On 05/03/13 14:07, Alexander Berntsen wrote: http://www.gnu.org/licenses/javascript-labels.html Yes, it would be. I expect the generated page to be insanely huge, but if LibreJS loads a page so big that blocks your browser, it's not our fault at all :) I see however that it tries to confirm that the source js matches the minified version, which may be quite hard. Furthermore, the resourceloader can multiple modules in one request, producing apparently different urls, so if we had to create all possible urls, expect a factorial growth.

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 05/03/13 14:38, David Gerard wrote: I did not call anyone antisocial. Furthermore I am not affiliated with GNU LibreJS. - -- Alexander alexander(a)plaimi.net http://plaimi.net/~alexander

On 3/5/13 5:53 AM, Helder . wrote: Yes, that's really the case. We took JSMin out of MediaWiki because of it's stupid evil license. Ryan Kaldari

I don't see the purpose of adding a licence string back on to JavaScript post-minification. Any recipient wanting to create a derivative work or redistribute those files is going to go back to the much more readable source files. It would be good form to add licence information to all the JS files in the same way we do for all the PHP files. Many or all of them are missing that now. Given they have a consistent licence, making that clear in each file is just grunt work. I don't see the need for that to survive minificaiton though. If somebody wants to auto verify licence status with software, they can run it on the original JS source before it get's minified. As others have implied regardless of whether you think satisfying the FSF is important, satisfying an automated tool is a concern that can be delegated to the tool owner. The licence status of on wiki user JavaScript is a separate issue, and possibly much more complicated. CC-BY-SA-3.0 is not an ideal licence for software, and it seems likely that there will be code pasted into some user JavaScript pages that is licensed under an incompatible licence. Luke Welling On Tue, Mar 5, 2013 at 10:57 AM, Mark Holmquist <mtraceur(a)member.fsf.org>wrote;wrote:

I would just like to note that while it may be "silly" or "useless" to insert licenses into minified JavaScript, it is nonetheless *legally required* to do so, regardless of the technical aspect of it. And it is not a question of whether we want to support some labeling program that reads JavaScript licenses; both the GPL and CC licenses have requirements that when you convey source code or binaries through any medium that the license be prominently displayed. I strongly doubt that a company is going to sue the WMF for something like this, but even so it's not a good idea to specifically ignore legal requirements for a third-party software. *--* *Tyler Romeo* Stevens Institute of Technology, Class of 2015 Major in Computer Science www.whizkidztech.com | tylerromeo(a)gmail.com

Yes. There seems little value in unqualified people debating if it is legally required. The mainstream FOSS licences all predate minification and seem to have been written with compiled languages in mind, not interpreted languages. Most have language that requires the licence in the source version, but not the binary version. Deciding whether minified JavaScript is technically or in spirit a binary form seems like something best left to experts. My conscience would certainly be clear if we only had a licence in our source distribution. Luke On Tue, Mar 5, 2013 at 12:25 PM, Marc A. Pelletier <marc(a)uberbox.org> wrote:

The licensing information is on the page itself, of which the minified js winds up a part. For every file or other object that makes up the page to all contain the licensing information would be pretty unusual. It's like taking a file out of a page and then complaining that it has no licensing information when said information was in the page text right under it. On 05/03/13 17:36, Tyler Romeo wrote: -- -— Isarra

On Tue, Mar 5, 2013 at 10:10 AM, Caroline E Willis <cewillismail(a)gmail.com>wrote;wrote: Yes. :) This was already on my radar, and I am following this discussion (which has been useful; specifically, I did not know about the bug already filed on the issue). For those of you who don't know me, I'm new to the foundation, but have been around foss and foss licensing for a while; a good backgrounder on me is here: http://www.mail-archive.com/wikimediaannounce-l@lists.wikimedia.org/msg0052… Luis -- Luis Villa Deputy General Counsel Wikimedia Foundation 415.839.6885 ext. 6810 NOTICE: *This message may be confidential or legally privileged. If you have received it by accident, please delete it and let us know about the mistake. As an attorney for the Wikimedia Foundation, for legal/ethical reasons I cannot give legal advice to, or serve as a lawyer for, community members, volunteers, or staff members in their personal capacity.*

