Wikipedia talk:Copying within Wikipedia/Archive 1

Archive 1

Archive 2

Archive 3

Purpose of page

Wikipedia's copyright policy sets out the conditions under which material may be reused, but there is no clear guideline explaining to Wikipedians precisely how to do that. We have Help:Merge and Wikipedia:Split, but sometimes those documents are not to the point: what if material is copied, rather than moved? The requirement for attribution does not change.

Note that this is not dealing with a theoretical situation. I encounter this frequently. --Moonriddengirl ^(talk) 10:43, 7 October 2009 (UTC)

Article history cleanup, future Revision deletions, take 2

I would suggest adding link to a edit that did such a fix by a dummy edit, i.e. what I want is an example of an edit summary for a clean up fix. --Stefan ^talk 12:31, 23 October 2009 (UTC)

There you go, this is pretty much the language I use when I do selective deletion + restore these days. MLauba (talk) 14:42, 23 October 2009 (UTC)

Interlanguage/Interwiki copies

At Village Pump, User:TheFeds suggests we cover interlanguage/interwiki copies from other Wikimedia sites. This is a very good suggestion, since I don't believe we have clear guidance on this anywhere. meta:Help:Transwiki proposes a list of contributors as well as a link back for transwiki. Looking at some articles listed at the talk page of Wikipedia:WikiProject Intertranswiki, I see a reference here, and a template {{Translated page}}. Perhaps we should include this under specific situations? --Moonriddengirl ^(talk) 11:49, 24 October 2009 (UTC)

Of the existing pages (WP:Copy-paste, WP:Reusing Wikipedia content), this seems like the best place to cover it. I think we may want to consider splitting off Help page(s), which would allow for more detailed instructions. Flatscan (talk) 03:31, 25 October 2009 (UTC)

All right; I have done. I personally would prefer to avoid splitting off into additional pages if we can to avoid sending people on a scavenger hunt for the precise instructions they need. :) I've subsectioned the "special" area to make it easier to find specific situations in the TOC and also to make it easier to link to them from talk pages or other policies/guidelines. I'll update at VPP. --Moonriddengirl ^(talk) 13:06, 25 October 2009 (UTC)

Signpost mention

This page was mentioned in the Signpost: WP:Wikipedia Signpost/2009-11-16/News and notes#New guidance on copying within Wikipedia. Flatscan (talk) 03:53, 18 November 2009 (UTC)

Guideline proposal

The following discussion is archived. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

The document has been promoted to guideline status. --ThaddeusB (talk) 03:54, 11 November 2009 (UTC)

---

rfctag policy placed 25 October, removed at promotion by ThaddeusB 11 November

Consensus needed to promote (or not) a proposed guideline to clarify conditions under which material on English language Wikipedia and other Wikimedia projects may be reused within EL Wikipedia itself in accordance with copyright policy and Terms of Use. --Moonriddengirl ^(talk) 13:28, 25 October 2009 (UTC)

Reasons for

• There is currently no clear guideline governing conditions under which material on English language Wikipedia and other Wikimedia projects may be reused within EL Wikipedia itself in accordance with copyright policy and Terms of Use. --Moonriddengirl ^(talk) 16:51, 26 October 2009 (UTC)

Support

• Support promoting to guideline. This really spells out some of the terms used on WP:C in a clear and accessible manner, it represents, in practice, a simple restating of an established policy to make it more readable. MLauba (talk) 10:38, 26 October 2009 (UTC)
• Support promoting to guideline. I run into copying within Wikipedia on a regular basis, and there is currently nothing to which contributors can be pointed that succinctly spells out the way such reuse is handled. --Moonriddengirl ^(talk) 10:41, 26 October 2009 (UTC)
• I also support promoting this to a guideline per User:Moonriddengirl's argument above. Theleftorium 10:56, 26 October 2009 (UTC)
• Support - much needing instructions. --ThaddeusB (talk) 15:12, 26 October 2009 (UTC)
• Support promotion to guideline – in this area, there is a widespread lack of understanding among users and a shortage of written guidance. Flatscan (talk) 03:50, 27 October 2009 (UTC)
• Support seems a reasonable explanation and provides the required guidance on the subject. MilborneOne (talk) 23:11, 5 November 2009 (UTC)
• Support- Much needed and well written! --Stefan ^talk 11:53, 6 November 2009 (UTC)
• Support - there's been too little guidance on this up to now. – Toon 23:03, 10 November 2009 (UTC)
• Well done. A handy guide, with an appropriate level of detail and background. TenOfAllTrades(talk) 23:18, 10 November 2009 (UTC)

Oppose

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

Objections at VPP

There were some procedural objections raised at WP:Village pump (policy)/Archive 69#Wikipedia:Copying within Wikipedia has been marked as a guideline archived Flatscan (talk) 04:46, 24 November 2009 (UTC) when this page was promoted a few days ago. I don't see any substantial opposition, but I thought it would be a good idea to keep a link around. Flatscan (talk) 01:49, 16 November 2009 (UTC)

Article history cleanup, future Revision deletions

Food for thought, when revisions are removed from article history, either through selective restore like today, or using a future revision deletion feature, it has to be made clear that deletion without attributing what is kept in the article introduces a new copyvio at the expense of removing one. As the discussion at WT:REVDEL shows, what the people used to copyright work take for granted isn't necessarily recognized as the best practice. MLauba (talk) 14:00, 7 October 2009 (UTC)

Good point. Hmm. --Moonriddengirl ^(talk) 14:37, 7 October 2009 (UTC)

I've started a section on it under "Specific situations." Please feel free to expand as seems appropriate, remaining mindful of the KISS principle. :) We want to be careful to avoid bloat. --Moonriddengirl ^(talk) 14:45, 7 October 2009 (UTC)

Done so, but it's probably WP:CREEPy now. ;) MLauba (talk) 15:48, 7 October 2009 (UTC)

I cut it down a little. I think the sweet spot for this page would be 2-3 sentences covering both page deletion and revision deletion, along the lines of "if the content was originally added in a deleted revision, it can't be used without additional attribution". More can be written, but I think it would be better suited at WP:REVDEL or a supplementary Help: how-to page. Flatscan (talk) 03:40, 8 October 2009 (UTC)

Considering the discussion at WT:REVDEL is highlighting that common admin practice is actually introducing attribution infringment quite often when a deletion + selective restore is done by an admin not well versed in copyvio matters, I'd be in favour of making sure the text is explicit enough in all relevant policy & guideline pages. The rewording you did up to this revision is actually fine for me, and below what I'd consider WP:CREEPy. MLauba (talk) 07:46, 8 October 2009 (UTC)

Although their problems and solutions are very similar, deletion (revision hiding) is not the same as copying, which makes coherent discussion difficult. I'm fine with rewriting and renaming this page to cover both in separate sections, but WP:Attribution is already taken. Flatscan (talk) 03:37, 9 October 2009 (UTC)

(undent) Oh, I finally see what you were meaning. You're right, but at present, attribution now represents the bulk of the proposed guideline. Perhaps creating WP:Contributor attribution and shortcut WP:BY (as in CC-BY-SA) would be feasible? MLauba (talk) 10:30, 9 October 2009 (UTC)

WP:Contributor attribution is much better than my idea of "Contribution attribution". The redirect left at WP:Copying within Wikipedia can be pointed at a relevant section. Flatscan (talk) 03:01, 10 October 2009 (UTC)

Are there any objections to renaming as proposed? Flatscan (talk) 02:22, 12 October 2009 (UTC)

Not from me :) MLauba (talk) 08:21, 12 October 2009 (UTC)

In case it wasn't clear from the lack of activity, I aborted the rename. I got stuck on a minor aesthetic issue in the examples that was mitigated by Moonriddengirl's organization into subsections. Flatscan (talk) 03:46, 27 November 2009 (UTC)

Putting this into practise

Time to put this new guideline into practise with a test case...

Please see this discussion about a project where material has been routinely copied verbatim to a large number of articles without attribution.

If you're really passionate about this topic, it would be awesome if you could help fix some of the articles with the necessary attribution! Many thanks, Knepflerle (talk) 12:46, 24 November 2009 (UTC)

Assistance is still required, if anyone is still around here... Knepflerle (talk) 14:06, 10 March 2010 (UTC)

Sorry. Personally, I would be glad to help if my plate weren't already so full. :( I also know the pain of trying to get other people to help out with this kind of work. WP:CCI is picking up more volunteers (yay!), but we have thousands of articles waiting attention there still, and I have yet to figure out a good approach to getting enough people involved to actually keep up with the workload; I don't even know where to ask. I hope somebody will respond to your need, though; FWIW, I think it's great that you're working on it. --Moonriddengirl ^(talk) 14:23, 10 March 2010 (UTC)

Peculiar deletion policy

I understand there is an attribution requirement, but I think that someone has gone a bridge too far with the policy about deleted articles. It is contradictory to say that an article can be properly attributed with a simple link in the edit summary, but then say that a deleted article can't be reused - because the article using a link-attribution might at any time be pointing at a deleted article.

Somehow Template:Copied, with its dire warnings that such-and-such article must not be deleted, has been elevated to a Wikipedia policy.

Is it really the belief of Wikipedia that a third-party site which displays material attributed to Wikipedia must immediately shut down should the content ever be deleted here? If so, the advice sometimes given that reusers can attribute the material by a simple link to Wikipedia is rather misleading - they could become Copyright Criminals at any time, without warning. Perhaps we should design some handy tools to spit out a list of a few thousand editors to an article and the articles from which parts of an article might have been copied, so that the reusers can easily append a few megs of stuff so they know that the free content they've copied can be used. (Hmm, but what if some of the contributors' accounts have been deleted or moved?)

But if not, then those sites are granted greater freedom than Wikipedia editors.

