Wikipedia to be Licensed Under Creative Commons

sla291 writes "Jimmy Wales made an announcement yesterday night at a Wikipedia party in San Francisco : Creative Commons, Wikimedia and the FSF just agreed to make the current Wikipedia license compatible with Creative Commons (CC BY-SA). As Jimbo puts it, 'This is the party to celebrate the liberation of Wikipedia'."

Strange...

I thought RMS was not a fan of the Creative Commons license.

Re:Strange...

I heard RMS give a talk where he criticized Creative Commons (the organization) because not all of the licenses they publish guarantee freedom. As he put it (paraphrasing from memory): "If you take the intersection of all the licenses offered by Creative Commons, you get nothing. There are no core freedoms that all the licenses guarantee."

Basically, RMS thinks that some of the licenses are great (the ones that allow redistribution, derivative works, and promote share-alike), but thinks others are terrible. RMS is famous for being careful with words, and dislikes the fact that when you say "this is available under a Creative Commons license" it basically means nothing (until you know which specific license is being used, you don't know what freedoms are being guaranteed).

Of course the FSF's intention is to promote freedom, whereas the Creative Commons organization has as its core mandate something more along the lines of "promote understanding of copyright law, and show copyright holders that they don't have to use a maximal, all-rights-reserved copyright, but that they can distribute under more permissive licenses, too." The creative commons organization emphasizes author choice instead of user freedom.

Still, all that having been said, there is some clear overlap between the CC licenses and the GPL. So, an appropriate license can certainly be compatible, and I'm fairly confident that RMS approves of those freedom-granting licenses.

Re:Strange...

GPL may not restrict commercial use and sale of software, but it sure makes it difficult. Like it or not, a large amount of software written today is written for financial gain. GPL makes it quite literally impossible to ensure you get paid for your work, because anyone who buys your software also gets the code and full right to give it out to whoever they want to.

It's good that some companies can make money on services and support (Red Hat etc), but that doesn't work everywhere.

There needs to be an open source license that gives you everything the GPL does, but only to people who paid for the software. Anyone who has a license for the software can modify it all they want, give out their modifications and even complete modified versions, but only to others who have a license for the software.

Re:Strange...

There needs to be an open source license that gives you everything the GPL does, but only to people who paid for the software. Anyone who has a license for the software can modify it all they want, give out their modifications and even complete modified versions, but only to others who have a license for the software.

Hmmm... This is a good thought here.

The closest that I've seen to something like this is what the old Borland did with the original Delphi source code. It still belonged to Borland, but they gave you the complete source code for nearly the entire compiler, and in theory you were permitted to share changes to that source code with anybody else who bought a licensed copy of that compiler. It in fact was a fairly common practice among Delphi programmers.

They also did something that was fairly unusual for compiler writers (although fairly common now): They explicitly granted license for the libraries for you to use in any way you wanted... as long as you gave them only others who paid for the license. And the binaries were completely free of copyright restrictions for the end-user (meaning the software developer/publisher). You don't worry about this with GCC, but other major compilers did have some pretty profound restrictions on how you were "allowed" to redistribute your software once it ran through their compiler... or be required to pay some sort of licensing fee for the library files needed to run your software.

An "open source" license like you are talking about still doesn't deal with what happens to abandonware, which IMHO is one of the real strengths of the GPL: If MySQL A.B. goes completely out of business, the MySQL software will still be available from many sources, and you will even be able to find people willing to make patches for the software and make future updates and releases. There is no GPL'd abandonware, other than the thought that a particular development group has disbanded (like the ReiserFS group run by Han Reiser).

Requiring somebody to purchase a license from a single entity of some sort gives a worry that the people collecting the money from the licenses may not be available if you don't have the license and really need the software. Under this sort of arrangement, how would you solve this problem without resorting to something like the GPL?

Re:Strange...

RMS is an idealist, whilst the people behind the CC organization are pragmatists.

I'll concede that both sorts of people are necessary, although I certainly know which one I'd put my money behind.

Re:Strange...

RMS is an idealist, whilst the people behind the CC organization are pragmatists.

I'll concede that both sorts of people are necessary, although I certainly know which one I'd put my money behind.

Me too. Years of observation has shown (time and again) that all those wacky things RMS warns about generally come true a year or two later. An idealist with a good sense of how human nature and "the market" works is a powerful powerful thing. Not all idealists are sitting in the meadow chasing dandelions.

