Wikipedia talk:Untagged images/Archive 3
![]() | 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 1 | Archive 2 | Archive 3 |
Images from Commons
Hi all. I was wondering how we should handle images that are from the Commons that only have a page here to categorize it on Wikipedia, such as Image:Lemur_huddle.jpg. Is there something I can tag it with that will prevent it from being listed again? ~MDD4696 03:23, 10 January 2006 (UTC)
Multipule warnings
So... dealing with Hbk3692003 (talk · contribs) and other people who repeatly upload picutres w/o copyright and source information, isnt that against Wiki rules... and eventually lead in a block. The only other warn template I could find was {{Image no source last warning}} which just says you have 2 more days before all their images are deleted... Can we do anything about these people? --Admrboltz (T | C) 23:59, 11 January 2006 (UTC)
- Well we generally don't block people unless they keep doing something they have been told not to. So generally only people who keep uploading after beeing told not to are blocked, meaning people are typicaly not blocked for whatever they uploaded prior to theyr first warning. If someone keep uploading unsourced and untagged images after beeing warned however, you should report it on either Wikipedia:Administrators' noticeboard/Incidents or Wikipedia:Administrator intervention against vandalism, use the later only in "slam dunk" cases where no debate is nessesary.--Sherool (talk) 01:08, 12 January 2006 (UTC)
- nods* so how do we warn people like Giantcn (talk · contribs)? --Admrboltz (T | C) 05:53, 12 January 2006 (UTC)
- Well personaly I would add {{image source}} and/or {{image copyright}} only once each, those messages include instructions to check and fix any other uploaded images too, so no need to clutter the talkpage with a dozen identical messages for each of the images, it just tend to overwhealm the user. So just make sure they know that they have at least one image that will be deleted if they don't fix things and then tag the rest of theyr problematic uploads. If they then keep uploading more unsourced images I'd suggest leaving a fairly sharp (though not hostile) message telling them that if they upload any more images without the nessesary information they will most likely be blocked. I don't think we have any handy templates for that kind of thing so you'll have to write the message yourself (point to WP:VAND and the message on the upload page itself). In this particular case I don't think any further action is required though. True he was technically warned in August and still uploaded such images, but it's been over a month since he uploaded anyting and he seems inactive anyway. If he comes back and keep doing it he should be nabbed right away though, but blocking an inactive user for something that happened a month ago would just be a waste of time IMHO. --Sherool (talk) 06:52, 12 January 2006 (UTC)
No information
Wouldn't the simplest way to deal with these untagged images be to have a bot tag them all as {{no licence}}, and have another bot notify - it would save a whole lot of editor hours.--nixie 12:37, 15 January 2006 (UTC)
- Well, it would result in a ton of images getting wrongfully deleted as well. Quite a few of the images I've run into are perfectly fine for inclusion, but they just don't have the proper tag. ~MDD4696 17:50, 15 January 2006 (UTC)
- Depends on whether we're getting ahead or falling behind. If manual labor is not keeping up with the uploads, then stern measures may be necessary. At commons we're shooting on sight, and it just barely keeps up with the copyvios. Stan 20:25, 15 January 2006 (UTC)
- That's a valid point... so far it looks to me like we're falling behind, unfortunately. ~MDD4696 03:55, 16 January 2006 (UTC)
- Yea... perhaps another modification of the upload screen requiring a source? It would be a bit daunting for the n00bs, but I think it is needed... --Admrboltz (T | C) 04:43, 16 January 2006 (UTC)
- But then you start getting a bunch of improperly tagged images, because people either guess or they just pick one so they can upload the image. ~MDD4696 04:46, 16 January 2006 (UTC)
- I think a decision should be required on the upload page in the license dropdown, even if the choice is something like, "I need help picking one". Currently if you don't choose anything it still uploads. - cohesion★talk 02:08, 17 January 2006 (UTC)
- That's definitely reasonable. The thoughtless can still upload without a real license, but at least are forced to consciously indicate that. Superm401 | Talk 07:20, 17 January 2006 (UTC)
- I think a lot of the problem is people just not realizing that one of the tags really has to be selected, it's a complicated decision, and if it seems unnecessary people will just skip it. I think if there was some warning that one wasn't selected people would (reluctantly) look through them at least. - cohesion★talk 10:12, 17 January 2006 (UTC)
- I think a decision should be required on the upload page in the license dropdown, even if the choice is something like, "I need help picking one". Currently if you don't choose anything it still uploads. - cohesion★talk 02:08, 17 January 2006 (UTC)
- But then you start getting a bunch of improperly tagged images, because people either guess or they just pick one so they can upload the image. ~MDD4696 04:46, 16 January 2006 (UTC)
- Yea... perhaps another modification of the upload screen requiring a source? It would be a bit daunting for the n00bs, but I think it is needed... --Admrboltz (T | C) 04:43, 16 January 2006 (UTC)
- That's a valid point... so far it looks to me like we're falling behind, unfortunately. ~MDD4696 03:55, 16 January 2006 (UTC)
- Depends on whether we're getting ahead or falling behind. If manual labor is not keeping up with the uploads, then stern measures may be necessary. At commons we're shooting on sight, and it just barely keeps up with the copyvios. Stan 20:25, 15 January 2006 (UTC)
Progress on these is not great this time, the list has been up for a few weeks and there is only about 10% done. I think the possibility of using a bot needs to be investigated, and mabye a default {{no licence}} tag should be added to uploads if information is not provided.--nixie 13:06, 17 January 2006 (UTC)
- Having manually done a hundred untagged images last night, I'm all in favor. Also have the bot spam talk pages, after about 50 of those long notices on a talk page, the uploader might get the message... Stan 14:26, 17 January 2006 (UTC)
- Yes, I think something needs to be done at upload to make it more obvious to people that a tag is required, as of right now if people don't pick anything there isn't even a warning about it. - cohesion★talk 19:26, 19 January 2006 (UTC)
- I think there should be a line for source, it should be required (or it auto tags it with {{no source}}) and if they dont select a license tag (i dont alot because I will tag it with one that isnt listed), it should tag it with {{no license}} not the fru fru "i dont know where i got it from" tag or add the 'Image with unknown license as of DD-MM" like {{no license}} does so its easier for us admins to clean them out. --Admrboltz (T | C) 19:41, 19 January 2006 (UTC)
- The advantage of offering people with an "I don't know" tag is that you can then conveniently find and delete them all at the end of each day; better than trying to push uploaders into lying about the image's status. Stan 20:30, 20 January 2006 (UTC)
- I dont mind the tag, but I think it needs to have the normal no source category added into it to make it simpler. --Admrb♉ltz (T | C) 20:35, 20 January 2006 (UTC)
- The advantage of offering people with an "I don't know" tag is that you can then conveniently find and delete them all at the end of each day; better than trying to push uploaders into lying about the image's status. Stan 20:30, 20 January 2006 (UTC)
- I think there should be a line for source, it should be required (or it auto tags it with {{no source}}) and if they dont select a license tag (i dont alot because I will tag it with one that isnt listed), it should tag it with {{no license}} not the fru fru "i dont know where i got it from" tag or add the 'Image with unknown license as of DD-MM" like {{no license}} does so its easier for us admins to clean them out. --Admrboltz (T | C) 19:41, 19 January 2006 (UTC)
- Yes, I think something needs to be done at upload to make it more obvious to people that a tag is required, as of right now if people don't pick anything there isn't even a warning about it. - cohesion★talk 19:26, 19 January 2006 (UTC)
Image tagging tools?
Has anyone developed any tools for image tagging? I looked briefly at Wikipedia:Tools but didn't see anything. Perhaps I will develop one for us to use... ~MDD4696 05:01, 16 January 2006 (UTC)
- I thought of one before, I think that is a good idea. --Jaranda wat's sup 05:20, 16 January 2006 (UTC)
- Well I "hacked" the "Wikify" tab user js script to make myself one click "no source" and "no license" tabs when editing images. If anyone is interested feel free to pillage my monobook.js script. Just copy everyting except the Lupin popup thing at the top to make it work (I don't have any easy to install "plugin" page). It automaticaly adds the relevant tag and save (so make any other edits before you click it) all in one click. I guess it could be even more time saving it you didn't have to open the edit page first, but this way you at least have the option to do a manual edit before you fire up the "auto-tagger". Notifying the uploader is still "manual" but I find it somewhat useful. --Sherool (talk) 08:25, 16 January 2006 (UTC)
- Well I had to redo my whole monobook.js but it seems to be working for me now with no source, no license, purge this page, the popups and the force edit summaries. --Admrboltz (T | C) 16:14, 16 January 2006 (UTC)
- I use mac text clippings, and just drag them over, a low tech solution, but it works pretty well for me. It's very easy to add new ones too, and change them based on the date etc. - cohesion★talk 08:56, 21 January 2006 (UTC)
If anyone is bored, they can check out my (still under development) image tagging tool by copying the code from my monobook.js. The addlilink, addTab, and addToolboxLink functions are required, so copy them too if you don't have them already. Just remember to refresh your own monobook.js once you've edited it.
