Wikipedia talk:Files for deletion/Unverified orphans

Much of the following discussion was moved here from Wikipedia talk:Images and media for deletion.

Hasty Deletion, Orphans
I'd like some clarification of certain deletion practices. The following classes of images should not, in general, be deleted:
 * Orphans which seem to be free-content
 * Untagged images which are likely to have been taken by the uploading user

Orphans are often useful content which have been removed from their original article for one reason or another. Untagged images are often uploaded by the creator, who naturally think that the copyright-warnings on upload are equivalent to them declaring their uploaded content to be GFDL. +sj +  04:31, 14 Mar 2005 (UTC)

Similarly, the general practice of informing a user about IfD as the image is being put up for deletion, in combination with a fairly quick deletion process, is unfortunate. Many users don't log onto WP every week. A separate deletion-page with a longer wait, for images waiting on a response from the creator, might be in order. +sj +


 * I disagree, and I know others do as well. I'll start a poll below all this. – Quadell (talk) (sleuth) 16:29, Mar 14, 2005 (UTC)
 * Correction: Actually, I think a separate page with a longer wait for unverified orphans might be a good idea afterall. – Quadell (talk) (sleuth) 18:27, Mar 14, 2005 (UTC)

Deleting excellent images
Re: the deletion of 'unverified orphans' :
 * "I've been doing this for about a month, and it's been generally well received." -- Quadell

Please work twice as hard to check the images you list, list 25% fewer, and wait a few weeks after notifying people on their talk pages before deleting on-the-fence contributions.

In skimming over the currently-listed IfD entries, it looks as though about 1 in 5 of these bulk-listed unverified orphans are, in fact, Moreover, some of these are truly excellent images, not simply fuzzy snapshots which have already been replaced with something better. That the Wuerzburg panorama was listed on IfD is truly scandalous.
 * contributed by their creator, or free-content in some other fashion
 * useful and encyclopedic, even if not currently used in an article
 * It would be, except that it's a copyright violation. – Quadell (talk) (sleuth) 16:21, Mar 14, 2005 (UTC)
 * You've got to be kidding. That's a small, grey, fuzzy image; the other is large, sharp, and vibrant with light and shadow.  No need to be defensive; just be more careful in your work.  +sj  +  05:45, 15 Mar 2005 (UTC)
 * That's Google's scaled-down cached version. Look close. It's the same image, dude. – Quadell (talk) (sleuth) 16:33, Mar 15, 2005 (UTC)
 * They are different images. They are taken at a different time of the year for a start (look at the trees) -- sannse (talk) 18:39, 15 Mar 2005 (UTC)
 * They, um. . . hm. . . Well, that's embarrassing. – Quadell (talk) (sleuth) 14:24, Mar 16, 2005 (UTC)

Due to the problems above, and efforts to tag images quickly, a number of good images have no doubt already been deleted. Here is an abbreviated list of images on the current IfD page that I could see with a quick scan should probably be kept -- at least until the original uploader can come along and help place in an article. Being an orphan does not make an image useless.

Please do not irreversibly remove someone else's contributions, simply to "reduce the size of the database" or to save yourself a few minutes of tracking down its attribution.

Not quite orphaned
Image:XcutMOSFET.png (used in talk pages, recently cut out of an article)
 * It was cut out of an article because it wasn't useful there. – Quadell (talk) (sleuth) 16:25, Mar 14, 2005 (UTC)
 * There was some question about that. It is explicitly not the job of someone cleaning up images to make the subject-level decision of whether or not an image is useful on a particular page.  If there are well-meaning contributors who feel that it is useful, and there is a discussion about that, deleting the image does nobody any good. +sj  +  05:45, 15 Mar 2005 (UTC)
 * I don't believe there are any well-meaning contributors who feel it's useful. I do think it's my job, as an editor, to make decisions about whether an image is useful. The IFD page is for these sorts of discussions.

License info available
via the user page of the contributor, or info in the caption
 * Image:Zilverr.jpg
 * What license do you think this is available under? I don't see one. – Quadell (talk) (sleuth) 16:27, Mar 14, 2005 (UTC)
 * Looking on the nl: page as suggested (replacing en with nl in the base URL), turns up the same image with "open bron/ parkservice" in the description. Sounds like a class of licensed photos to me.  Luckily, I don't have to know, since someone asked the image uploader, GerardM, about it back in December.  He clarifies (from his talk page): "The license is that it can be used by wikimedia projects as long as the original photographer is recognised. (see nl:wikipedia. GerardM 18:42, 14 Dec 2004 (UTC)" +sj  +
 * "open bron/ parkservice" sounds like useful license info to you? What country's park service? I don't get much insight from that. As for the user's talk page, is that Permission (which would mean that, according to Jimbo, it will be deleted at some point)? Or is it okay for any user so long as credit is given?
 * Image:Irontriangle.png
 * However, this may well be OB
 * This is, in fact, Obsoleted. – Quadell (talk) (sleuth) 16:27, Mar 14, 2005 (UTC)

