Commons:Village pump/Proposals/Archive/2018/08

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

Add syntax highlighter gadget

This section was archived on a request by: —⁠andrybak (talk) 17:03, 24 August 2018 (UTC)

Previously wikEd was proposed, but rejected. I propose enabling Syntax highlighter on Commons. It would be very useful, for example for editing complex templates, like Template:Aircraft cat/layout. —⁠andrybak (talk) 21:26, 23 August 2018 (UTC)

 Comment We already have the other syntax highlighter made by WMF, but Remember the dot's version isn't bad to use. --George Ho (talk) 04:56, 24 August 2018 (UTC)
 Comment I previously gave the "other" highlighter a try but switched it off soon because it does not allow for middle-click pasting of text on Linux (which is a killer bug for me but not in the power of WMF developers to fix because it occurs somewhere upstream). This one doesn't seem to have that issue, so I guess I'll keep it enabled for a while to test it. --El Grafo (talk) 08:49, 24 August 2018 (UTC)
George Ho, I didn't know about this button: . That's exactly what I needed, thanks! —⁠andrybak (talk) 17:03, 24 August 2018 (UTC)

Deletion of the photograph on the wikipedia page

We are creating a Wikipedia page for a spiritual master from India Dr. D G Vinod, aka Maharshi Vinod. When we tried to upload his picture on the Wikipedia Commons, the Krdbot deleted it as it found the same picture on the website http://www.maharshivinod.org/ . This deletion has happened twice and we have received an email informing us that it is not a good practice to recreate the deleted content and repeated violation of this Wikipedia policy from our side may result in our loosing editing rights.

Kindly note that we work for the institute ‘Maharshi Vinod Research Foundation’, based in Pune, India and the website http://www.maharshivinod.org/ belongs to us. Recently, we have taken a new initiative to create a Wikipedia page for Dr. D. G Vinod aka Maharshi Vinod as we have been told that through Wikipedia we will be able to reach many more people than through our website. Kindly note that we are not violating any copyright as all the photographs are our own property.

Can anyone help us on how to overcome this problem? — Preceding unsigned comment added by Shantimandir mvrf (talk • contribs) 10:05, 3 August 2018 (UTC)

Hello, I think that it is best to resolve this via COM:OTRS, this way no deletions will be taking place in the future. ℺ Gone Postal ( ) 10:36, 3 August 2018 (UTC)
When you write that in English Wikipedia, remember to acknowledge your conflict of interest on the relevant talk page, cite sources for your facts (and even more for anything that amounts to an opinion, e.g. remarks on his importance, greatness, etc.). - Jmabel ! talk 20:37, 3 August 2018 (UTC)

Restrict usage of Flickr2Commons

I've seen the Flickr2Commons tool abused fairly often lately, where users upload large numbers of files without categorizing them and without checking for Flickr washing and derivative works. The tool is also unmaintained and uploads many duplicates. I propose that we restrict the rate of uploads via edit filter 207. Extended uploaders, license reviewers, and admins would be exempt from this. If someone has a good reason to rapidly upload using this tool, they can request extended uploader status, and we can then revoke it if it's abused. Guanaco (talk) 12:00, 1 August 2018 (UTC)