My code adds an "Image Autotagger" link to your toolbox and a "Tag" tab to image pages. Most of it isn't functional yet, but if you are on an "Editing image:" page, you can use the Tag tab to append "No source" or "No license" automatically, the same as what Sherool's code does. You can use Image:TestImage.jpg for testing, just revert your edits when you are done.
If you do check it out and have comments or ideas, please let me know. ~MDD4696 04:24, 30 January 2006 (UTC)
- Oh, and it's Firefox only so far. Sorry! ~MDD4696 04:26, 30 January 2006 (UTC)
Maps
There are a lot of maps that look like they were made with the same tool example:Image:LocationOceania.png are these being created with some tool where the output can be gfdl'd? They are used in a lot of key articles, so I didn't want to move too quickly on most of them. Many are already on the commons and tagged correctly, but the remaining ones aren't and I don't know the source. - cohesion★talk 19:48, 21 January 2006 (UTC)
- I found the answer, I will post if here in case other people come across the issue, they are from a PDF the world fact book came out with apparently, so I am tagging them Template:PD-USGov-CIA-WF - cohesion★talk 04:08, 22 January 2006 (UTC)
Minor admin edit
Can an admin please edit Image:Logo.gif, Image:Logo.jpg, and Image:Logo.png to include the tag {{PD-user|Janke}}? I'll assume someone will do that and go ahead and remove it from the list now. Thanks :D - cohesion★talk 04:36, 22 January 2006 (UTC)
- Done for tagging, dont know what list its in, thats up to you :) --Admrb♉ltz (T | C) 04:47, 22 January 2006 (UTC)
- Thanky very mush sir :) - cohesion★talk 01:17, 23 January 2006 (UTC)
Some questions
I've been trying to help out with this project but got some questions.
What should you do when you encounted a fair use poster/cover which is high resolution? Plus, how can you easily check that. Just look at the size of the image in question?
- Good question... --Admrb♉ltz (T | C) 23:31, 22 January 2006 (UTC)
- I wouldn't spend time on it right now. First, we don't have an agreed number specifying what "low resolution" is supposed to mean (some seem to think 100x100, clearly too small), and second, once images are tagged, it is easy to go into fair use categories and sort images by size, address them all at once. Stan 23:48, 22 January 2006 (UTC)
- Makes sense. That's going to be a nice project btw. Garion96 (talk) 00:25, 23 January 2006 (UTC)
When you tag an image with {{or-fu-nr|~~~~~}} since it's obviously copyrighted and not in use. Should you then still contact the original uploader? Garion96 (talk) 21:08, 22 January 2006 (UTC)
- I dont, mainly because its not used in any articles. --Admrb♉ltz (T | C) 23:31, 22 January 2006 (UTC)
- Depends - sometimes images appear orphaned because the article using them has been vandalized (happened to one of my images just today). If I can identify what it's supposed to illustrate (such as when the article is linked from the desc page), I go look at it to see what's up. Stan 23:48, 22 January 2006 (UTC)
- That's what I thought. Tnx for the responses. Back to some more tagging. Garion96 (talk) 00:25, 23 January 2006 (UTC)
Tagging rate
Hi all. I've extrapolated a completion date based on how many images we've tagged since this round of tagging started. Of course, I expect things to move a little bit faster as more people join the project, but at the current rate, we will have completed tagging all 38318 images on July 13, 2006. That's a ways off.
I'm currently working on a specialized tool to make things go faster, but development on it will likely be slow (as in, a beta in a few weeks). In any case, does anybody have ideas to get this project moving faster? ~MDD4696 23:08, 25 January 2006 (UTC)
- My user:Admrboltz/monobook.js has a No copyright button and a no source button as well as an IFD button, it only does the image, it doesn't notify the uploader thats still manual but it helps.
- Based on the most recent data, the projected completion date has moved up to June 25, 2006. ~MDD4696 01:22, 27 January 2006 (UTC)
- That's a great script. I hope you don't mind me using it. Superm401 - Talk 00:47, 29 January 2006 (UTC)
- What's really bad is that we're probably not keeping up with the creation of new untagged images at this rate. Perhaps this needs to get more prominent on the community portal. Also, at the risk of getting my wrist slapped, I've been summarily executing the worst images (unused & no description & contributor gone), which speeds things up. Stan 01:53, 27 January 2006 (UTC)
- I also speedy anything that meets the criteria for speedy deleting images. A notice on the recent changes page header might not hurt.--nixie 03:14, 27 January 2006 (UTC)
Looks like the {{no license notified}} usage has been changed so that needs to be reflected here. Thanks. —Wknight94 (talk) 01:50, 26 January 2006 (UTC)
Tags on Templates for deletion
Some tags have been listed on templates for deletion which may be of interest. The templates listed are Template:PD-AUGov, Template:Webimage, Template:PD-USGov-Congress-USBG, Template:QualificationsandCurriculumAuthorityCopyright, Template:LandRegistryCopyright, Template:TeacherTrainingAuthorityCopyright, Template:Ordnance Survey Copyright, Template:UNPhotoArchive, Template:PD-old-50. Template:CanadaCopyright is waiting to be orphaned. Does anyone involved in image tagging ever look at templates for deletion? Secretlondon 03:39, 26 January 2006 (UTC)
- Any of the "x for deletion" pages make my eyes bleed, so I generally avoid. :-) Not so important anyway, the great majority of the untagged images I see are speedy-delete orphans or unsourced not-quite-so-speedy deletes. Stan 06:38, 26 January 2006 (UTC)
Don't forget to check the history
When you find an untagged image, you might want to check the article history: maybe the uploader provided a suitable licence but the licence has been deleted by a vandal. It would have been a shame if Image:How e-mail works.png had been deleted. Gdr 13:49, 2 February 2006 (UTC)
Long-term plan
Even while this is going on, I'm sure new untagged images are piling up, and so I'm thinking we need a strategy that is more of a steady process. For instance, a bot could go through frequently and and put each untagged image into some sort of a "untagged" category for human review. This could also catch blanked existing images that Gdr mentions, so one would want it in addition to any of the upload changes suggested a while back. Stan 20:34, 5 February 2006 (UTC)
- (Note that my reply kind of goes off on a tangent) I think that in general we have a huge problem with regards to images that are either untagged or tagged with templates that don't describe the copyright/license status of the image, or are depreciated. I wonder whether there's any benefit in setting up some sort of project or something that attempts to ensure that all images are tagged with a valid, encouraged template. Right now WP:UI is ensuring that all untagged images as of last December are getting tagged. However, some of the contributors continue to tag images with depreciated templates such as {{fairuse}} or dubious templates such as {{coat of arms}}, which just moves the problem around, and doesn't eliminate it. There isn't any project that's trying to re-tag all images tagged with, say, {{coat of arms}} with a valid template (or delete them), or to ensure that all current uploads are tagged with a template that's not depreciated or dubious, or to ensure that people don't re-tag existing images with dubious templates. Maybe having some sort of bot that checks whether an image has been tagged with a template in WP:ICT#Public_domain, WP:ICT#Free_licenses, or WP:ICT#Fair_use might be a start. JYolkowski // talk 22:59, 5 February 2006 (UTC)
- I see nothing wrong with changing the script we're using now to put images in a category instead of on pages as they are now. It's also appropriate to include images tagged with non-copyright tags. However, this won't solve the problem of us being overwhelmed by untagged images. Superm401 - Talk 01:10, 6 February 2006 (UTC)
- You can get an indication of the number of new untagged images being added by looking at [1]. The current rate seems to be about 15 per hour, or about 1000 over a three day period. Luckily many of these get caught and fixed before they hang around for too long. The advantage of dealing with the newest images is that you may be able to properly instruct uploaders before they continue to upload hundereds more untagged/copyright infringing images. --Martyman-(talk) 01:27, 6 February 2006 (UTC)
- To me that strongly suggests a daily category, a la the no-source images. Psychologically I suspect that people will work harder if doing "just a few more" empties the category, vs making a tiny dent, plus the list of dated subcategories clearly when things start to fall behind. Stan 13:45, 6 February 2006 (UTC)