On 5 March 2013 11:55, Matthew Flaschen <mflaschen(a)wikimedia.org> wrote: I think the point is that https://upload.wikimedia.org/wikipedia/commons/4/48/Magnolia_%C3%97_soulang… have any licence information in it either, though https://commons.wikimedia.org/wiki/File:Magnolia_%C3%97_soulangeana_blossom…, and this is analogous to the output of load.php not having licensing information in it, but the composited page having it. (And licensing of Gadgets is a complete mess, but that's somewhat orthogonal to the point.) J. -- James D. Forrester Product Manager, VisualEditor Wikimedia Foundation, Inc. jforrester(a)wikimedia.org | @jdforrester

On 03/05/2013 12:29 PM, Luke Welling WMF wrote: I am referring to Isarra's comment: "The licensing information is on the page itself, of which the minified js winds up a part." As far as I can tell, that is not true for the *code* license(s) for core and extensions. Matt Flaschen

On 03/05/2013 02:33 PM, Platonides wrote: Do you really expect people to find that? We're basically talking about what is visible in the "binary" version of the site. We all know they can get license information from the source by doing git clones. I don't think it's realistic that people will successfully guess they can visit that /w/COPYING url. And not all the code is under GPLv2 anyway, though it should all be free on WMF sites. Matt Flaschen

This is based on a flawed reading of the GPL. The GPL covers the distribution of program code. The license specifically states that “The act of running the Program is not restricted”. (Furthermore: “Activities other than copying, distribution and modification are not covered by this License; they are outside its scope.”) The terms you are all referring to relate to the distribution of the software, not the running of the software. Wikipedia.org, does not distribute the software, that is MediaWiki.org's job. If Wikipedia wanted to, we could remove all licensing information from the software and it would still be completely legal. The GPL *only* comes into effect once you start distributing the software. This is why other licenses such as the Affero General Public License have been written, to stop people using and modifying software like Mediawiki, but failing to release their modifications back to the community. The current method of distributing Mediawiki via Mediawiki.org is perfectly complaint with the GPL. -- Chris

On Mar 5, 2013, at 6:22 PM, Tyler Romeo <tylerromeo(a)gmail.com> wrote: Sure, but it depends on your definition of "prominently displayed". First off, I agree our javascript files should have license headers in form of a code comment on top of the files (like we do for PHP files). But only to clarify their license, not because it is required. Because we already have a general LICENSE in our distribution, which if I recall correctly explicitly states that unless otherwise indicated, all is under said license. We don't have a license header in our release notes, in jpg, png, svg, sql files etc. A good example (to make it more complicated) is our README where we mention certain PNG file (cc icons) and JS files (sajax) are from a different author and license. We don't alter their PNGs and JS files, instead we mention it elsewhere (whether it belongs in README is another question). However I don't think it makes sense in any way for this to be sent to the browser. A few examples. ## Media in wiki pages We don't display the license or attribution of images inside the article near to the image. You go the the image descriptions page (by clicking the image) and there it is. ## Content of wiki pages We do display the license on the bottom of every page (which is about the wiki content, not the software). But not the authors. You go to the History page of the current article and find a list of contributors there. Note that the user doesn't click on the content here, but on the "History" tab. ## Server-side code in the software Any program code in our software that is not sent to the client. But its result is sent to the client. Everything you see on the wiki is the result of executing that server-side code. And if you consider HTML to be part of what you "see", then there's actually a significant amount of server-side code being sent to the client, because that is literally or abstractly (Html.php) explicitly written in the code. ## Client-side code in the software Any program code in our software that is sent to the client (css, javascript). These are commonly combined and minified, which means HTTP headers are not an option (unless you'd implement offsets or delimiters correlating to the content). ## Media in the software Any interface images and icons in our software. These are commonly embedded as base64 encoded data, which means HTTP headers are not a feasible method for delivery of information. ## Media in print A photograph used in a magazine or print. It might have the license/attribution over top of the image or closely to it, but it isn't uncommon for there to be a dedicated page for it. That then refers back to the images (by page number, position and/or by title) to disclose the license and attribution. If you'd look at any single spread (e.g. open it on page 3, you see page 3 and 4) you wouldn't have a complete legal picture. The same if you take out a page and "access" it directly. And even more so if you were to take scissors and take out an individual photo, in which case you'd lose the info even if it was printed right next to it. ## Conclusion So let's take the extremes and sum them up: * A page can contain multiple pieces of content from different sources (software interface, wiki page content, wiki media embedded) that can all be from different authors under different licenses (some might even be non-free, e.g. when embedding fair use, though lets avoid that can of worms for now). * Our wiki text source does not have license headers. Instead the platform on which they are primarily displayed (accessing html pages) has a footer. When accessing it from the API you're circumventing the main portal and are expected as a consumer to "check out" the primary access point to find out the license and author. * Like wise, accessing a multi-media file[1][2] directly does not give you attribution or license information in the file itself or in its headers, not even a link to it. I presume the rationale here is similar to the "Media in print" example. One might argue that because it is accessible over a separate http request it needs to be standalone, but I'm not sure thats justifiable. It is an implementation detail of how the web works. You can't require everything to be in the same web request (imagine MediaWiki ajax loading of article contents, the footer would be there always and you'd be loading the actual content over a separate request). You could also consider an individual page of a book to be a separate "request", again see the "Media in print" section above. * We certainly aren't going to embed the GFDL legal text in every http request… So given all that, whilst not having a clue whether all that is legal – I'm assuming so since that's practically how every website in the world operates (both free and non-free websites) – I think it is acceptable for our program code to follow similar guidelines as multimedia and text (since code is text). So it ought to be legal for our software to deliver individual bits and pieces to the browser that are not a complete package with license and all (like pages in a book). Instead one is expected to know about the colophon page. If you are in a position where you're legally required to have permission to do what you're about to do (e.g. copy our javascript), you go back up the chain and access the complete package. Find the "Powered by MediaWiki" button on the bottom of the page the code was bundled with (the "colophon"). Then, after looking up MediaWiki's license, go and find that code again in the original MediaWiki book and find the helloWorld.js in all its glory on page 42. Not sure how well that analogy flies, -- Krinkle [1] https://upload.wikimedia.org/wikipedia/commons/thumb/e/eb/Baantjegracht_Dok… [2] https://upload.wikimedia.org/wikipedia/commons/e/eb/Baantjegracht_Dokkum_20…