  •  Support This tool caused a lot of issues yet and has been abused multiple times. The proposal is reasonable. --Steinsplitter (talk) 12:09, 1 August 2018 (UTC)
  •  Oppose Low rate limits, like 5 files a minute, for accounts with under 2,000 pre-existing uploads would be a good idea. Restriction of full use defined by other rights and groups seems too like a club membership model of governance. -- (talk) 12:17, 1 August 2018 (UTC)
  •  Oppose The limit proposed by the OP won't solve the purported problems; but will hinder the work of those of us who use the tool properly. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:36, 1 August 2018 (UTC)
  •  Oppose Why deny those (new or even long term contributors) who have valid reasons to use it, just because a small minority abuse or misuse it? Bidgee (talk) 12:56, 1 August 2018 (UTC)
  •  Comment To be clear, everyone who has commented here would be unaffected by this. Extended uploader is almost always granted upon request, and it does not have a particularly high standard: "Extended uploaders are supposed to be familiar with the project scope, and especially with the file license policy." When it's been denied, it was because the user had recent copyvios, had almost no edits, or didn't actually need the rights for anything. I expect anyone hindered by this would ask to be an extended uploader, and we'd summarily grant their request in most cases. Guanaco (talk) 13:21, 1 August 2018 (UTC)
  • Last time I looked, I only had "Image reviewer", it was something I had to ask for and those who have been on Commons for awhile may not even have the "standards" (rights) you're proposing and again why treat all new contributors as evil? Bidgee (talk) 13:46, 1 August 2018 (UTC)
    Image reviewer and license reviewer are the same thing, and "extended uploaders, license reviewers, and admins would be exempt from this." I don't think this would be treating them as evil, but rather asking them to check with someone before running a potentially disruptive script. Guanaco (talk) 15:04, 1 August 2018 (UTC)
  • @Guanaco: Would you agree that a minimum of 1,000 or 2,000 pre-existing uploads would be a fair way of constraining use of the tool, with users below this count expected to put in a request for an exception? -- (talk) 14:14, 1 August 2018 (UTC)
    @: I don't see a problem with enabling it for users with that many uploads. Unfortunately upload count doesn't seem to be an edit filter variable, so we'd have to develop an updated tool. This could be the best solution overall, as a new tool could provide better guidance in the interface and also check for duplicates and incomplete uploads. Guanaco (talk) 15:04, 1 August 2018 (UTC)
  •  Oppose This is a very bad idea. Every time somebody makes the proposal to do something in the name of stopping abuse there needs to be a question "How will this impact those users, who do not abuse the tool?" clearly such a question was not asked here. "[T]hey can request extended uploader status" is not a sufficient solution. Wiki was originally the place that "everybody can edit", but we are becoming more and more like English "Wiki"pedia. Restricting good users to punish a few that happen to abuse the tool (or maybe not even abuse, but simply make a mistake, let's not forget AGF) is counterproductive endeavour. ℺ Gone Postal ( ) 13:50, 1 August 2018 (UTC)
    I think everyone can and should be able to edit and upload to Commons, and I see many problems with bureaucracy and instruction creep at English Wikipedia. However, high-speed bot-like uploading is something that should be regulated to some degree. I don't want to punish anyone, but the reality is that one clueless user's good-faith mistake means someone has to sort through dozens of files, create DRs, then someone else has to close them. This cleanup can waste many hours of volunteer time. Guanaco (talk) 15:04, 1 August 2018 (UTC)
     Weak support But only if we find some sensible limits. I do agree that bot uploads can and should be monitored more closely than personal uploads because they have the potential to do more damage, but they also can improve something (there was a reason why such tools were created in the first place). ℺ Gone Postal ( ) 19:12, 1 August 2018 (UTC)
  •  Support, I have seen far too many duplicates and laundered files pass through this tool, and so far developer @Magnus Manske has been unwilling or unable to stop the duplicates per these issues. See also Commons talk:Flickr2Commons#Flickr2Commons does not check for duplicates (or ignores them) and Commons talk:Flickr2Commons#duplicates.   — Jeff G. ツ please ping or talk to me 14:37, 1 August 2018 (UTC)
And now User talk:Magnus Manske#Flickr2Commons does not check for duplicates (or ignores them). I can't believe there is *nobody* who can fix this. Even I might be able to figure out the culprit, but even if I did, I wouldn't be able to apply a patch. - Alexis Jazz ping plz 17:22, 1 August 2018 (UTC)
I might be able to figure stuffs out if it wasn't the intersection of bitbucket, php, and complexity. I wonder, if this change will be implemented, should F2C be aware of it and operate just on the margin of the rate limit? --Zhuyifei1999 (talk) 19:00, 2 August 2018 (UTC)
  •  Oppose the proposed limit of 3 files per minute (if I understand correctly),  Support something like a limit of 200 files per hour. That would allow a user to import a whole shoot, then categorize it before moving on to the next. @Guanaco: "To be clear, everyone who has commented here would be unaffected by this." I would be affected by a limit of 3 files per minute. - Alexis Jazz ping plz 17:22, 1 August 2018 (UTC)
    @Alexis Jazz: I meant everyone who had commented here at the time. If you want extended uploader, let me know because I think you are qualified. Anyway, 200 files per hour is problematic because it's likely to break in the middle of a large batch upload, and 100 unchecked images per hour from one user is still disruptive. Guanaco (talk) 23:04, 1 August 2018 (UTC)
  •  Support If updating the tool was possible that would be best as it would allow us to update and patch the tool since that seems to be an obvious issue. If not I still support. It isn't hard to be added to the extended uploaders group. That is what I believe is meant by anyone here wouldn't be affected. -- Sixflashphoto (talk) 17:39, 1 August 2018 (UTC)
  •  Support, I come across many files are out of scope that are from a mass upload from it.--BevinKacon (talk) 20:57, 1 August 2018 (UTC)
  •  Comment, restricting the bot won't solve the duplicate issue (which is very annoying, how could we fix this???). Since its a really useful tool for uploading from flickr it would be a shame to limit its usage. (and a filter which blocks uploads of sunsets and sunrises would be great /s) Amada44  talk to me 21:10, 1 August 2018 (UTC)
  •  Support I have spent far too much time cleaning up mass uploads from numerous users too lazy to name, describe, and categorize their uploads. F2C is an extremely powerful tool - it's essentially a bot - and using it for unchecked mass uploads is abuse. It is exactly the kind of tool that should require advanced permissions - if users cannot show that they apply useful filenames, useful categorization, and useful descriptions on small upload runs when they have plenty of time to do so, why the hell should we allow them to do sloppy uploads en masse? Pi.1415926535 (talk) 22:27, 1 August 2018 (UTC)
  •  Support I did one crappy F2C run and spent a week and a half cleaning up my mess. We should at least be able to vet users to the point where we know they're the type of people who will clean up the mess if they leave one, and not leave it for someone else to do. GMGtalk 22:55, 1 August 2018 (UTC)
  •  Comment we should wait for the filter to generate a few more hits, but it looks like only a handful of users are going to be seriously affected by this. (and many occasional uploaders of a few dozen images that never caused any problem anyway) Some of those will be granted an extended uploader flag, leaving even fewer users. Wouldn't it better to educate (or block..) those users instead instead of creating an abuse filter? - Alexis Jazz ping plz 00:29, 2 August 2018 (UTC)
Users that refuse to properly clean up their mass uploads generally have no interest in putting in the effort, no matter how much they are reminded. (That includes some otherwise very respected editors, but I digress.) It is better to prevent them from accessing extremely powerful upload tools until they demonstrate that they will use them properly, than face a nasty choice between blocking them or continuing to allow them to abuse those tools. Pi.1415926535 (talk) 00:39, 2 August 2018 (UTC)
We could try keeping the filter as is (log only), and grant extended uploader to those who often use the tool correctly. Those who abuse it could have it disabled by username with a different filter. Guanaco (talk) 01:02, 2 August 2018 (UTC)
Guanaco, this is what I wrote before reading your reply: "Since we seemingly have no control over Flickr2Commons, how about using the abuse filter to target those problematic users instead of everyone? And maybe target uncategorized Flickr2Commons uploads from anyone else." I have no issues with using the filter to find problematic users.
DSC 4536, this is a bird.
DSC 4535, this is another bird.
Pi.1415926535, I think it just won't work. Look at these birds for example: uploaded by @Rudolphous: . Photo: quite poor, not really likely to be used in an article unless there really is nothing better. Filename: useless. Description: useless. Categories: Category:Birds in Artis (Artis is a Dutch zoo). The photos don't show anything really typical for that zoo, so the only useful information here is "birds". That's very nice, I'll surely find them next time I search for "birds". This abuse filter will be of no help for this because Rudolphous is an administrator. You are probably thinking "well surely this is an exception". You be the judge: User:Alexis Jazz/Great images with captions. - Alexis Jazz ping plz 04:15, 2 August 2018 (UTC)
Point taken. I improved the Artis pictures, hopefully that shows some willingness. Rudolphous (talk) 06:29, 3 August 2018 (UTC)
  •  Support without fix or limitation, this is an spam tool. -- User: Perhelion 21:15, 2 August 2018 (UTC)
  •  Support F2C is a useful tool but Flickr is not a problem-free repository. In addition to Flickrwashing, there are other recurrent problems such as uploads with only PD Mark 1.0 (but no explanation of why a file is in the public domain). The concerns expressed by some users above and below are understandable but we need to see some progress on this issue. Green Giant (talk) 11:23, 13 August 2018 (UTC)
@Green Giant: but this filter (if it would actually limit actions, tagging is fine) won't solve the problems. One problem is that Flickr2commons doesn't check for duplicates properly. While the uploader has some responsiblity here as well, it's a bug. If there is a hole in the street, you don't fine people for falling into it, you close the hole. The fact that this is seemingly impossible should really worry everyone around here. If for whatever reason Flickr2commons breaks entirely, there simply will no longer be Flickr2commons because there is nobody who can fix it and Magnus Manske is on permanent holiday or something. Magnus is allowed to do that, but with nobody to take his place we have an issue that nobody seems to be getting the gravity of.
The other is that problems are not exactly limited to newbies. Even some admins are not using the tool properly. I'll analyze some of the filter hits below, it's clear this filter should not limit actions. - Alexis Jazz ping plz 15:35, 13 August 2018 (UTC)
@Alexis Jazz: Whether or not an upload tool has bugs, the responsibility for checking and fixing uploads lies entirely with the uploader. However, some people seem to think their only responsibility is to upload huge numbers of files, leaving clean-up to others. I don’t particularly like real-world analogies for wiki-problems but let’s say we think of F2C as a knife. If someone new to knives accidentally cuts their finger, we don’t blame the knife manufacturer. Instead we take safety measures to try to reduce accidents. The filter is a safety measure to reduce the number of poorly-managed uploads. To be honest, I believe other powerful tools like Cat-a-lot and VFC should also be limited-access because there is potential for large numbers of bad edits. Green Giant (talk) 16:17, 13 August 2018 (UTC)
@Green Giant: I agree in part with you, but when the knife manufacturer refuses to fix problems with their knives, we shouldn't ban knives. The filter isn't sufficiently selective for bad uploads. Too bad the knife manufacturer more or less has a monopoly.. - Alexis Jazz ping plz 16:21, 13 August 2018 (UTC)

Example

Commons proudly holds over 600 exciting and high quality photos of Ontario Highway 401.--BevinKacon (talk) 19:27, 12 August 2018 (UTC)