- You can get an indication of the number of new untagged images being added by looking at [1]. The current rate seems to be about 15 per hour, or about 1000 over a three day period. Luckily many of these get caught and fixed before they hang around for too long. The advantage of dealing with the newest images is that you may be able to properly instruct uploaders before they continue to upload hundereds more untagged/copyright infringing images. --Martyman-(talk) 01:27, 6 February 2006 (UTC)
Here's some food for thought. It's a list of tags used on images uploaded between about 24 and about 48 hours ago (tagging status as of last night). There were almost 2,000 uploads in total, but some were duplicates or were speedied. Anyway, there were about 200 untagged images uploaded, so assuming this is typical and based on the current progress rate, keeping up doesn't look like a problem. It looks like there might be a problem that's at least as big with uploaders tagging images incorrectly (e.g. there's a lot of junk tagged with {{CopyrightedFreeUse}}, I find it hard to believe that 10% of the images on Wikipedia are described by {{Promotional}} and every reason I saw in a {{PD-because}} tag was incorrect), but that's not a problem we strictly need to worry about here. Anyway, here's the list in case anyone's interested:
Type of licence | Associated tags | Count |
---|---|---|
GFDL | ({{GDFL}}, {{GFDL-self}}, {{GFDL}}) | 160 |
Creative commons licenses | {{cc-by-2.0-map-of}}, {{cc-by-2.5}}, {{cc-by-sa-2.5}} | 38 |
Generic any-purpose copyright (a lot of which appear to be incorrectly tagged) | {{CopyrightedFreeUse}}, {{CopyrightedFreeUseProvided}}, {{NoRightsReserved}} | 200 |
Other free licences | {{FAL}}, {{GPL}}, {{CrownCopyright}}, {{Sejm}} | 4 |
Self-made public domain images | {{PD-self}} | 96 |
U.S. Government public domain images | 49 | |
Depreciated public domain tags | {{PD}} | 24 |
Other public domain | 56 | |
Promotional images | {{Promotional}} | 188 |
Album covers | {{albumcover}} | 90 |
Logos | primarily {{logo}} | 164 |
TV and film screenshots | {{film-screenshot}}, {{tv-screenshot}} | 156 |
Video game screenshots | {{game-screenshot}} | 59 |
Depreciated fair use tags | {{fair use}}, {{fairuse}} | 5 |
Other fair use tags | 247 | |
Unknown copyright status tags | {{Coat of arms}}, {{FOTWpic}} | 16 |
Wikimedia copyright | {{CopyrightByWikimedia}} | 1 |
Permission-only | {{AllRightsReserved}}, {{Copyrighted}} | 3 |
Listed for deletion | {{db}}, {{ifd}}, {{imagevio}}, {{redundant}} | 9 |
No license | {{no license}}, {{no licence}} | 87 |
No source | {{no source}} | 42 |
Orphaned fair use | {{or-fu}} | 3 |
Uploader unsure | {{Don't know}}, {{Don't know 2}}, {{Some web site 2}} | 55 |
Untagged | 191 |
JYolkowski // talk 00:45, 9 February 2006 (UTC)
- Interesting! Since the current pile of untagged images is about a six-month accumulation (right?), at 200/day that adds up to 36,000, or pretty close to the starting pile size, suggesting the rate hasn't changed much. To put it another way, if this initiative manages to finish the current pile in four months, there will be a new pile of 24,000 untagged images accumulated in the meantime. That's why I think we need to do something semi-automatic to truncate the incoming stream. Stan 01:23, 9 February 2006 (UTC)
- I think it's a good argument for doing so. Based on your arithmetic, it sounds like we'll continue to be several months behind for a very long time at this rate. Furthermore, I think that it's good to catch untagged images as soon as possible, before the uploader's been gone for several months. Anyone have thoughts as to how to best deal with the new untagged images? JYolkowski // talk 00:57, 10 February 2006 (UTC)
- See also my comment below (shortly). JYolkowski // talk 20:34, 11 February 2006 (UTC)
- I think it's a good argument for doing so. Based on your arithmetic, it sounds like we'll continue to be several months behind for a very long time at this rate. Furthermore, I think that it's good to catch untagged images as soon as possible, before the uploader's been gone for several months. Anyone have thoughts as to how to best deal with the new untagged images? JYolkowski // talk 00:57, 10 February 2006 (UTC)
Image Autotagger (Firefox only)
I've created a brief demonstration of my Image Autotagger tool if anyone would like to preview it. The code is available from my monobook.js, but I've got to polish it a bit, make it a little more flexible, and of course write documentation. You can test the Fast Tagger tab on Image:TestImage.jpg.
Features:
- Image Autotagger - A popup tool window for many images
- Rapidly tag a list of images
- Select license from a drop-down and have it inserted automatically
- Automatic edit summaries
- (To do) Automatically notify uploader
- Fast Tagger - A popup "tag" tab for one image
- Tag an image in just one click
- Quickly tag an image with No License, No Source, or IFD.
- Automatic, overridable edit summary
I'd really appreciate any comments or ideas! ~MDD4696 23:52, 5 February 2006 (UTC)
- I gave this a try and couldn't figure out how to use it. How do you let it know which image you want to tag? Also it has a few layout issues in Firefox for me. --Martyman-(talk) 00:12, 6 February 2006 (UTC)
- I should've put more of an explanation up, because it seems that YouTube takes a while to "process" videos you upload. Currently, you must use a list of images as your starting point. For example, type User:Mdd4696/Sandbox into the box in the upper right, and then click "Load". This loads a list of images for you to tag (you can also used the Wikipedia:Untagged Images lists, like Wikipedia:Untagged images/2005dec13-2). Double click on any of the images in the list to load it and edit it. You can automatically insert a tag from the drop-downs, or you can edit by hand. Does that help Martyman? I'll look into making it more intuitive from the get-go. Also, could you take a screenshot and post it at ImageShack (free, no registration)? ~MDD4696 00:32, 6 February 2006 (UTC)
- Ok, thanks for that, that makes much more sense. I will give it another try. I have uploaded a screenshot at [2]. For anyone interested I just added the follow text to my monobook.js to include Mdd4696's javascript, once theis is up and running properly it would be nice if it could have it's own page that could be included like this. --Martyman-(talk) 00:41, 6 February 2006 (UTC)
document.write('<script type="text/javascript" src="' + 'http://en.wikipedia.org/w/index.php?title=User:Mdd4696/monobook.js' + '&action=raw&ctype=text/javascript&dontcountme=s"></script>');
- Alright, if anyone's interesting, the Image Autotagger now has a page: User:Mdd4696/Image Autotagger. ~MDD4696 05:28, 15 February 2006 (UTC)
Stop the problem before it starts
Let's try to stop the problem of untagged images before it starts and keep images out of the mass dumping of images by getting them tagged early. Try to watch the upload log and look for untagged images. If you can't determine the tag or other information leave one of the no source or no copyright status tags and alert the user. Thanks. Dustimagic *\o/* (talk/contribs) *\o/* 00:20, 11 February 2006 (UTC)
- That's very inefficient though - as the above statistics show, only one in 10 of new uploads are untagged, and the other 9 clicks on an image are wasted time. Stan 13:48, 11 February 2006 (UTC)
- Not nessesarily, just look for the ones with a blank upload summary, those will at least not have any source info (though they might be albumcovers and such), also when you find someone who uploads blank images check theyr contributions list, chances are they have several other such images to theyr name as well. One solution would probably be to have a bot running on the toolserver, simmilar to what the anti-vandalism people have, that will inspect all new uploads and automaticaly tag images with no license template and notify the uploader within minutes of the file beeing uploaded. However that won't solve the problem with people playing the "license rulette", there are probably at last as many mistagged images as there are untagged ones, so it would not hurt to have humans sanity check the upload log, the upload rate is not that huge, so with a handfull of people and a bit coordination it should be doable, if a bit boring (I must confess I check it too rarely myself). --Sherool (talk) 14:16, 11 February 2006 (UTC)
- It's also easier to find untagged images if you watch images uploaded by newbies. ~MDD4696 18:45, 11 February 2006 (UTC)
- As an experiment, I've created User:JYolkowski/recent untagged images and I'll update it regularly for the next little while until I see whether there's any value in this or not. Anyone who'd like to help tag the images listed there is welcome to do so. Here's what I do:
- Select the "edit" link next to each of the links (I open each link in a new tab on my browser).
- Add {{subst:nls}} to each page, unless it's obvious what the licence should be. This template transcludes {{Uploader did not tag image}} and {{No licence}}.
- Remove the links from the page when done.
