Once a work is released into the public domain, it is essentially out there for anybody to use for whatever purpose – there’s no need to get permission, follow attribution guidelines, or pay royalties. The flip side is that the original author no longer has any rights over the released work – even if the work is resold and ends up making a fortune, for example, the original author can receive no recourse for it.

What, then, is the appeal of the public domain that results in so many people releasing works into the public domain in recent years? Part of the reason is similar to the reason why free licenses are so popular today: licenses are becoming increasingly more visible on the Internet, and furthermore, organizations such as the Creative Commons and the Free Software Foundation have devised ways in which licensing becomes no longer an issue exclusively for law professionals but for the average Internet user too.

What is the legal backing behind releasing works into the public domain? Basically, the public domain constitutes anything that isn’t copyrighted: works in this category include formerly copyrighted works that have passed their copyright expiration date (which depends entirely on the country), works that were created before copyright was created, and works that were never in copyright (certain government-created materials, and materials that the creator released into the public domain). It depends on country to country whether releasing one’s work into the public domain is actually codified or not. In many countries, including the United States, releasing works into the public domain itself is not codified, and is instead treated as an abandonment of property (since copyright is automatically assigned under the Berne Convention) under common law.

Public domain is basically the most radical copyleft license there is; the author reserves no right to it the licensed work, and licensees are free to do whatever they wish with the work, whether that involves plain copying, modification, or redistribution, for profit or otherwise. The number of Internet users subscribing to a copyleft approach as radical as this is still very low. This is completely logical, as users will feel less well compensated by a public domain work than, say, a work licensed under the only slightly less radical Creative Commons Attribution license.

Because of this, I do not feel that there is any specific need to press for authors to release their works into the public domain. It is far more important to encourage more individuals, and more organizations and companies, to use a license that offers real freedoms (as do the Creative Commons licenses), but still offer the copyright owner some protection.

Created and maintained by Richard Stallman and the Free Software Foundation, the same organization that maintains the GFDL, the GPL allows licensees to freely distribute copies and modifications provided a number of requirements are followed. One major restriction is that any derivatives or verbatim copies must be licensed under the GPL too (the GPL has a viral clause, similar in spirit to the GFDL). This means that companies cannot take free software, make a few modifications to it, and then re-release it as proprietary software. Another restriction is related to patents. The GPL forbids licensees from putting patents on GPL-licensed software, citing the fact that “any free program is threatened constantly by software patents.”

The last restriction, which is perhaps the most important, is that all licensees of the software must be given access to the source code. Licensees must either receive the source code along with the compiled version of the software, be able to request it and receive it from the developers for at least three years, or be given an Internet link where they can download the code (perhaps the most oft-used approach).

What’s interesting is the way in which the GPL makes itself enforceable. Because users of free software do not sign licenses (and often are not even given the opportunity to click the click-wrap “I agree” buttons that are common in proprietary software), it is very plausible for a user to say that they did not see the license when they started using the software. The GPL notes this, saying, “You are not required to accept this License, since you have not signed it.” (§ 5) However, it goes on to note “nothing else grants you permission to modify or distribute the Program” and that any modifications or redistribution imply that the licensee has read and accepted the license.

This is interesting in that it notes two layers of rights: the default layer, which is restricted by law, disallowing any modification or redistribution of the software, and the GPL layer, the one with many of those restrictions waived. The GPL layer comes into existence only when the licensee has read and accepted the license; therefore, if anybody claims not to have read the GPL but engages in any redistribution or modification that would have been allowed under the GPL, they would simply be breaking the copyright laws present in the default layer.

The GPL is not the only popular free software license. Many others exist, including the MIT License. The MIT License, developed at the Massachusetts Institute of Technology, is even “freer” than the GPL: it allows licensees to create derivatives and license them in any way, including for profit.