On 3/5/13 1:03 PM, vitalif(a)yourcmc.ru wrote: I license all of my MediaWiki extensions under an MIT license since I want people to be able to reuse the JS code on-wiki, but some people have claimed that even MIT isn't compatible with CC-BY-SA [1]. I've been thinking about switching to CC-Zero instead. It's funny how most "free software" is so burdened with inane incompatible restrictions that we can't legally use it in many situations. What do people think about using CC-Zero as a license? Now that's free software! 1. https://en.wikipedia.org/wiki/Wikipedia_talk:Copyrights/Archive_15#CC_BY-SA… Ryan Kaldari

On 5 March 2013 22:08, Tyler Romeo <tylerromeo(a)gmail.com> wrote: People will say any spurious bollocks in a licence discussion. (You've been on Commons, right?) This is why we have proper lawyers on hand :-) I appreciate it would be *nice* to put the licence in the JS, Mako makes the point as nicely in the bug as the original poster didn't in this thread. But there must be a method that isn't operationally insane. - d.

On 6 March 2013 15:20, Jay Ashworth <jra(a)baylink.com> wrote: WTFPL of course! - d.

On 2013-03-05 6:29 PM, "MZMcBride" <z(a)mzmcbride.com> wrote: I wonder how osi would feel about https://github.com/avar/DWTFYWWI license. -bawolff

On 2013-03-05 9:17 PM, "Antoine Musso" <hashar+wmf(a)free.fr> wrote: domain: The ocaml tool does security verification from what I understand. The actual rendering is done by TeX.(I think) Also I didnt think the license of a tool extended to its output. I can make non gpl images in the gimp, etc -bawolff

MZMcBride wrote: https://github.com/lcrocker/OneJoker It seems Lee is alive and well and still waiving his rights. :-) MZMcBride

The need for minification suggest that maybe the web needs a bytecode format for css / javascript / xml, one designed to save space. I know text is the tradition in unix, but anyway. -- -- ℱin del ℳensaje.

