After some recent FFD DRV issues, one aspect that came up was low participation in these. I believe part of this is based on the current advice where, while the file page and uploader notification are mandatory, the notification on article pages that use the image is optional. Given that we are likely dealing with images where the uploader has long and gone, this basically means these notices are falling on deaf ears.

Making the article talk page notification mandatory would at least avoid anyone later saying "but I never knew the image was up for deletion" if they were using it for their article. It would also encourage more participation at FFD which, I believe, many closers would appreciate. (Obviously, orphaned files would not need this). It might be more work on the nominator, but it would help in the long run. --MASEM (t) 16:59, 12 December 2012 (UTC)[reply]

You're spot on, Masem. We already notify AfDs at the article page; that followers of the article don't see the nomination for the included images is a mistake caused by a technical limitation. I wholeheartedly support this rule. The best way to get it working would be a Bot that created automatic notices at all articles where the image is used, so the regulars at FfD don't have any extra work to comply. Diego (talk) 19:10, 12 December 2012 (UTC)[reply]

Definitely should be done by a bot, or/and Twinkle. I absolutely refuse to do it manually. There can be no excuse for imposing yet more obligatory paperwork on nominators, in a matter that should be as easily automatized as this. Fut.Perf. ☼ 17:10, 20 December 2012 (UTC)[reply]

I definitely think we need a bot to do that, which shouldn't be too hard (just has to watch the current FFD page). --MASEM (t) 17:13, 20 December 2012 (UTC)[reply]

I agree that notification on articles, rather than user talk pages, is more appropriate. And I think a bot to do it automatically would be wonderful, not least because the bot's messages would be neutrally worded.

At the same time, we would need to remind nominators that it is inappropriate to orphan an image and then nominate it for FFD, and closers who notice such things should default to keep rather than delete. — Carl (CBM · talk) 17:05, 20 December 2012 (UTC)[reply]

No, orphaning an inappropriate image is absolutely not inappropriate in any way. Never has been, never will be. It is a simple application of WP:BOLD and will therefore always remain legitimate. Fut.Perf. ☼ 17:10, 20 December 2012 (UTC)[reply]

Orphaning an inappropriate image is fine. Nominating an image for deletion is fine. But orphaning it and *then* nominating it immediately for deletion is misleading, because it prejudges that the image was inappropriate, which is exactly what the FFD is supposed to determine. If an editor is nominating an image for deletion they should leave it alone so that the discussion can accurately consider the way the image is being used. If the consensus finds that the image is inappropriate, the image will be deleted, after all, so a short delay is fine. The same holds for categories and templates; it distorts the deletion process to orphan something just before it is nominated for deletion, and is not a legitimate application of "bold editing". — Carl (CBM · talk) 17:15, 20 December 2012 (UTC)[reply]

I'd have to agree that it is bad form to orphan an image and put it to FFD where context can't be evaluated. Only if the image fails one of the CSDs for files (obvious copyvio, obvious free replacement) should the image be removed before further discussion, if that is needed. I do think that if a bot can review the articles that an image was used on in the past 24 hrs, then this would still catch those cases and notify editors on those articles, but this still wouldn't make the practice right. --MASEM (t) 17:20, 20 December 2012 (UTC)[reply]

I suppose we are all primarily thinking of non-free images here, right? (Free images that are not orphaned are very rarely brought to FFD in the first place). If it's a non-free image, it shouldn't be necessary for an FFD voter to even look at the article at all – the FUR must tell them which articles it was used in, and the FUR ought to tell them enough to understand how it was used and for what purpose. If the FUR doesn't give you enough information to judge the appropriateness of an image, then the FUR is bad and the image ought to be deleted for that reason alone. Apart from that, everybody knows how to use the article history to find out what context an image was used in prior to orphaning, if the need should arise. Fut.Perf. ☼ 17:24, 20 December 2012 (UTC)[reply]