I think that the actual license requires attribution to the original article, but says nothing about whether that article is "deleted" or not. Deletion isn't really deletion anyway - someone interested could ask an admin for attribution information, for example. Wnt (talk) 16:56, 27 February 2010 (UTC)

Hopefully others will comment, as I'm not intimate with cross-site reuse. I believe that the situation you describe – an external site attributes via hyperlink only, then the article is deleted on Wikipedia – the site is in technical violation, same as on Wikipedia. I think that (I thought it was a page about "transwiki", but I can't find it) copying the page history or a condensed list of authors is recommended when copying a page to a wiki without access to Special:Import. Flatscan (talk) 05:22, 28 February 2010 (UTC)

The authors of the text MUST be credited (failure to do so is plagiarism). That means the history or list of authors must be copied to the receiving site. This isn't a Wikipedia rule; it is the terms of the licensing agreement (not written by Wikipedia), which is legally binding and not open to community consensus. --ThaddeusB (talk) 02:07, 9 March 2010 (UTC)

That is required only if the page has been deleted, correct? Flatscan (talk) 05:14, 9 March 2010 (UTC)

According to Wikimedia:Terms of Use, the list of contributors is one way to attribute; the hyperlink is another, either to the Wikipedia page or "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" --Moonriddengirl ^(talk) 11:41, 9 March 2010 (UTC)

To clarify my reading, the hyperlink is acceptable while the WP page is visible, but deleting the page – hiding the history/list of authors – makes the hyperlink insufficient. It's very similar to on-wiki copying. Flatscan (talk) 05:10, 10 March 2010 (UTC)

Yup. So long as the history is maintained here, the hyperlink is sufficient, according to WMF. --Moonriddengirl ^(talk) 11:41, 10 March 2010 (UTC)

If a list of authors is needed for a deleted article, WP:Requests for undeletion should be able to provide it. Flatscan (talk) 05:09, 22 March 2010 (UTC)

Attribution for removed content

My understanding is that content that has been removed (not deleted) must still be attributed, as it is easy for any user to restore it from a previous revision. WP:Redirects for discussion/Log/2010 March 4#List of pop culture references to the 69 sex position concerns content that was merged, removed, restored much later, and removed again. The redirect was deleted as a result of the RfD. Flatscan (talk) 05:09, 22 March 2010 (UTC)

Sure, it must, in order to comply with CC-BY-SA, section 4(c) and GFDL, section 4-I. If we do not retain the history, we must retain the full list of contributors (though it can be filtered to remove very small contributions). --Moonriddengirl ^(talk) 10:46, 22 March 2010 (UTC)

Oh, wait. I thought you meant removed (not deleted) as in merged. Pardon me; I haven't had my first cup of coffee yet.:D If the material isn't currently published, I don't think the history is necessary, but there is danger of accidental restoration. I wonder if a history merge would provide the best solution here. I'll ask the deleting admin. --Moonriddengirl ^(talk) 10:50, 22 March 2010 (UTC)

The trouble with a history merge is that the two versions developed in parallel for a considerable time (18 Nov 05 to 24 Apr 06), and we would have the situation described in WP:HISTMERGE#A troublesome case:

If someone then page-history merges pages A and B using the method described above, the result will sequence the versions of A and B strictly by time, with the result that various versions of A will be interleaved between versions in the page history of page B (and/or vice-versa). Inspecting this merged history without means of distinguishing between the two overlapping progressions (since nothing in this history indicates which version belongs to which sequence) invites severe confusion.

My thought in closing the RfD was that if in future the material were brought back into the article, the redirect should be undeleted at the same time, thus bringing back the history; but I suppose there is no easy way to ensure that some future bringer-back is aware of this requirement. Maybe restore the material briefly and remove it again with an edit summary "If you restore this material you must also request undeletion of the redirect List of pop culture references to the 69 sex position which contains its attribution history"?

If you don't think that would be enough, I suggest the method suggested in WP:HISTMERGE - undelete the List with all its history, move it to a subpage of the article talk page, and write on it "This sub-page preserves for attribution purposes the content of the page List of ... which was merged to the article after this AfD and subsequently removed."

JohnCD (talk) 18:52, 22 March 2010 (UTC)

Based on my limited experience requesting histmerges, I agree that one should be avoided here. Flatscan (talk) 04:55, 23 March 2010 (UTC)

Either of those could work; we could also copy the list of contributors (as it is relatively brief) and place it on the talk page, restoring the content briefly with a "see talk page for attribution", linking to the specific subsection. The subpage and talk page attribution methods have the advantage if somebody should restore an earlier point in history without seeing the edit summary note. Occasionally, for instance, I have to restore much older versions of articles when it turns out that extensive copyright infringement was introduced in the history, and I can see that if a history list were long enough I might miss such a note myself. --Moonriddengirl ^(talk) 18:55, 22 March 2010 (UTC)

Thanks for the responses. My preferred solution in this case is the subpage method, also described at WP:Merge and delete#Move to subpage of talk page. An advantage is that {{Copied}} works with it. My opinion is that – in the absence of alternate attribution – the source page and the content (via WP:Selective deletion or WP:Revision deletion) should be deleted and undeleted together. Users should not be allowed to easily restore content lacking attribution, either by keeping the history/list of authors regardless of the current state of the content or by technical measures like deletion. Flatscan (talk) 04:55, 23 March 2010 (UTC)

I have implemented the talk sub-page solution - see Talk:69 (sex position)/List of pop culture references to the 69 sex position - with an explanation on top and all the history beneath. Have a look at that - I think it is probably sufficient, or should I do another restore and delete of the material in the main article so as to leave a marker in an edit summary? JohnCD (talk) 16:37, 23 March 2010 (UTC)

Thanks for taking care of it. I placed {{Copied}}s to be sure – they can be removed or hidden if necessary. Flatscan (talk) 05:15, 25 March 2010 (UTC)

Proposed move of this page to Wikipedia:Attribution

I have proposed Wikipedia:Attribution be moved to Wikipedia:Attributing statements (or something) and this page (Wikipedia:Copying within Wikipedia) be moved to Wikipedia:Attribution. Please comment at Wikipedia talk:Attribution#Vacate this page. –xeno^talk 19:30, 28 April 2010 (UTC)

How to attribute

There's been some confusion recently about what degree of attribution is needed for material copied with Wikipedia: specifically, does the attribution have to be recorded in the edit summary?

The essay as it stands discusses points (b) and (c) of the acceptible means of attribution in the Foundation's Terms of Use, but does not cover point (a), that is attribution through a hyperlink to the original article: on WP, of course, this would be a wikilink.

The purists might say that a wikilink can be deleted in subsequent editing, and so is not sufficiently permanent to ensure attribution in the future, but I think this is an Aunt Sally. Let's look at some cases where a wikilink is used as sufficient attribution:

• the "Today's featured article" section on the Main Page consistes of material redacted from the lead section of the article, and clear links back to the article: are people really suggesting that a link in the edit summary is necessary or even helpful?
• the "Did you know?" section on the main page, and similar sections on many portals, use the same method of attribution, through a clear link back to article.
• sections in articles which are hatted with {{main}} are obviously summaries of the material in the article(s) linked through the template; these summaries will often (but not necessarily) include material which is attributed through that link.
• every single CC-BY image used on WP is attributed through a wikilink; the Wikipedia Manual of Style specifically discourages any other form of image crediting (see WP:CREDITS)

This last example is particularly telling. If we are to insist on edit summary attribution for all copying within WP, we are insisting on better treatment for those people of choose to edit Wikipedia under a CC-BY-SA license than for other authors under that license whose work we use in the encyclopedia. This is clearly and completely unacceptable.

There are, of course, cases where an edit summary attribution is necessary; specifically, those cases where there is no clear wikilink which indicates the source of the material:

• when material is merged into an article without a link such as {{main}} or {{more}}, especially if the original source article is left as a redirect to the new one
• when material is demerged out of an article to create a new article
• when material is translated from a foreign-language Wikipedia or otherwise copied between Wikimedia projects

Any thoughts, dear colleagues? Physchim62 (talk) 14:33, 5 November 2010 (UTC)

• Just wanted to note that I think this is very worth ironing out. Although I think the image point is unrelated as images are credited differently per WMF:Terms of Use and WP:Reuse, we do need to figure out how to clearly attribute text content where source is implied—particularly with cases such as TFA and (as was brought on my talk page), Portals (the example provided: Portal:Law of England and Wales/Selected article/1). Per the letter of the guideline, edit summary is required. Should an explicit exception be made for these? I'm not myself convinced that {{more}} or {{main}} are sufficient attribution, since such tags can be placed after content is split or after content is merged or even when two articles develop independently. (More on images: I'm not sure how expecting people to view the "edit history" and then click a link to an article and then click its edit history is better treatment than a single click through to an image description page? I'd see that point if they were credited on the article's face, as we currently do with external CC-By-SA sources, such as are credited via Template:CCBYSASource or Template:Citizendium.) --Moonriddengirl ^(talk) 18:29, 5 November 2010 (UTC)

• You seem to misunderstand the image point and the status of the Terms of Use. By clicking on "save" to make the above comment, you agreed to certain interpretations of the CC-BY-SA-3.0 license which are contained in the Terms of Use. That's fine, both of us are old enough to know what contributing to Wikipedia entails. But the person who uploads a CC-BY-SA image to Flickr hasn't chosen to be part of our happy community, and may well know nothing of the Foundation's Terms of Use. We will use the image regardless and enforce our interpretation that a click-through attribution is satisfactory. And yet there are people who think that a click-through attribution is not enough when the user has actively signed up to a set of Terms of Use which state that a simple hyperlink "to the article or articles […] contributed to" is sufficient attribution. I find it unacceptable to protect the Wikimedian more than the third party. Physchim62 (talk) 23:22, 5 November 2010 (UTC)

• That would seem to be an issue with the way images are handled, not the way text is attributed. It doesn't seem to me to follow that if we are enforcing our interpretation on Flickr uploaders that we need make it even harder to attribute text contributions out of parity. --Moonriddengirl ^(talk) 23:29, 5 November 2010 (UTC)

• I think we're talking at cross-purposes here. We have a method for attributing the autors of images (either Wikimedians or third parties) through the IDF. I'm not sure that that completely complies with the letter of the licences, but it's a bit late now, Wikipedia has effectively imposed an "industry standard" on their treatment and only copyright wonks like myself bother to worry about it. What is certain is that I am not required to type the attribution for an image into the edit summary when I add an image to an article, because the attribution is seen as obvious (through the link back to the IDF). Are we clear up to that point? Physchim62 (talk) 23:47, 5 November 2010 (UTC)

• Yes, images have their own attribution avenue. Attribution for text contributions is maintained in the article's history; for images, on the image page. For good or for ill, this is how it's done, and, yes, we are clear that this is how images are attributed. --Moonriddengirl ^(talk) 23:51, 5 November 2010 (UTC)

