Moved from Wikipedia:Deletion review/Log/2022 January 18

I would like to see this revision deletion reverted. It was done per WP:RD1 to remove a small amount of copyrighted text. Given that the copyvio was noticed relatively late, this resulted in the deletion of a large number of intervening edits and obscured the provenance of a decent amount of newly added content.

This is against current policy: see Wikipedia talk:Revision deletion#RD1, attribution and intervening edits for the details, including a caveat about a novel interpretation that – if adopted – could result in a change to those policies. But even if that were to happen – so far there has been zero indication of that – this revdel would still likely appear as disproportionate. That's because in order to completely expunge less than 0.7 kB of copyvio text, it resulted in the deletion of about 60 intervening edits and so has erased traceability for the numerous changes introduced in them, as well as for the 2.5 kB of text that they added. – Uanfala (talk) 13:56, 18 January 2022 (UTC)[reply]

• I don't think there's anything we can do about incorrect use of revdel: per WP:REVDEL#Appeal and discussion of actions, "[s]uch a review should take place at the Administrators' Noticeboard". I do agree that deleting 61 revisions is a questionable large-scale use of revdel, although it may well be within discretion nonetheless. Extraordinary Writ (talk) 18:05, 18 January 2022 (UTC)[reply]
  • WP:XRV might also be a reasonable location--it's not listed on the things that belong there or the things that don't (I don't think). But yeah, DRV is the wrong place I'm afraid. Hobit (talk) 18:13, 18 January 2022 (UTC)[reply]

• This is an example of one of several ways in which Wikipedia's policies relating to copyright are in fact unworkable and self-contradictory. If I find text which infringes my copyright, I can legally require all versions of the page in which it appears to be removed from public view. However, Wikipedia policy is that history should be kept, to enable us to know which editor wrote which part of the page in question. It isn't possible to satisfy both requirements. Personally, if I see a relatively small copyright infringement which would require the loss of dozens of revisions in order to remove it, I normally leave it alone, unless it seems to me that there's some reason why it's particularly problematic. However, once the content has been removed, I absolutely would never restore it, no matter what Wikipedia policy or consensus among editors might say, because if I did so I would be knowingly breaking the law. Also, I wouldn't recommend that any other administrator do so either. The law has to take precedence over what a group of Wikipedia editors think. That doesn't mean that I'm happy with the situation, but it seems to me that that's how it is. JBW (talk) 19:27, 18 January 2022 (UTC)[reply]
  • I've no idea what was there, but copyright law (in the US and in general) isn't black-and-white. Unless this was a huge portion of the original work it is very likely it falls under "fair use". I feel like I should give a Wiki-seminar on the issue some day. Or maybe I'll write an essay. But no, it's almost certainly not breaking the law. Hobit (talk) 20:22, 18 January 2022 (UTC)[reply]

Are you lawyer and are you giving legal advice? nableezy - 22:06, 18 January 2022 (UTC)[reply]

It's possible to offer a well-informed view about copyright on a Wikipedia discussion page without being a lawyer or giving legal advice. Hobit does in fact know what he's talking about.—S Marshall T/C 22:27, 18 January 2022 (UTC)[reply]

I wrote "if I see a relatively small copyright infringement..." Perhaps I should, to be strictly accurate, have said "if I see a relatively small element which I think is likely to be a copyright infringement", but my statement was premised on there actually being a copyright infringement. It is perfectly true that many so-called "copyvios" on Wikipedia aren't actually copyright infringements under United States law, but that is a different issue. JBW (talk) 23:07, 18 January 2022 (UTC)[reply]

@Hobit: I have thought very carefully about this, and I am confident you are mistaken. "Fair use" in United States copyright law applies to a fairly limited set of circumstances, and it pretty clearly doesn't include telling everyone in the world that they are free to reuse the material for any purpose whatsoever, including purposes which aren't covered by the concept of fair use, which is what anyone posting it to Wikipedia is doing. That is why Wikipedia's copyright policies make an exception for "fair use" only for images, not for text, for which Wikipedia's licensing terms are significantly different. JBW (talk) 16:39, 19 January 2022 (UTC)[reply]

@JBW: Sorry, been away. I don't have time right now, but I'll jump back in soon. But no, fair use isn't a fairly limited set of circumstances, it applies quite broadly.. But [1] gives an okay overview, [2] is better. But fair use is rarely a black-and-white thing and while I'm not a lawyer, small quotes of text are often much safer to grab than an image because of point 3--it's harder to grab a useful portion of an image than it is a useful part of text. Now if you are copying a novel and putting the whole thing on the web to make money while adding no commentary, fair use is pretty much out the window (fails 1, 2, 3, and maybe 4 if there really is a market for it). But things here are almost always going to do well on point 1. Gah, sorry, really don't have time for this right now, but it's more fun than what I should be doing. Oh, last thing, as I recall, the reason we have fair use stuff for images is partly (mostly?) so commercial enterprises using our work can figure out what they should pull out. Because images are a lot harder to justify wrt fair use. I'll find links to that later. None of this is hugely relevant--it's generally good that people are really careful with copyright here. Again, INL, but I do teach some of this so I've had to learn the basics. Hobit (talk) 01:36, 20 January 2022 (UTC)[reply]

Keep in mind, US Fair Use law is a defense, not an allowance. The scales are weighted against the entity that reused material to prove it meets fair use, it is not a presumption that it is allowed. As we should strive to keep WP out of copyright problems, we generally edge on making sure we're tighter on copyrighted materials. --Masem (t) 06:09, 20 January 2022 (UTC)[reply]

