Commons:Bureaucrats' noticeboard

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

Shortcut: COM:BN

Bureaucrats' work area (archive)
Requests for bot flags
request | watch
To request a bot flag.
Requests for GWToolset and translation admin rights
request | watch
To request to become a GWToolset user or a translation administrator.
Other resources: Need administrator assistance? See the administrators' noticeboard. Need help? Try the FAQ, or the Help desk! Have an idea or suggestion? Tell us at the Village pump! Need a checkuser? See the CU request page!

This is a place where users can communicate with bureaucrats, or bureaucrats with one another. Please refer to the links above for specific bureaucrat requests.

Filing cabinet icon.svg
SpBot archives all sections tagged with {{Section resolved|1=~~~~}} after 7 days.

Archives: 1, 2, 3, 4, 2015, 2016, 2017


Public Domain Mark (PDM)[edit]

After Flickr introduced the CC Public Domain Mark (PDM), a lot of people use it to mark their works as free to reuse. But we had rejected it as an applicable license here after a short discussion which was not reviewed/closed by a crat. It leads to endless discussions; many experts expressed a different opinion. It comes again VP now. Even if we don't like to accept, Flickr is a major source of the media files which portrays the opinions and interests of the people around the world. So neglecting this issue will be certainly a stopping block for the Wikimedia to collect such contents.

The current solution is that we need to contact individual Flickr users and request to change the license to CC0. We know it is not a smooth solution. we can't collect much contents that way as it is a slow procedure. Some people suggested in that recent VPC discussion that we may contact the legal for advice. I don't know how much it is practical as the legal had said many times that they can't advice the community. Any suggestions? Jee 04:04, 27 January 2017 (UTC)

The fact is that Public domain has a meaning only in the US, and other countries with similar laws. In France, for instance, you cannot even put something in the public domain; it doesn't mean anything for the judges. That's why we need CC0, which is a license contract. The issue is on Flickr side, for having badly formulated their "PD" version. --Ruthven (msg) 10:53, 27 January 2017 (UTC) PS: This discussion should be moved at Commons:Village pump/Copyright I reckon.
As far as I know, Flickr implemented it after a consultation with CC. If there is some mistake from their side, it can be resolved with proper discussions as there seems a good relationship with CC and WMF. That's why I bring it here as crats are supposed to be the elected members to lead and guide the community in such matters. I too can contact Mdennis (WMF); but I think it will be more effective if from the crats' side. (COM:VPC discussed this matter in depth, even recently. So what we need is some professional advice; not just another discussion.) Jee 11:49, 27 January 2017 (UTC)
(I corrected a dead link in Jee's post.--Elvey (talk) 01:37, 31 January 2017 (UTC))
I'm confused. It seems like the discussion was getting going and suddenly User:Josve05a ended it, prematurely. What came after the "Therefore" in the close doesn't follow from what came before. Was there discussion off-wiki? I'm not saying the decision was wrong, but it does seem premature. Looks like we have respected users whose views were run right over as if they weren't seen. Can anyone present a concrete instance where someone used a work that was released as public domain and was sued successfully for having used it? --Elvey (talk) 02:15, 31 January 2017 (UTC)
No comment about the "license" had been made in months. The only comments that had been happening whitin two month of closing were on how different external tools should implement the new licnese template. The RfC was closed after having been open for 7 months and 20 days. (tJosve05a (c) 07:37, 31 January 2017 (UTC)
Also, the bureaucrat's noticeboard really isn't the right place to discuss this. This has nothing to do with admin/bot management (i.e. crat tasks) and should be discussed on COM:VP. Sebari – aka Srittau (talk) 08:04, 31 January 2017 (UTC)
The one who opened this thread requested advice from the crats. This issue has been in a long debate, even now. Even admins are confused on what to follow. I don't see anything wrong on bringing the issue here. Poké95 11:03, 1 February 2017 (UTC)
The bureaucrat's noticeboard may be used to ask for leadership on an issue, in line with COM:Bureaucrats, and Bureaucrats may choose to offer their opinions here, or point to existing policy. However this noticeboard is not the place to reach a community consensus on a copyright interpretation or an interpretation of the precautionary principle, nor are Bureaucrats supposed to act like an elected Arbcom and make rulings that the wider community then must comply with. For these reasons if there is to be a proposal for new guidelines or the RFC is to be reopened, those normal procedures for consensus have to happen. Whether those procedures are lead by a Bureaucrat is up to the individual Bureaucrat, though there as absolutely nothing to stop someone else from taking the initiative. -- (talk) 12:08, 1 February 2017 (UTC)
Yes; my intention is very clear from my initial post and from my reply to Ruthven: it is a request to the crats to lead the community in gathering some professional advice on this matter, if possible. It is not an attempt to override the current status with another discussion. I don't even intended a discussion here. As we discussed a lot, I don't see any benefit in another community discussion unless we get some qualified advice. And for copyright matters, "any expressed consensus can be taken into account so far as possible, but consensus can never trump copyright law." So what we need to know is whether PDM is legally valid if issued by the copyright holder. Jee 15:27, 2 February 2017 (UTC)