• Ah! OK! maybe I see (maybe only one of) the misunderstanding(s). If you talk about "attribution" in WP:PLAGIARISM terms, then attribution certainly isn't contained in the edit history. The attribution is contained within the edit itself. Am I allowed to include a direct quote in an article if I include the correct citation? Of course I am, I am not required to repeat the citation in the edit history because the attribution is obvious. So am I allowed to dump the entire lead of an article onto the Main Page for 24 hours? Of course I am, because the attribution is obvious. Am I allowed to make a derivative work of another WP article under a {{main}} hatnote? Of course I am, because the attribution is obvious. Physchim62 (talk) 00:10, 6 November 2010 (UTC)

• Right, attribution in plagiarism terms has nothing to do with edit history; that's licensing attribution. If we follow your argument, however, then just as direct quotes require correct citation, so would copying from other Wikipedia articles--which in accordance with WP:PLAGIARISM means either putting it in quotation or attributing, as with {{CCBYSASource}}. You have just provided a link in your last note to WP:PLAGIARISM and to {{main}}: am I to look to them for attribution? Of course not. You've linked them, but you didn't copy from them. Neither does every {{main}} imply content has been copied. As I note above, it is sometimes split and sometimes developed independently. {{main}} can be placed well after content in an article has been developed; it carries no inherent presumption of copying, and even if it did it carries no indication whatsoever of which direction copying goes. Too, every article on Wikipedia is co-licensed under GFDL unless specifically excluded, and GFDL requires that we must "Preserve the section Entitled 'History', Preserve its Title, and add to it an item stating at least the title, year, new authors, and publisher of the Modified Version as given on the Title Page." Although our contributors do release the requirement of listing at least five authors by agreeing to a link, I think it's quite pushing it to say, "Well, we assumed you'd be perfectly comfortable with this link, which is not visible in the 'history' section at all and which may or may not be understood to suggest attribution...." --Moonriddengirl ^(talk) 00:25, 6 November 2010 (UTC)

• I was beginning to feel like Salviati in Dialogue Concerning the Two Chief World Systems, but you kindly throw me a lifeline at the end there! You say "I think it's quite pushing it to say, 'Well, we assumed you'd be perfectly comfortable with this link, which is not visible in the 'history' section at all and which may or may not be understood to suggest attribution....'" That is exactly what we do with third party authors of images, and that is why I think it is unacceptable to require anything more for editors who have expressly consented to being attributed by a hyperlink back to the original article. Physchim62 (talk) 00:43, 6 November 2010 (UTC)

• And back 'round we go again: that would seem to be an issue with the way images are handled, not the way text is attributed. :) I think we're going to have to wait for others to weigh in here, because I believe that you and I are at opposite ends of a divide. I don't see implied attribution in {{main}} (where there may be no attribution at all and no indication of which direction attribution goes even if there is) as at all comparable to attribution at an image page, where authorship is clearly expressed. One can click the link to the {{main}}, but it does not give any information as to which article came first or who wrote what. One could as well say that as they have expressly consented to being attributed to a hyperlink that a "see also" is sufficient or a standard wikilink in an article...these are also hyperlinks to the article and also make no implication of attribution. --Moonriddengirl ^(talk) 00:48, 6 November 2010 (UTC)

Right, I'm no copyright expert but happened to be discussing a copyright issue on Moonriddengirl's page and so saw the request for more input. I've got to say I agree with her point of view. The difference, in my opinion, between images and text, is that any reasonably experienced editor will know that an image has attribution information somewhere other than the article's history and so know they need to go and find it if they want to. In the case of text, unless there is clear attribution in the history or on the article, it is not clear whether all the attribution is in the history or some is elsewhere - the use of {{main}} or similar does not help in this regard as the section may or not need attributing to the other page. Hence in these cases I think we have a problem as it's not clear the link is also attribution. The situation on the main page and the like is less clear as, again, it will be clear, to any experienced editor, where to look.

My other concern with this is my use of "experienced editor". I've just realised that no where is it made obvious how to get to the attribution information for anyone new to the site just looking at one page. Personally I'd expect this information to be linked to in the standard footer (maybe on WP:CC-BY-SA) - sorry if I've just opened another can of worms. Dpmuk (talk) 01:36, 6 November 2010 (UTC)

I don't think the "reasonably experienced editor" is the right benchmark. Any copyright judge we're going to come up against is unlikely to be a "reasonably experienced editor" of Wikipedia (although I know one or two who are, and I'll keep their identities private for obvious reasons). Our policy has to be understandable to the "(wo)man in the street": I still don't have an answer as to why we should favour our own contributers over third parties in terms of copyright attribution. Physchim62 (talk) 02:22, 6 November 2010 (UTC)

Your comment about "reasonably experienced editor" is exactly the reason for my second paragraph - we need something either in, or linked to, the footer of each page that explains how we attribute in practice - not just the WMF terms of use that don't really help anyone. As to your main point I don't think there's a difference between how we treat our own contributors and how we treat third parties - the difference is in how we treat images and text. Images are attributed in the same way whether the original source is here or elsewhere and the same with text. The issue as far as I'm concerned is that whereas an image in an article implies attribution by link, and so the attribution chain is clear, not every use of {{main}} implies attribution by link (as the two articles may have been written entirely independently), and so the attribution chain is very far from clear. Dpmuk (talk) 02:34, 6 November 2010 (UTC)

I have been involved with edit summary attribution for a while (Help talk:Merging/Archive 3#Merge edit summaries, December 2008). I err on the side of too much rather than too little, adding in revision ids and other information.

• The "(more...)" links on TFA and portals indicate that the full article is available through the link. It's pretty obvious to me that attribution would be available at the full article. I'm undecided whether those pages must be made to conform with this guideline, but their standard formatting would make a bot or possibly an AWB run pretty easy.
• As others have commented, {{main}} does not imply copying or derivation. The pages may have developed independently, or – in the case of a split – the dependency could be reversed.
• The disparity between credit given for text and images is a long-running issue (e.g., WP:Requests for comment/Picture of the day photo credits). I'm not intimate with the dispute, but I disagree that the solution is weaking our attribution for text.
• I don't know whether a user unfamiliar with Wikipedia and seeking attribution will be able to find the relevant place. However, once a user realizes that the authors of text are listed in the page history, it is reasonable to expect him/her to look there consistently, not in the page text or on the talk page ({{Copied}}, {{translated page}}, etc.). This is the reason for placing the comment and link in the edit summary. Regarding {{Copied}} in particular, I think there are instances where the templates were placed without the required dummy edit (maybe a few by me), but a bot could repair them.

Flatscan (talk) 05:30, 6 November 2010 (UTC)

Signed comments in discussions

Do discussions containing comments signed with ~~~~ need additional attribution when copying or archiving? I've been assuming that they do not. See WP:Village pump (policy)#talk page attribution question. Flatscan (talk) 05:09, 12 February 2011 (UTC)

Archived to WP:Village pump (policy)/Archive 83#talk page attribution question. Flatscan (talk) 04:59, 23 April 2011 (UTC)

Attributing multiple usernames of same person

If a person contributes to a page with multiple usernames, then copies the content, is the username on the copying edit sufficient, or do all usernames used need to be attributed individually? Assume that the single person is publicly disclosed or known, e.g., WP:Sock puppetry#Legitimate uses. Flatscan (talk) 04:44, 19 June 2011 (UTC)

(off the top of my head response) The multiple usernames should be prominently linked at the top of the main userpage (is this explicitly required). I assume (without confidence) that such a disclosure is permanent (perhaps we should have a central location as a permanent record of linked usernames). Then, the different usernames are just aliases for the same person, and it is the person who must be attributed, so there is no issue.

I'm guessing that this most likely applies to a person who uses a main account on a safe computer, and a second account for unsafe computers? Ideally, the second account name makes it obvious that it is a secondary account for the first. --SmokeyJoe (talk) 06:26, 19 June 2011 (UTC)

The specific case is a person who originally edited under his real name, added a username for his website, and later phased out the first account for privacy reasons. WP:Changing username should have been used to rename the account, keeping all edits under one username. The connection is verified easily given the real name account (its user page redirects), but there is no link in the opposite direction. Flatscan (talk) 04:09, 20 June 2011 (UTC)

A user who began with a real name username and then moved to an anonymous pseudonym? I suspect that this is very common. The link in the other direction could be found from “what links here” filtered to userpages.

If the person doesn’t tell us, and doesn’t ask this tricky attribution problem, then we will blithely continue as if the two usernames are independent. This should not be a problem, unless the person continues to mix the contributions, such as by using one alt. account to alter edits by the other alt. account. If he does this, then he is being very casual in maintaining his newfound efforts of anonymity.

If the person does tell us, and wants to anonymise the contributions from the real-name username, then he should go now to WP:CHU. Keeping things technically simple, he could have it renamed to an obvious variant of the new pseudonym. Then, the two usernames can be explicitly linked without forgoing privacy. --SmokeyJoe (talk) 05:50, 20 June 2011 (UTC)

Renaming the real name account to a pseudonym variant is a good idea. I had some concerns about recording the real name permanently in an edit summary, e.g., content originally contributed by [[User:Real Name]] and [[User:Example]], but it can be revdel'd if there is a serious privacy issue. Flatscan (talk) 04:15, 21 June 2011 (UTC)

Originality of lists of Wikipedia articles

At WP:Articles for deletion/List of cat breeds originating in the United States, I commented that the list's contents (a bulleted list of article links) may be factual and non-creative, falling under WP:Copying within Wikipedia#Where attribution is not needed, "Simple, non-creative lists of information". Is my interpretation generally correct? Does this have any application to WP:Categorization? For this specific list, Polydactyl cat may present a wrinkle, as American Polydactyl and Maine Coon Polydactyl are not universally recognized as breeds. Flatscan (talk) 04:38, 2 August 2011 (UTC)

I believe you are correct. While it may not be complete, the list strives to be--it isn't a cherry-picked subgroup of "cat breeds originating in the United States" but intended to be a comprehensive, alphabetical list of factual information. In terms of the cats which are not universally agreed upon, I'm not sure that would constitute a copyright problem but think that they are probably a small enough element to be de minimis, even if they did; and it would certainly be easy enough to omit them in merging over the list. --Moonriddengirl ^(talk) 14:36, 2 August 2011 (UTC)

Thanks for your reply. For any future readers, I'd like to highlight that this list is very basic. If it contained additional content like a blurb for each breed, I would tend toward following normal attribution for copied content, depending on what was merged. Flatscan (talk) 04:36, 3 August 2011 (UTC)

I agree. If List of cat breeds originating in the United States is nothing more than a list of cat breeds originating in the United States, which a simple alorithm could extract from a database, accepting standard definitions of all terms, then it is not creative content. --SmokeyJoe (talk) 08:07, 3 August 2011 (UTC)

Required edit summary

There was a recent change to the lead that reduced the edit summary containing a reference to the source article from a required step to one of a few alternatives. I'm pretty sure that the longstanding text is correct. For example, see "(This step is required in order to conform with Wikipedia's licensing requirements. Do not omit it nor omit the page name.)" at Help:Merging#Performing the merger.