I suppose that after the amount of time I have spent on Wikipedia I should have known better than to think I could come to AN and make a statement relating to copyright which broadly expressed what I meant, without the risk of then being dragged further and further into having to more precisely define the terms I used. When I said that "fair use" applies in "fairly limited set of circumstances" I didn't mean in a closely defined or precisely delimited set of circumstances; I simply meant that there are some circumstances in which it applies, and many more in which it doesn't. Hobit, the first link you have given above says, amongst other things, that an example of the criteria likely to be used in assessing a fair use defence is "The purpose and character of the use, including whether such use is of commercial nature or is for nonprofit educational purposes". (My emphasis.) That is precisely the kind of thing I was referring to: someone may post something on Wikipedia in a way which in itself might be defensible as fair use, but announcing to the world "Hey, I give you permission to use this if you like, including commercial purposes to enrich your business" might well be seen as violation of the copyright holder's rights. I believe that statement to be substantially true, but even if a lawyer could find reasons why it is not exactly accurate, I don’t care, because I am 100% sure of the essential point, namely that using copyrighted material in a particular way is not the same as announcing that you have the right to give blanket permission to others to use the same material in any way whatsoever, for any purpose whatsoever, and that you are hereby granting such permission. I am confident that there are situations where the first of those would stand a chance of being accepted by a court and the second wouldn't, and I really don't care whether the particular way I have expressed it is precisely legally correct. JBW (talk) 17:05, 20 January 2022 (UTC)[reply]

JBW, which policy requires that "history should be kept, to enable us to know which editor wrote which part of the page in question"? WP:Copyrights#Re-use of text (policy) and WP:Copying within Wikipedia#Proper attribution (guideline) both mention "a list of all authors" as sufficient, despite that method being less precise than your statement. Flatscan (talk) 05:23, 19 January 2022 (UTC)[reply]

@Flatscan: The policy on revision deletion states that revision deletion for copyright reasons cannot be used if it would remove "any contributor's attribution". We could fill up a very large amount of space debating whether there is any way of making sense of that compatible with "any contributor's attribution" meaning no more than a list of contributors, since revision deletion of copyright infringing content doesn't remove contributors' names, but doing so would be totally pointless. To me that is a policy saying that we should not remove the record of what contribution one editor made because another editor infringed copyright, but if you prefer to mentally substitute some expression such as "the preferred practice, supported by a strong consensus of editors" for "policy" in my comment above then you are welcome to do so. JBW (talk) 17:22, 19 January 2022 (UTC)[reply]

@JBW: The longstanding wording of RD1 has consensus, but the disagreement centers on the meaning of "attribution". I believe that the RD1's weight of policy originates from WP:Copyrights, so these pages should have the consistent interpretation that "a list of all authors" – name only – is sufficient. Previous discussions WT:Revision deletion#RfC on Change RD1-wording (2017, 5-3) and WT:Revision deletion/Archive 4#RD1 wording (2014, 3-1) leaned in this direction. Flatscan (talk) 05:35, 20 January 2022 (UTC)[reply]

• The above has been moved from DELREV (diff). – Uanfala (talk) 22:05, 18 January 2022 (UTC)[reply]

• Deleting a revision is an administrative decision so I feel that XRV would be the preferred venue for this discussion.—S Marshall T/C 22:27, 18 January 2022 (UTC)[reply]
• I endorse the deletion of 61 revisions within two months as within discretion. Uanfala, are you looking for something specific in that history range? The majority of non-revert additions were by User:UnpetitproleX (then User:Unpetitprole). My assumption is that 2409:4060:2D98:9A3A:5C0B:B19A:F479:9256 added a {{Pie chart}} (which you removed) and updated some numbers. Flatscan (talk) 05:23, 19 January 2022 (UTC)[reply]

  The deleted revisions [3] added 2,542 bytes of text and made a number of modifications to the existing content. Deleting these revisions means that we don't know who contributed which content, and why each of the changes was made (see Wikipedia talk:Revision deletion#RD1, attribution and intervening edits for the various problems that this may cause). – Uanfala (talk) 12:52, 19 January 2022 (UTC)[reply]

  I am asking for specific revisions, content, or changes that you believe to be inaccessible or muddled. For example, if data or a source was lost when the chart was removed, an admin would be able to retrieve that. Flatscan (talk) 05:35, 20 January 2022 (UTC)[reply]

Roughly speaking, we have the following options:

1. No revdel, each edit is clearly attributed
2. Yes revdel, keeping the list of intermediate authors intact as attribution.
3. Revert to pre-revdel version, take each author in turn and do an edit summing up his/her cumulative edits, with attribution in the summery. This attribution, of course, won't change if the user is ever renamed.

93.172.243.103 (talk) 08:05, 19 January 2022 (UTC)[reply]

• I've linked to the WT:REVDEL thread with the explanations above, but I guess it may be of benefit to summarise the more salient points. The policy at WP:RD1 states that blatant violations of copyright can be redacted, though if doing so would remove any contributor's attribution, this criterion cannot be used. The how-to guide at Wikipedia:Copyright problems/Advice for admins further elaborates that [i]t may be a good idea to use revdel for copyvios, but [o]therwise, so long as the infringing text is removed from the public face of the article, it may not need to be removed/deleted permanently unless the copyright holder complains via OTRS or unless other contributors persist in restoring it. I don't know what others take away from here, but for me this clearly shows that revision deletion is not obligatory and that it is explicitly forbidden in cases like the one here. For those endorsing the deletion, it may be helpful to explain why it's within admin discretion to violate policy here. – Uanfala (talk) 12:52, 19 January 2022 (UTC)[reply]

  I disagree with your interpretation of "attribution", expanding it to incorporate "traceability", as I find it at odds with its meaning in WP:Copyrights#Re-use of text (policy) and WP:Copying within Wikipedia#Proper attribution (guideline). (I linked them in a previous comment above.) RD1 would be hamstrung if its use were restricted to a sequence of edits by the copyvio-introducing user interrupted only by reverted or reverting edits. Flatscan (talk) 05:35, 20 January 2022 (UTC)[reply]