On 06.03.2013, 2:01 Ryan wrote: My extensions are WTFPL;) -- Best regards, Max Semenik ([[User:MaxSem]])

On Tue, Mar 5, 2013 at 2:11 PM, Jay Ashworth <jra(a)baylink.com> wrote: True, but if it's legally required it's not like we have an option. *--* *Tyler Romeo* Stevens Institute of Technology, Class of 2015 Major in Computer Science www.whizkidztech.com | tylerromeo(a)gmail.com

On Tue, Mar 5, 2013 at 3:08 PM, Jay Ashworth <jra(a)baylink.com> wrote: If you read the licenses, it's pretty obvious. Also, popular libraries (such as Google's hosted versions of jQuery and others) always include license headers in the minified versions. And in the unlikely event that's not good enough, the Foundation may well But WMF getting a license doesn't help everybody else who uses MW. *--* *Tyler Romeo* Stevens Institute of Technology, Class of 2015 Major in Computer Science www.whizkidztech.com | tylerromeo(a)gmail.com

----- Original Message ----- Minification is a WMF cluster issue, not a MW software issue, is it not? Cheers, -- jra -- Jay R. Ashworth Baylink jra(a)baylink.com Designer The Things I Think RFC 2100 Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII St Petersburg FL USA #natog +1 727 647 1274

----- Original Message ----- Ah; thanks. Have not looked at internals lately. Since minification to me as a netadmin is a strategic "size of pipe" issue, I assumed it was something deployed on WMF sites, not something baked into the base package. Cheers, -- jra -- Jay R. Ashworth Baylink jra(a)baylink.com Designer The Things I Think RFC 2100 Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII St Petersburg FL USA #natog +1 727 647 1274

On Tue, Mar 5, 2013 at 9:16 PM, Tyler Romeo <tylerromeo(a)gmail.com> wrote: That's not what I see. If I look at jQuery as hosted by Google [1], it starts with the following comment (and nothing more): /*! jQuery v1.9.1 | (c) 2005, 2012 jQuery Foundation, Inc. | jquery.org/license //@ sourceMappingURL=jquery.min.map */ It does link to a license (though it doesn't even mention what the license is directly), but it certainly doesn't contain the whole license itself. And, as I understand it, that's what you claim is required and what others claim would be a waste of bandwidth [1]: http://ajax.googleapis.com/ajax/libs/jquery/1.9.1/jquery.min.js Petr Onderka [[en:User:Svick]]

On 05/03/13 21:55, Matthew Flaschen wrote: I just checked and there are 73 authors of the resources of MediaWiki core. More than I expected, but not unworkable. We could relicense our css and javascript as MIT, MPL, GPL-with-explicit-exception... Regarding GPL requisites, it seems clear that minified javascript is “object code” [1], which we can convey per section 6d [2], which is already possible if you know how the RL works, although we should probably provide those “clear directions”. Most problematic would be that you should also obey sections 4 and 5 (although I see a bit of contradiction there, how are you supposed to “keep intact all notices” where most notices are present in comments, designed to be stripped when compiled?) But are we conveying it? As javascript is executed in the client, it probably is. 1- «The “source code” for a work means the preferred form of the work for making modifications to it. “Object code” means any non-source form of a work.» - Section 1 2- «Convey the object code by offering access from a designated place (gratis or for a charge), and offer equivalent access to the Corresponding Source in the same way through the same place at no further charge. If the place to copy the object code is a network server, the Corresponding Source may be on a different server (operated by you or a third party) that supports equivalent copying facilities, provided you maintain clear directions next to the object code saying where to find the Corresponding Source. (...)»