Clearly states that it was contributed by the [active] creator

 * Image:Memorial_Hall_Spring.jpg
 * Image:Single-channel_Bell_test.jpg
 * You're right, but they're orphaned. The creator has been notified. If the [active] creator doesn't want them deleted, he can tag them. In fact, I always check to see if the image is still unverified and still an orphan before deleting. If you think it's useful, simply tag it and put it in an article. I personally don't believe either of these is useful to Wikipedia. – Quadell (talk) (sleuth) 16:48, Mar 14, 2005 (UTC)

Style & caption suggests it was uploaded by the [active] creator

 * For these. Could be. Or maybe not. No evidence either way. Again, if they're useful, then put them into an article and they won't be deleted.
 * Image:Narrow_lanes_2.jpg
 * Image:Orsistersdowntown2.jpg
 * Image:Wuerzburg_panorama.jpg
 * A *glorious* photo, with hi-res version. Orphaned just last month by an an anonymous edit.
 * And a copyright violation, as noted above. – Quadell (talk) (sleuth) 16:48, Mar 14, 2005 (UTC)
 * Seems unlikely, as noted above. +sj +


 * Image:Bushland.jpg
 * An ancient photo, which had been taken by one long-time contributor and sharpened by another.
 * The uploaded told me he got it off the Internet. It's been obsoleted. – Quadell (talk) (sleuth) 16:48, Mar 14, 2005 (UTC)
 * Okay. As long as you checked before listing it on IfD.
 * I did. I work hard at this. Assume good faith. – Quadell (talk) (sleuth) 16:33, Mar 15, 2005 (UTC)


 * Image:Xi%27an_city_wall.jpg
 * With hi-res version and detailed caption.
 * True. I wish we knew whether it was usable on Wikipedia or not. – Quadell (talk) (sleuth) 16:48, Mar 14, 2005 (UTC)
 * Me too. Note how we're not deleting images like the autofellation picture, which is far more likely to be a copyvio?  At least this one could reasonably have been a vacation photo. +sj  +  05:45, 15 Mar 2005 (UTC)
 * But we are listing that image, and having a discussion about it. You're arguing (if I understand you correctly) that they shouldn't be listed. Right?


 * Image:Zacazonapan.gif
 * With hi-res version.


 * Image:Img_0204.jpg
 * With hi-res version.
 * Image:Img_5910.med.jpg
 * Image:Imgp0340.jpg
 * Image:FranceLiberate.jpg

Style suggests it was contributed by the creator

 * Image:Kartal.jpg
 * Image:Parliamenthouse5.jpg
 * Image:NymphaeaRubraRoxb.Asia.jpg
 * Image:Pavillion.jpg
 * Image:Rahejatowers1.jpg
 * Image:UMass_campus.jpg
 * Image:Mcdonaldsdrivethrumenu.jpg
 * Image:Yasukuni.jpg


 * Plus a whole series of lovely images by User:Shakespeare:
 * Image:Frink_Park-2.jpg
 * Image:Frink_Park-3.jpg
 * Image:Fremont_Bridge-2.jpg
 * Image:Fremont_Bridge-1.jpg
 * This user might log on once a month... too late for his images: contribs
 * Several of these are duplicates of used images, so should be deleted. Others are simply not used. If you think they should exist on Commons, then by all means, upload them there. – Quadell (talk) (sleuth) 16:48, Mar 14, 2005 (UTC)
 * That's news to me. They're not actually duplicates.  Are you saying that "being similar to another image" is cause for outright deletion?  I can see asking the deleter to upload them to Commons instead... +sj  +  05:45, 15 Mar 2005 (UTC)

All in all, I don't see the problem. If an image isn't being used, and I'm reasonably certain of it's copyright status, then I consider whether I think it would be useful in an article or not. If it would be, I tag it and include it. I also search for it on the Internet, since I don't want to tag an image as GFDL if it's really a copyright violation. (This is quite a bit of work; note that it would be just as easy for me to find a different image on the Internet and upload it and include it.) If I don't think it would be useful, I list it on IfD as an unverified orphan and notify the uploader. When the week is up, before I delete it, I check to make sure it's still unverified and still an orphan. If not, then someone liked the image or found it useful, so I leave it alone.

The only possible problem is if a user uploads an image, fails to tag it, and does not include it in any articles, but doesn't want it deleted. The uploader would then have to not log in for the week that it was listed here.
 * More likely, a user uploads an image, assumes because of all of our references to copyright and the GFDL that s/he has already "tagged" it, adds it to a relevant article and leaves. A while later, someone else removes the link to it from that article; replacing it with something else?  Or just not liking how it formatted that part of the page.  The original user may not come back, and certainly shouldn't be expected to have to defend their contributions forever.  +sj  +