Alternatively you might pigeon-hole him as a very in-touch cynic. In that sense consider RMS's fanaticism of protecting Freedoms from the point of view of never underestimating the creativity and number of sleazy people out there ready to make to quick buck and rip you off if they can.

Re:Strange...

Years of observation has shown (time and again) that all those wacky things RMS warns about generally come true a year or two later.

It's funny you say that. I would have said exactly the opposite. My years of observation have shown (time and again) that the software world carries on developing new things for commercial, charitable and personal use purposes, without any great disaster happening because it isn't all licensed under something like the GPL.

Meanwhile, all the claimed benefits of the GPL have turned out to be rather shallow in practice. Forking of major projects happens relatively rarely. Most end users of most software don't really want to be able to hack the code (not least because most of them wouldn't even know where to start). Those who are willing to share their work with the community get tied up in silly arguments over the technicalities of the licences in question, and that problem has been made worse by the introduction of the GPL2 vs. GPL3 debate. The entire OSS world has not collapsed under the weight of patent claims, nor is it ever likely to with or without GPL3's help for the simple reason that many major businesses now rely on it and they have patents of their own to fight back with, just as everyone in the commercial world has done for years.

On the whole, with due respect to his past achievements, I'd say these days RMS talks a lot but often comes across badly and isn't particularly relevant in the modern software development world. I'd back the pragmatist every time.

Re:Strange...

Can you include a GPL code snippet in a Creative Commons document (e.g., to illustrate a concept)?

It's not a simple question. For instance, it is ironic to note that you cannot legally include GPL code in a document licensed under the GNU Free[sic] Documentation License.

Source code for images and audio

In prose, any derived work is also prose, so you would retain the Four Freedoms just by being granted the same license.

"Source code" means the preferred format for editing a work. Machine-readable text without DRM is pretty much the only format I can think of where "source code" is the document itself. The "source code" for a graphic work in PNG format might be a layered file or an SVG file. The "source code" for a sound recording in Vorbis format might be the original multitrack project.

Fantastic

I much prefer CC and use it in my own work frequently. I've contributed to Wikipedia many times, and think this is a great move. It will also boost CC, which deserves all the exposure it can get.

Won't make any difference

Presuming that the GFDL doesn't allow commutation with the CC licence then this change (if true, since the only source in the submission is a blog) won't make any difference unless wikipedia is wiped and they start again. Everything up until today on wikipedia is licensed under the GFDL, so that content will always be under the GFDL, because that is the licence the contributors agreed to when they submitted content (apart from a few who have made statements releasing their work of more restrictions, such as PD), and that licence can't be revoked or replaced by a CC-BY-SA licence without their permission (all hundreds of thousands, if not millions, of them).

Ahh, but they're actually changing the GFDL

If you RTFM, it's not that they're moving Wikipedia from GFDL to CC-BY-SA; rather, the GFDL is becoming compatible with CC-BY-SA. If the GFDL license in use had the usual "or any future version" clause in use, then the content was initially given with permission for relicensing under this new version -- so no problem at all.

Difference?

What are the differences between Creative Commons and their current GFDL?

Re:Difference?

Principally that the GFDL has some clauses that make odd but relatively minor requirements. It bars the makers of derivative works from removing any "invariant sections" from the original work (does not apply to Wikipedia). Distributing any GFDL work requires that you distribute with it a "transparent" copy of the entire license, which is impractical for a single printed Wikipedia article, for instance. But the core rights that the GFDL grants (duplication, derivative works, commerical or non-commercial use) are the same as those granted by CC-BY-SA. The GFDL just contains some "FSF-isms".

Appropriately enough, the Wikipedia article on the GFDL [wikipedia.org] includes a list of criticisms that cover this topic.

Re:Difference?

What are the differences between Creative Commons and their current GFDL?

GFDL requires that so-called "Invariant Sections" (talking about the author and their relationship to the subject matter) be carried forward into future versions unchanged. Wikipedia articles don't have Invariant Sections, but you could take a Wikipedia article, change it, and then add an invariant section; everybody who wanted to use your changes would then have to keep the invariant section intact.

GFDL also requires that the title of the work be changed after every modification, and that sections titled "Acknowledgment" and "Dedication" be kept intact. Nobody really cares about these clauses, and Wikipedia has long ignored them.

If you want to redistribute a (modified) version of a work, the GFDL also requires that you accompany it with a copy of the GFDL and list at least five of the principal authors of the work on its title page. That's also widely ignored, by Wikipedia and others.