- Any comments are welcome. P.S. If someone who has toolserver access wanted to write something to list recent untagged images, that would be even better. JYolkowski // talk 20:45, 11 February 2006 (UTC)
- Isn't that what Wikisense/Untagged Images does? I found it higher up on this page. Superm401 - Talk 19:40, 12 February 2006 (UTC)
- Cool... I did check that out before, but it seemed a lot less functional last time I did. At any rate, that seems to now have everything I'm interested in, so I'm doing to discontinue updating the page I mentioned above. JYolkowski // talk 01:13, 13 February 2006 (UTC)
- Isn't that what Wikisense/Untagged Images does? I found it higher up on this page. Superm401 - Talk 19:40, 12 February 2006 (UTC)
- As an experiment, I've created User:JYolkowski/recent untagged images and I'll update it regularly for the next little while until I see whether there's any value in this or not. Anyone who'd like to help tag the images listed there is welcome to do so. Here's what I do:
- It's also easier to find untagged images if you watch images uploaded by newbies. ~MDD4696 18:45, 11 February 2006 (UTC)
- Not nessesarily, just look for the ones with a blank upload summary, those will at least not have any source info (though they might be albumcovers and such), also when you find someone who uploads blank images check theyr contributions list, chances are they have several other such images to theyr name as well. One solution would probably be to have a bot running on the toolserver, simmilar to what the anti-vandalism people have, that will inspect all new uploads and automaticaly tag images with no license template and notify the uploader within minutes of the file beeing uploaded. However that won't solve the problem with people playing the "license rulette", there are probably at last as many mistagged images as there are untagged ones, so it would not hurt to have humans sanity check the upload log, the upload rate is not that huge, so with a handfull of people and a bit coordination it should be doable, if a bit boring (I must confess I check it too rarely myself). --Sherool (talk) 14:16, 11 February 2006 (UTC)
JYolkowski, that's great. Dustimagic *\o/* (talk/contribs) *\o/* 04:29, 12 February 2006 (UTC)
- I tried the untagged images tool, and it seemed effective - I cleared a five-hour chunk of 02-13 uploads. Downside is that tool's db is delayed, so you have to wait a while to see the list change. It's remarkable how many people are uploading images as one of their first WP edits, even before they get a welcome message, and how many don't actually use the image on any page after uploading it... Stan 21:20, 14 February 2006 (UTC)
The new problem
On the project page there is discussion of image pages that are only links to images on the commons Image:William-Adolphe Bouguereau (1825-1905) - Day (1881).jpg This is actually very common, and I think it happens for every image on the commons maybe. When you click on any image on en you go to an en page saying the image is on the commons, but there are no gfdl tags etc. There are *a lot* of images that are like this. Anyway, I don't really know what is going on and would like someone to explain it :) Is the parser not seeing that they are tagged, or are they pages that really should be deleted? - cohesion★talk 07:45, 12 February 2006 (UTC)
- Most images that are on commons get deleted here, just tag them for speedy or delete them --Jaranda wat's sup 08:01, 12 February 2006 (UTC)
- In most cases, there is no image page here and the text is just imported from commons; this is not a problem for UI. The problem is when there is an empty image page here, possibly with a duplicate image. The pages without any image should be deleted as I2; those duplicated on commons should be tagged as {{NowCommons}}, not speedied. Superm401 - Talk 19:28, 12 February 2006 (UTC)
or-fu-nr template
Possibly a stupid question, but what happened to {{or-fu-nr}}? Was the lack of a date format an issue? Dethomas 23:41, 17 February 2006 (UTC)
- Can you be more specific? I still use the template. --Admrb♉ltz (T | C) 03:27, 19 February 2006 (UTC)
- Nevermind... just used it... use {{orfud}}Admrb♉ltz (T | C) 04:48, 19 February 2006 (UTC)
Category:Uploader unsure of copyright status
Can we add the images in Category:Uploader unsure of copyright status to the list of things needing tagging at the next dump? It is getting out of control really, and doing it within a category is rough. - cohesion★talk 18:47, 20 February 2006 (UTC)
Alot of admins are mass deleting from that category, even though some of the images are valid like logos that were used in articles and even some PD images. I prefer that we look though the images first, then delete --Jaranda wat's sup 19:17, 20 February 2006 (UTC)
- Yeah, I agree we should look at them, even more so since someone asked for us to :D I would like them to be mixed in with the other images needing tags so we could look at them here too. This could be in addition to the category. - cohesion★talk 05:52, 21 February 2006 (UTC)
- I hack at it from time to time, I reckon that about 10% of what is in there is easily saveable. It really needs users to go through by hand: non admins can always add a {{nosource}} or {{orphanfairuse}} tag as appropriate so that they get hit by the mass deletions. Physchim62 (talk) 06:01, 21 February 2006 (UTC)
Yugolavian images
There are a number of untagged images (Image:Djilas.jpg is an example) that claim: "This image is freely available on the internet from various sources in the public domain. The same image was created in the former (socialist) Yugoslavia and as such does not enjoy any copyright protection."
There are similar templates for former Soviet images and for images from Poland. I haven't been able to verify the claim that images from the former Yugoslavia aren't entitled to protection. How should we best deal with these?
--SteinbDJ 14:53, 22 February 2006 (UTC)
- The images are still lacking a source, and should be tagged {{no source}}. --Admrb♉ltz (T | C) 17:27, 22 February 2006 (UTC)
Commons image with a nearly blank description here
What are we doing about images on commons that have a description page here used to categorise the images. I have been ignoring them and removing them from the lists, but this will result in them being flagged again the next time a full database search is performed looking for images lacking copyrights. An example image is Image:Regions of Czech Republic.png. Should there be some kind of {{commonslicense}} tag that could show up as a license tag in the database. --Martyman-(talk) 02:02, 3 March 2006 (UTC)
- I think it would be a good idea. If it has a cat I will leave it alone, but if it is just random text, especially if its the exact same text as on commons, I will just delete the page (if you arent an admin just use {{db|blank image page, image is on commons.}} --Admrb♉ltz (T | C) 05:02, 3 March 2006 (UTC)
- I agree, whatever the parser needs to not add these back to untagged lists, I don't think we actually need any on-screen information about these images that isn't already there though. For example a message saying "This image is on the commons and has been tagged there" will be redundant since that information is already provided. - cohesiont 19:42, 8 March 2006 (UTC)
get tagged images
I want to know what should be done once an image gets tagged with an image tag or no source tag. Should i remove or system removes it automatically. Thanks-- Shyam (T/C) 09:00, 3 March 2006 (UTC)
- remove it from the list, and update the count on the main page :) --Admrb♉ltz (T | C) 09:02, 3 March 2006 (UTC)
- When I tag an image with "no source", i am trying to notify the uploader with {{subst:image source|Image:Image name}} but their responses are very worst. Is there any other good tag with which we can notify the uploader. Thank you -- Shyam (T/C) 07:34, 4 March 2006 (UTC)
- Well, they're being nicer than they would have been had the images been actually deleted! :-) I think it shows the hazard of blasting through and applying the no source tag indiscriminately - if the pages say "my photo" and you say "no source", then you're calling them liars, but if you say "no license", then they're OK with it. Another thing is to look out for intent - "GFDL" is often given as [[GFDL]], which is sufficient to demonstrate intent, and just needs a small edit. Even with my slower approach to tagging, there were still a couple people who grumbled that I notified them(!). Welcome to the admins' world, where the mess-makers are free to criticize the people who clean up after them... Stan 14:04, 4 March 2006 (UTC)
- Thanks Stan for your solution. I have a confusion which tag should I use for an image if it has an url address using Fair use in...images tag. Shyam (T/C) 21:33, 4 March 2006 (UTC)
- If it has a source URL and {{Non-free fair use in}} its tagged as far as we are concerned. There is a WP out there for Fair Use. --Admrb♉ltz (T | C) 21:42, 4 March 2006 (UTC)
- Thanks for your suggestions, I am very new to tag the image. So i am not so expert and facing many problems. Here is one- If an image like Image:Chesapeake Bay Bridge small.jpg has a amall version of the original image Image:Chesapeake Bay Bridge.jpg. I don't know what's the need of another one which is smaller but it exists. Which tag should be used for smaller one? Thanks-11:44, 6 March 2006 (UTC)
- Keep the big one. Make sure everyything links to the big one and tag the small one with {{Redundant image|name of replacement image without the Image: prefix}} --Admrb♉ltz (T | C) 23:43, 6 March 2006 (UTC)
- Thanks for your suggestions, I am very new to tag the image. So i am not so expert and facing many problems. Here is one- If an image like Image:Chesapeake Bay Bridge small.jpg has a amall version of the original image Image:Chesapeake Bay Bridge.jpg. I don't know what's the need of another one which is smaller but it exists. Which tag should be used for smaller one? Thanks-11:44, 6 March 2006 (UTC)