I think it is required because people looking for attribution will be viewing the page history, where contributing users are normally found. They will not be looking at templates at the bottom of the article or on the talk page. Even if full attribution cannot fit in edit summaries (e.g. a long list of authors on the talk page), there must be a note with a pointer. A brief explanation would fit in WP:Copying within Wikipedia#Proper attribution. Flatscan (talk) 05:42, 5 November 2011 (UTC)

There is a previous discussion #How to attribute (November 2010) above. Flatscan (talk) 05:11, 6 November 2011 (UTC)

It is required and the changes are being made by a user whom I reverted because they didn't make the attribution. They are trying to change the guideline to bring it inline with their own personal opinion on what attribution is actually required instead of simply acknowledging that edit summary attribution is a requirement. Polyamorph (talk) 10:38, 5 November 2011 (UTC)

I don't believe it's actually required, at least by the terms of use. From the wikimedia foundation terms of use, which you agree to every time you hit save:

As an author, you agree to be attributed 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.)

You're notice that no where does it say it has to be in the edit summary, just that a hyperlink or url to list of authors is provided - this can just as easily be done on the article itself. That said including a note in the edit summary is best practice and I agree it should be in the various guidelines where it is relevant. Dpmuk (talk) 13:13, 5 November 2011 (UTC)

It's not required, it's one good way to satisfy the requirements, but inline hyperlinks where you refer to the original text is also allowed by the licensing.- Sheer Incompetence (talk) Now with added dubiosity! 13:37, 5 November 2011 (UTC)

It is required by the merge instructions This step is required in order to conform with Wikipedia's licensing requirements. Do not omit it nor omit the page name. When making a merge you must provide attribution in the edit summary or else you have failed to do the merge properly. It's pointless adding a hyperlink to an article if it has been made into a redirect. Polyamorph (talk) 13:52, 5 November 2011 (UTC)

The merge instructions are are a simplification - attribution is required and that is by far the simplest way to do it for a merge. It's the best way to deal with the licensing requirements but not the only way, as I point out above when quoting from our terms of use which clarify the more general attribution requirements in the license. It may be required by a policy or guideline but it's not required by our license. Dpmuk (talk) 13:59, 5 November 2011 (UTC)

If it's required by policy and/or guidlelines then it's required fullstop. Lets not split hairs here, the merge guidelines clearly state what is required. The only reason why these changes have been made are because I reverted the user's merge where they failed to make any attribution whatsoever, so in retaliation rather than acknowledging their mistake they decide it's best they just come here and make changes to the wording to suit their point of view. Polyamorph (talk) 15:14, 5 November 2011 (UTC)

I think Dpmuk has a point that the ToU do not specify among this guideline's various methods of attribution: edit summaries, templates, and lists of authors pasted to the talk page. The bold text on Help:Merging should be clarified to be more precise – that attribution is required by the license, and that the edit summary is required by this guideline. Flatscan (talk) 05:11, 6 November 2011 (UTC)

In addition to what is required, editors should consider what is desirable. All sorts of copying can be done while staying (just) on the correct side of legal requirements, but other considerations also apply at Wikipedia, namely ethics and convenience for other editors. Johnuniq (talk) 01:26, 6 November 2011 (UTC)

Any material contributed before this wiki adopted the dual GFDL/CC-BY licensing model was released under GFDL only. GFDL is pretty clear (OK, you might need a hacksaw to cut the words open, but it is clear) that the principal and/or complete author list must be shown in the History section of the Document. That means that the original authors must be credited in the article history. Editors making copy-paste moves should be complying with this requirement after they've been notified of it. The post-fix is to append the history to note prior copying and point to a talk page thread where it is detailed. Articles and portions of articles created after the dual-license cutover are licensed under CC-BY, which is not so definitive in language on exactly how attribution should be made. WMF worked with the Creative Commons folks on the dual-licensing thing, so perhaps more clarity could be found with an office query. Anything from before we switched over 2 years ago though, must comply with the GFDL provisions, so definitely needs attribution in the article history itsekf - at the least. Franamax (talk) 07:06, 6 November 2011 (UTC)

Thanks for the reminder, I had forgotten about the dual licensing changeover. I think that current contributions are co-licensed under CC-BY-SA 3.0 and GFDL. The exception is text imported under CC-BY-SA only, as described in the lead of WP:Copyrights. Flatscan (talk) 05:43, 7 November 2011 (UTC)

I agree with Franamax about attribution requirements under GFDL. Currently, our contributors' work is co-licensed. Reuse elsewhere has the option of complying with one, but reuse on Wikipedia must comply with both. The exception for single-license is only permitted at WMF:Terms of use when importing material from outside. (Even if it's our own material, we can't import it under single-license unless it has multiple authors.) To comply with GFDL, we have to maintain an accurate attribution in history. (This does raise a complex question for me, though: we currently have no guidance and no mechanism for what to do when copying content that is single-licensed from one article into another. Hmmm.) --Moonriddengirl ^(talk) 11:55, 7 November 2011 (UTC)

Using a uniform procedure in most cases would be simplest. IAR allows us to ignore this guideline (but not WP:Copyrights) if there is a good reason to provide only CC-BY-SA attribution for single-licensed content. A bot would be able to create a graph of {{CCBYSASource}} and {{Copied}}. Maybe {{Copied}} should have a CCBYSAonly parameter. Flatscan (talk) 05:52, 8 November 2011 (UTC)

This is complete madness. Who is going to sue and for what and how? We're usually talking about people copying material from one part of a free resource to another part of the same free resource, without making any claims that they wrote it, and very typically including links back to the original material.- Sheer Incompetence (talk) Now with added dubiosity! 04:07, 10 November 2011 (UTC)

The problem is that the people writing the license completely messed up, they should not have specified the mechanism for determining who wrote it, only that it be made reasonably accessible.- Sheer Incompetence (talk) Now with added dubiosity! 04:07, 10 November 2011 (UTC)

It's complete madness to comply with the legal contract we enter into with our contributors? How come? :) Wikipedia had the option to be public domain, in which case none of this would be required, but chose instead licensing schemes that require attribution. This is the legal reality we live with. --Moonriddengirl ^(talk) 11:15, 11 November 2011 (UTC)

Regardless of legal obligations it's also simply polite to acknowledge those editors whose work you have copied! Polyamorph (talk) 14:52, 13 November 2011 (UTC)

Copy of Google translate = copyright violation?

Editors often take articles from foreign language WPs, run them through Google translate, and dump the result into the en WP. That is a violation of Google's copyright on the translation, isn't it? Shouldn't that be mentioned under "Translating from other language Wikimedia Projects" and a speedy deletion tag added for it? El reggae (talk) 22:22, 4 July 2012 (UTC)

Hmm, does Google assert a copyright interest in their translations? My first take on this is that it would be a major stretch to say that a machine translation has any transformative value (as opposed to the actual software that runs the translation engine, which is very likely a creative expression of novel ideas, but we're not copying the software, just the arbitrary output) - so the copyright stays with the authors on the other-language wiki, and absolutely must be attributed to them. Pasting in a Google Translate session unaltered is pretty (IMO very) sloppy editing practice, but I don't see it as a copyvio of Google Inc.'s creative expressions. Others here may differ, so I'll be interested to read more opinions. Franamax (talk) 23:44, 4 July 2012 (UTC)

I would also say that the editor pasting in a direct machine translation forfeits any claim they may have to "own" that text, since their own creative input is zero. So it's probably worth noting in a null-edit summary or on the talk page (or both), just to be clear on the chain of attribution. Again, other views may differ from mine. Franamax (talk) 23:51, 4 July 2012 (UTC)

Sorry for duplicate discussion, discussion is now at Wikipedia talk:Pages needing translation into English#Copy of Google translate = copyright violation?. El reggae (talk) 11:18, 5 July 2012 (UTC)

Creativity of page names

Do page names contain enough creativity to require attribution? Moving a page makes a minor edit to record the move in the page history.

• Contentious WP:Requested moves are usually implemented by the uninvolved closer, not the original proposer. (A preceding move war would have created proper attribution.)
• A user does a cut-and-paste move. Their name will be in the page history, but the move will be attributed to the admin who fixes it.
• A user creates an incoming redirect. A different user finds the redirect, maybe from the search box or Special:WhatLinksHere, and requests that the pages be swapped using {{db-move}}. As with the histmerge, attribution will go to the G6-fulfilling admin.