The choice of whether to use the GPL, or an even freer license, such as the MIT license, is not one to be taken lightly: if the developer creates a product licensed under the MIT license, and later decides to reuse some portion of the code in a commercial product, he or she risks having competitors who have the same code and who can also resell their derivatives if they want. With the GPL, this is avoided – all licensed copies cannot be sold for profit, but the developer still has the right to use the code in any way he or she wants.

That is the beauty of free software licenses: while giving users and other developers new freedoms that were unheard of back when proprietary software was the only paradigm, they still manage to protect developers and their right to be compensated for their hard work. Many developers see the beauty in this, and decide to license their software under a free license, bringing us today to a world in which almost anything – from enterprise-level database software to the core software that powers our phones – is available gratis.

Photos are no different from texts in that they have an automatic copyright; however, I think they’re also different in that very good photographs can often be sold for a price (unlike, say, a contribution to a blog or an article on Wikipedia). This means that photographers are sometimes reluctant to give away rights to their photos as easily as their text contributions.

Despite this, however, there are a huge number of images on Flickr and the Wikimedia Commons that are licensed under lenient free licenses. In the case of the Wikimedia Commons, a free license is required — all media that are not licensed under the Creative Commons, GFDL, or a similar free license are deleted. Still, it’s amazing that so many people have licensed their photos and other artistic creations under free licenses that may even allow others sell their work for profit.

Part of the reason, I think, is the relative ease-of-use of the licenses, as I mentioned in a prior post. But some of it is thanks to these websites, who have made it incredibly easy for users to license their contributions under a free license. Users uploading images to the Wikimedia Commons are presented with a selection box that allows them to select a license of their choice; on Flickr, while the default license is “All rights reserved” (i.e., no free license at all), it is a matter of a few clicks to go into the preferences screen and change the default license for all past and/or future uploads.

Flickr default license selection screen:

Wikimedia Commons upload interface with license selection:

Compared to the relative “automaticness” of Wikipedia’s GFDL licensing, both of these websites make the user explicitly choose to license their media under a free license. Even then, freely licensed media are available in abundance — all this can only mean that the free licensing movement has caught on among the regular Internet crowd, even among those who would not have been interested in licenses before.

All of this is not without its downsides, though. There will always be companies that take such freely licensed media from the Internet (much of which is of sufficient quality to be used in professional settings) and use it for their own profit. While this is explicitly allowed under the free licenses that many of these media are licensed under, it has and will most definitely continue to generate controversy when it happens.

One particular example is of Virgin Mobile and Flickr. In 2007, Virgin Mobile used several freely licensed (specifically, CC-BY licensed) images from Flickr in their advertising campaign. Virgin duly attributed the images, printing in fine text the URLs of the Flickr profiles of the original photographers. So far, so good. The problem was with one of their posters that used a freely licensed image with a picture of a girl, Alison Chang, in it.

The problem ultimately was that although Flickr had every right to use the photo itself in their advertising campaign (as they had attributed it), they did not have the right to use Chang’s likeness in their campaign without a model release. Even worse, the poster’s tagline, “Dump Your Pen Friend,” is clearly derogatory and insulting to Chang.

In the end, Chang and the photographer sued Virgin Mobile, and surprisingly, the Creative Commons organization because “as the creator of [the CC-BY license], they have an obligation to define it succinctly.” It will be interesting to see what happens with this lawsuit. While it seems that Virgin Mobile appears to have committed some privacy violations (wholly unrelated to copyright), the plight of Creative Commons seems to be less easy to predict. I would guess that Creative Commons has no obligation to specify each and everything that licensees may do with CC-licensed content. Considering that CC is a free license that deals with copyright issues, those using the licenses should consider other issues, such as privacy and using the likeness of people, in parallel with copyright. Creative Commons should not be held liable for organizations and other entities that fail to do their due diligence in this regard.