- If it has a source URL and {{Non-free fair use in}} its tagged as far as we are concerned. There is a WP out there for Fair Use. --Admrb♉ltz (T | C) 21:42, 4 March 2006 (UTC)
- Thanks Stan for your solution. I have a confusion which tag should I use for an image if it has an url address using Fair use in...images tag. Shyam (T/C) 21:33, 4 March 2006 (UTC)
- Well, they're being nicer than they would have been had the images been actually deleted! :-) I think it shows the hazard of blasting through and applying the no source tag indiscriminately - if the pages say "my photo" and you say "no source", then you're calling them liars, but if you say "no license", then they're OK with it. Another thing is to look out for intent - "GFDL" is often given as [[GFDL]], which is sufficient to demonstrate intent, and just needs a small edit. Even with my slower approach to tagging, there were still a couple people who grumbled that I notified them(!). Welcome to the admins' world, where the mess-makers are free to criticize the people who clean up after them... Stan 14:04, 4 March 2006 (UTC)
- When I tag an image with "no source", i am trying to notify the uploader with {{subst:image source|Image:Image name}} but their responses are very worst. Is there any other good tag with which we can notify the uploader. Thank you -- Shyam (T/C) 07:34, 4 March 2006 (UTC)
Using the toolserver
The automated, up to date tool here seems very good; I hate to be harsh, but why are still keeping the mannual lists around? I just went through 80 items on the mannual list, and all of them had already been tagged... JesseW, the juggling janitor 11:02, 8 March 2006 (UTC)
- According to the tool, we actually have 22,807 untagged images as of right now, not 17,000. It can sort them by user(so you can find out that one user has 180+ untagged images, while most have about 14 images each), and by upload date; so we can do the recent ones first. Overall, it just seems like such a better way to do the work... JesseW, the juggling janitor 11:13, 8 March 2006 (UTC)
- I think the main reasons we still use lists are a lack of awareness of the tool. If it does what it says on the tin, then yes, probably we should be using it, and extending its use (if possible) to other active wiki fixun tasks. --Tagishsimon (talk)
- Cool. Well, added one link to a list of untagged images uploaded by a user who's uploaded 100+ of them; if no-one objects in a few days, we can de-link the lists, and change the page to point to the tool only... JesseW, the juggling janitor 12:31, 8 March 2006 (UTC)
- This certainly looks promising, but I think at least some of us use Image Autotagger. This is a nice way to sort and find out what is left but I wouldn't want to have to go back to manually typing the tag in every time. Maybe the autotagger can use these lists somehow, it looks like you can output at least a few ways. - cohesiont 19:57, 8 March 2006 (UTC)
- That, and the tool server is very very s...l...o...w... Admrb♉ltz (T | C) 22:31, 8 March 2006 (UTC)
- Having just learned about the toolserver, I have been thinking about how I might incorporate it with the autotagger. Hopefully I will have an update for everyone soon that will make the tool much more useful. ~MDD4696 22:47, 8 March 2006 (UTC)
- That, and the tool server is very very s...l...o...w... Admrb♉ltz (T | C) 22:31, 8 March 2006 (UTC)
- This certainly looks promising, but I think at least some of us use Image Autotagger. This is a nice way to sort and find out what is left but I wouldn't want to have to go back to manually typing the tag in every time. Maybe the autotagger can use these lists somehow, it looks like you can output at least a few ways. - cohesiont 19:57, 8 March 2006 (UTC)
- Cool. Well, added one link to a list of untagged images uploaded by a user who's uploaded 100+ of them; if no-one objects in a few days, we can de-link the lists, and change the page to point to the tool only... JesseW, the juggling janitor 12:31, 8 March 2006 (UTC)
- I think the main reasons we still use lists are a lack of awareness of the tool. If it does what it says on the tin, then yes, probably we should be using it, and extending its use (if possible) to other active wiki fixun tasks. --Tagishsimon (talk)
According to the tool, we now have 21,923 untagged images; we are actually making progress... JesseW, the juggling janitor 07:10, 10 March 2006 (UTC)
- How do you get the tool to tell you that, other than paging through 1000 at a time? --Tagishsimon (talk)
- Guess, enter that number in the URL param for ofs, then page around (maybe by editing the URL again, till you get the end). It's not particularly efficient. ;-) JesseW, the juggling janitor 03:24, 14 March 2006 (UTC)
- It's worth noting that the tool appears to ignore all sorts of tags ... the first four images it served to me just now were all long since tagged, but still listed. I'm not knocking it, but I don't have time to get involved with its owner to ascertain why tagged images are being listed for tagging --Tagishsimon (talk)
- It's because parts of the database schema are not updated often enough; the tool uses a particular table to identify possibly untagged images, then checks each one individually, but displays them even if the detailed check says that they are tagged. So you are right that the count is high, but I suspect it's still closer than the list numbers, as we get lots of new ones constantly, and the tagged ones do eventually get removed from the tool's list. JesseW, the juggling janitor 11:55, 10 March 2006 (UTC)
- As we get closer to the end, the lists will tend to have more and more images already tagged as people track down related images and editors fix up their own images after getting a first notice. I think they're still a nice way to keep track of progress until the toolserver list goes under 1,000 or so. (An interesting thing I notice about the toolserver list is that the older images it lists have almost always been vandalized...) Stan 20:58, 10 March 2006 (UTC)
Number of untagged images according to the tool (as of 2006-03-12 15:45, due to replication lag): 19,293. So there are about 9,250 images on the tool's list that are not on the mannual list. Not great, at all, but not too bad. And we are making progress. JesseW, the juggling janitor 03:24, 14 March 2006 (UTC)
- As of 2006-03-15 20:32 the tool says we have 17,692 untagged images (a decline of 1,601), meaning 9,010 are not on our mannual list. JesseW, the juggling janitor 02:18, 16 March 2006 (UTC)
- As of 2006-03-16 12:39 the tool says we have 17,297 untagged images(a decline of 395), meaning 9,688 are not on our mannual list. JesseW, the juggling janitor 12:59, 16 March 2006 (UTC)
- As of 2006-03-17 10:14 the tool says we have 16,979(a decline of 318) untagged images, meaning 10,107 are not on our mannual list. While we are still making progress, we are not keeping up with uploads; the # not on the list is going up. This is not good. JesseW, the juggling janitor 10:30, 17 March 2006 (UTC)
- You would expect the number not on the list to go up, since people are still doing untagged uploads by the hundreds each day. It does mean that the static list overstates progress - the decline of 318 is still going in the right direction, just not so impressive. That's why I suggested above to cauterize the influx by instituting a parallel process to identify and whack day-old images, get to the uploaders on the spot, so it's not months before we inform them how to do things right. Stan 13:31, 17 March 2006 (UTC)
- I agree with you. The only way the not on the list # would go down is if we fixed more of the recent uploads than the combination of the old ones we fixed and the new uploads themselves. Sorry, I misread it above. JesseW, the juggling janitor 14:03, 17 March 2006 (UTC)
- You would expect the number not on the list to go up, since people are still doing untagged uploads by the hundreds each day. It does mean that the static list overstates progress - the decline of 318 is still going in the right direction, just not so impressive. That's why I suggested above to cauterize the influx by instituting a parallel process to identify and whack day-old images, get to the uploaders on the spot, so it's not months before we inform them how to do things right. Stan 13:31, 17 March 2006 (UTC)
- As of 2006-03-17 10:14 the tool says we have 16,979(a decline of 318) untagged images, meaning 10,107 are not on our mannual list. While we are still making progress, we are not keeping up with uploads; the # not on the list is going up. This is not good. JesseW, the juggling janitor 10:30, 17 March 2006 (UTC)
- As of 2006-03-16 12:39 the tool says we have 17,297 untagged images(a decline of 395), meaning 9,688 are not on our mannual list. JesseW, the juggling janitor 12:59, 16 March 2006 (UTC)
As of 03:28, 19 March 2006 (UTC) the tool says we have 16,516 untagged images(a decline of 463), meaning 10,669 are not on our mannual list. JesseW, the juggling janitor 03:28, 19 March 2006 (UTC)
- I think we don't have more than 10,669 in our manual list because many of the images in manual list are tagged and deleted. Shyam (T/C) 09:07, 19 March 2006 (UTC)
- I'm not sure what you mean, but if you are saying that many of the items on the mannual list are already done, and so the 10,669 # is more or less meaningless, I agree with you. It's mainly just for curiosities sake that I've been updating this. JesseW, the juggling janitor 09:22, 19 March 2006 (UTC)
In any case, as of 15:13, 22 March 2006 (UTC) the tool says we have 15,304 untagged images (a decline of 1,212) which is quite nice, although 15,000 is still nowhere close to done; sigh. JesseW, the juggling janitor 15:13, 22 March 2006 (UTC)
- As of 04:39, 23 March 2006 (UTC) the tool says we have 15,081 untagged images (a decline of 223). JesseW, the juggling janitor 04:39, 23 March 2006 (UTC)
Based on the historical data I gathered above, I extrapolated, and if we continue to cut down the backlog at the current rate, we should have it dealt with by late April, i.e. about 30 days from now. JesseW, the juggling janitor 05:41, 23 March 2006 (UTC)
What should be done with this one?