• As someone who has done 1000s of uncontested RD1 revision deletions, I don't believe there is any issue with this revdel. I read "any contributor's attribution" as keeping the editors name in the page history; non-admins can see the user's name in the history and non-admins can see that they made an edit, they just can't see what was added in the edit. Therefore, attribution is still kept. That said there's no exact standard for what is revdeled for copyvio and what isn't. I tend to be more liberal with not revdeling some edits than others, but I am willing to delete 1000s of revisions if the plagiarism is that large and extensive. On the other hand, I think it is overkill to delete several revisions for some paraphrasing issues and try to not do so. Every admin has different standards and opinions on the matter, and I've been meaning to start an extensive RFC on the matter to get consensus on the scope of revdels. I would like a co-drafter, so if someone wants to do that please reach out. Moneytrees🏝️^{Talk/CCI guide} 20:04, 19 January 2022 (UTC)[reply]
  • Your reading of "any contributor's attribution" may well be right, but as I point out below, the actual wording of the license is that each contributor is to "indicate if You modified the Licensed Material and retain an indication of any previous modifications". I don't see how what we're doing here would succeed at doing that. Hobit (talk) 04:54, 20 January 2022 (UTC)[reply]
• A JBW points out, we are between a rock and a hard place. First of all, the Foundation's policy on copyright infringement is much stricter than US law in some respects. The entire purpose of RevDel isn't to remove copyright infringement "because we are nice guys". It is to remove any liability in a court of law, as we can demonstrate we took action to remove the infringing material from our archives and pages. When it comes to copyright and RevDel, it is probably better to err on the side of using too much, rather than using too little. There is still attribution to the editors caught in the middle, we just don't know what they attributed for. Keep in mind, Oversight/Suppression removes even more, although typically for different reasons. But it's the same problem. I've never heard of an editor needing the specifics of attribution to maintain their copyright on an edit caught in the middle of one of these RevDels, so weighing the potential damage versus the damage of not removing all instances of infringement, again, you err on the side of removing. So I endorse. Dennis Brown - 2¢ 20:26, 19 January 2022 (UTC)[reply]
  • Again, I don't know the details in this case and IANL. But doing what you propose--losing track of who contributed what, is probably more of a black-and-white violation. The CC BY-SA license we use indicates that Wikipedia can use other's work, but that you are to "indicate if You modified the Licensed Material and retain an indication of any previous modifications". [4]. I don't think contibutors are going to sue us over this, but if they did, as I read it they'd be in the right. One could even make an argument that anyone who edits such an article is in violation of the license because they aren't retaining an indication of previous modification. But this is all a bit silly--there are so many steps before anyone could really have a case against Wikipedia, worrying about legal liability for something like this is pretty far fetched. The DMCA's safe harbor provisions probably apply to Wikipedia, so the worst reasonable case is a takedown notice I'd think. TLDR: it's good to be careful with copyright, but attribution is required under CC BY-SA and if we are violating that license for a given contribution, we have no rights to use that contribution and any use of it is now a copyright violation. It's not a rock and a hard place, it's a rock and the same rock. But in most cases the situation is likely worse when we don't respect the rights of our contributors--it's more text and we're taking the whole thing they contributed which makes a fair use argument a lot harder. Hobit (talk) 04:51, 20 January 2022 (UTC)[reply]
    • According to wmf:Terms of Use#7. Licensing of Content, editors agree that a list of all authors (but please note that any list of authors may be filtered to exclude very small or irrelevant contributions) is sufficient attribution for their contribution. DanCherek (talk) 05:02, 20 January 2022 (UTC)[reply]
    • WP:Copyrights#Re-use of text (policy) and WP:Copying within Wikipedia#Proper attribution (guideline) also cover this. (I have linked them a few times in this discussion.) Flatscan (talk) 05:35, 20 January 2022 (UTC)[reply]
    • The confusion here stems from the fact that you're quoting CC-BY-SA 4.0, while we use the CC-BY-SA 3.0 Unported license, which has no similar language requiring to carry the specifics of any modifications or adaptations of the material of contributors. Unless memory fails me, the addition of the line you quote is one of the reasons the WMF has stuck to 3.0. MLauba ^(Talk) 16:37, 20 January 2022 (UTC)[reply]

    @Hobit to add on to what MLuaba says, we can't even use text licensed under CC-BY-SA 4.0, as the WMF has said it is not compatible with our CC-BY-SA 3.0 Unported license. Moneytrees🏝️^{Talk/CCI guide} 18:13, 20 January 2022 (UTC)[reply]

    And this is why I'm not a lawyer. Details like that are important. Thanks to those that pointed it out. Hobit (talk) 16:07, 21 January 2022 (UTC)[reply]

• I also have over a thousand (mostly) uncontested revision deletions. For me, this is an edge case. Copyvio stayed for two months in a high-traffic article, and several dozen revisions had to be deleted. Probably still a reasonable time period and almost a reasonable number of revisions. I would say it is still within discretion of the admin, but close to the limit.--Ymblanter (talk) 21:39, 19 January 2022 (UTC)[reply]
• I'm not an admin, but "discretion" is a nonsense argument. Administrators have no special discretion when it comes to implementation of WMF's terms of use and licensing. Fortunately for us, it's also an entirely unnecessary argument. As noted above by several users, Wikipedia's licensing requires attribution in the form of a list of authors. Revdel maintains the list, which is still accessible through the page history. We have no requirement to attribute every tittle of the text to each individual author on a byte-by-byte basis. AlexEng^(TALK) 11:34, 21 January 2022 (UTC)[reply]
• I'm a bit surprised this hasn't been mentioned yet, and of course IANAL etc but if we're talking about legal issues, putting aside what out terms of use say and what the CC BY-SA 3.0 says, we don't actually have to comply with CC BY-SA 3.0 in most cases. We only have to comply with one of the CC BY-SA 3.0 or GFDL since all contributors here have agreed to dual licence their contributions. The only case when we must comply with the CC BY-SA 3.0 is when we've added content from somewhere else that is only licenced under the CC BY-SA 3.0 or some other compatible licence which isn't compatible with the GFDL. While as a matter of policy, we require that our content is compatible with the CC BY-SA 3.0 hence you cannot import text which is only licenced under the GFDL, this does not affect ultimately we only have to comply with one of the licences to fulfill our obligations to our contributors, just like any re-user of our content. Nil Einne (talk) 16:17, 22 January 2022 (UTC)[reply]