The example of 600 photographs of Highway 401 highlighted by BevinKacon is a good one. A justification for considering the value of uploading lots of photographs of the same object that a casual viewer may think boring or redundant. The highway is over 500 miles long and the most widely used reference set for viewing highways remains Google streetview. The photographs we have in this collection on Commons provide a valuable alternative collective view of the highway over different years and are guaranteed to be freely reusable, unlike Google streetview images which can only be justified under Fair Use. Over time, such collections have increasing archive value, see Category:John Margolies Roadside America Photograph Archive, a historic collection of 9,000 photographs of American roadside buildings and places. -- (talk) 19:56, 12 August 2018 (UTC)

Your argument may hold true, if they all had geo data (they don't), and they were actually being reviewed, named and in put their correct category. Instead we get snow with nonsense file names, File:-6 365 06.01.2011- Snow! (5332045236).jpg, the sky File:Dramatic Sky (5608043948).jpg, blur File:Commuting Warp (6088536421).jpg, File:Ontario Highway 401 (27356726050).jpg a back of a sign.--BevinKacon (talk) 20:14, 12 August 2018 (UTC)
Snow track photo is a bit dark but fine if someone is looking to illustrate driving in snow. The dramatic sky is exactly what is says and is atmospheric. The Commuting Warp is arguable, but it is not an accidental blur but a deliberate effect by using long exposure, could be used to illustrate heavy nighttime highway driving. The back of a sign may seem bizarre, but it does tell us who has the contract for the signage on this part of the highway, the town the sign is placed within and confirms the year of construction. If these are the worst examples, it's quite an interesting set. -- (talk) 16:28, 13 August 2018 (UTC)
@: I added Category:Tire marks on snow to File:-6 365 06.01.2011- Snow! (5332045236).jpg, changed the white balance, cropped it and requested a rename. Far too much work for a photo that will never be used because File:Tire tracks in new snow at a parking lot.jpg and File:Impression left by snow tire rolling through snow.jpg are better. - Alexis Jazz ping plz 18:50, 13 August 2018 (UTC)
@Alexis Jazz: For what it's worth: voy:en:Winter_driving#Tires. —Justin (koavf)TCM 19:02, 13 August 2018 (UTC)
If you look at my editing history, you'll find that:
  1. I perform batch uploads, applying the most sensible top-level category for that upload (ie - one most applicable to all files being uploaded)
  2. Shortly thereafter (sometimes the same day, sometimes later) I inspect that category to more carefully sort the files into sub-categories
  3. I nominate for deletion the files that should have been filtered out during the Flickr2Commons upload, but that I didn't catch at the time (for example, File:Commuting Warp (6088536421).jpg mentioned above); although Flickr2Commons is useful, it has its failings (it doesn't state file size, whether the pic is a screenshot of a video, whether the file exists on Commons with a different name - it does catch same name uploads, ...)
Also note that files do not need to be geotagged in order to be more precisely sorted. I've refined the category for most files in Category:400-series highways (Ontario) by inspecting the signage on the pics, and correlating similar files by timestamp; you can see this in Category:Ontario Highway 400, such as on File:400 North (4394817913).jpg. The fact that I have not yet processed this current batch is immaterial, as there are no deadlines on Commons. (BTW: Highway 401 is over 800 km long, and I uploaded these files because they represent segments of the highway with no coverage on Commons; it shouldn't just be pics of the highway passing through a few major cities). As a final note, I'd appreciate being invited to a discussion about me, instead of having to find it on my own. Mindmatrix 13:40, 16 August 2018 (UTC)

Hi Fae, can you do these ones too? All media needing categories as of 2016 from Flickr. Thanks.--BevinKacon (talk) 21:09, 16 August 2018 (UTC)

Block individual users from flickr2commons

Given the concerns above, my opinion now is we should attempt a less restrictive measure. I've changed edit filter 207 to simply tag all flickr2commons uploads as such, and I created edit filter 208. This new filter allows users to be blocked from using the tool in response to disruption or misuse. When placing a restriction, an admin must always notify the user of the action on their talk page, and they should also log the date and reason for the action in the edit filter description.

This isn't as clean or as pretty as the upcoming partial block feature, but these two filters should let us easily observe and address misuse without resorting to a full site-block. Is there support for allowing admins to restrict individual users in this fashion? Guanaco (talk) 06:49, 2 August 2018 (UTC)

  • My concern about this is that it creates some potential nasty conflicts between experienced users. Based on Alexis's comments above, I would probably block Rudolphus from this tool, as well as several other admins. And while that would stop the disruptive use, I've now pissed off an admin. So any restrictions on F2C, while extremely necessary, need to be implemented in a way that allows admins and other otherwise-well-regarded and well-connected users to be blocked from the tool without resulting in wheel-warring or other arguments. Pi.1415926535 (talk) 23:06, 2 August 2018 (UTC)
    I think we have to handle this in the same way as blocks and other sanctions. If it's likely to be controversial, a noticeboard discussion should happen first, or immediately after if the problem is urgent. Also, we should be able to trust admins to respect consensus. If several people say "Stop what you're doing," an admin should be willing to listen. On the other hand, poorly categorized but in-scope files are arguably a net gain. It's similar to the old inclusionist vs. deletionist divide, and there aren't any easy, catch-all solutions. But in general when dealing with long-term good faith editors, discussion solves more problems than hostile admin actions. Guanaco (talk) 03:14, 3 August 2018 (UTC)

Analyzing some filter hits (it's more amusing than it sounds, trust me)

It looks like the Guanaco altered the filter to tag all Flickr2commons uploads, so one admin also ended up in this list.

Unfortunately the source doesn't have much of a description. // sikander { talk } 15:43, 13 August 2018 (UTC)
@Sikander: based on what you can see, you can write a slightly better description yourself. The description you added is a considerable improvement. I expected Category:Women wearing head-mounted displays to already exist, but it didn't, so I created it. - Alexis Jazz ping plz 17:36, 13 August 2018 (UTC)
@Howard61313: Flickr2commons shows the filename like "432 Park Avenue à Manhattan (20158908745).jpg" which you can change right there. - Alexis Jazz ping plz 16:26, 13 August 2018 (UTC)
This section was archived on a request by: This has revealed some interesting information, but there's clearly no consensus for my original proposal. We clearly must develop better tools, including duplicate detection and a reworked and properly maintained flickr2commons. Structured data and partial blocks may help with this issue. For cases where a particular user must be stopped from using flickr2commons, Special:AbuseFilter/208 is available. Guanaco (talk) 01:03, 14 September 2018 (UTC)

No longer allow GFDL for some new uploads

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

I'm aware of the odds for this proposal. I'm proposing it anyway, perhaps it'll pass, otherwise we may at least get some idea of where the community stands on this today. Any further restrictions as well as dealing with GPL may be done with other proposals as those things are more complicated. Last time When this was proposed in 2013, it resulted in too many words, so I'll try to limit the scope. If some people want to scream at me for this, I expect nothing less.

For those who are not aware, the GNU Free Documentation License is the license Wikipedia used before 2009. It was designed for software documentation. But Commons still allows uploads with nothing but a GFDL license today.

The main issue with GFDL is that it requires re-users to include 7 pages of text, which is generally not practical for individual pictures, video or audio. Because of this, this lead to the existence of {{GFDL 1.2 or cc-by-nc-3.0}} which is, in its very essence, a backdoor Creative Commons Attribution-NonCommercial license. After all, nobody is going to include 7 pages of text.

The proposal: GFDL alone will no longer be a valid license for uploads that meet the following requirements.