Image:Parisnice2006-profil300.gif? I'm not sure what to tag this as... JesseW, the juggling janitor 11:50, 8 March 2006 (UTC)
- Probably {{no license|month=March|day=8|year=2006}} since we know the source but not the license. --Tagishsimon (talk)
- So tagged. Thanks! JesseW, the juggling janitor 12:29, 8 March 2006 (UTC)
Unable to see the proect page
Can somebody tell me the reason for not able to see the project page. It shows starting article dialogue. Thanks Shyam (T/C) 00:53, 14 March 2006 (UTC)
- Have you tried clearing your browser cache - Shift Reload in Mozilla(Ctrl Reload in IE, I think), and purging the page (action=purge instead of action=edit)? Can you paste here the exact text that appears? JesseW, the juggling janitor 00:56, 14 March 2006 (UTC)
- It doesn't work for me either - all I see is the "Wikipedia does not have an article with this exact name.." message. but when I click edit this page, all the content is there. I tried to purge (shift+reload in firefox) but it didnt work -- Astrokey44|talk 02:40, 14 March 2006 (UTC)
- That sounds exactly like a purge problem. Try this URL. Let me know if this works. JesseW, the juggling janitor 03:27, 14 March 2006 (UTC)
Problem with some contributors
There are some contributors who look at this page, but put little effort into fixing the problems. Some contributors simply place templates on user's talk pages telling them about an untagged problem, and make no effort to find the necessary information themselves. One particular user did so to an imae which I uploaded, and I gave him the exact URL of the image, and he still would not apply the correct tag himself. This sends a negative message in a wiki environment and discourages contribution. I think people would have less of a problem with some contributors to the untagged project if they showed that they are willing to put more effort than just applying templates (Which I think are bad anyways). --MateoP 04:48, 17 March 2006 (UTC)
- Look further up this page - 17,000 images still messed up, after months of cleanup effort. It would be completely within WP rules to mark all the images with a bot and delete them en masse in a week. I've cleaned up several thousand myself, and let me tell you, at this point I've lost all interest in hearing people's excuses for making messes and not cleaning up after themselves. We wouldn't even have to have this project if you had done what you were explicitly and repeatedly instructed to do as part of uploading. Stan 06:04, 17 March 2006 (UTC)
- To play devils advocate for a second, some of the images are quite old, back when the image use policy wasn't clear, and we were allowing even non commercial images... Admrb♉ltz (T | C) 06:40, 17 March 2006 (UTC)
- According to toolserver, we've scrubbed everything up to early April 2005; all of those 17,000 images are now known to be less than a year old. Stan 13:20, 17 March 2006 (UTC)
- Irrelevant to the topic. This isn't a "who's to blame for what" topic. --MateoP 06:51, 17 March 2006 (UTC)
- While I sympathise with your point of view, which seems to be that merely alerting uploaders to the untagged status of their images discourages contributions, the basic point remains that, at this point, contributions of images without clear sources and licenses are not something we need more of. Discouraging such contributions is useful, sadly enough. When/if we don't have a 17,000 image backlog asking for more kindness and follow up is more likely to recieve a friendly answer. JesseW, the juggling janitor 08:00, 17 March 2006 (UTC)
- The part about MateoP's chutzpah I find really irritating is that in fact we have acted to save huge numbers of the images. It should only have been necessary to post one note to each active editor; they would then go and fix up all their own uploads, as the notice requests, and our focus could have been on inactives. Most active editors have cleaned up after themselves, but there are some incorrigibles. Stan 13:20, 17 March 2006 (UTC)
- The point, I would have thought, of alerting the uploader of the image problem is two-fold: firstly, so that that uploader can tag the untagged image as they should have initially; and secondly, so that the uploader learns about the importance of image tagging, and will hopefully remember to do so next time. MateoP says he informed a tagger of the source URL, and then expected that editor to update the tag accordingly, when clearly, MateoP should have done it themselves! Far from discouraging contributions, I think the warnings we leave encourage users to contribute, and cleanup and maintain their contributions where appropriate. --Canley 02:39, 21 March 2006 (UTC)
- Irrelevant to the topic. This isn't a "who's to blame for what" topic. --MateoP 06:51, 17 March 2006 (UTC)
I offer the following two templates to help
Template:Owned item and Template:Public place for whatever images they may apply to. knoodelhed 10:07, 21 March 2006 (UTC)
- Neither works as a licence tag, though. Not sure when they'll be of use. --Tagishsimon (talk)
- IANAL, but I don't think private property laws preclude taking photographs from publicaly accessible spots, so I don't really see the point of these templates... ~MDD4696 17:42, 21 March 2006 (UTC)
- To original creator - Please explain further what use you envision for these two templates, as it is not clear to other Wikipedians (witness the above). I have temporarily added a note to them asking they not be used until such a clarification has been made, so we don't have to revert stuff if they turn out not to be useful. I look forward to your response. Thanks! JesseW, the juggling janitor 02:03, 22 March 2006 (UTC)
- Explanation: Templates like these could be used in sub-classification, after an appropriate license has been claimed, to explain the image source, or to assert why the contributor believed he had the right to take or upload the picture (which could be thought important by some jurisdictions). Also would help for brevity of composing the description; e.g., {{owned item}} instead of the many versions of "This thing belongs to me", or {{public place}} in place of "I took this picture in the woods/at the beach/on the street.," etc. - knoodelhed 11:09, 23 March 2006 (UTC)
- Cool. That makes sense. However, I have some suggestions. It would be good to think out what sort of "source templates" we might need, and make many of them at once, rather than piecemeal like this; also, they should be much smaller amd more inconspicous if they are to replace simple text like "I took this in this park". Finally, I strongly think that this work, while certainly interesting and worthwhile, ought to wait until we at least have the licensing tags handled. Thanks for your response. JesseW, the juggling janitor 10:24, 24 March 2006 (UTC)
- Not cool. Beyond academic interest, I do not see that we get anything out of the new templates. Who cares that the spoon that I took a photo of belonged to me, rather than to my local cafe? Meanwhile images tagged with a source tag but not with a licence tag will, I think, evade the new tag detector software & make our search for untagged images that much more difficult. I realy don;t buy into the notion that it is important to justify "the right to take or upload the picture (which could be thought important by some jurisdictions)". Until otherwise convinced, I strongly oppose the proposal & the tags. --Tagishsimon (talk)
- Cool. That makes sense. However, I have some suggestions. It would be good to think out what sort of "source templates" we might need, and make many of them at once, rather than piecemeal like this; also, they should be much smaller amd more inconspicous if they are to replace simple text like "I took this in this park". Finally, I strongly think that this work, while certainly interesting and worthwhile, ought to wait until we at least have the licensing tags handled. Thanks for your response. JesseW, the juggling janitor 10:24, 24 March 2006 (UTC)
- Explanation: Templates like these could be used in sub-classification, after an appropriate license has been claimed, to explain the image source, or to assert why the contributor believed he had the right to take or upload the picture (which could be thought important by some jurisdictions). Also would help for brevity of composing the description; e.g., {{owned item}} instead of the many versions of "This thing belongs to me", or {{public place}} in place of "I took this picture in the woods/at the beach/on the street.," etc. - knoodelhed 11:09, 23 March 2006 (UTC)
Automatic tagging
I'm working on a bot to automatically tag some recently-uploaded images. Currently, it will tag images with no image description page, or images with a template but no additional text, as {{no source}}. Would either of the following be helpful for the image tagging project:
- Adding a {{no license}} template to images uploaded with an external link but no template.
- Adding a {{untagged}} template to images uploaded without a template applied, which would make finding untagged images possible in realtime, without relying on database dumps.