Then, if no one else thought it was useful and included it in an article, it would get deleted. Then the uploader would have to re-upload it next time he logged on.
 * Gosh, that's idealistic.
 * Not as idealistic as assuming untagged images are GFDL. – Quadell (talk) (sleuth) 16:33, Mar 15, 2005 (UTC)
 * Most uploaders are infrequent editors. They may not have a "next time" they log on.  They've forgotten their old p/w, make a new account, only read WP logged out; or no longer have their old files.  Or they log on, see that someone has deleted work of theirs, and in displeasure remove the rest of their work from WP (they may not be admins, but they can just unlink their images from articles, remove any license tags, and wait for the StreetSweeper to come around and delete them).  I can think of at least two historical examples. +sj  +

That's a small price to pay, in my opinion, for a process that has so far deleted hundreds (thousands?) of images that were only taking up space on the server.
 * This concern (about space on the server) seems somewhat out of proportion. The bulk of the database-size continues to be taken up by article histories. +sj  +  05:45, 15 Mar 2005 (UTC)
 * They've had huge problems the load on the image server in the last 2 weeks. See Image server overloaded.
 * NO NO NO. Image server overloaded does not mean image server nearly full. Plugwash 14:27, 16 Mar 2005 (UTC)
 * Right. . . Image server overload doesn't mean the image server is nealy full. I DIDN'T DIDN'T DIDN'T say it did. :) It means the image server can't handle all the requests being made on it. – Quadell (talk) (sleuth) 15:38, Mar 16, 2005 (UTC)

Reasonable people may disagree on this, but I want you to know that I'm not just including unverified orphans in a knee-jerk reaction. – Quadell (talk) (sleuth) 16:48, Mar 14, 2005 (UTC)
 * Yes, I could see that. And I'm glad you take the time to do this work, even if I don't agree about how it should be done.  +sj  +  05:45, 15 Mar 2005 (UTC)

Oh, one more thought. As I noted on the project page, when anyone uploads a file they are given the following warning: "Also, choose an image copyright tag and add it to the description page if you did not do so in the upload summary. If this is not done, and the copyright status of the file cannot otherwise be found, it may eventually be deleted." So if they are eventually deleted, I don't a great injustice there. (Yes, there are some who uploaded images before the warning existed, but they are in the minority.) – Quadell (talk) (sleuth) 18:17, Mar 14, 2005 (UTC)
 * This is why I pointed out when you were including older images in these lists. Like that panorama!  +sj  +  05:45, 15 Mar 2005 (UTC)

Poll on undeleted orphans
Poll closed. No consensus found. See next section.

Preamble
Several users have in the past few months listed "unverified orphans" to be deleted. Here are the rules of unverified orphans (UOs):
 * 1) An image is an unverified orphan if it is not used in any Wikipedia articles, and has not been given a valid image tag.
 * 2) An UO should only be listed if the submitter believes that the copyright status cannot be determined without further information from the uploader, or that the image is not useful in any Wikipedia articles.
 * 3) UO pages should be tagged ifd, and the user page of the uploader (if active) should be tagged with idw-uo.

Arguments both for and against this practice have been detailed above. It has also been suggested that a separate page be created for unverified orphans, with a longer wait. So should users list images to be deleted if they are unverified orphans, assuming the above practices are followed? And should they be listed here, or on a separated page?

Unverified orphans should be listed for deletion on IFD.

 * Note: this assumes the above rules are followed.


 * 1) – Quadell (talk) (sleuth) 18:26, Mar 14, 2005 (UTC)
 * 2)    &mdash; Zeimusu | Talk 13:08, 17 Mar 2005 (UTC)
 * 3) &mdash;Korath (Talk) 14:15, Mar 17, 2005 (UTC)

Unverified orphans should be listed for deletion on a separate page.

 * Note: this also assumes the above rules are followed.


 * 1) – Quadell (talk) (sleuth) 18:31, Mar 14, 2005 (UTC)
 * 2) - Plugwash i think we need a wait time of at least a couple of months AND the email this user for the uploader must have been tried. Thier userpage should also be checked for homepage and/or email links that may allow contact with the user. Plugwash 10:57, 15 Mar 2005 (UTC)

Unverified orphans should not be listed for deletion.

 * 1) +sj  +  10:43, 15 Mar 2005 (UTC)
 * 2) Arwel 16:04, 16 Mar 2005 (UTC)
 * 3)  &mdash; Xiong (talk) 14:03, 2005 Mar 29 (UTC)