A badly written NFUR is not a reason for deletion, it's a reason to fix the NFUR. A reason for deletion is that no sound FUR could be written - and that is something that has to be decided by discussion. Particularly if there was no NFUR at all, or a very bad one, if the image is orphaned it can be very hard to tell where it was used, and thus hard to tell whether a sound NFUR could be written. A good nomination would list the pages where the image was used, but if the nomination only says "No NFUR, replaceable" then it is very hard to see where the image might be applicable. — Carl (CBM · talk) 17:27, 20 December 2012 (UTC)[reply]

If the FUR is totally absent or so deficient that it doesn't even name the articles it is for, then the image is unlikely to come up at FFD at all; it will usually be just tagged for the speedy queues. Those that do end up at FFD are typically those that have a formally "valid" FUR template and all. They are very often totally deficient content-wise, but they do tell you what article they are for. Fut.Perf. ☼ 17:32, 20 December 2012 (UTC)[reply]

My real concern with orphaning is that I have seen it used as a way to avoid scrutiny. If we have a bot that starts placing notices on pages that use an image, allowing a nominator to orphan the image first allows the nominator to bypass that notification, which is counterproductive for getting full participation in the FFD. — Carl (CBM · talk) 17:28, 20 December 2012 (UTC)[reply]

@Masem: Is there some way I do not know to tell where an image was previously used? As far as I know (and I have done a lot of programming with bots and the toolserver database) there is no record of pages that previously used an image, category, or template, only a list of pages that currently use it. If the image is a CSD, I agree that it is fine to orphan and delete it immediately, but in that case there is no reason to open a FFD. — Carl (CBM · talk) 17:27, 20 December 2012 (UTC)[reply]

No, I don't think there is, short of going through the edit histories of the articles in question. Fut.Perf. ☼ 17:34, 20 December 2012 (UTC)[reply]

Hmm, I thought there was, but I guess not. There technically should be a #10c-meeting rationale on the page, and if it uses one of the templates we can ID the article it was used it ; it would be more difficult for non-template rationales since we don't require linkage to the article of use. --MASEM (t) 18:12, 20 December 2012 (UTC)[reply]

I've sometimes detected those articles by going through the file history and noticing that some FURs have been deleted, which included the articles where the image was used. Diego (talk) 18:32, 20 December 2012 (UTC)[reply]

Given that there is no easy way for an automated tool to track the original articles that an orphaned image/file was used in, I wonder if we need to consider more logic and process here. We could make it an onus on the person that orphans an media file and puts it to FFD to make sure all talk pages of all affected articles are notified of this discussion (in cases where the files remain but tagged for FFD, a bot could do this). Otherwise, we restrict when files can be orphaned (never before they are taken to FFD, and in the case of a di- tag, only after 7 days after the tag). If we make the act of "orphan then FFD" an improper process, then all we need to have more FFD input is to have a bot that notifies articles using the file of the issue - and this can be done for both FFD files and those tagged with di- problems. --MASEM (t) 04:28, 21 December 2012 (UTC)[reply]

I think the bot is useful either way, since not all the images will be orphaned. The bot could also post a note on the FFD to say where the announcements were placed, or to say that the image is orphaned; that would help the people in the discussion have better context without having to look up where the image is used. — Carl (CBM · talk) 14:10, 21 December 2012 (UTC)[reply]

Wait a second, people. Why are we talking about people first orphaning and then FFD'ing an image? Does anybody ever do that? Why would anybody ever do that? This seems quite academic to me. If you have orphaned an image, then you don't put it on FFD; you put it in the {{orfud}} queue. The only situation where I can think of a reason I would subsequently put something on FFD is if other editors had in the meantime insisted on putting the image back into the articles. On the other hand, the suggestion to prohibit orphaning outright is a total non-starter – making it significantly more difficult procedurally to remove a bad image than to add it would mean turning both WP:BOLD and WP:NFC upside down. To have a rule that effectively says you can never remove a non-free item without prior discussion would only be legitimate if there also was a corresponding rule that you can never add one without such discussion. Now, that might be a good rule to have indeed. Fut.Perf. ☼ 08:42, 22 December 2012 (UTC)[reply]

