lack separate articles. The third point might be the rub, however; if it's substantial enough for a standalone article, then why wouldn't it be split off into one? I don't think anything in Terence Reese bibliography qualifies, for example. postdlf (talk) 19:00, 21 July 2011 (UTC)

(the rub) Yeah, that's been a point of contention before. Where it's become most problematic is when people vote to merge a series of articles into one list or -ography type article, and include the images. I think, as you do (correct me if I'm wrong), that if a section of an article where that section covers a specific sub-entity of the article is substantial enough, and well sourced enough to support an article, it should. Period. If that's the case, then no argument sustains to include the non-free image that depicts it until the stand alone article exists. Thoughts? --Hammersoft (talk) 19:05, 21 July 2011 (UTC)
If you are just iterating elements in a list and providing nothing more than primary data about them (say, for a discography, the year, songlist, producer, recording studio, etc.) and that's all you can say about that element because of the lack of secondary information, then there is no allowance for a cover because we would never allow that album as an article in a first place. The use of cover art is intrinsically tied to notability of a work due to the fact that it presumes that part of the creation, marketing, and reception of the work is carried by the branging and marking of the cover. No chance of an article == no chance of a cover image. Of course, the other bound is that when we do have articles on these elements, the cover art can be found there and there's no need to use it in the list. This situation avoids any problems with the possible mix match of notable and non-notable elements in the same list whereby, where we could justify the covers for notable ones, the non-notables would be denied this - instead, we just don't allow any. --MASEM (t) 22:29, 21 July 2011 (UTC)

All reasonable observations. postdlf (talk) 17:44, 22 July 2011 (UTC)

100 Years

I just removed 100 Years...100 Movies from the examples of what not to do, since a discussion on the article's talk page established that AFI considers these lists public domain, merely requiring proper credit. --SarekOfVulcan (talk) 17:53, 21 July 2011 (UTC)

Police Service of Northern Ireland

File:Img psnibadge.png (edit | talk | history | links | watch | logs)
File:Flag of the PSNI.png (edit | talk | history | links | watch | logs)

Police Service of Northern Ireland has two nearly identical Non-free files in it's info-box, one (File:Img psnibadge.png) without a FUR for the page ? My question is : Other than the FUR is it appropriate for the article to have the two files ? Mt king ^(edits) 09:26, 23 July 2011 (UTC)

Looks like the top one is the Police Service badge, and the bottom one is the Police Service flag, which (lo and behold) consists of nothing more than the badge on a field of green, so basically two copies of the same thing. I would think it would be sufficient to just use the flag, and note in the image caption that the flag bears the image of the badge. When you discussed it with the article's editors, what did they say? postdlf (talk) 15:02, 23 July 2011 (UTC)

I wanted advice here as to what was the correct course of action before doing anything. Mt king ^(edits) 22:18, 23 July 2011 (UTC)

Is a cast photo replaceable by a composite image of free actors' images?

I found File:Parks and recreation season 3 cast.jpg on the list of orphaned fair use images and I added it back to the article of Characters of Parks and Recreation replacing the new image of File:Parks and Recreation characters.jpg. There is a discussion at Talk:Characters of Parks and Recreation#Image where I feel it has not been explained how a cast image is replaceable by one image that is a combination of free pictures of the actors given that numerous character lists use the cast photos and WP:NFLISTS suggests using them as opposed to individual images. Also, if this image was replaceable, three of the nine cast members do not have free images, so would the combination image be a replacement when some of the elements are not represented in the new image? I would appreciate anyone's opinions about these images either here or there. Aspects (talk) 21:41, 23 July 2011 (UTC)

So what are we supposed to do?

After months of kerfluffle regarding numismatics articles, we're now effectively back to where we started. I know the supporters of the use of these images would say we're right back to where we should be. Please, bear with me. I'm not trying to start another war.

Allow me to use a case example. As of now, Commemorative coins of Poland: 2008 is the #1 user of non-free content by article on the entire project. See this report (note: #1 is inaccurate, as those files have recently been deemed free license for our purposes). This article uses 44 non-free images. I note several problems with this article and the images it uses;

There is 44kb of text in the article. Do we really need 44 images to support 44kb of text? There's no policy or guideline on this of course, but there is no featured article on the project (based on sampling of WP:FA) with anywhere near this 1:1 ratio of non-free content usage.
22 of the 44 images have dimensions exceeded 1000 pixels. Per this template noting NFCC policy, the sizes need to be reduced.
17 of 18 of the external references used link to the Polish Mint (3) or the National Bank of Poland (14). These are primary sources. Per that policy, "Do not base articles entirely on primary sources". Sure, 1 of the 18 is not a primary source, but the article is effectively based on primary sources.
The only non-primary source is the Polish Numismatics Society. The only place that is used is to support assertions of the market value of five coins. Further, a numismatics society is a tertiary source. There's no secondary sourcing for this article at all.
All of the uses of images violate WP:NFTABLE, which discourages the use of non-free content in tabular format. The problem here is that there is no critical commentary of the designs in any respect. The entire meat of the article is comprised of descriptive summations of the 44 designs.
All of the images have a fair use rationale. To some, that would appear that it meets WP:NFCC #10c. But, on closer inspection every single "purpose of use" stated for the images is "Currency picture". A few have additional text (below the rationale box) of "Demonstrative of the layout of Polish currency". If that is a valid reason for use, then there would be absolutely no reason why we can not use non-free images for every single currency item in the world. Sure, you could update the rationale to something lengthy and sounding good, but in practice the only thing these images are being used for is displaying the coins. That's it.
Per WP:NOTGALLERY, Wikipedia is not a mere collection of photographs. But, that's what we have here, plus technical details. There's no encyclopedic content.
Per WP:NOTEVERYTHING, Wikipedia is not a complete exposition of all possible details. Yet, that's what we have here. There's no summary of accepted knowledge regarding these coins. It's just a reposting of technical details from primary sources. If I were to create an article on a company listed on a stock exchange that had the name, logo, and technical details of the company and nothing else it would likely be deleted. But, this doesn't?
Per WP:INDISCRIMINATE, Wikipedia is not an indiscriminate collection of information. But, that's what we have here; an indiscriminate collection of every technical detail of each of these coins.

As it stands, I could take a number of actions. I could;

Place the article for AfD. It would be kept, even though it's based entirely on primary sources which is prohibited by policy (the one tertiary source not withstanding).
Remove the images as violating WP:NFCC #8 and #3a. They would be restored by several people who insist it is appropriate use. This already happened.
Tag all the images as disputed fair use because the rationales are woefully weak, and without changing the way in which they are used in the article they will remain woefully weak. All those taggings would of course be undone, and I'd probably be accused of a WP:POINT violation. There might be an effort to lengthen the purposes of use, and almost certainly without changing the nature of use, leaving the rationales still woefully weak.
Rewrite the article to be more of an article rather than a reposting of technical details of these coins. It would be reverted for not following Wikipedia:WikiProject_Numismatics/Style/Currency_article#Coins.