  • Licensed as GFDL one month after this proposal, if passed, becomes policy with the option to extend this to two months if problems arise. This gives existing projects and users a little bit of time to decide on a new license. Anything licensed as GFDL before is not affected by this proposal. What counts is the license date, not the upload or creation date.
  • Main or only content is a photograph, painting, drawing, audio or video.
  • Is not a software logo, diagram or screenshot that is extracted from a GFDL software manual.

So uploading a GFDL software manual will remain acceptable, anything already licensed as GFDL is not affected, software logos, diagrams and screenshots extracted from manuals are not affected. And a dual license like CC BY+GFDL is also no problem. What will be affected are files like File:Wedge tail eagle flight Jan13.jpg, File:19870300-skoda-100-bw-011.jpg and File:National March on the NRA 8045524.jpg if they weren't already licensed as GFDL before this proposal comes into effect. So the eagle, Skoda and national march will stay, but this practice would no longer be allowed for licensors in the future. - Alexis Jazz ping plz 07:34, 7 August 2018 (UTC)

  •  Support By Foundation:Resolution:Licensing policy, we are expected to judge the freedom of licenses/works against the Definition of Free Cultural Works 1.0. This states that the four freedoms "should not be restricted by the context in which the work is used". When applied to a photograph, the GFDL prevents the photograph from being used in any context where including the whole license text is impossible, such as on postcards, t-shirts, billboards; so that GFDL can't properly be considered free for photographs. These requirements were designed and make sense for books, so continuing to upload genuinely GFDL books is fine. BethNaught (talk) 08:31, 7 August 2018 (UTC)
Take a look at something like this: File:Gzilla icon.png this is a 51x51 icon that requires attribution. I don't know how it is for you, but for me the author field currently is significantly larger than the image itself. This restricts the use of this file. It really does. I would not put it on a coin right now for example, it would fit, but the attribution wouldn't. If your thing is about finding something that is useful then read this same proposal of 5 years ago. I saw people saying there that what we need is the ability to easily filter content by licences. However, it is much easier to just say "I don't like it, so others must stop doing it" then to find a way to make something better. The way I see it we could have something like a current cat-a-lot javascript window, where you have "All licences" selected by default, but you can simply select or unselect some fields: "Requires attribution", "Viral", "Requires licence text", then "Forward compatible with ..." and "Backwards compatible with ...", as well as simply selecting a specific licence. This would make it possible for people to find files that are licenced as they need. ℺ Gone Postal ( ) 14:54, 21 August 2018 (UTC)
Bit of a stretch, but "Design: Larry Ewing and Raph Levien" will fit on a coin. - Alexis Jazz ping plz 21:50, 21 August 2018 (UTC)
You are unable to see the forest because of all of those blasted trees that get in your way. The exact image was an example. All the Attribution licences require you to list all the authors, what if the work was constantly improved that the list of the authors is now longer than GFDL licence. Does this mean that the work is now unfree? Of course not. The fact that it is somewhat inconvenient for some specific purpose does not invalidate the "freeness" (freenity, freeousity?) of some licence in general or file specifically. ℺ Gone Postal ( ) 01:34, 22 August 2018 (UTC)
That's one viewpoint, but nothing involving Gnu is ever "as simple as that".
You can also interpret this as, "If there's a valuable resource which is published under GFDL, no, you can no longer use it on Commons (or maybe even a Wikipedia)". That's quite restrictive. Commons is not merely a source of content, it's also a consumer of content. Should we limit its ability to consume GFDL material? Andy Dingley (talk) 11:21, 9 August 2018 (UTC)
If you want an example for that scenario: I could obtain a permission from the somewhat unwilling photographer of the then (2007) only photography of murdered Theo van Gogh, File:TheoVanGogh.jpg, only by offering the somewhat "unsuitable" GFDL license. --Túrelio (talk) 12:31, 9 August 2018 (UTC)
Túrelio, Andy Dingley your arguments aren't valid. We don't accept licences on the basis of "oh but we will miss out on X possible content". We accept the licences because we are a free content project that offers free content with as few strings attached as possible. There is plenty content with -NC and -ND restrictions, plenty where the photographer would wish for some mythical "Educational use only" licence or a "Wikipedia use only" licence. I've even known a photographer who wished to prevent creationists and socialists from using their photos. If we allowed all those, we'd be overflowing with images to illustrate Wikipedia, but would no longer be a free content project. The "consumer of content" idea Andy suggests is not our mission, and a bit of an odd way of describing the transfer of free images to Commons. There are millions of educational photos, including by large organisations, with -NC and -ND restrictions, whereas there aren't really any off-Wiki websites where GFDL is used for photos. So we already impose a massive restriction on our educational content. The fact that you had to go back 11 years to 2007 to find one example vs all the photos in the British Museum collection with a -NC licence shows this is a bad argument. -- Colin (talk) 14:06, 9 August 2018 (UTC)
They're not invalid arguments, it's a judgement call. But nor is it an obvious or one-sided one.
I always warn people off using GFDL. But, there is content which has already been licensed as such, and isn't viable to change. A policy change like this would exclude Commons from using such content.
By all means, advise people against choosing GFDL as a new licence - but to forbid it on existing content is to exclude content with a very small tradeoff (unlike -NC or -ND, which are much greater). Andy Dingley (talk) 14:13, 9 August 2018 (UTC)
I think the tradeoff is actually smaller when comparing to -NC and -ND. For many photos, there is no point in creating derivatives. I think a crop doesn't even count as a derivative, but I'd have to look that up. And most files are never used commercially, so it's not hard to argue that should be accepted as well. But it's not what Commons does.
"but to forbid it on existing content is to exclude content with a very small tradeoff"
Content that is already licensed with the GFDL is not forbidden. To allow people to license existing content (like, old unpublished photos) as GFDL in the future would be arbitrary and create a loophole. - Alexis Jazz ping plz 15:24, 10 August 2018 (UTC)
Andy it is an invalid argument and you again make it. There are millions of files licensed -NC and aren't going to change. I don't see you jumping up and down demanding that we start accepting -NC. There are like 3 photos off-Commons licensed with GFDL and probably 2 of those were an accident. Basically nobody uses it for photos apart from a small handful of Austrian anti-free-content Wikipedians. But that doesn't stop trolls recommending to GLAMS that it is a really bright idea. You aren't making a "judgement call" you are setting up a totally false argument that we should keep this license in case Commons loses out on a few images. So please don't keep repeating "would exclude Commons from using such content" because that problem is a rounding error from 0. The real issue is existing Wikipedians who experience has shown will actually use CC licences when they need to (e.g. to get a Commons FP). They just prefer not to. So the effect I see of this is that actually there will be more free content on Commons and fewer users going "Why does this photo of a mountain have a licence that refers to a text computer software manual?" Please point me to the huge collection of GFDL photographs off-Commons that we would miss out on uploading? -- Colin (talk) 17:37, 10 August 2018 (UTC)
  •  Support per Notafish and BethNaught. --De728631 (talk) 20:50, 7 August 2018 (UTC)