Flatscan (talk) 04:08, 27 September 2012 (UTC)

Names and titles are generally not protected by copyright. Also, article titles are in general not created by our editors since we prefer titles that are already in use in reliable sources. Jafeluv (talk) 06:15, 28 September 2012 (UTC)

Rewrote intro

This page began: "Wikipedia's licensing requires that attribution be given to the original author." I wouldn't state it like that. Not just the original author must be given, but all authors. Furthermore, the intro does not say anything about importing content from other Wikimedia projects. I think this is important enough to note in the lead, because other Wikipedias sometimes contain interesting stuff which we at en.wp do not have. I've rewrote the entire thing, see here. Cheers, theFace 15:31, 16 December 2012 (UTC)

I agree with you. :) I've made a few changes to the text. I think we must mention the edit summary requirement in the lead but felt that the specifics on the templates could be read in the section. --Moonriddengirl ^(talk) 15:55, 16 December 2012 (UTC)

Thanks, looks good! However, the one thing I don't get is the part which reads that one may provide a list of all contributors of a page in the edit summary. If a source page has 1 or a few contributors, that may be feasible. But if it has had 20 contributors, it would be rather tedious. In any case, I don't see the point of it, as all contributors are already listed in the page history. Cheers, theFace 18:32, 16 December 2012 (UTC)

The one major advantage to listing contributors in the edit summary is that if the original article is deleted or moved, tracking it back by the link may become complicated. (A smaller advantage, I suppose, is that those who choose to reuse the content but attribute through a list will find it easier if the names are used than if they have to follow the link back to look at the list there.) But, yes, it's really going to be limited application, as there aren't too many articles with only a couple of contributors. My mention of the list in the lead is only so that it accurately reflects the guidance given in the body. --Moonriddengirl ^(talk) 11:49, 17 December 2012 (UTC)

Need to make this more known

I am no expert on copyright, but know that if I take something from the another website or source and add it here verbatim I am most likely creating a copyright violation. However, I was not aware for a long time that if I split, merged, forked or otherwise an article on Wikipedia that I was doing the same thing (and even now I am still probably not fully aware of what is required to avoid this or which situations it applies to). These practices are needed to maintain the encyclopaedia and all done in good faith. I am not the only one as I have seen many experienced editors copy and paste information around the encyclopaedia, almost all resulting in an overall benifit. I would guess that this is one of the more abused and least realised copyright processes. I think it needs to be either made more visible or the license changed to freely allow copying withing Wikipedia. One thing that could be done is to use a bot similar to Madmanbot to advise editors when information is copied within Wikipedia, or if possible one that automatically adds {{copied}} template when this occurs. AIRcorn (talk) 00:39, 14 January 2013 (UTC)

Upgrade to Policy

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.

---

Copying within Wikipedia was made a guideline in 2009. As the guideline covers an important area of copyright compliance, I propose that it should be upgraded to a full policy. Monty₈₄₅ 18:38, 27 December 2012 (UTC)

Discussion

• Support as proposer Failure to abide by the instructions in the guideline will often result in a violation of the CC-BY-SA and GFDL licenses under which contributions are made. Without proper attribution, the content is not being used under a license, and is therefor a copyright violation. Committing copyright violations is not the sort of thing we should only generally discourage, but should be strictly prohibited, and that requires a policy, not a guideline. Monty₈₄₅ 18:38, 27 December 2012 (UTC)

Note, in the interest of full WP:CANVASS disclosure, I mentioned that I was starting this RFC in a related discussion here. Monty₈₄₅ 18:52, 27 December 2012 (UTC)

• Support Per Monty. Darkness Shines (talk) 18:46, 27 December 2012 (UTC)
• Support Indeed. Of course, a bit of more polishing will be needed to avoid ambiguities (if any), but it doesn't weaken its potential. — ΛΧΣ²¹ 20:46, 27 December 2012 (UTC)
• Support. I like the idea, generally, of making it increasingly clear that Wikipedia expects all editors to understand the spirit, as well as the letter, of why plagiarism is unethical, even for material that is CC-BY-SA 3.0. --Tryptofish (talk) 20:54, 27 December 2012 (UTC)
• Support, although this page is still very 'guideline-ish', containing things like: "At minimum, this means..." and "Content reusers should also consider...". Still, the legal foundation behind this document is quite important. I'm not a fan of copyright, but it may be a good idea to follow it where possible, so that would make this a policy. Cheers, theFace 14:16, 28 December 2012 (UTC)

PS: there is a Template:Policy and a Template:Legal policy. I think we should use the first one here. - theFace 14:21, 28 December 2012 (UTC)

Retracted my support. Changed my mind. This page covers a trivial issue and it's good as it is. - theFace 22:29, 19 January 2013 (UTC)

• Support Seems obvious to me. Dougweller (talk) 21:52, 2 January 2013 (UTC)
• The overarching policy is WP:Copyrights, and we need a guideline-level page in this area. As pointed out by Face, this page is written at a level of detail appropriate for a guideline. It describes best practice, and not everything it recommends is required. While persistent refusal to use appropriate edit summaries or templates might lead to sanctions, a user wouldn't be blocked for using {{Merged-to}} and {{Merged-from}} (as directed by WP:Merging#How to merge) instead of {{Copied}}. I have occasionally wished that people would take this page more seriously, but I disagree with promotion overall. Flatscan (talk) 05:05, 7 January 2013 (UTC)
  • Good points. Indeed, this page contains a lot of recommendations. Then again, every policy contains recommendations. The line between the two isn't firm. What it boils down to is that it is required by our licence itself to state where info came from. If you merge or split articles, a small note in an edit summary is the absolute minimum. This is mandatory, and this page should make that clear. Cheers, theFace 13:19, 8 January 2013 (UTC)
    • The prose begins with a piped link to WP:Copyrights, but I think that increasing its prominence would add weight. Custom header boxes based on {{supplement}} ("intended to supplement") or modifying the nutshell might be a piece of the puzzle. Flatscan (talk) 05:50, 10 January 2013 (UTC)
  • I glanced at the discussion that preceded this RfC (Talk:Human rights abuses in Kashmir#Lock). Repairing the attribution is preferred over removal unless the insertions are disruptive, so the edit-warring wouldn't be supported by this page, regardless of its classification. Flatscan (talk) 05:34, 9 January 2013 (UTC)
• I see the RfC tags and the mention at WP:Wikipedia Signpost/2012-12-31/Discussion report. I suggest listing at WP:Centralized discussion also. Flatscan (talk) 05:05, 7 January 2013 (UTC)
  • As for WP:Centralized discussion: feel free to list it there. - theFace 13:19, 8 January 2013 (UTC)
    • Done, but proponents of this proposal should have done this to start with. Failing to do so, then refusing to do so and making someone else do it, simply encourages the view that this is poorly thought out, and that broad community attention is being avoided. — SMcCandlish  Talk⇒ ɖ^⊝כ^⊙þ Contrib. 20:27, 9 January 2013 (UTC)
• Support Credit where credit is due is a fundamental part of sound editing practices. MLauba ^(Talk) 13:22, 8 January 2013 (UTC)
• Strongly oppose per Flatscan. A large number of points in this guideline are guideline material and do not rise the level of official policy considerations. The proper way to address this is to get consensus at WT:Copyrights to add something concise to WP:Copyrights that encapsulates the crucial, and only the crucial, points of this guideline, thereby importing them into policy, without sucking all of this page's miscellaneous details and micromanagement into the policy for no reason. — SMcCandlish  Talk⇒ ɖ^⊝כ^⊙þ Contrib. 20:20, 9 January 2013 (UTC)
• Oppose Copyvio and plagiarism are serious problems here . To some extent they are such difficult problems because we have a tendency to concentrate on the minutia rather than the essence. Copying within WP is just a matter of giving the attribution in the correct manner, and if someone neglects to do it, it is usually because they haven't figured out the proper format, and all that is really necessary to do is to add the necessary link, and point the person to the guideline so they will do it right in the future. We do have a difficult problem preserving continuity with moved articles, but it is not because of any desire to violate ATT,it's because of the various technical difficulties (I confess I am among the admins who have never really figured out how to fix copy-paste moves--after getting it right only about half the time, I've left it to the more careful and clever among us). McCandish has it right--I think we are getting at the same thing though we said it differently. What we really need is clearer directions and a less censorious attitude. There are major real copyright and plagiarism problems from the earlier years that affect entire articles, and these are what we need work on finding and removing. The Copyright Problems board is weeks behind, and that's what needs attention. I'll do as I say, and I'm going there right now to fix at least one of them. DGG ( talk ) 02:50, 10 January 2013 (UTC)
  • I agree that WP:Copyright problems is more critical, but preventing the creation of new WP:CWW issues is also important. They are easy to fix minimally, but identifying non-compliant copies and fixing them to best practices are somewhat (less than WP:CP) difficult and time-consuming. Users need to be aware of and to respect this page. By the way, this page is the "clearer directions". Flatscan (talk) 05:50, 10 January 2013 (UTC)
• Oppose per Flatscan, WP:Copyrights is the policy and this page is for "guidance" on best practice to meet the policy. Congratulations especially to Flatscan and Moonriddengirl for their excellent work. --SmokeyJoe (talk) 04:40, 10 January 2013 (UTC)

I should note that WP:Copyrights does not contain any info about copying within Wikipedia. I've made a proposal for some new text at the talk page, see here. Cheers, theFace 19:26, 10 January 2013 (UTC)

• Comment I was asked for my thoughts here and just wanted to note that I consider this really a process page. The attribution requirement is part of the wmf:Terms of Use. You can't get much more "policy" than that. :) For that reason, I'm not really worried about whether we label this page policy. I'm comfortable with it's guideline status. And I would be appalled to find us treating contributors who accidentally fall afoul of it harshly. It's a very common error and easily rectifiable. The only time I ever have problems is when people persistently ignore it (after good faith explanations), and that's covered by plenty of other policies, including WP:C ("Never use materials that infringe the copyrights of others.") --Moonriddengirl ^(talk) 14:46, 11 January 2013 (UTC)
• Oppose: copyright policies do not need adding to, and, in general, we should not upgrade guidelines because they are widely agreed with or we want to make them better enforced; it just does not work that way. This is essentially a supplementary to existing copyright policy, albeit a useful one. Add or improve WP:C in this area if necessary with a minimum of words and fuss. Grandiose (me, talk, contribs) 16:25, 11 January 2013 (UTC)
• Oppose as redundant, as Moonriddengirl has stated. Also, as the writer of documents licensed under CC 3.0 BY-SA, I retain the copyright and can copy my own text whenever I want. Others must attribute such copying to the original article. This distinction is ignored in this guideline. (It is strange that WP lists itself as the author of works, rather than noting the first or primary writers, when it generates APA or MLA or Chicago bibliographic entries....) Kiefer.Wolfowitz 11:50, 13 January 2013 (UTC)15:37, 18 January 2013 (UTC)
  • When the author and re-user are the same is covered in the second sentence of WP:Copying within Wikipedia#Where attribution is not needed. There is some confusion on this point: recently, a user drafting new articles in her user space and copying to article space was contacted (WP:Administrators' noticeboard/Archive243#Help me explain, December 2012). Flatscan (talk) 05:06, 14 January 2013 (UTC)
• Oppose - As DGG said, "Copying within WP is just a matter of giving the attribution in the correct manner, and if someone neglects to do it, it is usually because they haven't figured out the proper format, and all that is really necessary to do is to add the necessary link, and point the person to the guideline so they will do it right in the future." --Sue Rangell ✍ ✉ 20:11, 13 January 2013 (UTC)
• Support. The guideline is well-written and addresses an important copyright issue; it should become policy. It can still be edited as needed, with consensus. I don't get the argument that only WP:Copyright problems matters, so we shouldn't pay attention to anything else. Both matter. Superm401 - Talk 19:05, 14 January 2013 (UTC)
• Oppose per Flatscan and DGG. Policycreep. KillerChihuahua 17:57, 15 January 2013 (UTC)
• Oppose policy creep. Smallbones_(smalltalk) 18:40, 16 January 2013 (UTC)
• Oppose - I think this is sufficient as a guideline, per DGG, and on the whole my reading of it is more of a guidance page on how to deal with a common but easily fixable problem, rather than as a rigid policy. I agree however there should be some reference to it at Wikipedia:Copyrights, which serves as the main policy page. CT Cooper · talk 13:37, 19 January 2013 (UTC)