As already stated the topic was discussed at multiple venues already. Could one volunteer to compile a summary page of the results? --Krd 15:32, 2 February 2017 (UTC)

I'll make a try.
1. The RfC was closed with a rational, "Images uploaded from Flickr under the Public Domain Mark 1.0 needs to be tagged with the template {{Flickr-public domain mark}} (substituted as it currently is coded) given that we have no idea why it is under Public Domain. This PD Mark is not a license, nor can someone release something under it. Our current Public Domain-templates, such as {{PD-self}} has a secondary clause that states: "I grant anyone the right to use this work for any purpose, without any conditions, unless such conditions are required by law." which is needed as a "fallback license" in case releasing to the public domain has no legal meaning. We can't relicense something from PD Mark 1.0 to another PD-license/template, since they have different legal text and the Flickr user has not agreed to those terms."
2. At Commons:Village_pump/Copyright/Archive/2016/12#Creative_Commons_Public_Domain_Mark_on_Flickr, those arguments are questioned in case if the Flickr user is the copyright holder. There are other cases where Flickr user is not the copyright holder which is not considered. There Carl Lindberg stated that "To me, we should allow it if it's pretty obvious the Flickr account is the copyright owner. CC0 is definitely better, but I think it's the equivalent of PD-author, which is good enough to be beyond COM:PRP for me. Custom licenses or PD statements may have theoretical issues in some countries (it may not be legally possible to waive rights, or a country may allow such a statement to be revoked, etc.) but I think that is splitting hairs to disallow such works where the intention is fairly clear. Such things should probably go through a license review so someone else confirms that the Flickr account seems to be the author, but at that point I think it's OK. Carl Lindberg (talk) 18:39, 20 December 2016 (UTC)"
3. I endorsed Carl's argument. Then Ghouston questioned about the lack of "fallback clause". "They haven't agreed to the fallback license on {{PD-author}} though. There should be a template that retains "In some countries this may not be legally possible" and leaves it at that. --ghouston (talk) 02:47, 23 December 2016 (UTC)" But this is an issue for all files that have been released to the public domain by non-Commons users. See Commons:Village_pump#Template_for_generic_public_domain_dedications. (Sorry for the late response; I was away.) Jee 04:19, 6 February 2017 (UTC)
That's part of it, PD-author has often been misused. But if the Public Domain Mark on own work is accepted, I think it should be with a custom tag that describes the situation exactly, not just a generic public domain template. --ghouston (talk) 05:50, 6 February 2017 (UTC)
Yes; and I've no problem for a new tag. What I said is "fallback clause" alone is not a reason to decline as we already accepted many files that have been released to the public domain by non-Commons users published elsewhere without a "fallback clause". Jee 06:14, 6 February 2017 (UTC)
You fail to mention the point which others brought up after the RfC, which weren't included in my closing of it, that it isn't a release, but a label or "mark" of something to be believed in the public domain.
Notice Notice: Public Domain Mark 1.0. What is it, and what are the legal implications?
“The tools also differ in terms of their effect when applied to a work. CC0 is legally operative in the sense that when it is applied, it changes the copyright status of the work, effectively relinquishing all copyright and related or neighboring rights worldwide. PDM is not legally operative in any respect – it is intended to function as a label, marking a work that is already free of known copyright restrictions worldwide.”
“It is a statement without any legal effect. The creator [Flickr user] can at any point change their mind and remove the PDM, and that it was previously applied means nothing, since they have not actually given up their rights, or licensed the work. PDM is not a legaly binding release which is non-revocable, which is needed to be stored on Commons. If someone changes a work from PDM to ARR, any use of it by us, or anyone else, is a blatant copyright violation.”
“It is a label. I think so, Creative Commons think so and it clearly says so. It s not a release of copyright. Our discussions if it is similar to other licenses or {{PD-author}} or not, is a non-question, since it is a revocable label. That's it.”
“With this announcement Flickr users will be able to choose from among our six standard licenses, our public domain dedication, and they will also be able to mark others’ works that are in the public domain.”