  I would add that if we put aside what the ToU says and what the licences say and the advice we've been giving for a long time for how we ensure copyright compliance; if we really couldn't delete revisions without violating our contributor's copyrights, we'd have a major problem. Let's put aside revdeletions of copyright violations, there are sometimes other reasons why a lot of revisions have to be deleted, especially BLP or privacy issues. Remembering also that even deletion of revisions on this page and other talk pages is a problem unless you argue that a list of revisions isn't but a list of revisions along with talk page signatures is enough. Again IANAL etc so perhaps an "end justifies the means" argument could be made that to protect against libel but it seems a bit odd to me that we can violate someone unrelated's copyright for this reason. But in any case, there is also a lot of stuff where there's no risk of libel e.g. real names and other personal details of article subjects or even family members or other people, and also of contributors (sometimes including outing) and it seems even more dubious to claim we are allowed to violate contributors' copyright to rev-delete in those cases. Also, as mentioned my Moonriddengirl and others, it isn't just random advice but it's the pratice we've been following too when the issue arises. (Nowadays we generally try to ensure the article is undeleted somewhere whenever possible and if it's translation link to the translation, but I'm fairly sure you can find cases where all we have are a list of contributors.) So if we didn't already cover this in the ToU or it wasn't enough, we really need to fix this in some way. I mean we could somehow craft an exception that covers those cases only, but why? It makes much more sense that we ensure we legally can rev delete whenever, and decide by policy when we should do so. Nil Einne (talk) 17:46, 22 January 2022 (UTC)[reply]

Let's ask the WMF

I just sent this email to Legal:

As a result of this AN thread (permalink) I'm asking a question about whether revdeling copyright violations, in particular revdeling several revisions over a long period of time, is consistent with Wikipedia's creative commons license.

In particular, the RD1 criteria for copyright violations says "If redacting a revision would remove any contributor's attribution, this criterion cannot be used". There has been some debate over what this means; my reading is that, as long as the usernames of the editors are kept in the history, then attribution is still kept, even if non admins cannot see it. Therefore, revdeletions over a long period of time are ok. In doing larger revision deletions for copyright, I believe most admins adhere to this particular reading.

I'm asking if my interpretation of keeping attribution is correct and in line with Wikipedia's creative commons license and the related attribution requirements. If it is not, I ask what should instead be done in the case of large scale revision deletions. I will post this email to the AN discussion, in order for the community to know what I have exactly asked, and so that this question can have a public answer.

Also pinging Moonriddengirl who may have something to say. Moneytrees🏝️^{Talk/CCI guide} 05:28, 21 January 2022 (UTC)[reply]

Moneytrees, I am one of those admins ("attribution" to me means "who edited" not necessarily "what each person added"), so please either ping me here or otherwise let me know of Legal's reply to your query. Primefac (talk) 10:59, 21 January 2022 (UTC)[reply]

It will definitely be good to have that issue clarified, though I would like to point out that the notion of attribution is not confined to its legal aspect. There's also one other thing that we'll need to ask the legal team about: whether revdelling is necessary in the first place. One of the competing assumptions in this discussion is that having small-scale copyright violations in the history of pages will pose a legal risk to the foundation. I think we need to find out if that is indeed the case. – Uanfala (talk) 14:10, 21 January 2022 (UTC)[reply]

I do have something to say although it won't be as useful as what Legal will have to say. Nota bene that I have not read all of the discussion below. Somebody pointed out to me that I had been pinged, and since I don't have a ton of time I've only read the above. Also, I am not a lawyer. I work for WMF Legal, but copyright is not part of my job other than leading the team who executes on the attorneys' directions in cases of DMCA. I'm speaking in my volunteer hat totally. :) Legal has spoken to the issue of copyright violations in page history in the past, but it's always important to remember that legal opinions will change according to evolving issues of precedent. See m:Wikilegal/Copyright Status of Wikipedia Page Histories. (Frankly, knowing that the Foundation's attorneys are not able to provide legal advise to volunteers, I've always found that a bit opaque - but I understand why it is.) However, for me, revdeletion of copyright violation was never primarily about the issue of hosting copyvio content in a past edit; it was primarily about the many times those of us who worked on copyright infringement saw material inadvertently returned. When page histories become complex, people do not necessarily notice even ALL CAPS edit summaries saying that copyrighted content has been removed and must not be restored, and in cleaning up present copyright issues by reverting to earlier versions, I have myself returned major infringing material without realizing that the older edit was problematic. It has long been Wikipedia policy, well before I began working in copyright, that a list of authors was sufficient to meet the requirements of the license. Before revdeletion was a thing, guidance in some cases was to restore the content without the copyvio and include an attribution list of authors on the talk page. This suffices according to Wikipedia:Copying_within_Wikipedia#List_of_authors. As the Terms of Use notes, an attribution may be supplied in any of the following fashions: "in any of the following fashions: a) through a hyperlink (where possible) or URL to the article or articles you contributed to, b) through a hyperlink (where possible) or URL to an alternative, stable online copy which is freely accessible, which conforms with the license, and which provides credit to the authors in a manner equivalent to the credit given on this website, or c) through a list of all authors. (Any list of authors may be filtered to exclude very small or irrelevant contributions.)" Just as we have supplied lists of contributors in the past on talk pages, I myself have considered edit histories a list of contributors even if individual edits are not identifiable (as they are not, through lists of authors). However, I have also always yielded to Wikilegal. --Moonriddengirl ^(talk) 16:12, 21 January 2022 (UTC)[reply]

So it turns out we already have some guidance from the Legal team. Even though it doesn't give us the sort of "OK"–"Not OK" answer that we would have liked to, it certainly provides a lot of useful context. And it does seem to throw out of the window the notion that the foundation could get into legal trouble because of unredacted copyvios in old revisions.