Comments

 * Please leave comments here, and not in the voting section. – Quadell (talk) (sleuth) 18:26, Mar 14, 2005 (UTC)
 * m:Polls are evil, but I support what Quadell's been doing. dbenbenn | talk 00:41, 15 Mar 2005 (UTC)
 * Yes, polls are evil. But I'm voting while it's here... Deleting UOs is dangerous unless you...
 * {a) add an exemption for images uploaded before November 2004,
 * b) add some kind of clause regarding the quality of the images; one should be much more loathe to dispose of great images than blurry thumbnails, and
 * c) add a clause regarding getting feedback from the uploader.  A fairly regular editor, who edits once a month or so, should be given a goodly amount of time to see and respond to a request for clarification.
 * Deleting images is not in the same category as deleting articles. It is one of the few irreversible actions any admin can carry out, and so carries a significantly higher burden of care and responsibility.  Most users will not simply "upload their images again" if they were mistakenly deleted.  +sj  +  10:43, 15 Mar 2005 (UTC)
 * I'm surprised you didn't support the 2nd option, Sj. You had brought the idea up, and I'd thought it might be a workable compromise. – Quadell (talk) (sleuth) 17:55, Mar 15, 2005 (UTC)
 * I don't see any pressing need to delete orphans in any case - someone may choose to use them later. -- Arwel 16:07, 16 Mar 2005 (UTC)
 * Orphans have long been candidates for deletion, and should remain so. These images can't be used in articles. The only protection we have in keeping these images is fair use, and fair use claims without credit are weak. All these images do is pollute the namespace, sit on disk and invite law suits. The proper place for discussing media for deletion is here, not on a separate page, so I don't support that option.  &mdash; Zeimusu | Talk 13:20, 17 Mar 2005 (UTC)
 * Zeimusu said it. However, it might be politic if the nominator of an image were to avoid deleting it himself, as is (supposed to be) done for VfD. &mdash;Korath (Talk) 14:15, Mar 17, 2005 (UTC)
 * That's fine by me. The thing is, I've been deleting 99% of the images here for the last month. If I didn't delete them, there would end up being quite a backlog here. But if someone else wants to do the actual deletion, I'll gladly step aside. – Quadell (talk) (sleuth) 12:43, Mar 18, 2005 (UTC)

So what do we do about those UOs? Request for comments.
Well, the poll above did not reveal a consensus. So let's build one. Currently we're in a sort of deadlock, with the unverified orphans starting to take over the page. We could delete them, quarantine them on a separate page for a set amount of time, or leave them unverified. This is a rather important decision I think. Many of us have made our opinions clear, but now its time to try to hash out a compromise. I'm all ears. – Quadell (talk) (sleuth) 01:40, Mar 21, 2005 (UTC)


 * I was unaware of the poll while it was open. My own view is that unverified orphans should receive the lowest possible priority: there is no point investing time in digging up sources (through the Image Tagging and/or Image Sleuthing projects) if they are not being used. The best solution for isolating them would be by software. If we were to create a category along the lines of Category:Orphaned media files with unknown sources, it would require unnecessary work to update it if an orphaned image does get used (plus image categories are hard to navigate, etc.). Perhaps the easiest solution would be to use a bot to create a list of images that are either untagged or tagged as and that are currently orphaned.


 * Regarding deletion, I'm not opposed. If we have a list of unverified orphans, another bot run could post messages on the uploaders' talk pages, saying, in polite terms, "tag it or forget it". (I know, there are problems with that which need to be ironed out.) Then after a reasonble delay, they can be deleted, IMO. --MarkSweep 03:26, 21 Mar 2005 (UTC)


 * There are probably thousands of untagged orphaned images out there. Maintaining a queue and keeping track of all those requests to tag images while you wait for the reasonable delay to expire would be incredibly labour-intensive. I think first we should put a message on the image upload page warning people that untagged images will be deleted. There is one like this on commons, the one on en is significantly less harshly worded. Then, we should post a notice to the announcements page and the village pump announcing that all untagged images will be fair game for speedy deletion in some long time&mdash;3 or 6 months maybe. Since something like this would affect all of Wikipedia, we'd want to solicit opinions to ensure everyone agrees with this plan before we implement any of it. Except maybe the harsh warning on the upload page, we need that.  04:56, 21 Mar 2005 (UTC)


 * That's a good point. Even though I'm not directly involved in image deletion, I can imagine quite well that it would be an impossible task to give proper notice to uploaders, keep track of reactions, etc., which is why I initially suggested enlisting bots to do some of the work. So is the consensus now that UOs should be considered fair game? If so, then the steps described by Foobaz are very good suggestions: start by amending the upload text and bring the matter to general attention.  I'm just afraid that if we have to put this matter to a vote that requires 80% approval it will be hard to communicate the sheer magnitude of the task to everyone and that some will insist that we can only delete UOs after individual notification of uploaders. --MarkSweep 08:59, 21 Mar 2005 (UTC)

Since it was asked elsewhere, and I was not aware of the poll earlier, I think it only makes sense to delete all untagged images that are not being used within, say, a month of their being uploaded. If they aren't being used there's no point to hold onto them to risk lawsuits and take up disk space. They should probably be listed somewhere as being deleted in the next week, so that anyone who wants to look through them can tag any that they think are worth saving. If nobody cares enough to tag them then there's no reason for them to be here at all. DreamGuy 04:50, Mar 21, 2005 (UTC)