(tJosve05a (c) 23:01, 7 February 2017 (UTC)

And we forget to mention the simplest option is to inform the Flickr user that their tag choice is unlikely to be correct/useful and that CC0 would be a better choice for their own work. That way the "licence" details are fixed at source, somebody learns something useful for future images, and Commons is in no doubt as to the free nature of the image. But no, Commons users would rather argue among themselves than talk to a real photographer. -- Colin (talk) 08:06, 8 February 2017 (UTC)
  • PDM is not a license, and when a work is published without an explicit license or without an enough explicit statment from the copyright holder, it is considered by us, and also likely legally, as not free. But if it is not a license, it may stay the question if it is a relevant statment. But honestly "This work has been identified as being free of known restrictions" worry me a bit, as this mean all and nothing. Identified by who? by the same kind of persons who upload here a CD cover saying "it's my cd, this is not a copyvio" or by the same kind of persons who says "This is free, I found it with google...". Seriously, a work must be licensed by it's copyright holder, that's all. And all the persons who know really what is a license and what is the meaning of a license don't use PDM. Then we should not accept that. Christian Ferrer (talk) 17:31, 8 February 2017 (UTC) Furthermore a statment coming from a person who don't really know what is exactly a license can not be considered as an "enough explicit statment", not even as a relevant one. Christian Ferrer (talk) 17:38, 8 February 2017 (UTC)
With regard to the last part of your viewpoint, if this were true then valid OTRS releases would be ignored as irrelevant for anyone that any administrator feels does not sufficiently understand copyright licenses. That's not how copyright law works, nor is it how Commons should work. -- (talk) 18:00, 8 February 2017 (UTC)
LoL! when there is a mail statment this begin to become to be relevant, PDM alone not. Christian Ferrer (talk) 18:12, 8 February 2017 (UTC)
And we come to Colin suggestion that is when there is good content with PDM, we should contact the account owner for that they put a relevant license or for that they give an OTRS explicit agreement to us. PDM alone is not enough and IMO has even not very much to do with copyright law. It is even the opposite " do what you want with this content but it is up to you to take responsibility if there is legal concerns...." re-LoL Christian Ferrer (talk) 18:40, 8 February 2017 (UTC)
I don't believe there is anything to discuss here, though I would rather not be laughed at. My statement stands. Thanks -- (talk) 18:57, 8 February 2017 (UTC)
That's not me who put OTRS and PDM at the same level. PDM is the anti-license, it's the anti-OTRS... indeed it's not funny at all. Thanks to you too. Christian Ferrer (talk) 19:27, 8 February 2017 (UTC)