  •  Oppose This makes no sense. Let's say we have two persons A and B, currently A can upload A's (own) work under GFDL and can upload B's work that is uploaded elsewhere and licenced under GFDL; B can similarly do the same. If this proposal passes, these two persons cannot upload their own work, but can still with no problem upload each other's work to the project as long as they first place it within a manual ("You can scan this picture (see pic 1) into Paint and save it with File->Save. The end.") The point is that there are free licences are that not useful for some purpose, what we need is the ability to search for content by the licence and the clear understanding that if the files are licenced differently they cannot be considered duplicates of one another, since for some purposes they cannot be substituted with the other. ℺ Gone Postal ( ) 21:48, 7 August 2018 (UTC)
Can you give examples of this actually happening? Commons is probably the only place on the internet where anyone is still uploading images that are GFDL and somehow not CC-BY-(SA). The tiny potential for license incompatibility issues is a much smaller problem than the current abuse of GFDL. Pi.1415926535 (talk) 22:33, 7 August 2018 (UTC)
(Edit conflict) The manual exception is limited to software logos, diagrams, and screenshots. This wouldn't include photographs, which are the majority of our content. I do see a small loophole for logos and diagrams, but I doubt anyone would actually use it except to troll. We can deal with such rare nonsense case by case. With Commons:Structured data (and even now, with PetScan, though this tool can be hard to use), searching by license will be feasible, but it doesn't solve the problem. I believe if you see something here, you should be able to use it! This is our purpose, the thing that truly distinguishes us from other media-sharing sites. If someone wants to share their content under a license that doesn't allow free reuse, there are plenty of options. Guanaco (talk) 22:42, 7 August 2018 (UTC)
There are 4 freedoms that are necessary to consider a licence "free", all 4 are there when it comes to GFDL. In the contemporary world, where adding 7 pages is as simple as putting a QR code somewhere these arguments that are presented here are pure nonsense. Nowhere in the licence does it say that the licence must be printed on paper, just that it must be provided. QR would be a reasonable way to add it to something like a t-shirt or a coffee mug. Problem solved. But, of course, it's much easier to make a huge problem out of it and to demand that other people licence their work under different terms. I have had instances when I needed a free image of some cities and I had no way to provide the authorship info at all. If I were to come here ranting and raving and demanding that all the Attribution licences get banned, I would be (rightfully) told to get lost. ℺ Gone Postal ( ) 04:54, 8 August 2018 (UTC)
@Gone Postal: the "reasonable attribution" thing from Creative Commons is not a part of the GFDL. It just says:
"You may copy and distribute the Document in any medium, either commercially or noncommercially, provided that this License, the copyright notices, and the license notice saying this License applies to the Document are reproduced in all copies,"
In response to "QR would be a reasonable way to add it to something like a t-shirt or a coffee mug", just for kicks-and-giggles, I was going to upload the GFDL license in QR form. Turns out I can't! The GFDL license text isn't licensed as GFDL. Derivates are prohibited, so I can't upload the GFDL QR version to Commons because a license like CC BY-ND isn't allowed here. So I uploaded it to imgur: GFDL 1.2 license text in QR codes. That'll be a fun T-shirt. - Alexis Jazz ping plz 16:56, 8 August 2018 (UTC)
@Alexis Jazz: Let's take a look at the image that is actually used in starting this discussion: File:Wedge tail eagle flight Jan13.jpg. From there "[...] any reproduction of this image, in any medium, must appear with a copy of, or full (hyperlinked) URL of the GFDL license [...]" To me this doesn't say "the author is trying to pressure people into using non-commercial licence by making everybody else reproduce pages of text for no reason". The thing is that this boogieman is completely made up. It is possible that somebody somewhere will try to do that, but it is not the reason to stop everybody. ℺ Gone Postal ( ) 20:22, 8 August 2018 (UTC)
@Gone Postal: that text is from User:Fir0002/credits. The statement is actually incompatible with the GFDL, I'm not sure this kind of thing is possible. It doesn't really matter, T-shirts and coffee mugs don't support hyperlinks, so even by the somewhat loosened requirements of Fir0002 you'll have to include a copy of the full license text. - Alexis Jazz ping plz 20:44, 8 August 2018 (UTC)
A URI linking to a copy of the license isn't itself a copy; it's just a pointer. Compare the full text of CC BY-SA 3.0, which specifically says in section 4(a) that you can either include a copy of the license or the URI for it. clpo13(talk) 21:22, 8 August 2018 (UTC)
But a file name in a directory isn't really a full licence text either, it's just a pointer to something other place on your drive. I think that you are trying to find restrictions in the licence, that would be unenforceable in the contemporary court. I know that we are not talking about the Netherlands, but in there you have a thing called "giggle test", it means that if you can expect people to begin giggling in court room when explaining your argument, then probably it won't fly. You keep bringing up T-Shirts as if that were the main reuser of free images. I believe that it isn't, the main (over 98.37% I would say) reuser would be other websites, and there you have no argument that a link to a licence with the text isn't the text of the licence. Even when Wikipedia used GFDL, they didn't attach the text of the licence to the articles, but rather they had a separate page that was linked to from each article. So even if I grant you that it will be slightly difficult to print the t-shirt it is the very definition of the edge case. Let's imagine that we will have a licence that says that you must provide a full name of the licence, and the licence name will consist of 90 characters and somebody will say "But you cannot fit that neatly into a Tweet", will we have to abandon the licence just because of that? Free licences can and do have restrictions, and if the restrictions are not to your liking, then you select a different work, but you do not make a false claim that the licence isn't free. ℺ Gone Postal ( ) 04:50, 9 August 2018 (UTC)
Ok, this issue is steamrolling to succeed. Fair enough, the community does have the right to set limitations. But can we at least agree that this limitation should only apply to images because transmission of the licence is too difficult for them. Video, for example, can first of all just use a few frames to display the licence text, or if that is too difficult it is trivial to place text inside the file itself (in the tags of OGG or as an attachment of WEBM since it is just a slightly changed MKV). ℺ Gone Postal ( ) 15:09, 13 August 2018 (UTC)
Video is an even more stupid format. If you broadcast it, you'd have to include all 7 pages of GFDL in your end credits. Embedding the licence as metadata is also unlikely to satisfy anyone since you can't ensure everyone will be able to see it. Honestly, has anyone, in the history of humankind, outside of Wikipedia, ever created a GFDL-licensed video? Other than if extracted from a GFDL software manual, there is zero good faith reason for anyone to use GFDL on Commons for any media. The only people using it are doing so in order to try to restrict usage outside of Wikipedia. And that's not really what we're about. -- Colin (talk) 15:23, 13 August 2018 (UTC)
And here's another thing. You cannot say on one hand that GFDL is a non-free licence and then argue for grandfathering old images. If GFDL is not free and we cannot find the author to get them to relicence, then we must delete those images. Commons is a repository of Free Content and the licence either falls within that category or not, and it matters not what date of the publication was. ℺ Gone Postal ( ) 18:08, 20 August 2018 (UTC)
  •  Support GFDL is fundamentally unsuited for the vast, vast majority of images on Commons. Those who are using it to upload their own photos are going against the spirit of Commons. Pi.1415926535 (talk) 22:33, 7 August 2018 (UTC)