 * I tend to agree with Dreamguy. Along the same lines, once an article in a foreign language is spotted on WikiPedia, it is tagged for translation. But if it isn't translated within a certain amount of time, it is nominated for deletion instead since apparently noone cares about it. And, of course, if in the future someone does care, the article can always be recreated from scratch. In English this time.
 * If it were stated at the Village Pump that all untagged images would be deleted after three months, this may well spark interest in them and cause users to come and tag images to prevent them from being deleted - which seems like a Good Thing. And if nobody cares about a certain image, it will end up deleted, which seems equally good. Radiant_* 08:45, Mar 21, 2005 (UTC)
 * What Radiant said. [!] --Theo (Talk) 09:51, 21 Mar 2005 (UTC)
 * Yah it seems like a concensus if forming here. I've been involved in the image tagging / sleuthing projects and finding sources to save the vast number of these images – let alone working with absentee uploaders – is a monumental task. If the images aren't in an article now *and* don't have a source, honestly I can't think why they would be worth keeping. Are there any good examples of images that are UO that *would* be worth keeping? So yes, I am in favor of deletion, with a lengthy (months) waiting period. --MaxPower 14:41, 2005 Mar 21 (UTC)

There are users who upload many pictures of flora/fauna or of places while on tour. These are more appropriate for commons, but I'd assume some upload these pictures in en.wikipedia anyway. Such pictures (at least not all that were uploaded) usually aren't used in articles. Some users have a user page including all such pictures. IMHO such images shouldn't be considered orphans and deleted (does the policy allow anyone to move images to commons without contacting the original author?). I'd also assume there are some users who don't include those images even on a user page. What about them? -- Paddu


 * If the uploader has given some indication that they were taken by him or her, and not just taken from a random page on the internet, then we should tag them GFDL and not list them here. If there's no indication that the pic was homemade, then we won't know if they're legal or not, so we couldn't move them to Commons anyway. (Incidentally, anyone can move pics to Commons, with or without the original uploader's permission. But if there's no source info, they'll be deleted from Commons.) – Quadell (talk) (sleuth) 12:35, Mar 23, 2005 (UTC)


 * Long long ago, the image upload page didn't say so many things. Uploaders just had to check a box saying they have permission to upload. So almost no one bothered to say what copyright the image was under. In fact I'd assumed all content was under GFDL (a year ago or so) until I looked at some of the non-GFDL tags. I was even scoffed at by a wikipedian who I know in real life for saying that his/her images would be deleted unless he/she tags them. All of those old images shouldn't be deleted unless someone proves they are copyvios. The "image server" must be optimised for frequently used images rather than deleting orphans just because the "image server" isn't able to handle, or whatever. There are many online forums (e.g. /.) where most content are copyright contributors. Many might be violating copyrights while commenting on /., but they don't seem to delete those, unless if approached by the copyright owner probably. Similarly in wikipedia deleting images listed in WP:CP is fine, but other images shouldn't be deleted just because we aren't able to get much info. If asked to choose between having a slow "image server" and a losing lots of images and contributors, I'd choose the former (which would eventually be set right once wikimedia gets more funds & hence better h/w). -- Paddu 13:24, 23 Mar 2005 (UTC)


 * IFAIK, the image upload page has always asked you to give the source of the image. Simply "I took this photo" or "(c) me" is enough. It's been our policy for a while to delete suspected copyright violations. See WP:PUI for example. – Quadell (talk) (sleuth) 14:32, Mar 23, 2005 (UTC)


 * From c. Dec. 2003 to 20 Feb. 2004, this is what the upload page showed: (note that the stuff there is HTML and not wiki-syntax so for example what your browser shows as foobar sdfdsf is actually foobar sdfdsf . Even on 20 Feb. 2004, only fair use images were asked to be tagged as such  and the rest were said to be assumed to be GFDL. Before Dec. 2003 there is no way the upload page would have said anything more about citing sources (since otherwise those messages would have been copied in the initial MediaWiki:Uploadtext). The text before Dec. 2003 could possibly be got by browsing the MediaWiki CVS (what's now implemented through the MediaWiki: namespace were all implemented by a php script in those days).


 * It took really long (2 years+) for the image tagging, making the upload page require citing sources, etc. to be implemented. Those who uploaded in 2002-2003 didn't know about image tags, heck there wasn't even a template namespace. So please assume good faith w.r.t. old images. Or else the situation of Anthere's images would keep recurring.


 * BTW I'm here since Jan. 2003. I couldn't find when you joined since for large offsets Special:Contributions chokes nowadays. I'd ask those involved with image cleanup to please consider the fact that image tagging wasn't always the same, especially in 2002-3. Thanks! -- Paddu 19:38, 23 Mar 2005 (UTC)


 * What I meant was this: "I would like to ask those involved with image cleanup to please consider the fact that image tagging wasn't always the same, especially in 2002-2003. Thanks!".