The question then arises, why do we do revdel for small-scale copyright infringements in the first place, intervening edits or no intervening edits? Moonriddengirl's comment above points to the expedience of preventing the restoration of the infringing content by editors digging through the page history. But as hinted in this subsection, revdel'ing may also make it more difficult for the watchers of a page to recognise it if the infringing material is added again. – Uanfala (talk) 15:26, 22 January 2022 (UTC)[reply]

To avoid having to drag either the well-meaning newbie through 3RR, or the volunteer who removed the infringing material to get dragged to there. To signal that the content was problematic and should not be readded. To ensure that future editors don't look at the history, notice that 2k text got suddenly removed a few months ago and just paste it back in without paying attention to the edit summary that removed it, or the copyclean notice on the talk page. To follow the precautionary principle in all copyvio matters. It's far from perfect, but it's the least imperfect system we have, even when it comes at the cost of occasional wall of text ANI discussions and confusion between license versions. MLauba ^(Talk) 16:09, 22 January 2022 (UTC)[reply]

Well, if the current practices are all down to these minor – and, as pointed out above, not uncontroversially relevant – practicalities, then it seems really difficult to see justification for them. Unless someone points out relevant new legal considerations, then I think the best way to avoid such walls of text at AN in future is to have a big RfC reconfirming our policies and then making sure people stick to them. – Uanfala (talk) 17:40, 22 January 2022 (UTC)[reply]

• Uanfala, you state And it does seem to throw out of the window the notion that the foundation could get into legal trouble because of unredacted copyvios in old revisions. No it doesn't. She opined as an editor, talking about some of the motivation. She did NOT say there is zero risk of legal action. You need to read more carefully. If you had read what she linked, you would have seen "Section 108 of the Copyright Act[2] allows libraries and archives to reproduce and distribute copyrighted works under certain circumstances. However, this statute likely does not apply to archives with only a digital presence, such as Wikipedia page histories, and therefore probably does not exempt the use of copyrighted material on Wikipedia from infringement liability. Fair use may be invoked as a defense in the event of copyright infringement in Wikipedia page histories, but the judicial outcome of asserting fair use is generally unpredictable. " which is the official position, and certainly is NOT saying there is no liability. In fact, is clearly saying we are not immune to action by a copyright holder, we would have to invoke Fair Use, and that may not stand up in court. So you are mistaken, there is a legal consideration when we RevDel infringing material. It isn't always the only reason, but it is always one of them. Dennis Brown - 2¢ 13:37, 23 January 2022 (UTC)[reply]
  • It was precisely because of reading that document that I posted above. The passage that's most relevant here is the following: In the case of Wikipedia page histories, [if a fair use defense fails in court,] the potentially liable party would be the user who uploaded the copyrighted work, as that initial upload is what allegedly violates the copyright owner's exclusive rights. Additionally, downstream re-users of the work may risk liability if their own use of the work is not permissible under fair use or another defense. Is there anything I'm missing here? – Uanfala (talk) 14:13, 23 January 2022 (UTC)[reply]
    • What you are missing is the fact that you don't want Wikipedia to be dragged to court to begin with. Even if you win, it is very expensive and time consuming. You avoid this by removing obvious infringement from the archives. Because Wikipedia is hosting that infringing material, and they don't qualify as a common carrier (the standard in US law for exemption), the infringed party could still tie up a lot of time and money. So we RevDel. The only loss is the rare listing of specifics of the edit someone is credited for, which isn't required for attribution under the current license. So no, it does not throw out the window the idea that it could cost the Foundation/Legal time, money, effort, hassle. Dennis Brown - 2¢ 15:04, 23 January 2022 (UTC)[reply]
      • So we're going to gear all our practices towards the eventuality that someone may one day spot one paragraph from their website in an old revision on Wikipedia and then instead of asking nicely or sending a takedown notice would go straight for a court case that they know they won't win? If this is indeed the case, then I would really like to see that spelt out by the legal team, because there's nothing either in our policies or the documentation presented so far to suggest that revision deletions like the one Himachal Pradesh are necessary.
        All this matters because the losses from such deletions are not as insignificant as they're made out to be. It's not about satisfying the legal requirement for attribution, that's a red herring (though yes, even here it would be nice, for example to be able to re-use content by citing only the actual contributors rather than be forced to enumerate everyone, including all the vandals, who has touched the page). It is about our ability to look after articles. Revision deletions erase the link between the edit, the user who made the edit, and the explanation for that edit. This makes it a range of editing activities more difficult, like selecting which parts of an article are likely to require attention, or cleaning up after disruptive users (including copyright violators!), and it makes it more likely for errors corrected in an intervening edit to get repeated.
        The sort of basic accountability that revdel erases is essential for dealing with any moderately developed article, and I don't think it's reasonable for the community to be actively taking that away from itself because of a hypothetical danger than no-one legally competent enough to comment has claimed to exist. – Uanfala (talk) 15:44, 23 January 2022 (UTC)[reply]
        • This feels like FUD more than anything. Every issue listed (which I mostly disagree with, but don't want to waste more space arguing them) applies to every form of redaction; should we stop hiding BLP issues per RD2 because it will make cleaning up after disruptive users more difficult? Should we stop hiding disruptive links per RD3 because selecting which parts of an article are likely to require attention will be somehow harder?
          None of the arguments you've just made have to do specifically with RD1/copyright revisions (in fact, you're specifically avoiding that issue), and since regardless of whether someone will sue us for it, we should take preventative steps to avoid it, you should either expand your concerns to all of Revision Deletion or bring copyright-specific concerns with the use of it. Primefac (talk) 16:42, 23 January 2022 (UTC)[reply]
          • I was simply pointing out some of the benefits of having article histories (which seemed to be dismissed by the previous comment). The WMF Legal document itself acknowledges them. Primefac, I don't understand your demands here. I came here to challenge an RD1 deletion that was a stark violation of our policies; it would have been equally unacceptable if that deletion had been done under any other non-oversightable revdel criterion (and this is already covered by the policy). People were endorsing the deletion because of an understanding that otherwise we risk getting into legal trouble, but no credible indication has been given that this is the case. I'm not buying the principle that we should go out of our way to avoid all conceivable vexatious lawsuits. If we did that, then our articles would never contain criticism of public figures because of the risk the article subjects could take us to court for libel. – Uanfala (talk) 17:44, 23 January 2022 (UTC)[reply]
            • That is the problem. You still call it "RD1 deletion that was a stark violation of our policies" when in fact others have demonstrated it was not a violation of policy. You raised the question, you didn't like the answer you received, you keep bludgeoning the issue. Dennis Brown - 2¢ 20:59, 23 January 2022 (UTC)[reply]
    • Uanfala The other thing you're missing here is that you, actually, are a downstream reuser of an unauthorized copy of some third party's work. Granted, you didn't modify the text in question, but the introduction of the copyvio content in the article turned it into an unauthorized derivative work. Every edit after the introduction of the copyvio is a fruit of a poisoned tree. Sure, we can all reason all day long that no reasonable owner, or judge, would ever treat the rest of the article as any material component to a copyright infringement case. Except there's precedent out there, for instance from the music industry. Part of any defense the WMF would have to mount if any copyright infringement case were to go to a court is that both the WMF but also the volunteer editors took all reasonable measures to address issues we are made aware of. RD1 is one of an arsenal of many small scale reasonable measure that we can point to to demonstrate that we take such issues seriously. And if that comes at the cost of someone not being able to see exactly what you did to the article by pruning the language section, I'm sorry, but that's a very small price to pay for everyone's benefit, including your own. And even if you don't care for it, you are not alone being shielded by these RD1s. All of Wikipedia is. MLauba ^(Talk) 22:17, 23 January 2022 (UTC)[reply]