Now would be a good time to close this thread. There doesn't seem to be anything actionable for the crats to do. We all understand that the PDM is not a licence, and must not be used as such. --MichaelMaggs (talk) 03:16, 9 February 2017 (UTC)

+1 --Krd 06:32, 9 February 2017 (UTC)
Checkmark This section is resolved and can be archived. If you disagree, replace this template with your comment. --Krd 08:57, 22 February 2017 (UTC)

Upload project spam blacklist exception 'right'[edit]

After finding that images were being rejected from upload during the Portable Antiquities Scheme project, I have created T157897 - Spam blacklist by-pass right for agreed batch upload projects. As I have suggested that agreeing an exception could require Bureaucrat approval, views on Phabricator from Bureaucrats might help decide whether there may be a realistic and volunteer-time inexpensive solution. Thanks -- (talk) 13:15, 12 February 2017 (UTC)

MediaWiki:Spam-whitelist exists, but is scarcely used, if that helps anybody. Nick (talk) 13:51, 12 February 2017 (UTC)
Just in case the task is not all that clear, in the example, we want to allow uploads and text from finds.org.uk, and when that text includes blacklisted links we want to allow them; this does not mean we want to whitelist the links so we do not add them to the spam-whitelist. E.g. there are bit.ly links in the example, bit.ly should remain blacklisted, but when the upload is part of an agreed batch upload, we want to make an exception. It's surprisingly hard to put this in a pithy way. -- (talk) 14:00, 12 February 2017 (UTC)
Ah yes, I understand now, 'we' won't know what the blacklisted links are and don't want the upload batch to fail due to the existence of unspecified blacklisted links. I'd support an exception to the blacklist, but the blacklisted links need to be recorded so we can then fix them (as obviously the existence of blacklisted links prevents subsequent editors saving the description page). Nick (talk) 15:58, 12 February 2017 (UTC)
I also agree with both the ability to bypass the blacklist and the recording of the exceptions. Reguyla (talk) 16:12, 12 February 2017 (UTC)
I'm unsure how the recording of exceptions would work. It would be neat if the uploader did not have to detect these (which means making every upload quite a bit more complex) but instead if the blacklist filter process added a category to the page where there was a link filter match, or maybe we could create a new automated page error tag. One benefit of the error tag might be to retrospectively identify past pages using blacklisted links, for example File:Marcha_das_Mulheres_Negras_(23137414611).jpg has a bit.ly link in it which was accepted in 2015. -- (talk) 17:17, 12 February 2017 (UTC)
@Nick, Reguyla: Unfortunately Phab:T157897 is being actively parked, meaning it will never realistically happen. It looks as though we might need to run a full scale RFC to allow non-programmers to be able to by-pass the blacklist for approved upload projects and in the meantime this will require special parsing of uploads to expand any possible mention of a URL that might hit the blacklist (or more likely, just drop everything that looks like a URL). I encourage more views on the Phabricator task, as technical RFCs on Commons usually attract very little attention from the general community. -- (talk) 12:38, 14 February 2017 (UTC)
Thinking outloud, but I think there's a few ways to work round this .
Where blacklisted links aren't expected in bulk, batch uploads could be undertaken in some form of a two stage process - first the image descriptions scraped, external links extracted from that data, checked against the spam blacklists, a list of files with problematic URLs generated and manually rectified - either remove the URL or replace with an appropriate substitute. The upload proceeds in the meantime with the URL removed.
URL shorteners could probably be dealt with automatically in some form, so the bit.ly type link replaced with the full URL during the upload process. If the bit.ly URL is masking a blacklisted link, then go to back to my first option - record the data, strip out the URL from the description and upload.
Just some thoughts, the exact sequence and process would need to be fine tuned, but hopefully something could be done to solve this if we're not going to get a spam blacklist bypass feature of some description for bulk uploads. Nick (talk) 12:52, 14 February 2017 (UTC)
There are lots of possible technical solutions. The issue is that this is currently a burden on volunteers rather than built into the system. There is nothing to stop the WMF funding improvement to the blacklist process, so that 'shortners' are automatically expanded, then re-checked against the blacklist. The uploader/tool user would not have to write special adapting scripts just to compensate for the blacklist.
As mentioned in the task, the objective is to ensure that untrained volunteers and GLAM professionals can support Commons with mass uploads, not just programmers and tool-maintainers. I had thought this had become a WMF strategic objective, if someone knows of a relevant link that might be useful to justify the task. -- (talk) 12:58, 14 February 2017 (UTC)
The problem with automatically expanding and substituting the underlying full URL is, looking at the example given File:Marcha_das_Mulheres_Negras_(23137414611).jpg the bit.ly URL expands to http://www.marchadasmulheresnegras.com/#/!manifesto/c15t1 bit.ly URLs (and others) linking to inactive hosting holding pages. I'm sure others go to 404s and yet more go to defunct domains. I don't know what the options are there, really, though, given we don't really care whether full links added to descriptions are live or not, but it does seem wasteful trying to fetch the underlying full URL only for it to be potentially useless. Thoughts ? Nick (talk) 13:05, 14 February 2017 (UTC)
I disagree, a fully qualified url clearly states where it is at the time of its creation, and gives better hope of finding successor sites. Urls shorteners will die and have no evidence on site. Dead urls we manage and understand and can be updated, we have no control over a url shortener. I would always prefer a full dea url than a dead url shortener.  — billinghurst sDrewth 13:19, 14 February 2017 (UTC)
Thinking further about this - I'd agree with having a full dead URL is better than a dead URL shortener, and perhaps 404s and dead sites could automatically be linked to relevant Internet Archive copies of the URL. I'd be interested to know how many uploads with dead URLs we've got going, and if the dead URL was simply duplicating the image data/description which accompanies the image being uploaded from its current location, or if it's different and additional information. Nick (talk) 13:38, 14 February 2017 (UTC)
Yes, I think it is clear that we do not want these shortened URLs, and would even prefer to see the dead one (until editors decide to remove it). So if shortened URLs are the main thing the blacklist is pinging you on, then please try to catch & deal with them. If you have other examples, I'd be interested to see what you think should be allowed through. --99of9 (talk) 22:58, 14 February 2017 (UTC)
  • Pictogram voting comment.svg Comment this should not be a discussion just put in front of bureaucrats. This is a general discussion for the whole community, and it seems disrespectful to the community to put it away in a back-corner.