 * I'd like to add to that period te first half of 2004, when most uploaders might not have known that the upload page text had changed and might have just clicked submit without reading what all it said. Hopefully these people would have realised image tagging, etc. by late 2004. In short, for old images for which the uploader hasn't been around for a while, be a bit more reluctant in deleting. Delete if there's evidence of copyvio, but in case it is possible for the person to have taken the picture or it is possible for the image to be considered fair use, please don't delete. I would like the image-deletion folks to ask other (probably more trustworthy than me) old-timers before simply dismissing my claims about the "recentness" of image-tagging and image-source citation. -- Paddu 20:01, 23 Mar 2005 (UTC)

We really need a bot here. Is anyone capable of making a bot that could tag unverified orphans, list them on a page, and sort them by original upload date? – Quadell (talk) (sleuth) 19:57, Mar 23, 2005 (UTC)


 * Well, whenever the deletion of an image is considered, just check if the original upload date as shown on the image description page is earlier than . I'd guess a threshold of around May 2004, but some consensus on the threshold might be easily reached, e.g. by posting on WP:VP or something so more people respond. -- Paddu 20:08, 23 Mar 2005 (UTC)


 * Right. But the thing is, there are tens of thousands of untagged orphans. That's a huge job. I've been listing unverified orphans here for months, going through them manually, spending 1-6 hours per day on it, and I'm only up to the Js. If there were a way of using a bot to make this easier, it would be quite welcome. – Quadell (talk) (sleuth) 20:33, Mar 23, 2005 (UTC)


 * OK, so you just want a bot to generate the list of all UOs? I had thought after so many images got tagged, just looking at Category: unverified would give all UOs, and you want a bot just for sorting by upload date. Probably I was too dumb to understand ;). -- Paddu 21:17, 23 Mar 2005 (UTC)

This entire effort, not to mention the discussion of it, is a Black Hole, a zone into which Wikipedians pour time and energy, and from which nothing of value emerges. It just doesn't matter what you do with something of this nature. It costs nothing sitting there, and costs much to winkle it out, fight over it, and kill it; there's a risk of going too far and killing the wrong thing. Instead of arguing over UO, spend some time creating images and uploading them. &mdash; Xiong (talk) 14:12, 2005 Mar 29 (UTC)

Two thoughts to deal with some of the images.
I just had a thought. What if I listed these on Possibly unfree images? I'll exclude any that were uploaded before before, say, June 2003. Like all images listed there, they'll be listed for at least a month before being deleted. Any objections? And would anyone be willing to help me with that? – Quadell (talk) (sleuth) 20:56, Mar 23, 2005 (UTC)

Another thought: for images that have already passed their seven days, if they're obviously not available under a non-free license (e.g. Image:MarilynChambers.jpg or Image:Pbimg 441.jpg), it seems to me they can be deleted. They could only be kept under a fair use claim, and since they're not used at all, there's no use to be fair. They've already been listed with no specific objections. Does anyone object to this? – Quadell (talk) (sleuth) 20:56, Mar 23, 2005 (UTC)


 * Listing for one month must be good enough, but I'd hope the threshold date is increased. Probably instead of thresholds we should be looking at the uploader's contributions, general behaviour, etc. Many "good" wikipedians might have contributed many images (legally) in 2003 (I'd say 2004 but Quadell still only accepts 2003 ;^) ) & left before realising they've to cite the sources. From an image that they uploaded we can get to the user page, etc., see that they've contributed a lot of images but the user page history doesn't seem to talk of any negative behaviour/bans/etc. so probably all the images uploaded by this particular person should be deleted only if confirmed as copyvio. So for each image we consider the uploader's "status"/behaviour/trustworthiness in wikipedia and decide whether to list all their images for deletion (e.g. their might be some who have contributed nothing but N images of which at least a few are copyvios, then certainly we have a strong suspicion that the others might also be copyvios & hence these can be deleted) or not. I simply have no idea how many such "good for the encyclopedia" UOs are out there but due to the tremendous amount of info. in wikipedia, I'm guessing there might be lots & I don't want wikipedia to lose them.


 * Also, it might be easy to find out & consider if on deletion, getting back the same or an equivalent image would be easy. e.g. if someone uploaded a simple diagram even way back in 2002, instead of wondering what the copyright is, it might be better to delete it after placing a request for a similar diagram in Requested images.


 * BTW Wikipedia_talk:Copyrights/archive3 (and similar older versions of related pages) has some insight into what was happening in Feb-Jul 2004. Basically those days images were of two kinds -- fair use (this must be mentioned according to the unofficial policy at Image use policy, though nobody might have cared as it was mentioned clearly that it was unofficial), public domain/GFDL (assumed by default so the uploader wouldn't care to mention that it is GFDL, public domain images can obviously be restricted by the uploader to GFDL).