You bet there's people doing exactly that. Images with perfect Purpose rationales will have them deleted with comments like "fails NFCC#8" or "can be explained with words", removed from the articles where they're used, and then tagged as orphan. That's why requiring a discussion at least for the cases where the status of the image is changed for worse is mandatory. This shouldn't affect images that have never had valid rationales or that have not been used in articles; only when one single editor is judging the image all by himself. Diego (talk) 12:14, 22 December 2012 (UTC)[reply]

You evidently didn't understand what I was saying. Of course people can orphan images and then tag them as being orphaned; that's what the orphan tag is for. This has always been legitimate, and always will be. What I'm asking is: why would anybody first orphan an image and then take it to FFD? Fut.Perf. ☼ 12:56, 22 December 2012 (UTC)[reply]

Ok then, to answer your original question see here that yes, there's people placing orphaned images at FfD, so this is not an academic debate. At the very least it provides a centralized record of what has been deleted for this reason. It's a good practice to review it and find out which images had a legitimate encyclopedic purpose before being nominated, and which ones are simply vanity photos not used in any article. Diego (talk) 15:16, 22 December 2012 (UTC)[reply]

Those are a totally different kettle of fish. Almost all images that get taken to FFD for being orphaned are low-quality free files that have been sitting around unused for a long period. People find these unused and nominate them for being not useful. It would be very unusual for somebody to find such a free file in use in an article, orphan it themselves and then nominate it at FFD. But we here were talking about non-free files, were we not? Fut.Perf. ☼ 16:12, 22 December 2012 (UTC)[reply]

@Fut.Perf: "Why are we talking about people first orphaning and then FFD'ing an image? Does anybody ever do that?" Indeed; and the last person who did that and was spotted ended up desysopped, at ArbCom and with several discretionary sanctions. Thats why it is discouraged to orphan an image and then bring it to FFD.

@Masem This is a very good idea, but we need to establish which articles would the message be left on. My bet is to make the bot leave it in the talk of articles into which the image is currently placed, and report back in the case it is orphaned.

— ΛΧΣ²¹ 02:04, 31 December 2012 (UTC)[reply]

"...the last person who did that..."? Refs or it didn't happen. Fut.Perf. ☼ 21:17, 1 January 2013 (UTC)[reply]

As a recent event, I guess that everyone new about it (the ArbCom case was just opened), but here's the ref: [1] — ΛΧΣ²¹ 22:04, 1 January 2013 (UTC)[reply]

And where are the diffs showing SchuminWeb first orphaning and then FFD'ing non-free images? (not to mention: where are the diffs showing that Arbcom considered those actions wrong, if indeed he did so?) Fut.Perf. ☼ 22:50, 1 January 2013 (UTC)[reply]

Ugh. The diffs date back to 2007 until 2011. I am too lazy to find them but if I remember well, some threat at ANI must have them. And well, ArbCom accepting the case and passing a motion is enough grounds for it; although the issue there went beyond, because he not only orphaned them and then took them to FFD, but sometimes even deleted them himself. But that's kind of irrelevant. What we should know is that orphaning then taking to FFD is disruptive and should not be done. I am against excessive use of non-free images because they represent a danger to the site (and the foundation), but if somebody orphans an image right before it's taken to FFD, then some context will be lost when assessing the information. — ΛΧΣ²¹ 00:52, 2 January 2013 (UTC)[reply]

I browsed not only through the Arbcom page but also to the preceding RfC/U. I do not find any reference there to any claims that somebody first orphaned an image and then immediately took it to FFD. Diffs or it didn't happen. I very much doubt it did. This thread is pointless as long as people keep muddying the waters with confused claims about what the issue is and mixing up one process with another all the time. You are now the second person to do so. Fut.Perf. ☼ 08:02, 2 January 2013 (UTC)[reply]