    The use of url shorteners is actively discouraged by the global community. The counter argument in the ticket is that if the upload tools resolved the urls back to the base urls, prior to upload that this becomes a non-issue; it also aligns with the current practice and expectation here and through Wikimedia wikis.  — billinghurst sDrewth 13:15, 14 February 2017 (UTC)

What would you consider a suitable community consensus, a discussion on the Village Pump, or something else? It would also greatly help if you could help us out by drafting a proposal. I am in no doubt that there would be mismatches in expectations if I were to draft it, and this would probably result in any consensus being rejected downstream as unclear or misworded.
In terms of this thread, it was created so that the "official" status of requests was agreed as being a Bureaucrat task. Implicitly it was not created by me as a way to cut out the Wikimedia Commons community by holding a discussion in a "back-corner", so could you do me the courtesy of not painting it that way? Thanks -- (talk) 13:23, 14 February 2017 (UTC)
The proposal you brought here was about getting people/bots whitelisted where a url on the blacklist is utilised. That is totally presumptive that it is okay, and your whole proposal is framed in the manner that you should be able to upload that way, let us make solution so your (collective) uploads are not inconvenienced. I dispute the foundation of that argument. Most discussions should start before the broadest likely community.  — billinghurst sDrewth 10:48, 15 February 2017 (UTC)