 * OK I think I'll stop now (it's 3:30 am here & I've to sleep :) ) and let others interested in defending old images fight the case henceforth. -- Paddu 21:56, 23 Mar 2005 (UTC)


 * You make some very good points. Would you be willing to help with this? – Quadell (talk) (sleuth) 00:26, Mar 24, 2005 (UTC)


 * Alas, I'm too lazy & can't find time. I seem to be slow at simply everything [my total no. of edits (some info. at my user page) & my rate of edits to this page "yesterday" (while doing no other work) should be enough indication]. And most of the time, I have other work to do so my rate is even less. In comparision to your contributions, anything I could contribute to image-tagging, etc. would be negligible (BTW I believe I've tagged a few images but my total image namespace edits is < 50). -- Paddu 05:48, 24 Mar 2005 (UTC)

I asked this at Wikipedia:Village pump (policy)#Unverified orphans and got only one response until now. User:Xiong responded opposing deletion of UOs. -- Paddu 20:52, 27 Mar 2005 (UTC)

Image server load
I presume the load on the image server is due to requests & not due to disk space. So why does anyone get the notion that there are a lot of requests to UOs? To reduce the load articles and (especially) templates should use fewer images. To that end I see that images have been commented out from the frequently used templates. Removing UOs isn't going to reduce the image server load in any way. Removing them due to proven copyvio, obfuscatingly low quality, redundancy (e.g. due to an uploaded image that is a larger version of the same) are all OK. But removing every unverified image, especially those uploaded in 2002-2004 is unacceptable. I would like pointers to where I should register this complaint so it would be visible enough & enough people would discuss and there can be a consensus. No offence to anyone meant, but IMHO the hardwork wasted on deleting images no one cares for can be better used to fight vandalism and real proven copyvios.

Thanks! -- Paddu 08:18, 29 Mar 2005 (UTC)


 * I don't know how much of a load the unverified orphans are placing on the server. What do you think about using the GFDL-presumed tag on whichever UOs you think are most likely GFDL? – Quadell (talk) (sleuth) 14:11, Mar 29, 2005 (UTC)


 * Wow! Probably a minor tweaking is necessary. Since we only presume it is by the uploader. It could be either GFDL or PD. Thanks for this tag.


 * BTW I just sent a mail to WikiEn-l hoping we could make more people participate in the discussion about what to do with unverified images (which become UOs in the course of time), especially those uploaded in 2002-2004. -- Paddu 20:46, 29 Mar 2005 (UTC)


 * Note that as I am not a member, a moderator would have to approve before the message reaches the WikiEn-l subscribers. -- Paddu 21:24, 29 Mar 2005 (UTC)


 * I was bold and created a new category for these images at Category:Presumed GFDL images so they don't get mixed up with those that are confirmed to be GFDL. -- Paddu 21:24, 29 Mar 2005 (UTC)

Still unverified untaggued orphans
Paddu explained in length the issue. There are ten of thousands of untagged orphans. Some were recently uploaded and do not follow current rules which make tagging mandatory. However, many were uploaded before this became mandatory and in their uploaders mind, they were gfdl, since they agreed to follow gfdl when they uploaded it.

A couple of assumption should be clarified. It is not because an image is untaggued that it is necessarily a copyvio. We currently keep copyrighted images on Wikipedia. We assume we can use them under fair use doctrine. It sometimes happen we are wrong, and in this case, the board or anyone else receive a complaint from the copyright holder. The solution is generally to promptly delete the image and I can tell that in 4 years, all cases have been fixed to the satisfaction of the copyright holder. We should assume good faith from most uploaders and only delete untaggued images when someone raise a complaint about them.

In case the image is untaggued and is indeed a copyvio, it being an orphan reduces the risk of any legal complaint.

Another assumption is that an orphan image is "not interesting". It is not so. It may be that several images of the same topic already exist and not all may be put in the article. Which does not mean the ones non linked have no value. So, deleting them for having no value is not very wikilove.

Third point : we have room on our servers. If we were lacking room, I would agree we should not keep what is uncertain. But keeping these images is not hurting the whole system from a storage point of view.

Fourth point : the only case where the orphan untaggued would be likely to be problematic is when the whole content of the db is downloaded, possibly to be put on a cd rom. Here, if the image revealed itself a copyvio, we would not be able to do anything.

So, here is what I support
 * tag all them untaggued orphans.
 * ask uploaders to tag them
 * if no answer in xx weeks, change the tag to "verified but still untaggued orphans" (or any name suitable)
 * have all "verified but still untaggued orphans" excluded from downloads.

Anthere


 * Question: In your opinion, does it matter when the image was uploaded? What I mean is, if an image was uploaded in January of 2003, the uploader may not have been instructed to list the source and license information. But if it way uploaded last month, the uploader was definitely warned that if he/she didn't tag the image, it might be deleted. Should we treat the two cases differently? If so, what should the cut-off date be? – Quadell (talk) (sleuth) 20:31, Apr 4, 2005 (UTC)


 * You are correct. The cases are different I think. Though in both cases we should assume good faith imho, the recent uploader is supposed to know about the rules. And contrary wise to the old uploader, he did not agree per default. I would say we should inform both, but right, probably delete the recent images... This is in particular true for images uploaded these days, as the uploader is supposed to still be around. Date of cut off ? I suppose it would be good to look at old versions of the upload page... I remember setting up the french image policy with help of Alex756 in october november 2003. At that time, there was no tagging policy on en. Fr was the first one really. I do not know when tagging became mandatory. When Yann launched the tagging campaign ? 6 months ago ? I am not sure. Anthere 18:44, 5 Apr 2005 (UTC)