- Sounds like a good idea :) Admrb♉ltz (T | C) 03:03, 22 March 2006 (UTC)
- First of all, I want to thank you for the great work you've done with regards to OrphanBot and your other work in this area. It is greatly appreciated. However, I strongly oppose adding {{no source}} or {{no license}} tags by unmonitored bot. As someone who does a large number of the actual deletions of images marked as {{no source}}, I make the assumption that an actual human being made at least a minimal review of the image before tagging it. Having a bot do so does not seem right to me. However, I would not oppose a distinct tag, say {{no info}} being applied by a bot to images that lack an image description page; it would be fine if that tag used the same categories as the no source tag, but I would want to be able to distinguish bot-added tags from human added ones. Regarding adding {{no source}}-like tags to images with a template but no additional text - depending on the template, this would be OK. Many fair use templates are generally viewed as not requiring a specific statement of the source (i.e. {{logo}}, {{album cover}}, etc.); it would obviously be unacceptable to tag them as {{no source}}. However, some tags (such as {{Non-free fair use in}}) do explicitly require additional text, and tagging {{Non-free fair use in}} images as {{no license}} if they lack additional text would seem OK to me. AFAIK, a {{untagged}} template is not necessary, as we can find untagged images in real time via the toolserver tool (assuming the toolserver is not lagged); a template could not be as up-to-date as that. I would also not be happy with bot-tagging images with a external link but no tag as {{no license}}; I assume that images are only tagged with that tag after a at least minimal attempt has been made to see if the external link provides a license we can use, so bot-tagging them is not a good idea. Responses appreciated, of course. JesseW, the juggling janitor 03:39, 22 March 2006 (UTC)
- Currently, the bot considers {{PD-self}}, {{GFDL-self}}, {{logo}}, {{MoviePoster}}, and any of the "cover" fairuse tags to indicate that the image or tag is self-sourcing. I'm fine with using {{no info}} instead of {{no source}}: the bot doesn't care, and I can format the tag in such a way that OrphanBot handles it properly. --Carnildo 04:07, 22 March 2006 (UTC)
- That seems like a fine list of the self-sourcing templates, so that part seems fine. {{no info}} seems to be already set up as a redir to {{no source}}, so you can just use it instead, and I'll remember that if it says {{no info}} it was added by a bot, and look more carefully when deleting them. Considering Stan's comments below, I suppose an {{untagged}} template wouldn't hurt, although I still think it wouldn't be particularly helpful, so I'm neurtral on that. JesseW, the juggling janitor 05:10, 22 March 2006 (UTC)
- Currently, the bot considers {{PD-self}}, {{GFDL-self}}, {{logo}}, {{MoviePoster}}, and any of the "cover" fairuse tags to indicate that the image or tag is self-sourcing. I'm fine with using {{no info}} instead of {{no source}}: the bot doesn't care, and I can format the tag in such a way that OrphanBot handles it properly. --Carnildo 04:07, 22 March 2006 (UTC)
- I would favor just having a {{untagged}} template. Many of the images at issue have all kinds of info on their description, it's just not in the idiosyncratic form we're asking for. The template is somewhat redundant with toolserver, but touching the image, along with posting a notice on the uploader's talk page, will get more immediate attention - with the current in-your-face style of the upload page, anyone who's still not following instructions is going to need some heavy-handed application of the clue-by-four. Untagged template also facilitates the creation of per-day categories, while toolserver just runs it all together. (Don't get me wrong, toolserver is great, I'm working from it all the time now instead of the lists.) Stan 04:50, 22 March 2006 (UTC)
- What I'm trying to do with this bot is automate the easy cases: images that can be automatically identified as falling into certain categories.
- Based on my testing, and on an earlier check of a day's uploads, about 5%-7% of all uploads have a blank image description page, which is a clear-cut case for identifying it as {{no info}}. Another 7%-10% have a description but no tag, which would be suitable for {{untagged}}, while less than 1% have a tag but no description. --Carnildo 05:30, 22 March 2006 (UTC)
- I will say, it is worthless to tag no source automatically. If we tag the images in this manner then OrphanBot will orphan the images automatically and there is a risk of deletion after doing so much even there is written some information about copyright status that may be useful to tag easily manually. See some images, like: Image:Cuatro.jpg and Image:Arpallanera.jpg and i have got a request from uploader to give some time to tag them properly. I oppose this automated tagging. Suggest further. Shyam (T/C) 18:43, 22 March 2006 (UTC)
- Both of the images you point to were uploaded with a description of the image. The bot would not have tagged either of them as "no source" -- it's only going to work on the most obvious unsourced cases of "no image description at all" and "copyright tag but no supporting information". --Carnildo 19:06, 22 March 2006 (UTC)
- So, any image that had a tag at all wouldn't be acted upon? That was what I thought you meant at first, which seems great, but then after reading some of the other responses I'm not as sure. - cohesiont 19:49, 22 March 2006 (UTC)
- If it's got a tag and any other information at all, or if it's got one of the self-sourcing tags, it won't be acted on. If the image description page consists entirely of a tag such as "{{fairuse}}" or "{{PD-USGov}}" (plus possibly section headers), then it will be tagged as "{{no info}}", although this is a fairly rare case -- most image description pages include a text description of what the image is depicting. --Carnildo 20:18, 22 March 2006 (UTC)
- I tend to think of {{PD-USGov-CIA-WF}} type tags to imply the source, but I have never honestly counted how many are only tagged that way, if you say not many then this seems like a really good idea. It might help a lot of good-faith people uploading at least realize there is a problem in a reasonable amount of time. - cohesiont 05:38, 23 March 2006 (UTC)
- I've added {{PD-USGov-CIA-WF}} to the list of tags recognized as self-sourcing. The number of uploaded images with only a tag is very low -- it seems to be well under 1%. It's much more common for images tagged after the fact, particularly those tagged {{GFDL}}, {{PD}}, and {{Fairuse}}. --Carnildo 07:08, 23 March 2006 (UTC)
- I tend to think of {{PD-USGov-CIA-WF}} type tags to imply the source, but I have never honestly counted how many are only tagged that way, if you say not many then this seems like a really good idea. It might help a lot of good-faith people uploading at least realize there is a problem in a reasonable amount of time. - cohesiont 05:38, 23 March 2006 (UTC)
- If it's got a tag and any other information at all, or if it's got one of the self-sourcing tags, it won't be acted on. If the image description page consists entirely of a tag such as "{{fairuse}}" or "{{PD-USGov}}" (plus possibly section headers), then it will be tagged as "{{no info}}", although this is a fairly rare case -- most image description pages include a text description of what the image is depicting. --Carnildo 20:18, 22 March 2006 (UTC)
- So, any image that had a tag at all wouldn't be acted upon? That was what I thought you meant at first, which seems great, but then after reading some of the other responses I'm not as sure. - cohesiont 19:49, 22 March 2006 (UTC)
- I've created a preliminary version of {{untagged}}. Feel free to modify it as needed. Also, if someone could come up with an appropriate wording for a message to post on the talk pages of people who've uploaded untagged images, preferably with suggestions on where people can go for assistance in tagging images, that would be appreciated. --Carnildo 07:36, 23 March 2006 (UTC)
- I've done a couple of trial runs with OrphanBot. It's examined about 500 uploads and tagged 18 of them as {{no info}}, 4 as {{untagged}}. Images tagged as {{untagged}} are placed in the category Category:Images with no copyright tag, with subcategories for each day. Any suggestions for improving the wording it uses to notify people about image tagging? The current message is at User:OrphanBot/untagged. --Carnildo 07:36, 28 March 2006 (UTC)
- I sort of accept a lot of fair use tags as implying the source as well, like {{movieposter}} and all that type. Maybe I'm being too lenienent with that, what do other people think? Image:Ozma wars03.png is an example, I think some people will get pretty angry if they see that sort of activity. - cohesiont 08:09, 28 March 2006 (UTC)
- For the purposes of applying {{no info}}, OrphanBot recognizes {{movieposter}} as being self-sourcing, on the assumption that the name of the movie is visible on the poster. It also recognizes any tag that explicitly gives a source ({{GFDL-self}},{{PD-self}}, {{GFDL-OpenGeoDB}}, and {{wikipedia-screenshot}}), and tags where the source of the image should be obvious from looking at it (the various cover tags).