  •  Support I think that GFDL is actually unfavorable for even use on Wikipedia articles. If I want to legally print and distribute a single article with a GFDL image, I have to print seven more pages per copy. This increases my printing costs and can make my small pamphlet excessively bulky. The basic principle of COM:L is "Wikimedia Commons only accepts free content, that is, images and other media files that are not subject to copyright restrictions which would prevent them being used by anyone, anytime, for any purpose." I understand there are nuances and technicalities always, but if copyright restrictions make something unsuitable for most purposes, how can we call it free? Guanaco (talk) 22:42, 7 August 2018 (UTC)
  • I agree with the spirit of this proposal but I have to ask. If as you say "The main issue with GFDL is that it requires re-users to include 7 pages of text, which is generally not practical for individual pictures, video or audio. Because of this, this lead to the existence of ... a backdoor Creative Commons Attribution-NonCommercial license. After all, nobody is going to include 7 pages of text." With the way FAL (Free Art License) is written "You can make reproductions and distribute this license verbatim (without any changes)." How is an FAL license any different from a GFDL in this problem? -- Sixflashphoto (talk) 23:52, 7 August 2018 (UTC)
    @Sixflashphoto: That quote is from the end of the license, authorizing verbatim copies of the license itself. For FAL-licensed works, a link to the license text will meet the requirement. See Section 2.2 and the user guide. Guanaco (talk) 00:01, 8 August 2018 (UTC)
 Support Ah. I originally saw that as applicable to the work under the license not necessarily the license text. 2.2 cleared that up, "...or indicate precisely where the license can be found". Thanks, easy support -- Sixflashphoto (talk) 00:17, 8 August 2018 (UTC)
@Ghouston: you can't rename the GFDL. The license text is protected by copyright and derivatives are forbidden. (the GFDL license text is not licensed as GFDL) As far as I know, you also can't just create valid licenses yourself on Commons. - Alexis Jazz ping plz 06:35, 9 August 2018 (UTC)
You may be right about renaming the GFDL. But Commons accepts any license that meets Commons:Licensing, as far as I know. There are various custom templates for particular wordings used by institutions. --ghouston (talk) 06:37, 9 August 2018 (UTC)
Commons:Licensing will be changed anyway as it's currently discouraging using GFDL for images. As far as I know (but I'm not 100% sure because I never tried to create a license), a new license needs to be marked as such by an administrator before the system will recognize it as a valid license. This proposal also doesn't rule out making changes to Commons:Licensing, but let's try to do one thing at a time. - Alexis Jazz ping plz 06:53, 9 August 2018 (UTC)
Commons:Licensing just mentions the need to be a free license, in line with http://freedomdefined.org . By this logic, we should disallow GPL works too, but we often find useful images in software packages in the wild which are licensed that way, and allow those. If we find something GFDL out there, we could still allow uploads. And yes technically, someone could create their own license, and if meets the "free" requirements, we would allow it. The GFDL, and GPL, are and would remain free licenses. It would be interesting to see if a court would upload the must-include-license-text requirement if a possible infringement came down to that issue -- probably not a guarantee. But, it is aggravating that people do upload works based on that particular provision of making it hard to re-use -- it is against the spirit of the site and the "free" concept. I won't vote either way on this, as unsure if this is a good idea. On the other hand, would prefer to still be able to upload GFDL/GPL works if they occur naturally elsewhere, as they are still free licenses. Carl Lindberg (talk) 01:16, 15 August 2018 (UTC)
Indeed. GFDL was never intended for use with images. Instead, FSF "designed this License in order to use it for manuals for free software, because free software needs free documentation... But this License is not limited to software manuals; it can be used for any textual work, regardless of subject matter or whether it is published as a printed book. We recommend this License principally for works whose purpose is instruction or reference." (v. 1.1 and later) De728631 (talk) 07:24, 9 August 2018 (UTC)
Please note, that this proposal is not about recommending that somebody uses Free Art License, I use this licence myself and I advice most people to use it for their photographs or other images. I don't think that you would get any arguments against recommending such a thing. Unfortunately this proposal attempts to make it impossible to upload a subset of free images here because the licence is not recommended. Imagine that for anything else and you will see how ridiculous it is. ℺ Gone Postal ( ) 14:34, 21 August 2018 (UTC)
  •  Support but wish this had been discussed before a poll opened. Commons:Requests for comment/AppropriatelyLicensed tried to avoid the issue Ghouston raised about other licences (though, truly, nobody is going to rename GFDL to avoid policy, and if they did, I reckon that would be fairly clear grounds for a block, on a website dedicated to free content). The GFDL explicitly licenses "textual works" and not appropriate due to the embedded license text condition. There is a risk someone might turn to GPL, which is even less suitable for images as it refers to "executable code". It is worth noting that the only people using "GFDL only" licences for photographs are those who deliberately wish to prevent their images being used off of Wikipedia. For what it is worth, User:Eloquence who helped found Commons and write The Definition of Free Cultural Works, also supported the earlier attempt to restrict GFDL: "GFDL is a silly license to use for a media file". A question was made about about "Are multiple wiki communities aware". They don't need to be. This is a decision for Commons. English Wikipedia, say, could continue to accept images with such licences if it wishes to do so. -- Colin (talk) 07:47, 9 August 2018 (UTC)
  •  Comment Thank you to clpo13 and De728631 for referencing my blogpost, written a long long time ago and still very much portraying the opinion I hold today. I just wanted to point out that the title of this poll may be a bit misleading, as the proposal is much more focused than No longer allow GFDL for some new uploads and really is about No longer allow GFDL ONLY for new uploads, if I understand it well. -- notafish }<';> 09:58, 9 August 2018 (UTC)
    • If there had been a discussion, we could have ironed out these wrinkles. All that is needed is to say that GFDL is no longer a valid licence for X kinds of images. Users have always been able to add as many invalid licences as they wish but must include at least one valid licence. So focusing on "alone" or "only" is potentially confusing if a user uploads with "GFDL & CC BY-SA-NC-ND". Also don't see any need for a one month grace period, which would just give those affected by the decision a month in which to moan about it. They know what they are doing, and don't need any help transitioning. -- Colin (talk) 10:08, 9 August 2018 (UTC)
@Notafish: I initially misread what you said: "No longer allow GFDL ONLY for new uploads" as "No longer allow GFDL, but ONLY for new uploads" but you probably meant "No longer allow GFDL ONLY-licensed files for new uploads". Either way, it's not correct. Uploading/licensing software manuals and textbooks with only a GFDL license will still be allowed. So you could read the title as "No longer allow GFDL [as a valid license] for some [only files that meet the requirements] new [existing files are not affected] uploads". But the more you turn such things into a legal document, the harder they get to read.
@Colin: so let them moan. At least on enwiki the bot(s) that automatically tag(s) files to be moved to Commons will need some adjustment. (or alternatively, they need to pass a similar proposal) The grace period doesn't really hurt but does address some worries that might exist otherwise. Better to have and not need than to need and not have. - Alexis Jazz ping plz 16:52, 9 August 2018 (UTC)
@Alexis Jazz: Indeed, GFDL only is what I meant, and yes, not for all uploads, but mainly for pics. -- notafish }<';> 06:55, 10 August 2018 (UTC)