 * I'm sure there are many still around in Wikipedia who know how the image upload policies evolved, but as I said the problem is that none of them seem to know about our discussions. After so many attempts at trying to bring more people to talk about this, less than ten are discussing the issue. I'm sure if the issue is somehow made "more visible" we would get a better estimate of what the cut-off should be and probably a better way to deal with the images on either side of the cut-off. How were so many people mobilised to vote on the speedy delete criterion changes? -- Paddu 08:29, 14 Apr 2005 (UTC)


 * On the other hand, there are relatively few people deleting unverified orphans. I think we've reached a preliminary agreement, at least for now. There seems to be a consensus that UOs should only be deleted if (1) they're particularly new uploads, or (2) there's some other reason for them to be deleted (e.g. it's a copyvio or obviously not useful on Wikipedia). We should treat this as the consensus, afaic. – Quadell (talk) (sleuth) 19:31, Apr 14, 2005 (UTC)

We can't debate away the law
There are 'unverified' images here that were clearly the result of some Google Image search and are clearly too new to be copyright lapsed, and are clearly not a case of fair use. We have no right to use these images, and no amount of debate, discussion, or consensus building can magically create that right. In particular there is an image that I went through a good deal of effort to create a much higher quality and clear-copyrighted free replacement, and frankly I'm rather annoyed that it has not yet been deleted. Wikipedia is a huge project and there is no image that we can not afford to lose, and in many case we would be better off with images created specifically for our needs, even if that means we must go without for a while. If an image is indeed freely available then it will return someday, and hopefully carry with it the the right background information so that it can be used by others with a clear conscience and without taking on undue liability. --Gmaxwell 05:40, 12 Jun 2005 (UTC)


 * If a specific image is clearly not free, it can be listed on Copyright problems. If you suspect an image is unfree, but aren't sure, it can be listed on Possibly unfree images. But if you don't have a clear reason to believe it's unfree, it's best to leave it alone, at least for now. Here's why.
 * There are several images that were 'unverified' and that I deleted, but that turned out to have been created by Wikipedians and uploaded. For example, there was a beautiful flower image created by User:Anthere, and she hadn't tagged the image since it was uploaded before image tagging existed. The image was released under the GFDL, but there was no way for me to know this; I assumed it was found on a Google search. So I listed it for deletion, and then deleted it after a week. Well, it turns out that that was the only copy of Anthere's image. Image deletion can't be undone either, so she was justifiably upset. Anyway, someone found a way to restore the image (it was on a Wikipedia mirror, luckily), so all's well that ends well. But the lesson is, there's no good way to know if an image is appropriate to delete or not, so we decided it's best to leave those sorts of images alone, at least for now. – Quadell (talk) (sleuth) 11:23, Jun 13, 2005 (UTC)


 * But the problem is that images listed on copyright problems are being moved *here* with no notice to the person who did the listing. 'The only copy' is a complete strawman argument as it is exceptionally unlikely that any image that is important will only be on wikipedia, and in any case many people mirror the content from our rsync dumps, including me. This is the second time I've run into you pushing an agenda that seems to imply that you think it's okay to break the law as long as we're not caught (the first was where you insisted that content we gained access to via an email that said "this may be used on nutrition articles on wikipedia...if you hyperlink to us" == GFDL), and frankly I find it offensive: I'd rather the original work I have created for wikipedia not be seen by the public as just more petty thievery by wikipedia.  The loss of Anthere's picture would have been bad, but this isn't the solution... Better tagging by uploaders, and notifying uploaders of concerns should address this for the most part... and at the end of the day life goes on if an image or two is lost forever, but by flagrantly ignoring copyright law we devalue the project for those that need some certainty that the content we call free is actually free, and we potentially put the project at risk from legal and public image attacks. So, what I am going to do is wait 48 hours for any further discussion then make a copy of the images here and then delete all of them.--Gmaxwell 20:35, 13 Jun 2005 (UTC)


 * You really shouldn't take unilateral actions like that.
 * You also should calm down: your comments above sound close to a personal attack against me. I'm the one who made this page. I deleted hundreds of untagged orphans - maybe a thousand. So far as I know, I'm the only one who who has systematically deleted untagged orphans. And more than half of the people who commented thought this was a bad idea. So I quit doing it.
 * I'm certainly not "pushing an agenda" about whether untagged orphans should be deleted. I'm just informing you that when I, six months ago, decided on my own to delete all untagged orphans, I was informed by several other users that there was no consensus to do so. There still isn't.
 * I don't think it's okay to break the law, but in these cases, you don't know whether the law is being broken or not. As I said above, if you know a given image is a copyright violation, then you can list it on WP:CP. If you suspect it's a copyvio, then you can list it on WP:PUI. But if you have no reason to think an image is a copyvio, other than that it's not tagged, then unilaterally deleting the image isn't enforcing the law. It's deleting Wikipedia's images outside of Wikipedia policy. – Quadell (talk) (sleuth) 15:04, Jun 24, 2005 (UTC)