- I sort of accept a lot of fair use tags as implying the source as well, like {{movieposter}} and all that type. Maybe I'm being too lenienent with that, what do other people think? Image:Ozma wars03.png is an example, I think some people will get pretty angry if they see that sort of activity. - cohesiont 08:09, 28 March 2006 (UTC)
- I've done a couple of trial runs with OrphanBot. It's examined about 500 uploads and tagged 18 of them as {{no info}}, 4 as {{untagged}}. Images tagged as {{untagged}} are placed in the category Category:Images with no copyright tag, with subcategories for each day. Any suggestions for improving the wording it uses to notify people about image tagging? The current message is at User:OrphanBot/untagged. --Carnildo 07:36, 28 March 2006 (UTC)
- For the purpose of removing no-source images from pages, it recognizes {{PD-self}} and {{GFDL-self}} as self-sourcing tags, flagging any images tagged with them for human review. With anything else, it's assumed that the person applying the {{no source}} tag knew what he was doing. --Carnildo 09:08, 28 March 2006 (UTC)
- Just to doublecheck, this is only checking recent uploads, right? Some vandals seem to find it fun to blank image pages; the toolserver list turns up a couple incidents each day. Stan 13:54, 28 March 2006 (UTC)
- This might turn out to be a really good idea, maybe stopping the people uploading 50 untagged images sooner rather than the usual month-later notification from human editors :) - cohesiont 18:20, 28 March 2006 (UTC)
- Just to doublecheck, this is only checking recent uploads, right? Some vandals seem to find it fun to blank image pages; the toolserver list turns up a couple incidents each day. Stan 13:54, 28 March 2006 (UTC)
- It's checking the past 24 hours once a day right now; once I get the bugs worked out, I plan to change that to checking a much smaller batch every hour or so. --Carnildo 19:05, 28 March 2006 (UTC)
Source had an image previously
I am in confusion, what to do with an image if a source had it previously but it has been removed from the source? Should image also be removed? I assume this would be unfair with the image. But how can someone believe that this image was there or not? Please suggest. Shyam (T/C) 12:37, 22 March 2006 (UTC)
- We should specify the source as it was. Fullstop. If the image is later withdrawn from the source, that has no impact for us whatsoever. Of course, the image must be usable under one of the acceptable licence tags. --Tagishsimon (talk)
- There are a lot of images where the "source" is a gallery of unsourced images. If the image disappears from such a website, it's no longer even possible to determine whether there was any source info to begin with (99% of time not, but sometimes we're surprised). In such cases status is totally unverifiable, and "no source" is appropriate. Copyright holders are people or organizations, not websites, and the real purpose of sourcing is to identify the copyright holder; only when you have that does the website's condition no longer matter. Stan 14:22, 22 March 2006 (UTC)
New experiment
Inspired by the "group by user" report of toolserver, I'm trying a new experiment with some of the bigger uploaders (10+ images) who are still actively editing, namely to post a single note that includes the URL to the toolserver result so they can see their own state, and asking them to clean up the images. No results to report yet. Stan 06:34, 23 March 2006 (UTC)
- SimonATL has cleaned up all his images nicely. Yay! Stan 14:50, 23 March 2006 (UTC)
I'm not sure what to do about this uploader. The user has uploaded about 50 images, apparently all from the same old document. Any suggestions? images JesseW, the juggling janitor 06:47, 23 March 2006 (UTC)
- There are two sets of the same images, uploaded with different names. I amused myself with a bit of research, and the reference to Vaderland in Image:Rsl 15.jpg dates it definitely to 1914 or earlier (see [3]), so PD-US is appropriate. Stan 14:50, 23 March 2006 (UTC)
Orphanbot help
So, as many of you know orphanbot notifies the uploader after we tag the images as no source or no license. Part of the message sends the user to a page if they have any questions. Right now it's sending people to User talk:Carnildo/images where a few of us help answer questions. I was wondering if people would be ok with this page being a subpage of Wikipedia:Untagged images or some other place. I would like to get it out of user talk space and into wikipedia or even help space. This would be a centralized place for people to ask about images they were notified about (same as it is now, just new location). Any thoughts about where the page should be, etc? I have asked Carnildo already and he is ok with moving it. - cohesiont 10:05, 24 March 2006 (UTC)
- That sounds like a great idea. We can put it pretty much anywhere; i.e. Wikipedia:Image legality questions would be a fine name. JesseW, the juggling janitor 10:21, 24 March 2006 (UTC)
- That sounds good, I didn't think of that, but really it makes sense to have one centralized question area for all these notifications, orphanbot, roomba (does it notify people?) user added templates {{image source}}, and {{image copyright}} could all point to it. - cohesiont 19:49, 24 March 2006 (UTC)
- It would also prevent a lot of the personal attacks against Carnildo. The JPS 23:56, 25 March 2006 (UTC)
- Ok, I started the page, nothing links to it yet, I thought I should probably give it some time to see what people want to change, I copied a lot from the Wikipedia:Newcomers help page but if anyone has any ideas change away :) - cohesiont 09:21, 26 March 2006 (UTC)
- It would also prevent a lot of the personal attacks against Carnildo. The JPS 23:56, 25 March 2006 (UTC)
- I've modified OrphanBot's notification messages to point to the new page. People should start showing up in the next day or so. --Carnildo 00:10, 27 March 2006 (UTC)
- Ok, I will change the {{image source}} and {{image copyright}} to point there also. There is a a link to include instructions for people helping answer questions, if anyone wants to put helpful tips there go ahead, if not I might just delete it, I don't know if we will really need instructions for the helpers. - cohesiont 06:07, 27 March 2006 (UTC)
- That sounds good, I didn't think of that, but really it makes sense to have one centralized question area for all these notifications, orphanbot, roomba (does it notify people?) user added templates {{image source}}, and {{image copyright}} could all point to it. - cohesiont 19:49, 24 March 2006 (UTC)
Real progress
Just so people don't get too happy that we've finished the static lists(even though that is great...) as of now, according to the up-to-date tool, we have 14,151 untagged images left. JesseW, the juggling janitor 03:25, 26 March 2006 (UTC)
- Wow, thanks for cheering us all up! :P haha - cohesiont 09:08, 26 March 2006 (UTC)
As we are now down to 400 items, I'm going to make new static lists; don't anyone else start, please... ;-) JesseW, the juggling janitor 01:08, 27 March 2006 (UTC)
- I'm formatting them back to the old way so people can use the image auto tagger. Does anyone have strong feelings that they should stay the new way? - cohesiont 10:21, 27 March 2006 (UTC)
- No, there's no need for the format to be changed. I only used the new format because that was what was easily available from the tool; we should just ask the maker of the tool to add an option to output in the old format. Thanks for fixing it. JesseW, the juggling janitor 10:59, 27 March 2006 (UTC)
As of 2006-03-29 23:14, the tool says we have 13,664 (a decline of 487). Slow, but we are at least still making progress... JesseW, the juggling janitor 02:05, 30 March 2006 (UTC)
- As of 2006-04-01 16:27(replication lag), we have 13,029 (a decline of 635). Continuing to make progress. JesseW, the juggling janitor 23:49, 2 April 2006 (UTC)
- Of course, there are also the 151 images in Category:Images with no copyright tag. --Carnildo 04:27, 3 April 2006 (UTC)
- Please add a note about that to the main page, so we can keep it small... JesseW, the juggling janitor 06:07, 3 April 2006 (UTC)
- Have you started the autotagging thing in production yet, or still testing? - cohesiont 06:21, 3 April 2006 (UTC)
- Large-scale testing right now, doing around 500 images a day. I'm trying to build up a fairly comprehensive list of self-sourcing/non-sourcing/other templates before I fully automate it. --Carnildo 06:59, 3 April 2006 (UTC)
- Sounds good, we might need to think about 2 things, when the category really starts filling up what should we do about it? Hopefully the tag itself will help a lot of people know they need to do something else, but what if they don't? Also, I don't think the untagged image tool will say they are untagged anymore if they have the {{untagged}} tag on them. (that sentence sounds crazy) maybe we could talk to meta:User:Duesentrieb and see if he could add it to the list of tags to ignore. - cohesiont 07:17, 3 April 2006 (UTC)
- All the images upto 2005 are tagged now. We have untagged images under 10,000 according to our manual list. Shyam (T/C) 20:00, 6 April 2006 (UTC)
- That is very good news. Congratulations! JesseW, the juggling janitor 23:06, 6 April 2006 (UTC)
- All the images upto 2005 are tagged now. We have untagged images under 10,000 according to our manual list. Shyam (T/C) 20:00, 6 April 2006 (UTC)
- Sounds good, we might need to think about 2 things, when the category really starts filling up what should we do about it? Hopefully the tag itself will help a lot of people know they need to do something else, but what if they don't? Also, I don't think the untagged image tool will say they are untagged anymore if they have the {{untagged}} tag on them. (that sentence sounds crazy) maybe we could talk to meta:User:Duesentrieb and see if he could add it to the list of tags to ignore. - cohesiont 07:17, 3 April 2006 (UTC)
- Large-scale testing right now, doing around 500 images a day. I'm trying to build up a fairly comprehensive list of self-sourcing/non-sourcing/other templates before I fully automate it. --Carnildo 06:59, 3 April 2006 (UTC)
- Have you started the autotagging thing in production yet, or still testing? - cohesiont 06:21, 3 April 2006 (UTC)
- Please add a note about that to the main page, so we can keep it small... JesseW, the juggling janitor 06:07, 3 April 2006 (UTC)
- Of course, there are also the 151 images in Category:Images with no copyright tag. --Carnildo 04:27, 3 April 2006 (UTC)
As of 2006-04-10 06:54 (replication lag) we have 10,978 according to the tool(a decline of 2,686). We are still making clear progress. Cool! JesseW, the juggling janitor 19:49, 10 April 2006 (UTC)
- Hm. As of 2006-04-18 11:43 (massive toolserver lag) we have 15,284 - but all the spot checks I made had already been fixed, so I suspect technical problems on the toolserver are producing the oddly high number. In any case, it's alarming. (5,000 more than we had last time I checked...) JesseW, the juggling janitor 02:43, 25 April 2006 (UTC)
- Replag seems to be hovering between five and seven days. --Carnildo 06:48, 25 April 2006 (UTC)
Is up and running, the tags {{image source}} and {{image copyright}} link there as well as orphanbot, so if anyone feels like helping people that's the place :D - cohesiont 06:33, 28 March 2006 (UTC)