A work licensed under CC-BY-SA can be relicensed under any later version of CC-BY-SA and also under any license deemed equivalent by Creative Commons (since CC-BY-SA 3.0). A work licensed under GFDL can only be relicensed under a later version if the licensor explicitly added a clause to that effect; the Wikipedia license agreement contains such a clause, but a downstream distributor could remove it.

Re:Difference?

GFDL requires that so-called "Invariant Sections" (talking about the author and their relationship to the subject matter) be carried forward into future versions unchanged. Wikipedia articles don't have Invariant Sections, but you could take a Wikipedia article, change it, and then add an invariant section; everybody who wanted to use your changes would then have to keep the invariant section intact.

GFDL also requires that the title of the work be changed after every modification, and that sections titled "Acknowledgment" and "Dedication" be kept intact. Nobody really cares about these clauses, and Wikipedia has long ignored them.

If you add an invariant section, the legal requirement for keeping those invariant sections is only to those whom you distribute that new version of the content after this modification. It doesn't apply to earlier versions...and Wikipedia would as a matter of custom delete any invariant sections and material that would have to be kept.

But on the whole, you are largely correct that Wikipedia does ignore this section of the GFDL by simply prohibiting as a matter of policy the creation of any invariant sections. There may be some GFDL'd content that was added to Wikipedia which contained invariant sections... and that content would either have to be deleted, or be in technical violation of the terms of the GFDL. The problem here is that there is, comparatively speaking, so little actual content outside of Wikimedia projects written using the GFDL that this is usually not a problem for copyright violation situations.

If you want to redistribute a (modified) version of a work, the GFDL also requires that you accompany it with a copy of the GFDL and list at least five of the principal authors of the work on its title page. That's also widely ignored, by Wikipedia and others.

I've complained about how the terms of this requirement might actually be met using the current interface on Wikipedia and the MediaWiki software. All of the raw information necessary to meet this requirement is kept on the servers, but it is not very easy to access and a pain to try and obtain. There is also no simple mechanism to distinguish between a vandal whose edits have been completely removed, and a serious contributor who has added some very real meat to the articles. Most lists of authors on Wikipedia articles include not only the "principal authors" but also vandals, crackpots, sysops (who clean up the mess from vandals), and people stopping by to fix the spelling of just one or two words.

The GFDL is also very weak in its formal definition over what might even constitute an author at all, and it is very possible that Willy on Wheels (look it up on Wikipedia if you don't know him) could get equal credit with RMS on the article regarding the Free Software Foundation. But that is a problem with the GFDL, not Wikipedia.

This is, however, something I've paid careful attention to when I've distributed Wikimedia content outside of the Wikimedia projects themselves. And that is something I have done... not just talked about.

Where are the source links?

Who else has published this news? I'm all for it if it's true, because it has the potential to immensely increase visibility for the Creative Commons project. I'd just like to see some other sources for verification.

As far as the direct impact of Wikipedia's usability from a sharing-of-content perspective, I don't really see it making a huge difference. I'm not a lawyer, though, and haven't spent time comparing the two licensing models in detail. Anyone care to comment?

No-one asked me.

I'd say that I'd have liked to be asked before my stuff was re-licenced, but everything I worked on and cared about on Wikipedia was deleted for not being notable enough, or something. Oh well.

Yes they did.

When your posted your stuff, they asked you if it was alright to license your contribution under the GNU FDL, and you agreed. The GFDL allows users to choose either the existing version of the GFDL or any future version, which they pointed out at the time. Now that the FSF has modified the GFDL, users (including wikipedia) can choose to use it for you contributions if they wish.

Good luck with that....

...considering that every contribution made to Wiki was made under the GFDL, not CC. Are they going to get permission from all of the past contributors to change the license, or are they going to throw it all away and start from scratch? They don't own the content (it was licensed to them under GFDL) so they just can't change the license.

Re:Good luck with that....

Trust me when I say that the shit has just hit the proverbial fan right now...I'm going to have to unsubscribe...my contributions may be removed from Wikipedia

I don't know if you're a big muckety-muck in the Greater Wiki Community; maybe you are, in which case I risk making a huge ass of myself. (I tried Googling for you, but all I kept coming up with was your many UserPages.) And, of course, it's always sad when people feel slighted or disenfranchised. That said:

I feel fairly certain that anyone who, by comparison to his own views, considers Richard Stallman and the FSF to be a bunch of money-grubbing, compromising, unprincipled corporate hacks is someone whose writing I'm not going to miss.