  •  Support -- Bwag (talk) 10:01, 10 August 2018 (UTC)
  •  Support Multilicensing has been the way to go for years now. Daniel Case (talk) 14:59, 10 August 2018 (UTC)
  •  Support --B dash (talk) 03:35, 14 August 2018 (UTC)
  •  SupportKwj2772 (talk) 09:08, 15 August 2018 (UTC)
  •  Support. -- Geagea (talk) 09:27, 15 August 2018 (UTC)
  •  Oppose As for Carl Lindberg. This proposal has strong bias. If the GFDL shall be banned, we need to change Commons:Licensing. Btw: the GFDL license text does not need seven pages printed, it fits perfectly readable to one page A4. The conditions for mobile phone contracts are generally much less readable and will be accepted at any time. We can not ban a license from commons which is considered "free" according to commons:licensing only because it is somewhat unhandy. CC-BY-SA and FAL are also rather unhandy (at another level), because they are viral. Are these to be banned in a next step, because someone feels they are unwieldy? You will already find many voices among professional (re)users who strongly advise against using images from Commons. But this has nothing to do with the licenses used, whether GFDL or any other licenses permitted here, but with contributors who use the platform "commons" for their business. But these contributors will continue to do so also with cc-by-sa et al. GFDL is a perfectly free license, and this proposal is simply not the right place and right way to tackle a problem that is social in nature. --Smial (talk) 10:43, 15 August 2018 (UTC)
First: this proposal is not "banning GFDL" by a long shot. Text, existing files and some other things are not affected. Changes to the wording of COM:L may be considered in a future proposal. CC BY-SA and FAL are not being discussed here, that is a slippery slope argument. Limiting the use of BY-SA and FAL would require another proposal, which would have a snowball's chance in hell of being passed. This proposal is merely dealing with a license that is actively being abused as a CC BY-NC backdoor. For the number of pages I simply clicked "Printable version" on this page. You could print the entire license on a postage stamp if you had a really good printer, but that's not the point. - Alexis Jazz ping plz 05:23, 21 August 2018 (UTC)
But what is the point of this proposal. It first starts by saying "GFDL is not a free licence in some cases", but then it says "but we should still allow old files". But if it is not a free licence, then we shouldn't. Before CC-BY-SA there were things like CC-SAMPLING for short clips of audio from copyrighted music. We wouldn't allow even a very old file under that licence, because it is not free. So the question is: Is GFDL free or not. Not even the original proposer can answer that simple question. If we determine that it is not free, then we cannot keep old files, because that goes against the very spirit of the project. But the truth of the matter is that no matter how much somebody doesn't like GFDL and how much it is not suitable for some people, it is non-the-less a free licence. Do I use it for my uploads? No, of course I do not. Would I ever use it? Maybe if it were a document or if I were to write a book or something like that. You are saying that limiting FAL (that I do use) would "have a snowball's change in hell of being passed", but when I read the discussion about GFDL banning just 5 years ago it is clear that it had no chance of being passed then either... look at what is happening today. People are saying "I don't use it, I don't like it, so nobody should be using it". ℺ Gone Postal ( ) 06:55, 21 August 2018 (UTC)
That's my point, thank you. In fact, if you consider future uploads with the GFDL as only free license as illegal, you must consider existing files with this licensing also illegal. Btw: I use "GFDL 1.2 only" as an additional option to offer compatibility with old works or documents that use this old GFDL. None of the other licenses allowed on commons supports this. --Smial (talk) 13:23, 21 August 2018 (UTC)
That "illegal" argument seems to confuse community policies with laws. Besides, grandfathering is a thing in legal situations too. And per the below, offering "GFDL 1.2 only" as an additional option will remain possible. Regards, HaeB (talk) 23:35, 26 August 2018 (UTC)
  • If I understand correctly, dual-licensing (i.e. {{Self|GFDL|CC-BY-SA-3.0}}) will be fine (after the turnover), right? If yes, this proposal is fine to me. — regards, Revi 15:09, 15 August 2018 (UTC)
    Yes. In fact, multi-licensing Copyleft licenses with other non-compatible Copyleft licenses including non-commercial license is good and encouraged as it will give more adaptation opportunities. Otherwise works with different non-compatible Copyleft licenses (like CC BY-SA and GFDL) can't be mixed in adaptions. Jee 02:53, 16 August 2018 (UTC)
@-revi: Correct: "And a dual license like CC BY+GFDL is also no problem". Jkadavoor is also correct, you can add any number of nonfree licenses to your files, provided you offer at least one free license. GFDL is now often combined with CC BY-NC which is not free either. You could also combine, for example, CC BY-SA (free) with GFDL and CC BY-NC. Non-commercial derivatives wouldn't have to be shared alike in that case. - Alexis Jazz ping plz 05:23, 21 August 2018 (UTC)
Which of the concerns about the GDFL that were mentioned above do you consider FUD, and why? Regards, HaeB (talk) 23:35, 26 August 2018 (UTC)
Even though the question wasn't for me, I can answer this. Anything that says that GFDL is not a free licence is FUD by definition. You could say that it's not the best licence for some cases, and you would be correct, but anybody who compares it to non-commercial or other proprietary licences is either seriously mistaken or is lying. In order for the licence to be considered free, it must insist allow 4 general freedoms (and hopefully insist on keeping those 4 freedoms down the line): Freedom to use, Freedom to adopt, Freedom to copy, Freedom to distribute altered versions. Nothing there says that there cannot be any restrictions on those freedoms (otherwise only CC0, CC-PD, WTFPL would be "free"), and GFDL places restrictions reasonable within most contexts. ℺ Gone Postal ( ) 11:35, 11 September 2018 (UTC)
Are you suggesting that people uploading work that they created under a free licence is a "longstanding problem"? ℺ Gone Postal ( ) 11:37, 11 September 2018 (UTC)
  •  Support Ronhjones  (Talk) 19:07, 27 August 2018 (UTC)
  •  Support Nobody that wants to use the free content should have to deal with the GFDL License terms. They are hopelessly outdated for free content. Best regards --Neozoon (talk) 19:51, 27 August 2018 (UTC)
  •  Comment 28 support (counting the proposal and @-revi: as support votes), 4 oppose votes so far. (87.5% support) - Alexis Jazz ping plz 20:04, 8 September 2018 (UTC)
  •  Support I would go further and disallow homebrew licenses. I support limitations on licensing for Wikimedia Commons, and think that files in Commons must have community-approved licenses. In the past it was not certain whether Creative Commons licenses would be popular and so we wanted to leave options open for other organizations to establish licenses. Also we wanted to be open to many individual organizations making their own licenses. As it happened, Creative Commons became a market leader and has defined the thought on free and open licenses for media. I do not mind files having additional licenses tagged just so long as those licenses can be overlaid with a Creative Commons license. Licenses which are less free than the standard Wikimedia Commons choice of Creative Commons license are a problem and we ought to discourage them now and prohibit them soon. We get great advantages from having a body of compatible licenses. GFDL is not intended for media. Various other homebrew licenses are often applied in error or to the confusion of users. Wikimedia Commons does not benefit from permitting endless odd experiments in licensing. Blue Rasberry (talk) 21:20, 10 September 2018 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
This section was archived on a request by: The proposal has wide support, and there is no active debate on the core proposal. I am closing this as successful, and I have updated Commons:Licensing to include the new rule, effective 15 October. Guanaco (talk) 07:56, 14 September 2018 (UTC)