Regarding your last sentence, see: Wikipedia talk:Copyrights#Wikipedia:Copying within Wikipedia. Cheers, theFace 22:29, 19 January 2013 (UTC)

• Oppose - Not everything needs to be a policy, and this is just fine as a guideline. In practice the two aren't that different anyway, but policy should be for main topics only, and this is not one of them. DreamGuy (talk) 00:44, 21 January 2013 (UTC)
• Comment With 6,914,330 articles policy creep is not a valid argument. -- Alan Liefting (talk - contribs) 09:22, 21 January 2013 (UTC)
• Oppose, per both Flatscan and Moonriddengirl. Copying within Wikipedia is an important issue. However, the key points are well covered in the terms of use which are linked from every page and the copyright policy that we already have. The Copying within Wikipedia page is more about the process than the principles. It's not necessary to elevate it to policy.—S Marshall T/C 20:01, 29 January 2013 (UTC)
• Support per nom. It's already a guideline; this is barely policy creep, if at all. See also WP:NOTCREEP: "'WP:CREEP' is not a substitute for actual arguments." If attribution is unimportant enough that this shouldn't be policy, I'd be very interested to hear why. --BDD (talk) 03:44, 16 February 2013 (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.

Other reasons for attributing text

I have added a section called "Other reasons for attributing text". This is because aside from the copyright issues there are also practical reasons why attribution is desirable. I have given two examples in which I have been mixed up:

1. Talk:Whiggamore Raid#Copyright problem removed
2. At the start of this month (August 2013) I can across an edit which I realised was a copy and a breach of WP:SAYWHEREYOUREADIT (see here) This has led me to realise that a handful or editors about five years ago regularly copied text and short citations between Roman Catholic articles without including the long citations in the References sections. When they made these copies the editors did not mention that they were doing so let alone from where the copied text came (which would now all be a breach of this guideline). Without the minimum copy information as laid out in this guideline, trying to find the entries for the references sections to go with the short citations is very time consuming (see for example here and here).

So I think it helpful to give guidance and encouragement to attribute copies within Wikipedia for reasons other than just to meet copyright requirements. -- PBS (talk) 16:14, 26 August 2013 (UTC)

I agree with the advice and attribute copies whenever possible, but I think that its placement interrupts the explanation of policy-level concepts. I suggest that it be moved to the end of the page. Flatscan (talk) 04:19, 10 September 2013 (UTC)

how to resolve old issues?

Hi. I came across a problem and reported it to the offending user at User talk:23 editor#mass_edits_to_World_War_II_persecution_of_Serbs_in_March_2012. They have failed to respond to my query to correct this problem in almost a month now. What can be done to correct this problem? --Joy [shallot] (talk) 18:57, 4 January 2014 (UTC)

@Joy: I have identified the source and attributed and spoken to the contributor in the thread you opened. Generally, I look for text matches on Wikipedia and then check the dates to see which article had it first. It can be difficult when somebody "moves" content from one article to another, rather than copy-pasting and leaving it in both locations, but in that case what I do is check their contributions at the time of the paste. Usually, you will find that they removed a chunk of text from an article at around the same time they pasted it. Thanks for finding the problem and for following up on it. :) --Moonriddengirl ^(talk) 20:34, 4 January 2014 (UTC)

Thanks for fixing that. In the meantime I found several more cases. Now I'm thinking, maybe we need a more stern wording in {{uw-copying}}? This is not optional... --Joy [shallot] (talk) 09:03, 13 January 2014 (UTC)

Users who have put their work in public domain

Some users, such as myself, agree to place their changes in the public domain. How does this affect attribution requirements? --Zzo38 (talk) 07:05, 10 November 2013 (UTC)

I am not familiar with handling public domain content. The lead of WP:Public domain requires "proper attribution", but the "relevant policies" do not include WP:Copying within Wikipedia, which covers CC-BY-SA/GFDL. Following this guideline would make it easier to identify the original and verify its release into the public domain. The reuser could also copy any attribution templates along with the text. (I looked in Category:Public domain copyright templates for an example, but they appear to be image templates.) Flatscan (talk) 05:34, 12 November 2013 (UTC)

Yeah if you release your content in to the public domain there's no legal requirement for attribution (so there's no copyvio if it doesn't happen), but it's still a good idea for practical reasons. In any case, unless you are the only contributor or someone makes sure all contributors have release their content in to the public domain or otherwise without an attribution requirement, we would need to attribute for the other contributors just in case. Nil Einne (talk) 23:31, 1 March 2014 (UTC)

Proposed change to Copying from other Wikimedia Projects section

I am proposing the wording of the Copying from other Wikimedia Projects section be changed to this:

If copying from another Wikimedia project that is licensed under CC-BY-SA (such as Wikimedia Commons, Wiktionary, and Wikispecies), you must satisfy the attribution requirement in the edit summary of your edit by either providing a complete list of the original content, or a direct link (such as an InterWikimedia link) to the original material.

Even if you satisfy the attribution requirement by providing a complete list of authors in the edit summary you are encouraged to include the direct link in case it becomes necessary to access that history in the future. The template {{Interwiki copy}} is available to provide additional notice on the article's talk page.

Although most Wikimedia projects are licensed under CC-BY-SA and require attribution consistent with the Foundation's Terms of Use, some projects are handled differently. For example, content from Wikinews is licensed under CC-BY and may be reused with attribution only to "Wikinews." (See Wikinews:Copyright.) It is the responsibility of the editor importing content to determine the license that applies and ensure that attribution is satisfied.

The changes are intended to make it clear that the attribution must be supplied in the edit summary, which makes the section consistent with the lede and the licenses. I have also removed the information about moving an article. According to the banner at Help:Transwiki transwiki is a "former standard procedure". If there is a current standard procedure it probably belongs in its own section since it presumably provides attribution in an entirely different way (by moving the entire history page over). ParacusForward (talk) 01:25, 4 March 2014 (UTC)

Is the proposed new wording contradicting itself? It says that the person adding the content must satisfy the attribution requirement, but it also says that the person adding the content only needs to provide attribution through the edit summary. Wikipedia distributes PDF versions of articles through the "Download as PDF" option in the "Print/export" section in the menu, and the PDF files do not contain edit summaries nor do they otherwise attribute users whose names only appear in edit summaries. Therefore, attribution through edit summaries would appear to be insufficient, as this doesn't give attribution in the PDF copies of pages. --Stefan2 (talk) 23:02, 6 March 2014 (UTC)

I think we've been sticking our heads in the sand a bit with that one, although it's something I've thought about over the last couple of weeks. I think the only really sensible answer I can come up with is to have a list of editors associated with a page, but aren't in the edit history, so that they could be added to PDFs etc. I suppose we could template the article itself but that could get quite messy. Of course the problem with such an idea is that it would involve a software change and probably some discussion about how it would work (who could edit the list etc). Of course the cynical me thinks we will never got the devs to agree to such a change as their only priority seems to be stuff that will "attract" new users and I doubt they'd put much priority on something as unimportant as meeting our legal requirements. Dpmuk (talk) 00:02, 7 March 2014 (UTC)

One way to associate editors to a page when content has been copied from elsewhere on English Wikipedia itself is to use Special:Import with XML files (requires user group "steward" or "importer"). However, this trick causes problems if multiple pages with concurrent history are merged (action=history becomes very messy), if content is copied from other Wikimedia projects (due to WP:SUL conflicts attributing edits to other people with the same user name, for example this edit was made by me on English Wikipedia but now attributed to another user on German Wikipedia) or if content is copied from external sources (the author must be attributed to a Wikipedia account which the author might not have). Therefore, this trick often only gives a satisfactory result when one Wikipedia article is split in two.

The PDF option is also broken when images are copied from another project to Commons, and here I don't see any solution. If I move an image from Wikipedia to Commons, then the PDF file attributes the image to me, although I don't have anything to do with the authorship of the image.

Big problem: If you use CC-BY-SA 3.0 content without proper attribution, then the CC-BY-SA 3.0 licence is automatically terminated, and you are no longer able to use the content. If reusers use PDF files with incorrect attribution, these reusers suddenly become unable to use certain Wikipedia articles, presumably without knowing it. Therefore, I think that this needs an urgent solution. --Stefan2 (talk) 00:29, 7 March 2014 (UTC)

It is my belief that a link to the to original article in the edit summary is all that is required to fulfill the license requirements. To support this position I will quote the lede of the article: "At minimum, this means a link to the source page in an edit summary at the destination page." I'm not sure how PDF exports figure into this. I had never tried to do one until just now. If providing attribution by edit summary is not sufficient to satisfy the attribution requirement then both the proposed wording and the current wording of the section are insufficient to direct behavior that is compliant with the license. ParacusForward (talk) 05:59, 7 March 2014 (UTC)

The PDF export function has some interesting attribution features. Try generating a PDF file for Wikipedia talk:Copyrights/Archive 15, for example. The PDF file says that all text was written by User:Manifestation, but it is obvious that this isn't the case. Therefore, I do not think that providing attribution in edit summaries is sufficient. It might be sufficient for the HTML version of Wikipedia, but it definitely doesn't seem to be sufficient for the PDF version of Wikipedia. Extensive use of history merging and Special:Import could fix this in some cases but not in all cases, and it would also have some unwanted side-effects (see e.g. WP:HM#Parallel versions). --Stefan2 (talk) 15:32, 8 March 2014 (UTC)

I generated the PDF as suggested. It contains https://en.wikipedia.org/w/index.php?oldid=596702336 – full URL for printing and hyper-linked for convenience – between the page name and the list of contributors. This method of attribution is described by Terms of Use 7.b.i.. Flatscan (talk) 05:07, 11 March 2014 (UTC) I see that ParacusForward and Stefan2 already covered this. Flatscan (talk) 05:15, 12 March 2014 (UTC)

Stefan, it is my intention to only clarify the existing guidelines, not to make any changes to the guideline. Do you agree that a link in an edit summary as attribution would be acceptable under both the existing and proposed guidelines? It seems backwards to try to retrofit guidelines to fit the style of attribution found in documents created by the PDF download functionality.