Modifying licenses

The whole notion of "or any future version" of the license, as is commonly used in GPL and GFDL licenses, has always worried me. IANAL, but from a legal standpoint, it seems odd that you can agree in a binding way to something which is yet to be defined.

Plus there's the (seemingly vanishingly small, at present) risk of the FSF being co-opted by some faction which changes the licenses in ways which make them entirely different in spirit to the current versions. That wouldn't mean the content wouldn't still be available under the current versions of the licenses (you can't un-license it once it's out there), but it could mean that forks could be made which were non-free. How do we know that, in say 40 years, the leadership of the FSF will be as principled and uncorruptible as the current leadership?

Re:Modifying licenses

With my small army of rebels I take over the FSF and I create GPL v4 which is the equivalent of a public domain license. I fork all projects that are GPL v2 or any later version. I change the license of my forks to be GPLv4 because it still is in the scope of the original license (because of the later version clause). Now I use all my code for free! Yeah!

Re:Modifying licenses

How exactly do you propose that something licensed under GPL v2 (or v3) could be forked to non-free, even under a future version of a license developed by an evil future-FSF? The politics of the FSF do not factor into the license chosen by a particular author.

If you include the "or any future version" clause, the politics of the FSF categorically do affect the licensing of your software, because it is the who FSF define the future versions of the license. Say I release SuperWidgetApp under GPL v2 with the "any future version" clause. Also say that down the line, bad people take over the FSF and make GPL v27, which has no requirement to release the source code. BadCorporation could then take the source to SuperWidgetApp, invoke the "any future version" clause and apply GPL v27; they can then make trivial changes, release it as HyperWidgetApp and not release the source (because under GPL v27, they don't have to). The GPL v2 - v26 versions would still be Free, but the modified version would not be.

Re:GPL by proxy

NO, its absolutely in the same spirit as the GPLv2. The GPL's purpose is to foster user freedom- the user is free to do what he wants with his software, including alter it, redistribute it, etc. GPL3 removes some things people were using to circumvent the GPL- patents and hardware lockouts. The spirit of the GPL is better served by GPL3 than GPL2.

Please clarify...

'This is the party to celebrate the liberation of Wikipedia'.

I guess the question that arises is..."Liberation from who?"

"compatible" with cc, not "switching to"

the op is no longer correct. the article has been updated to say that wikipedia will be cc-compatible, not that it will switch to it. to quote:

Contrary to the old title of this post (thanks to Larry for the clarification) Wikipedia is not switching to CC. It actually made a deal allowing the community to relicense the content of the wikis under a BY-SA license. So it's now up to the Wikipedians to choose whether they do or not.

this is a bit of legal-hair-splitting (standard ianal disclaimer), but it does mean that there there shouldn't be any legal issues with converting prior content.

also it seems that the cc by-sa license http://creativecommons.org/licenses/by-sa/3.0/us/ [creativecommons.org] is basically equivalent to the gfdl. it is not "public-domain"ing the content, nor is it "bsd"ing the content. it just seems to make it a less-software-centric license. (anyone else, please feel free to correct.)

About time

As a Wikipedia, this is great news. The GFDL is too cumbersome. They need to do it right though. They need to freeze Wikipedia, make a dump, make that dump permanently available under the GFDL, and then open up Wikipedia with the new license. Legally they probably don't have to, but this should help others who want to fork based on the license change.

Scraper Site

Wait, you mean all of those scraper sites will become legal? Aiding the development of even more scraper sites? Sweet!

Great news for open educational resources!

I'm on the Advisory Board of WikiEducator [wikieducator.org], a project sponsored by the Commonwealth of Learning and other international development agencies to build open educational resources primarily for the developing world. We use CC-BY-SA, and this news is phenomenal for us because it means that all of Wikipedia's content, which has up to now been licensed in a way that's philosophically identical but legally incompatible, is now available for us to use.

There's no reason that content should be in separate BY-SA and GFDL silos. It's critically important to be able to remix it together.

Big deal

So now the people at the bus stop say that anyone can repeat their ramblings if they want. Whoo Hoo.

TWW

Not yet, anyway....

As I write this, there is no official new version of the GFDL. It would have to be announced by the FSF, and it isn't. The FSF website says nothing about this, and the Gnu Project website lists the 2002 version as the latest.