On 03/06/2013 07:30 AM, Platonides wrote: Don't forget the 58 other authors of skins/ (although some commits touching that path might not be to CSS or JS): $ git log --format=format:%an --no-merges resources/ | sort -u > ../resources.txt $ git log --format=format:%an --no-merges skins/ | sort -u > ../skins.txt $ comm -23 ../skins.txt ../resources.txt Adam Miller Ævar Arnfjörð Bjarmason Alex Shih-Han Lin Alex Z Anders Wegge Jakobsen ankur Arne Heizmann Benny Situ Charles Melbye Daniel Cannon Daniel Kinzler Erik Moeller Evan Prodromou Gabriel Wicke Guillaume Blanchard Guy Van den Broeck Huji Ilmari Karonen isarra Jack Phoenix Jan Luca Naumann Jan Paul Posma Jens Frank Jimmy Collins Jon Harald Søby Jure Kajzer karun Katie Filbert Laurence Parry Leon Weber Lisa Ridley Lupin Magnus Manske Marcin Cieślak Matt Johnston Michael Dale Mohamed Magdy Nicholas Pisarro, Jr Nick Jenkins Nimish Gautam Patrick Reilly Philip Tzou Platonides Purodha B Blissenbach Remember the dot River Tarnell Rob Church Robert Stojnić Rotem Liss Ryan Schmidt Shinjiman SQL Tobias Tom Gilder Tpt Victor Vasiliev X! Zheng Zhu -- Wikipedia user PleaseStand http://en.wikipedia.org/wiki/User:PleaseStand

----- Original Message ----- We note that you have spoken. Alas, the other 153 people who own copyright in the code in question haven't and, no offense, Jack, assuming they will have the same outlook you do -- when it's on the record that developers' opinions on this range to both ends -- is probably not a safe enough bet for the foundation. :-} Cheers, -- jra -- Jay R. Ashworth Baylink jra(a)baylink.com Designer The Things I Think RFC 2100 Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII St Petersburg FL USA #natog +1 727 647 1274

I don't see how the copyright of MediaWiki's code is bike-shedding at all. As a volunteer, I'd like to be damn sure MW is actually an open source project. There's a reason copyright licenses exist, and it's to provide freedom for developers and users. If MW were completely licensed under the WTFPL, others could copy MW, change it, and then make it proprietary, whereas with the GPL there is a restriction on that. When I contribute my code to this project, I am fully aware and happy with the fact that it will *never* be used in a closed source product. Just because some people don't care enough about how laws exist in this world and we have to operate under them doesn't mean everybody else should be screwed over. So if we could actually get back on topic rather than bitching and complaining about doing things some of us don't necessarily enjoy. *--* *Tyler Romeo* Stevens Institute of Technology, Class of 2015 Major in Computer Science www.whizkidztech.com | tylerromeo(a)gmail.com

Non-lawyers arguing over how to interpret licenses, uses, and other stuff with the minimised code doesn't prevent such screwing over either. It is undoubtedly an open-source project; the question is the legal one of where all things need to be attributed and cited, and at the end of the day pretty much none of us are qualified to answer that in any full capacity. Some speculation can be fine and help people get an idea as to how to proceed, but this is indeed to the point of bikeshedding. This is not bitching. This is a legitimate complaint that this thread is getting out of hand with little productive value. On 06/03/13 21:42, Tyler Romeo wrote: -- -— Isarra

----- Original Message ----- While I disagree as to whether minified code needs a human readable embedded license, I don't think it's reasonable to characterize the discussion as bikeshedding, Chad. I care about the licensing on my code. I'm not alone. Cheers, -- jra -- Jay R. Ashworth Baylink jra(a)baylink.com Designer The Things I Think RFC 2100 Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII St Petersburg FL USA #natog +1 727 647 1274

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Since this thread is slowly moving over to a debate as to whether it constitutes bikeshedding or not (and people can't seem to agree on that either), I'm going to unsubscribe to this mailing list by the end of today (in 15 hours or so) as I get way too much email already. I have made my concerns clear. Thank you for verifying that your intentions are to use free software. If anyone were to CC me in any important emails (decisions regarding licence headers), I would be grateful. If someone has suggestions for how to make GNU LibreJs[0] accept Wikipedia, please email me about this. The only solution I can think of is using a whitelist -- but this may be dangerous, as developers can just go "oh I don't bother with licences, you can just whitelist my page". [0] https://www.gnu.org/software/librejs/ - -- Alexander alexander(a)plaimi.net http://plaimi.net/~alexander

----- Original Message ----- Perhaps. But HTML is also "executed" in the client, and some legal decisions have gone each way on whether the mere viewing of a page constitutes "copying" in violation of copyright (the trend is towards "no", thankfully. :-) Cheers, -- jra -- Jay R. Ashworth Baylink jra(a)baylink.com Designer The Things I Think RFC 2100 Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII St Petersburg FL USA #natog +1 727 647 1274