If we consider the process of first removing an image from a page and then nominating immediately at FFD is "out of process" and improper behavior, then we easily can get a bot to be made that tags FFDs that are non-orphans to drop messages on all talk pages for articles it is presently being used in, making this "mandatory" aspect trivial to do. (Note that I can see cases of where Editor A removes the image, Editor B readds, and then Editor A takes it to FFD to get more input - that's ok. It's if Editor A removes the image, and then there's no intermediate edits that restore the image, before Editor A FFD's it.) --MASEM (t) 21:27, 1 January 2013 (UTC)[reply]

Another thing to keep in mind: partial orphaning immediately prior to FFD will always remain legitimate in cases where a non-free image is used on multiple pages and some of these uses are blatantly and incontrovertibly wrong, especially where some of them lack a FUR. In such cases, I do indeed remove the image from those obviously wrong pages and then use FFD for the remaining ones where there is at least a formally valid attempt at a justification. I do not accept mandatory notification for those obvious cases, be it automated or manual – if the persons who put the images in couldn't even bother to write a rationale, why should I bother notifying them? Fut.Perf. ☼ 21:52, 1 January 2013 (UTC)[reply]

This is the case I am a bit worried about, not saying the approach is wrong. Clearly images failing #10c need to be resolved. We are trying to recreate a bot to do that as well, and like this, that notification should fall on the pages using the image w/o an obvious rationale; and then of course automatic removal from such pages after sufficient time should happen. So this is fine. But I can't think of any other claimed "automatic" removals at the present time (short of #9 problems of non-main space usage) that this would matter. (Eg, if an article is a discography and using lots of image covers, that's not an automatic removal allowance even though that it outlined as a problem - it's better to handle that by greasing the wheels than fighting it.) So as long as thats the only route that you're removing the images, that would be fine. But I would like to get this #10c bot set up as well in addition to this one to help. --MASEM (t) 00:20, 2 January 2013 (UTC)[reply]

A discussion related to the "orphaning before FFD" thread above has just started at WP:VPR#Restricting people from orphaning non-free images, unless free replacement is found or person still lives?. Anomie⚔ 13:08, 2 January 2013 (UTC)[reply]

Comment from an uninvolved user: After reding the proposal & discussion, I think this is a very good idea. "...to make the bot leave it in the talk of articles into which the image is currently placed" > favorite solution. Jesus Presley (talk) 21:45, 2 January 2013 (UTC)[reply]

• Where images nominated at FFD are currently used in an article, then I absolutely support making it mandatory to inform editors of those articles by way of a message on the talk page and/or tagging of the image caption. I have no issues with a bot doing this, but I do not consider it an undue burden to require it to be done manually (I'm pretty certain that there are various way this could be set up to require just a single button press). If it is possible, then I equally support linking to the discussion from the talk pages of articles that formerly used images - I can't conceive of any harm this could do. Thryduulf (talk) 22:30, 6 January 2013 (UTC)[reply]
  • The problem with the last point is that there is no way - unless clues are left on the image file - of where the image was formerly used before without a broad search of every article's history - its effectively impossible for an average user and even for bots. This is why we have to understand the mechanisms of removing an image so that we can identify that there are reasons to remove an image without having to FFD it (either objectively failing NFCC, or as a first BOLD edit in the BRD chain). And we can only make it as "mandatory" as any other notification requirements - which (to my surprise) aren't mandatory (you only have to tag the article or file, not anyone else). --MASEM (t) 22:38, 6 January 2013 (UTC)[reply]
• Comment ISTR seeing instances where a bot added a note to a caption in the relevant article, but I can't find it now. The notice was much less prominent than the AfD header, but it did something. Am I imagining things/dreaming? -- Trevj (talk) 14:34, 16 January 2013 (UTC)[reply]
  • I think this used to happen, but now a file that is at FFD has a notification automatically transcluded about the discussion - which means nothing to the user just watching the article itself since this doesn't ping the watchlist. --MASEM (t) 14:37, 16 January 2013 (UTC)[reply]

Was it {{Deletable image-caption}}? If a bot could do that, then it could place a more prominent header at the top of the mainspace page instead. -- Trevj (talk) 14:39, 16 January 2013 (UTC)[reply]

It looks as if these bot approvals may be relevant:

• Sambot 12
• Chris G Bot (6th request)
• SoxBot 19

-- Trevj (talk) 14:43, 16 January 2013 (UTC)[reply]

• Because I do, and used to do more, image tagging for various copyright issues I recall one of the scripts, which was triggered by Twinkle, or maybe it was Twinkle itself, automatically adding the {{Deletable image-caption}} tag to images in articles but it broke at some time a few years ago and has not been fixed or maybe could not be fixed at that time. I don't recall the details but seem to remember at that time it broke the actual image caption leaving neither the caption nor the notice being displaying. Maybe someone knows a better way now. It was and still is a great idea to tag the deletable images in the articles as that notifies the article watchers who may still have it listed. There will almost certainly be more that the on uploader who may be long gone so tagging them is often totally ineffective. Some article watcher may still respond or try to do something about it. ww2censor (talk) 23:55, 19 January 2013 (UTC)[reply]
  • For a bot to add stuff to image captions in articles is essentially impossible, because so many images are used in infoboxes and similar templates, and there are just so many different ways images and their captions are encoded in those. Some boxes have a simple "|image=[[File:...|caption]]" syntax, some have captions automatically inserted by the template, some have separate "|caption=..." parameters; the parameter names are in no way standardized; so it's basically impossible for a bot to figure out where to add such a note. Fut.Perf. ☼ 22:17, 20 January 2013 (UTC)[reply]
    • A few rules would seem simple enough to me: If the image has a caption in the "|caption=..." format, then append the template to that caption; if it is not inside a template and doesn't have a caption then create a caption in this format with the template. If the image is used with an "image=", "imagen=" or "image n" and is followed by a "caption=", "captionn=" or "caption n=" then append the template to the end of the caption parameter or create the caption in the right format if this is known. If neither of these are possible, then post the message to the talk page. Other common formats can of course be added when known. Thryduulf (talk) 23:29, 20 January 2013 (UTC)[reply]
      • Except that there is no reliable standard that "caption=" parameters will actually be called "caption=", nor that "image" parameters will be called "image=" (they might be "img=", "map=", "cover=" or any other arbitrary string), nor that the naming of a "caption" parameter will match that of the corresponding "image" parameter in any predictable way. Think of a monster like {{Infobox settlement}}, which has up to six different slots where images can be placed. The image parameters are called "image_skyline=", "image_flag=", "image_seal=", "image_shield=", "image_map=" and "pushpin_map=". Now try writing a script that automatically figures out that the corresponding caption parameters are "image_caption=", "map_caption=" and "pushpin_map_caption=" and that the remaining three slots don't have any caption parameters at all. Plus think of the fact that in many instances, even if you find the right syntax for the caption, adding any notice to it will horribly break the layout of the box. It's just not worth the trouble – much better to go to the talkpage straight away. Fut.Perf. ☼ 23:48, 20 January 2013 (UTC)[reply]
        • Which is a really good argument for requiring human editors to add image captions. Thryduulf (talk) 00:20, 21 January 2013 (UTC)[reply]
          • Nope, it's an argument for using the talkpage instead. As a human editor, too, I'm not going to learn the documentations of a few dozen infobox templates by heart before I start nominating images for deletion. But in any case, as I said earlier, I will emphatically refuse to do any manual extra work on notifications, no matter where and how. Fut.Perf. ☼ 00:30, 21 January 2013 (UTC)[reply]
            • And I will repeat that I consider that refusal an entirely unreasonable attitude contrary to the spirit of Wikipedia. Thryduulf (talk) 01:06, 21 January 2013 (UTC)[reply]
              • Shrug. Tell that to the people who fill Wikipedia with crappy uploads. I do a lot of work cleaning up image uploads, and frankly, eighty percent of the work I have to do in this area is caused by people who are either lazy, illiterate, irresponsible, or downright malicious. It's bad enough there is as much work to do about bad uploads as there is; any additional mouseclick to do is one too many. Fut.Perf. ☼ 06:54, 21 January 2013 (UTC)[reply]