This seems fishy in other ways. I could say the FSF was diligent in soliciting comments for GPLv3, but "diligent" seems like too soft a word. It would seem odd to change the GFDL with no advance notice whatsoever.

Not to mention, it seems unlikely that this is the third best moment of Lessig's life (after two things involving his wife, which I don't think we need details on). (Yes, I did RTFA, what there was of it. Wanna cancel my /. license?)

This smells like a hoax or prank to me. However, I'm going to look at www.fsf.org [fsf.org] and www.gnu.org [gnu.org] next week and see if there is anything to this.

What about the plagiarized material?

Lots of wikipedia content is straight copies from other sources. Does Wikipedia have the right to license that material under Creative Commons without the original authors' consent? I know that Wikipedia has been essentially providing that material for free anyway, but it seems strange to "formalize" the giving away of plagiarized content under a particular license that the original author might not agree with.

Why not let the users choose their licence?

As a user and contributor [wikipedia.org] (and donor) of Wikipedia I prefer GFDL. Not that I don't like CC. But my first preference is GFDL, and CC is my second preference, that's all. Oh, and I actually dislike the idea of "any future versions" even in GFDL/GPL although I do see practical advantages. However, I also see practical advantages in GFDL-CC compatibility, as now many people will be able to mix Wikipedia content with CC-only content which is a good thing. What would be a BAD thing would be a total CC switch by Wikipedia and the departure from GFDL.

So, I essentially do welcome this compatibility but only marginally... In fact I don't want to see the Wikipedia community getting away from the FSF and the GNU's focus on idealism and purity. I'm an FSF Contributing Member as well, so maybe I'm just a bit biased, but that's just how I feel. Perhaps the future will prove that the GFDL-CC compatibility is more good than bad.

What I don't understand, however, is why it's the Wikimedia or a group of admins who get to choose licences, and not let the users themselves one-by-one do it. Wiki articles emerge after a series of edit wars and vandalisms, and yet they are readable and useful. Meaningfull and useful articles emerge even when large groups of trolls try to bring chaos. What if each wiki article had its own licence decided by the initial contributor? Trolls would surely use this to bring more chaos, and users with no knowledge would also do stupid things, but in the end I believe that useful articles would still emerge, and the licence would be the choice of the community as a whole rather than a few people with lots of social capital or prominence in the wiki community.

I believe a wiki must be built by its users rather than by a core admin team... that's the spirit of the wiki. So, why on earth should the admins force users to either accept a predefined licence or not contribute? This idea led me to allow my users on my wiki [wikinerds.org] to choose the licence of their choice for the pages they create. Yeah I know at some point we will have a crazy mix of incompatible licences, but it is up to the users and their collective intelligence to decide how to use the feature of licensing choices. In the end I believe users as a community will make intelligent choices. That's the spirit of the swarm intelligence, after all, which is also the field of my academic research for my Master's... Give users some guidance, some rules of behaviour, apply the minimally possible central administration and let them free to do as they like.

I'd welcome the idea of letting users decide the licence they would like to be implemented in Wikipedia as well. Perhaps this could help more people to understand what licences are, and also see themselves how unreasonable the current copyright laws are, so perhaps more citizens could start demanding their representatives to start thinking about copyright reform or its total eradication... in my opinion copyright could be replaced by laws built on top of moral rights of authors where everyone is allowed to copy anything but only if the original author is prominently cited and credited. The more ordinary citizens get exposed to the silliness of copyright, the more they will demand changes from their governments.

Wikipedia could start doing that right now very easily. It just needs to remove the site-wide GFDL notice or add an "except where otherwise indicated" note after it, and then apply individual copyright notices on each article that is not GFDL. Of course, to maintain the freedom and the spirit of copyleft, Wikipedia and other wikis willing to use this approach could accept only a specific set of licences that meet certain criteria. For example, articles could be allowed to be either under the GFDL, the CC-By-SA, or the Free Art Licence, or any lcience in the spirit of DFSG, etc..

Re:They can't

Oh but contributors agreed to...

10. FUTURE REVISIONS OF THIS LICENSE

"[...] If the Document does not specify a version number of this License, you may choose any version ever published (not as a draft) by the Free Software Foundation."

That's why the FSF is involved.

Re:Wikipedia party

Jimbo is a party animal, that's for sure. The reputation of Wikipedians as boring people is undeserved.