I have an interesting perspective on this because I consult on copyright infringement cases. Wikipedia is much more thoughtful than most websites and it would be a bad business decision for a lawyer to go after a non-profit that makes an occasional good faith error. We should not spend so much angst speculating about the law, because we're not lawyers for WMF. WMF legal will let us know if they are getting takedown notices or if they sense a problem. If we need to change our process, WMF legal will give us advice. I move to close this thread. Jehochman ^Talk 21:09, 23 January 2022 (UTC)[reply]

I intend to try to clarify policy at WT:Revision deletion#RD1, attribution and intervening edits after discussion ends here. I found a similar discussion in 2021 and one in 2020, so this has been a recurring issue. While the licensing piece is not in dispute here, I created WP:Attribution does not require blame as an essay for future reference. It is very straightforward, so I expect that WMF Legal's response will not affect it. Flatscan (talk) 05:26, 24 January 2022 (UTC)[reply]

Primefac Uanfala Justlettersandnumbers, Legal got back to me on the 26th, I was busy then and only noticed just now. They say they agree with my interpretation, and cite the part of section 7 of the TOU, where it says "Through hyperlink (where possible) or URL to the article to which you contributed (since each article has a history page that lists all authors and editors)". They explicitly tell me that it does not require that each contribution made is readable, so use of revdel and oversight does not cause attribution problem under Wikipedia's CC license and is compliant with the TOU. They note at the end that this is their perspective (which is that they agree with me), but it is not legal advice. Moneytrees🏝️^{Talk/CCI guide} 05:24, 30 January 2022 (UTC)[reply]

Can some kind admin close this thread? The consensus is clear and discussion has died. TrangaBellam (talk) 08:43, 20 January 2022 (UTC)[reply]

Now archived. TrangaBellam (talk) 19:54, 23 January 2022 (UTC)[reply]

As a note, while the need for a reclose is indeed crystal clear, the admin will need some spare time to redo the original close - I was going to handle until I realised I wouldn't have time to do it justice --Nosebagbear (talk) 15:06, 27 January 2022 (UTC)[reply]

Hi all,

This is no slight against the admin who initially responded to the situation---I believe they act in good faith.

A subsection of of the Sandy Stimpson page about a politician referring to local articles as first-hand sources was declared to be breaking the rules by an administrator and I'd like a follow up on it. While this is a politician, and necessarily will be subject to a variety of viewpoints, I attempted to continue the inclusion of a section which described acts caused by their son. The following rule was invoked: WP:ONUS.

Specifically, I'm asking if this rule applies in this particular case. I personally believe that the politician's reaction and relation to events was covered by this rule. Recently I've recognized that thsi rule does not apply to this article given the present wording. I may be interpreting this wrong, in which case I'd hope this rule was updated to be more specific. A person publicly reacting to events (in that public citations exist and are relevant) should be included in Wikipedia articles. I want to make sure myself and others are able to easily follow this rule moving forward.

Thanks so much for considering this! Apologies if this is the wrong place to bring it up---it's what I found that made sense, and let me know if it should be posted elsewhere!