After some thought, I don't think the existence of the PDF download functionality necessitates any change to the guidelines or policy. I have found it difficult to interpret legal documents, but the CC-BY-SA 3.0 license requires satisfaction of the attribution requirements "be implemented in any reasonable manner." We do this by providing a link in an edit summary. So, when someone wants to see the credits on the Wikipedia website he or she clicks on History and is shown a list of pseudonyms along with, possibly, one or more links to other webpages. If he or she follows one of those links to another Wikipedia page, he or she can click on that page's history and see a list of pseudonyms, and possibly links to other webpages, etc...

The PDF created by PDF download contains a "Article Sources and Contributors" section with a list of pseudonyms along with an URL to a Wikipedia webpage where you can click on History to see a (possibly slightly larger) list of pseudonyms and, possibly, some links providing additional attribution.

I don't know that this is sufficient to satisfy the CC-BY-SA license, but I presume the style of attribution included in the document created by the PDF download functionality has been approved the WMF's lawyers who are more qualified to judge this than I am. The text in the PDF created by the PDF download functionality is certainly consistent with the Terms of Use which state that users have agreed to attribution "[t]hrough hyperlink (where possible) or URL to the article to which you contributed." ParacusForward (talk) 14:47, 9 March 2014 (UTC)

Attributing people by listing every author is permitted, but only as long as every author is listed. If some authors are missing, as is the case with the PDFs, then you have a problem. The wording in the PDF file suggests that the listed people are all authors, which is not the case. Attribution through links may be a problem if you have to click on links in multiple steps (first follow the link in the PDF file, then links from the linked article's history page). --Stefan2 (talk) 16:03, 10 March 2014 (UTC)

With this: "Attributing people by listing every author is permitted, but only as long as every author is listed. If some authors are missing, ... then you have a problem." You seem to be saying that any list of authors needs to be entirely complete, but I don't think this can be strictly true or we would not be able to provide attribution by link at all. Current policy allows the Wikipedia website to have a partial list of authors and additional attribution by link.

Regarding your statement that "[t]he wording in the PDF file suggests that the listed people are all [of the] authors," I will agree that there is certainly a difference between the website, which intermingles the links with the authors, and the PDF, which always contains one or more authors and exactly one URL. For the PDF, it is the case that loading the URL in the "Article Sources and Contributors" section may or may not lead you to more non-trivial authors. I don't know if this difference amounts to a legal problem. I've added this ping to @Mdennis (WMF): here to see of WMF has any input.

I also agree that "Attribution through links may be a problem if you have to click on links in multiple steps" (italics mine) because the license doesn't make it clear, and I am not aware of any relevant case law. However, since the Terms of Use allow re-users to provide attribution "[t]hrough ... URL to the page or pages that you are re-using (since each page has a history page that lists all authors and editors)" I am happy to assume this is not a problem. ParacusForward (talk) 02:28, 11 March 2014 (UTC)

Hi. Before I trot off with this to the legal team, let me be sure I have the list of questions you'd like answered -

• Does the practice of attributing copies by placing a link to the article in edit summary create attribution issues for PDFs, which generate a list of contributors only to the local document?
• While PDFs do include a link to the local document, would that alone be adequate attribution if the reader then must find the link to the other document in history?

Is that it? (Please ping me any replies - I fear I might miss them otherwise. :)) --Maggie Dennis (WMF) (talk) 10:23, 11 March 2014 (UTC)

Thanks, Mdennis. That sounds good to me, but I've added a ping to Stefan2 who also had some questions about images in PDF's (I may have been misunderstanding Stefan when I added "[of the]" to "listed people are all [of the] authors" in my last comment. He may have a question about crediting the wrong person/bot.) Also, a ping for Dpmuk who expressed skepticism with the current system above to see if he or she has anything to add. ParacusForward (talk) 13:59, 11 March 2014 (UTC)

Yes, Mdennis (WMF), I think that's what I was wondering about too. There is also one variant of the first question. When I generate a PDF file out of Wikipedia talk:Copyrights/Archive 15, then I see this:

“ Article Sources and Contributors
Wikipedia talk:Copyrights/Archive 15 Source: https://en.wikipedia.org/w/index.php?oldid=596702336 Contributors: Manifestation ”

This suggests that there are two pieces of information:

• The source: https://en.wikipedia.org/w/index.php?oldid=596702336
• The contributors: Manifestation

Is there a problem if there also are other contributors which you can only find by looking at the source? The use of the word "contributors" suggests to me that the list is complete, so I'm wondering if there is a problem if the list isn't fully complete.

The image problem appears if you try to generate a PDF file for the article Greta Hall, which contains File:Greta Hall.jpg. The PDF file lists the following:

“ Image Sources, Licenses and Contributors
File:Greta Hall.jpg Source: https://en.wikipedia.org/w/index.php?title=File:Greta_Hall.jpg License: unknown Contributors: Jkadavoor, Stefan4 ”

That is, it says that the image was created by Jkadavoor and me. However, if you look at the file information page, you will find that this was made by William Westall (1781-1850). I merely moved the file from Wikipedia to Commons, while the other user made a minor edit to the file information page. The attribution given in the PDF file is, as you can see, completely wrong. PD images do not require attribution, so there is no big issue here, but what happens if an image which requires attribution is used with incorrect attribution? --Stefan2 (talk) 00:07, 12 March 2014 (UTC)

While the image issue is certainly of concern, I believe Stefan2 may be misreading the output of the archive pdf. If you look at the history of that talk page archive, Manifestation is indeed the only human contributor doing manual archiving. If you wanted to have a proper attribution record, you'd need to make a pdf of the talk page instead of its archive (the place where the contributions were made), which would retrieve from the history the name of all the contributors. The way we archive by copying content instead of moving pages makes it impossible to get a true record of contributions to an archive.

Now granted, that makes pretty much every single talk archive page technically in violation of the attribution requirement. It is, in my opinion, an utterly irrelevant issue, however. The two questions related to article PDFs is a more interesting one. MLauba ^(Talk) 00:41, 12 March 2014 (UTC)

(edit conflict)Having thought about this some I think my concerns go wider than that, although I do share your concerns. I also have concerns that about our "Printable version" option which seems to be even worse of than the PDF version. I'll list my concerns and whether they apply to print, PDF, or both print and PDF. Oh and then I've thought of a few general questions that apply across the board.

1. Print. There is neither the text of the license or an URI to it which seems to be a clear breach of the CC-BY-SA license.
2. Both. If there is any image that is under a different version of the CC licenses then again there is no text or a URI to it.
3. Print. The Print version contains no information about the images. Once a user prints it then even the link to the image page will be lost. Although it could be argued that this a re-user problem I think describing it as a print version makes us somewhat at fault (at least morally).
4. PDF. (Only not print as we don't have a list of image contributors). Unless I'm much mistaken the image problem is worse than just moves to commons, it's also going to apply to any image that was uploaded by anyone other than the author as I assume the software gets the author list from the uploaded not any license tag added to the description - anyone with more image experience want to confirm this.
5. Both. Even if we have the URI of the article in the print/PDF version and that is deemed enough to include any comments in the edit history what happens if the article is later deleted - I assume already created books would be kept with the deleted article still in them. And that's without the issues for any re-users that may have kept a printed or PDF version (although again this could be argued to be an end-user problem).
6. General One which I hope the legal bods have looked at before. It appears to me that when we modify other people's work that their "credit will appear, if a credit for all contributing authors of the Adaptation or Collection appears, then as part of these credits and in a manner at least as prominent as the credits for the other contributing authors." It's certainly debatable whether even a link in the edit summary does that given how hidden that can be. This gets worse in the PDF/Print version where we have a list of authors that modified the page but those for copied in work (either from wikipedia and sister projects or else where) get at best a link or URI to the article.
7. General. Where do we stand with OTRS permission, especially when the person giving permission is not the person who added the material. I'm pretty sure the release we use means the copyright owner still has a right to attribution if they want but in most cases the most we end up is a template on the file/talk page. It's quite plausible that in those circumstances we won't end up with an author, or a link to somewhere the author is named, anywhere and certainly not in the list of any contributors.
8. General. If we include information from an external site, link to the external site as attribution and the site goes away then we no longer even have a dubious round about link to a list of contributors.

Gone of on a bit of a tangent there but this who thing has got me thinking about it. Some of the answers may be in the licenses themselves but I've done my best to check. Oh and ping Maggie. Dpmuk (talk) 01:44, 12 March 2014 (UTC)

MLauba - why is it utterly irrelevant? Without getting too beansy a long term editor could certainly cause a LOT of problems based on that if they wanted to. Dpmuk (talk) 01:51, 12 March 2014 (UTC)

And, although not that important, for future reference I happily identify as a him. Dpmuk (talk) 04:16, 12 March 2014 (UTC)

Discussion archives generally comply with WP:Copying within Wikipedia, although its validity may be in question considering the path of this discussion. Continuing the examination of WT:Copyrights/Archive 15, the bots link back to WT:Copyrights in their edit summaries (history), the breadcrumb and {{talkarchive}} box also link, and the majority of edits are comments signed by their authors. See also #Signed comments in discussions above and WP:Village pump (policy)/Archive 83#talk page attribution question from February 2011. Flatscan (talk) 05:15, 12 March 2014 (UTC)

Dpmuk, would you consider replacing your * bullets with # numbers for ease of reference? Regarding your points about deleted or removed sources, #Peculiar deletion policy (March 2010) has some discussion of the case where an external re-user uses a Wikipedia article that is later deleted. I think that your last point can be addressed proactively by saving a list of authors to the talk page at the time of use. My impression is that CC-BY-SA content with many contributors tends to be from wikis, with the majority of usable content on other Wikimedia sites. Flatscan (talk) 05:15, 12 March 2014 (UTC)

Numbered points done - I didn't realise I'd have that many when I started! I'd agree that the last one may have a solution like that - assuming a solution like that is acceptable. I think, per point five, that the argument could easily be made that a list on the talk page doesn't meet the attribution requirements of the license. That said it is at least better than what we normally do at the moment. Dpmuk (talk) 12:45, 12 March 2014 (UTC)

• Some comments about User:Dpmuk's numbered points above:

1 & 3: Strictly speaking, WMF only distributes an HTML page. That page contains clickable links both to the image pages and to the licence, so WMF might be covered. Of course, clickable links are useless in a printed copy of the article, so people who print out the article on paper may end up in trouble in a lot of situations, making the printable version useless.