Upload Wizard feedback

I have just uploaded 70 or so shots using Upload Wizard. The tech boys have done a great job- with respect can I add a few thoughts for improvements.

  1. On the select images to upload button- could we also just drag and drop selected images on the button as well as directory select.
  2. on a shakey line- the tab crashed while I was adding description- Firefox asked did I want to reload the tab?- obviously yes- and I am thrown back to the adding images page (where I was 40 minutes earlier)! Could we tweak the code so I am thrown back to where I was on the description page. It was a bad line. I made the same mistake by attempting to refresh the page to display some images that weren't thumb nailing.
  3. now a bigger one- I use shotwell as an image editor- could we add the facility to add simple image editing facilities to the wizard- crop, straighten principally or a link out to our favourite editor.

CheersClemRutter (talk) 21:00, 7 August 2018 (UTC)

This template takes one parameter - the page that has been re-uploaded. In general, most of the time I find that the uploader will change the file name (obviously in the hope that it will be missed...). I would like to suggest that it be changed to have a second (optional) parameter (the original file name), so to remind the user what the page was called, and suggest he requests undeletion of the original. So for example one would post {{subst:dont recreate|File:Martin Poch.jpg|File:Martin Poch (2017, foto Jana Plavec).jpg}} to the user's talk page Ronhjones  (Talk) 19:51, 28 August 2018 (UTC)

After blocking

Currently, the eponymous section of the policy provides that the sysop must

I propose an amendment that these points are not mandatory for any account of a known or evident serial abuser where this account has no positive contributions on Commons. “Accounts of a serial abuser” include (but are not limited to) those globally lockable by stewards as LTA (long-term abuse) or similar reasons.

Many sysops don’t follow these procedures wrt sock puppets for many years. For block-evading vandals I deem {{Indefblockeduser}} or similar unnecessary, since such worthless accounts only clutter categories, hindering possible review or research. Incnis Mrsi (talk) 07:43, 12 August 2018 (UTC)

  •  Support.   — Jeff G. ツ please ping or talk to me 03:38, 13 August 2018 (UTC)
  • I support the general point made, that all blocked accounts should receive a user block template which advises why the block has happened and how to appeal. It is less important that an administrator is required to watch the user page, however there has to be a system to ensure that appeals are not overlooked by accident. I disagree that accounts which are thought to be run by a LTA may opt out of having a formal notice or an appeal process, regardless of who is doing the blocking or the evidence, there are many years of evidence showing that accidents and misunderstandings have often lead to wrong accounts being caught up in an anti-spam or anti-sock campaign. -- (talk) 08:17, 13 August 2018 (UTC)
    First of all, it is indeed important that “an administrator is required to watch the user page” – it discourages some of them from hiding behind others’ backs. Misidentifying an account as a “known or evident” LTA must be qualified as a sysop error. Where your data about frequency of this kind of error come from? If a sysop has too bad judgement wrt blocks, then is it not a problem specific to user_talk. Of course, you can pick several purely accidental mishaps off myriads of LTA blocks, but would it be a significant problem in the context of where this account has no positive contributions on Commons? This and this are excluded by a huge margin. It is a very low-probability misfortune for a good-faith user to have his/her account identified as LTA before any positive contributions appear. And even in this case an appeal is still possible (Meta-wiki, posting as an IP, Email…). Incnis Mrsi (talk) 11:09, 13 August 2018 (UTC)
  • A long time ago I was on a forum and I got angry and wrote whole bunch of mean things repeatedly, in the morning I realised that I probably have overreacted, but was not able to even contact anybody to tell them that I apologise for this occurrence. I know that here you are distinguishing a single act of anger from the "long-term" abuse, but often these things tend to be left to interpretation. Sometimes I see that people equate abuse to uploading media to a category that others consider well saturated (categories of sunrises, roads, body organs... especially sexualised ones, etc). So a person who is honestly trying to contribute may easily be first threatened with the block, and after that individual (quite rightly, in my opinion) will decide to just move on and not engage is flame war and simply continues uploading content can be blocked with no way to even demand the explanation or ask how to upload educational media differently. Now with sockpuppets or throw-away accounts the issue is quite different, but then again, is there really no way somebody can be mislabeled as a sock? ℺ Gone Postal ( ) 09:37, 13 August 2018 (UTC)
    I don’t understand how this forum memoir is relevant. Again, the amendment is not about “abusers” in general, but about 1) serial abusers, 2) accounts thereof having no positive contributions. Also see my comment above. Incnis Mrsi (talk) 11:09, 13 August 2018 (UTC)
    Did you read past the first sentence? Because you have simply written the first part of my second sentence in different words. ℺ Gone Postal ( ) 15:54, 13 August 2018 (UTC)
  •  Oppose, the term "long-term abuse" has no actual meaning and every user (except for those banned by the Wikimedia Foundation) should be able to appeal or at least understand why they're blocked and how and where they can appeal. I agree with Gone Postal 📯 about the mislabeling, in fact a good faith user who uploads aviation related images could be mislabeled as a Russavia sock (something INeverCry did a lot in their day). Maybe we should split the categories and create "Category:Commons sockpuppets blocked in August 2018" for example to make sure that "main account" categories don't get overpopulated. --Donald Trung 『徵國單』 (No Fake News 💬) (WikiProject Numismatics 💴) (Articles 📚) 20:44, 20 August 2018 (UTC)
    @Donald Trung: indeed, the bulk of blocked accounts—on Commons or elsewhere—pertains to sock puppeteering (including a common block evasion). Again, the proposal is not about puppets in general, but serially manufactured vandalistic filth. Incnis Mrsi (talk) 07:38, 31 August 2018 (UTC)

No longer allow OFL as the only option for symbol-only uploads

While the community is bending backwards trying to prove to itself that GFDL is unfree we apparently have absolutely no problems allowing Open Font Licence (text here). This licence in its preamble clearly states "The OFL allows the licensed fonts to be used, studied, modified and redistributed freely as long as they are not sold by themselves" and then the condition section repeats that limitation "Neither the Font Software nor any of its individual components, in Original or Modified Versions, may be sold by itself." As such it is a non-commercial licence when used for font elements (which it often is).

Now I do not argue that the images should be outright deleted. Some of them (for example File:Libertine sample.svg and File:Blason ville fr Laurac-en-Vivarais (Ardèche).svg]) are also available under the free licence. Some of them are quite complex images simply utilising elements of the font and licencing the font as a part of the whole, in such case it becomes the same argument as COM:DM or COM:Stamps, where some of the individual elements maybe non-free but the work as the whole is non-the-less free. However, images like File:Font Awesome 4.7.0 icons.svg where the only licence is {{OFL}} and the only content is actually the presentation of the font would be disallowed.

I propose to add the following text to the licence template: "Note: This licence is for the font within the work and does not apply to the work as the whole. OFL is a non-commercial licence and you are not allowed to extract the elements of the font for the purposes of using them if you intend to use them commercially. The work as the whole must also provide its own licence."

We should also recategorise this licence as an unfree licence. ℺ Gone Postal ( ) 02:17, 28 August 2018 (UTC)

  • To be transparent, some of the points of the opposing opinion are listed on the talk page of the template. I am copying them here
I saw the Deletion requests. I think the prohibition of selling the font itself doesn't necessarily mean it is non-free because:


Do we even allow direct uploads of fonts? Pretty sure we would only have usages of those fonts in other works, so not sure those usages would matter either way for that clause of the license. And if that many other organizations think the license is free, not sure we should be different. Carl Lindberg (talk) 17:20, 2 September 2018 (UTC)
  • There is a big difference between a working font and a collection of glyphs extracted therefrom. Selling a graphic that uses a font, or even a complete specimen, is not selling the font: none of the ‘machinery’ is included that selects & positions the characters, and would optimize their rendering on a device in which the font is installed. Fonts are generally considered a form of software, which I would argue is out of scope here. (But it would be great if the WMF had better font support on the SVG-rendering servers, in both breadth and depth, so to speak.) Even commercial fonts usually allow embedding of their output in other works, such as PDF documents. At any rate, I think a limitation that applies solely to ‘live’ fonts is irrelevant here.—Odysseus1479 (talk) 04:04, 9 September 2018 (UTC)