--Marx01 ^{Tell me about it}

• Sandy Stimpson (edit | talk | history | protect | delete | links | watch | logs | views)

This appears to concern Talk:Sandy Stimpson#Wrt Stimpson's son's arrest from October 2020. The edit was diff which was added then removed by Marx01. It's easy to use Wikipedia articles to pile-up muck concerning a living person (see WP:ADAM) but that is not the purpose of Wikipedia. In particular, an article about a person is not available to coatrack negativity about family members. Johnuniq (talk) 08:40, 23 January 2022 (UTC)[reply]

@Marx01: I also see two edits from last year you made to the article that were reverted by User:Neutrality who you should have notified. Doug Weller talk 08:44, 23 January 2022 (UTC)[reply]

@Marx01:We can't tell what or who you are talking about. Please provide diffs, and you are required to notify the editor in question. You have not touched the article in five months, so we are really guessing here. If this is about the 2020 attempt to mention the subject's son's legal problems, I agree with its removal. Meters (talk) 08:58, 23 January 2022 (UTC)[reply]

Why, yes. If this is the edit in question, I should have been notified of this discussion. I removed negative BLP about a third party that had no bearing on the content of the article. There was no indication of it's relevance beyond the personal relationship involved. --Deepfriedokra (talk) 17:54, 23 January 2022 (UTC)[reply]

Thank you both for your responses! Apologies for the confusion; I haven't gone through this process before and missed that I should have notified the editor in question, though reflecting that makes complete sense. I was mainly bringing this up after thinking about it again for the first time in quite a while, I've left the article alone as a result of worrying about this rule. I think this clears up my concern, though I'm still a little confused about the line drawn by this rule. --Marx01 ^{Tell me about it} 20:15, 23 January 2022 (UTC)[reply]

@Marx01: simply put, just because a Wikinotable person's son commits a crime, it doesn't mean that the Wikinotable person is guilty by association. We do not tar the parent with the son's brush. Mjroots (talk) 17:37, 24 January 2022 (UTC)[reply]

Hello I want to point out that some registered users in particular Slay-12345678910 and Lorenzo 1109 who are repeatedly removing the DragRaceProgressTable/14 template on the page RuPaul's Drag Race (season 14), as usual, by inserting tables that do not respect the neutrality of the platform quoting the "Can yall give credit to the queens work .... like only showing who won the btm and elim can completely make someones track seem better or worse. like crystal methyds track looks way worse than it actually is. This is why people don't use wiki for track records anymore", which I find IMO nosense since we report the data that are actually announced during the episode (WIN - BTM - ELIM).

Of course, I've already helped re-adjust the table to standards, but ut there is an urgent need for blocking changes for both users cited and anonymous IP addresses that persist in modifying the progress table. --- « Ðømīnīk • Cåpuån » 16:58, 24 January 2022 (CET)

WP:AN3 is the place to deal with edit-warring (as I suspect even if these users are blocked now there will be others). Primefac (talk) 11:10, 25 January 2022 (UTC)[reply]

It is looking possible that sometime this year, the country currently known as Turkey will be acknowledged to have changed its English name to Turkiye or Türkiye (matching the spelling in Turkish). However, it has not yet done so to the satisfaction of the UN, the AP, or other sources that Wikipedia follows. There is likely to be a steady stream of ECP editors who want to change this name on articles such as Turkey before there is consensus to do so. Could admins please watchlist that page (and related articles such as Name of Turkey) to watch for any disruption that may occur, and to shut down untimely RM discussions? User:力 (powera, π, ν) 18:03, 25 January 2022 (UTC)[reply]

Even if it does change its name, we should retain its current name until such time as the common spelling changes. -- Rockstone^{Send me a message!} 03:35, 27 January 2022 (UTC)[reply]

There are multiple IP editors that are in an active edit war on the article. I did not want to create multiple edit war reports, so I think this is the best way to alert the admins to it. Here is a list of the IP’s involved:

1. User talk:76.69.7.202 — Main instigator
2. User talk:184.146.205.69
3. User talk:108.176.101.253
4. User talk:72.43.250.158
5. User talk:24.97.254.222

List of edits involved: [5], [6], [7], [8], [9], [10], (18 edits (in a row) by 76.69.7.202 between 21:09 - 21:28 January 21), (Revert of the 18 edits). [11], Smaller instances between January 22 to today's last edit for this discussion [12]. It is hard to keep track of all the edits, so instead of one very long/large confusing list, I recommend viewing January 14–17, 2022 North American winter storm's revision history to see the edit war.

I do not know if all of the IP’s have broken the 3 revert rule, but the edit history shows the crazy ongoing “war” between them. User talk:76.69.7.202 is the main reverter. Elijahandskip (talk) 16:29, 26 January 2022 (UTC)[reply]

Ridiculous accusation calling me the "Main instigator" on a non-existent edit war. I have clearly summarized each edit and used the talk page. 76.69.7.202 (talk) 20:46, 26 January 2022 (UTC)[reply]

That was not meant to be a "ridiculous accusation", as you are correct that you did use the talk page and edit summaries, you are the one who appears to be the main reverter (By reverting the edits of the other editors). Based on all the reverts done by other editors, I would estimate you broke it "unofficially" multiple times already, by reverts over the course of a few days (Stated below by Fram). While I believe you did the correct thing, at least to start with, you did continue o revert the edits instead of asking other editors to communicate on the talk page. That continuous revert, without asking for talk page communication, just added fuel to the fire to keep this edit war going. Elijahandskip (talk) 21:15, 26 January 2022 (UTC)[reply]

Uh, 3RR only applies when you do three reverts in 24 hours. The 76 IP made one revert the 26th, none on the 25th, one revert the 24th, 2 reverts the 23th, 2 reverts the 22nd, ... If you want to accuse people of "officially" and "unofficially" (???) breaking 3RR, you better provide evidence or retract your claims. Incorrect accusations only reflect badly on the accuser. Fram (talk) 08:48, 27 January 2022 (UTC)[reply]

Note that Elijahandskip changed their comment after I replied to it, and that their addition of "stated below by Fram" may give the impression that I somehow support their assertion of an editor "unofficially" breaking 3RR, while I have said no such thing and have no idea what "unofficially" breaking 3RR even is supposed to be. Fram (talk) 08:26, 28 January 2022 (UTC)[reply]