4: Yes, I only picked that image as an example. Many other files are also affected. For example, Main Page contains File:Ruby Loftus screwing a Breech-ring (1943) (Art. IWM LD 2850)-detail.jpg which was made by the British government, but the PDF file credits it to Materialscientist and Orlady. Although I can't exclude the possibility that those two Commons users worked for the British government back in 1943, I do not find it likely that they are the artists of the image.

8: Under Swedish copyright law, you would normally be allowed to continue distributing a copy of a work, if the copyright holder previously allowed you to distribute the work. Therefore, I would guess that Swedish law would allow you to continue distributing already created copies even after the linked websites go down. However, you would probably not be able to print a new edition of a book containing the article, and you may have to immediately terminate any online distribution of the article. Reusers in other countries may be affected in other ways. I am also not sure if it is a good idea to do things in gray areas. If it turns out that some of the things aren't permitted, then it will suddenly affect a large number of users and reusers who have had their licence terminated for non-compliance with CC-BY-SA and who will no longer be able to use it. For this reason, it might also be a good idea to upgrade to CC-BY-SA 4.0.

Someone thought that the talk page archive was a bad example. Actually, the same problem occurs with every page where content is copied from somewhere else. I thought of looking for an example with {{translated}} on its talk page, but found that many pages were articles with lots of contributors and were hard to check, and I thought that it would be easier to use a page with few contributors as it is easier to see that something is wrong. The talk page archive was found to have very few contributors.

By the way, I have noticed that the PDF feature attempts to omit bots from the list of contributors. As you can see, the archive bots are not credited for making the talk archive. However, this also means that other people whose names contain the letter sequence "bot" are unattributed. For example, no PDFs will attribute Niabot. Try generating a PDF for his user page: he's not attributed for creating that page, although he has made lots of contributions to it. This looks like a violation of his right to attribution. --Stefan2 (talk) 16:10, 12 March 2014 (UTC)

Good point on 1 - I thought it was unclickable as it doesn't appear in any way to be a link so you're right we may not have a legal problem there, but morally I think we do. I'd also have a concern that we may be creating problems for ourselves by labelling it as a printable version and not explaining that if you did print it in it's current state you've broken the license conditions.

Wrt 8 I suspect the problem may be the termination clause in the licenses. Now I'm not a lawyer but I doubt they'd put that clause in if it were unenforceable like you suggest. We definitely need guidance on this!

If that's how they're deciding whether it's a bot or not that's terrible. The code needs to be updated to rely on the bot flag. Dpmuk (talk) 17:55, 12 March 2014 (UTC)

For legal team review

Tracked in Phabricator
Task T29629

I started to put together a "How is this?" for legal, but this has gotten complex, and I can really use your help. :) Can we perhaps draft something for them together?

The English Wikipedia is discussing whether attribution generated by PDF and/or print versions of our content is legally sufficient.

PDF versions currently generate a "Article Sources and Contributors" section which includes a Wikipedia URL for the document which is being transformed into PDF and a (potentially partial) list of contributors to the document itself and to any images used in the document. However, this list does not take into account the following facts:

• Content is often copied into articles from other articles or other compatibly licensed sources (whether Wikimedia Foundation properties or not). Currently Wikipedia policies require that copied/translated content be acknowledged in the edit summary so that the history includes (in edit summary) either a link to the original text or a list of authors. Because the "contributor" output of the PDF only lists those who have edited the local document, it doesn't acknowledge these contributions.
• When articles are protected or editors have a conflict of interest, they are frequently encouraged to place proposed changes on the talk page of an article so other editors can evaluate these and place them in the article itself, if appropriate. The edit summary may attribute this content simply with a reference to the talk page or with the name of the editor. The "contributor" output cannot acknowledge these people.
• Uploaders and editors of images may not be copyright holders of the images. These people may find compatibly licensed content elsewhere and attribute it at the image description, but the PDF output "contributors" for image list reflects only (@Stefan2:, I'm not sure exactly what it reflects. :) When I tried a PDF version of the article Sami Yusuf, which includes an image I uploaded after modification but did not create, it got it right. But clearly it doesn't Greta Hall. Do you know?)
• Editors who have the sub-string "bot" as part of their pseudonym are not included in the credits whether they are truly a bot or not.

Print versions include the licensing information on the page, but no link to the license, and they offer no information at all on images - neither on licensing nor attribution.

The questions are this:

• Given the above, is the attribution generated for pdf versions of articles legally sound?
  • If not, what form of attribution would satisfy our TOU and US copyright law? Would it, for instance, be better to acknowledge that the contributors list is partial? Should it be omitted altogether?
• Does the license require that we include a license link or URL in print versions of the article?

Please, expand or change. (Specific shout-out via notification to Dpmuk, Flatscan, MLauba and ParacusForward, since i didn't mention them above. (Dpmuk — Flatscan — MLauba — ParacusForward))

The legal team is generally very good about answering these kinds of questions (although they can't always directly advise and must often do meta:Wikilegal reports), but it's not a lightweight process where I can go easily go back in and add or change the question, so it's best to have a neatly wrapped inquiry from the start. :) --Maggie Dennis (WMF) (talk) 12:21, 14 March 2014 (UTC)

Thanks again Maggie, I have made a few clarifying changes. Mainly around the fact the PDF doesn't include a clickable link, but just a URL. I also added the fact that people with the substring "bot" are not attributed. That seems like a clear bug that should be fixed, but it might help to bring it to legal's attention? ParacusForward (talk) 14:14, 14 March 2014 (UTC)

User:Mdennis (WMF): I don't know exactly how file attribution works. I have noticed several different variants:

• Greta Hall: The image is attributed to those who have edited the textual file information page.
• Sami Yusuf: Both images are attributed by showing the contents of the "author" field in {{information}}.
• Carnelli: The image is attributed to the correct person, but the namespace is included: "Contributors: User:Dtobias".
• Microsoft Windows: File:Windows logo and wordmark - 2012.svg gets "License: unknown Contributors: -". Other images are attributed using the other three methods mentioned above. Also, several images of old versions of Microsoft Windows are skipped in the PDF. While removing some images shouldn't have any legal consequences, it could potentially be confusing for readers if the text ask you to look at an image which isn't there.

I can't figure out how the PDF tool decides which attribution method to use. --Stefan2 (talk) 16:31, 14 March 2014 (UTC)

I'm going to guess here that it tries to be clever and detect the author from the license template (I seem to remember reading that somewhere) but I assume it gets confused when different / new templates are used etc. This is why, if it is what happens, it's such a terrible idea to try to use templates, which are so easily changed for something like this. Looks like you may have to be talking to the devs as well Maggie. I'll try to give a longer reply about the questions etc in the next day or so. Dpmuk (talk) 00:54, 16 March 2014 (UTC)

Bump, User:Dpmuk. :) I want to communicate this to the legal team when I can, but don't want to rush you! --Maggie Dennis (WMF) (talk) 17:33, 26 March 2014 (UTC)

I think that I'm broadly happy with it. I'd try to leave it as general as possible. I know this isn't great for legal but with something as important as this I'd hope they realise they should probably go out and be proactive and find out what actually happens - a description from us is never going to be the same as them seeing the same thing - as an example the print version does have links, they just don't look like links. I have no idea if that's a problem legally. (As an aside, new notification, pretty good, shame it can't tell you when you've looked at something and forgot to reply!). Dpmuk (talk) 16:52, 28 March 2014 (UTC)

---

At m:User talk:Mdennis (WMF)#For your attention, there was a request for a review of the questions to send to Legal.

Bullet points 1-2 and 4 ("Content is often copied", "When articles are protected" and "Editors who have") look fine.

Bullet point 3 looks messy, and the PDF tool seems to use multiple methods to identify the contributors. The legal team needs examples of articles where each method is used. I propose this wording:

• Uploaders and editors of images may not be copyright holders of the images. These people may find compatibly licensed content elsewhere and attribute it at the image description. The PDF output appears to use several methods to determine the contributors to an image. In some cases, the PDF output "contributors" for image list only reflects contributors to the text on the file information page. For example, the article Greta Hall contains an old public domain image from the mid-19th century, but the PDF output incorrectly credits the image to three Commons users.

However, I think that my wording also has problems: it would be better to give an example with a copyrighted image where attribution is required, and the text looks too long. Try to see if it can be improved somehow.

A couple of days ago, I was informed of this German court ruling, where a person was using a Commons image on an external website. Attribution was provided by linking to the file information page on Wikipedia (probably de:Datei:1986 Ulf Fink 800.jpg), and the court found that this didn't satisfy the attribution requirements in the GNU Free Documentation License. It seems that the image was used before {{cc-by-sa-3.0-migrated}} was added to the file, and it seems that the court therefore only evaluated licence compatibility with the GFDL. I may have misunderstood something in the court ruling as I do not speak German. It is unknown if courts in other countries would argue in the same way, and I'm not sure if {{cc-by-sa-3.0-migrated}} changes anything. According to the Terms of Use (wmf:Terms of Use#7. Licensing of Content), Wikimedia contributors agree that a link is sufficient attribution of text, but note that the terms of use list different rules for files. We import quite a lot of content from external sources (for example, images from Flickr or text from other freely licensed wiki projects), and old contributions predate the current terms of use. I would therefore like to add an extra question:

• Is there any difference if the author of the content has or hasn't consented to the Terms of Use in its current wording?

The other questions look fine. --Stefan2 (talk) 14:22, 1 July 2014 (UTC)