I've only been on wikipedia a month. I tried to add some useful text to the article AND to ensure the references given actually contained the numbers 76.69 was adding. I think she/he only found one reference for an area of Toronto getting 45 cm (but others say 50 or 55) after myself and two others kept saying there was no reference for 45 cm in Toronto. I didn't even make any changes for 4 days until today and just tried to mention 3 different measurements in parts of Toronto - downtown and Pearson Airport are the recongized En. Can weather stations. I note that even the reporter of this war (Elijahandship) tried to unrevert 76.69.7.202 reverts twice and 76.69 reverted his unreverts both times. Another editor, 72.43.205.158, even said on the talk page for the Jan. 14-17 storm that 76.69's conduct should be reviewed. User:184.146.205.69

Note Since this dispute is still ongoing(I assume) I have requested temporary full protection, so that this discussion may take place and the dispute gets resolved. 🌀CycloneFootball71🏈 |sandbox 21:47, 26 January 2022 (UTC)[reply]

• @CycloneFootball71: - are there extended-confirmed users involved in the edit war? Otherwise regular semiprotect would work just as well, without the huge side effects -- Nosebagbear (talk) 15:02, 27 January 2022 (UTC)[reply]

@Nosebagbear:, I did see 2 extended confirmed users who appeared in the conflict, however after looking at the edit history again they stopped conflicting just before they reached 3rr. When I first noticed the discussion, I looked over the history and then decided to request to fully protect it, at least until the discussion had been resolved so that there wouldn't be continued disputing at the same time. Unfortunately I didn't go more into depth on the history so I missed many key points most likely, and I would like to apologise for acting too quickly on that. My request was denied, so you're probably right to semi protect at least in the mean time to sort it all out. 🌀CycloneFootball71🏈 |sandbox 04:14, 28 January 2022 (UTC)[reply]

• Commen/Proposal - After a careful look at the edit history, I can say no one broke the 3RR rule. That said, multiple editors conducted 1-3 reverts every day. So it is one massive content dispute more or less. I do not believe any editors should be punished as no rules were actually broken, however, there is too much going on to leave it as is. I would support some level of page protection for a month or so too let editors discuss on the talk page rather than continuous reverts mixed with talk page discussions. CycloneFootball71 had a denied full page protection request, however, Nosebagbear suggested a semi page protection. I do not know which would be better, so I will let other editors do a small discussion on my proposal and see the next course to take. Elijahandskip (talk) 04:38, 28 January 2022 (UTC)[reply]

• Withdrawing as content dispute appears to have ended just after this AN began. Elijahandskip (talk) 23:26, 29 January 2022 (UTC)[reply]

In a process that began last year with WP:DS2021, the Committee is evaluating Discretionary Sanctions (DS) in order to improve it. A larger package of reforms is slated for sometime this year. From the work done so far, it became clear a number of areas may no longer need DS or that some DS areas may be overly broad. This discussion is intended to focus on those areas. Community feedback is invited and welcome at Wikipedia:Arbitration/Requests/Motions. Barkeep49 (talk) 17:03, 27 January 2022 (UTC)[reply]

Hi, several parts of WP:PERM are currently backlogged into something like 6 January, specifically the AP and AWB sections (full disclosure: I have a open request at PERM) -- Asartea ^{Talk | Contribs} 18:41, 27 January 2022 (UTC)[reply]

Thanks for the notice. I've granted three requests for AP and will examine some more as time allows today. – Muboshgu (talk) 18:58, 27 January 2022 (UTC)[reply]

Just to remind administrators here, that there is quite a big backlog at the request for AutoWikiBrowser page. Thanks. Itcouldbepossible^Talk 03:41, 28 January 2022 (UTC)[reply]

I've removed your header because it's the same issue as the preceding section. Primefac (talk) 08:04, 28 January 2022 (UTC)[reply]

@Primefac No problem, I did not know that there was a previous thread already. Itcouldbepossible^Talk 08:25, 28 January 2022 (UTC)[reply]

Akua Taylor was exempted from IP Blocks by zzuuzz but the situation still pertains.— Preceding unsigned comment added by Kinvidia (talk • contribs)

I've taken care of this. Kinvidia, try my talk page first next time :) -- zzuuzz ^(talk) 13:45, 29 January 2022 (UTC)[reply]

Hi zzuuzz apologies for not informing you earlier. Thank you so much for the quick response. Kinvidia (talk) 01:25, 30 January 2022 (UTC)[reply]

I noticed there’s an Article called Battle For Dream Island Which Doesn’t have a Talk Page.The Talk Page appears to be Locked under Administrator Access only.Can The lock be Removed so that way a Talk page for the article can be created? — Preceding unsigned comment added by Danubeball (talk • contribs) 23:04, 29 January 2022 (UTC)[reply]

Should be available to edit now. RickinBaltimore (talk) 23:20, 29 January 2022 (UTC)[reply]

2603:9001:5609:BE3F:CC19:9544:671E:8ED9 (talk · contribs) has made four edits which contradict cited sources. In particular, '8ED9 has changed the dates that the Sun enters or leaves astrological signs for the dates the editor prefers, with no apparent understanding that we should not put words in the mouth of a source that the source did not say. I have corrected three of these and warned the user, but to avoid contravening the three-revert rule, I ask that someone else look at this edit. Jc3s5h (talk) 01:06, 30 January 2022 (UTC)[reply]

Based on this I'd chalk it up to run of the mill number changing vandalism. Also 3rr applies on a per article basis, and while you shouldn't follow someone around referring their good faith edits based on disagreement, I don't think there's an issue reverting disruptive, unsourced number changing. ScottishFinnishRadish (talk) 01:13, 30 January 2022 (UTC)[reply]

I don't consider it vandalism. I guess the editor believes the dates are true, but doesn't understand how footnotes and citing sources work, and that if you attribute a statement to a source and the source didn't make that statement, it is a falsehood. Jc3s5h (talk) 01:22, 30 January 2022 (UTC)[reply]

Saying "I know a guy who acts like one astrological sign and not another, so the date has to be wrong" makes it difficult to assume good faith. ScottishFinnishRadish (talk) 01:26, 30 January 2022 (UTC)[reply]