While this plagiarism from Wikipedia may seem like child’s play, serious legal issues also play a part. As I’ve covered in this blog, Wikipedia content is not released into the public domain — by default, all content is licensed under the GNU Free Documentation License, and the copyright remains under the ownership of each contributor.

This is further complicated by the fact that some contributors choose to unilaterally license their contributions under a freer license, such as into the public domain. Some users have also chosen to “dual-license” their contributions, under, say, the GFDL and the CC-BY-SA. These license indications are not shown on the contributions themselves; each user who decides to do this usually places this information on their “user page.”

There have been a number of cases in which individuals or groups have lifted Wikipedia content without attribution of any kind, essentially committing copyright infringement. One recent case was the case of George Orwel (sic.), a New York-based reporter who pasted a substantial amount of content from a Wikipedia article into his book, “Black Gold: The New Frontier in Oil for Investors.”

What complicated this matter was that the user who contributed a large portion of the text, Ydorb, had licensed his contributions into the public domain. Therefore, it would seem that this act of “plagiarism” was, in fact, lawful (albeit troubling coming from an experienced reporter). However, the nature of Wikipedia’s collaborative and cumulative editing meant that portions that had been contributed by other users, who unlike Ydorb had not released their contributions to the public domain, were still under their copyright. This is an issue present in all wikis — because content is not simply written by one user and left at that, the ownership of specific bits of text, rather than entire articles, becomes important.

While it may be easy for people (even reporters) to become complacent about Wikipedia copyright, the bottom line is that every contribution to Wikipedia is still covered by the same copyright that automatically covers any other creative work. A free license is not a license to do whatever you want; there are certain bounds, and users need to be educated about this.

Immediate reactions on the Wikimedia Foundation’s community mailing list, foundation-l, seemed positive overall but rather confused. Users wondered whether a license like CC-BY-SA and the GFDL (both of which have a “Share-Alike” clause, which essentially says that copies and derivatives must be made available under an identical license), could be made compatible, and whether a license switch would be legally enforceable. The latter is not an issue, as both the CC-BY-SA and GFDL contain clauses permitting the controlling bodies to upgrade the licenses without notifying licensors. There is precedent for this; for example, the Electronic Frontier Foundation’s now-defunct Open Audio License designates version 2 of itself as the CC-BY-SA license.

Members of the community were further confused by the inherent compatibility of the two licenses; the GFDL allows invariant sections, for example, which, they argued, is against the spirit of the CC-BY-SA license. This hasn’t yet been adequately discussed — even if it is legally possible to create a variant middle-ground of the two licenses that is compatible with older versions of both, that risks estranging content owners who published their content under a previous license. (Those using the invariant clause of the GFDL, for example, may not appreciate having even those sections licensed freely under a CC-BY-SA-like license.)

Since the actual process of conversion is taking place within the Free Software Foundation and Creative Commons, it’s not immediately clear to the average Wikimedian what is going where. My personal hope is that the two organizations will be able to come up with a new license that legally transitions existing GFDL (with no invariant sections) materials to CC-BY-SA in the not-so-far future. I also hope that people who have contributed content to Wikipedia in the past will be happy with the change; this should not be too much of a problem, since (1) most people do not really care about licenses and (2) the people who do care, or at least those who have voiced their opinion on foundation-l, appear to be in favor of a change.

There are some outstanding issues that have yet to be resolved though — those detailed above, and some issues to do with the Creative Commons organization itself. More than one longstanding Wikimedian has expressed his displeasure at the organization, calling it an organization whose purpose is for lawyer collaboration and self-promotion more than for the spread of free content. While I cannot tell if this is true, having no personal dealings with the organization, Creative Commons is rather successful with its public relations, so it’s not hard to imagine them worrying a lot about their public image. However, the overall effect of the Creative Commons movement, it seems, is overwhelmingly positive, and Wikipedia’s joining forces with them seems like a promising move (and as some members of the Wikimedia community mentioned, it also means that the Wikimedia Foundation will have a lot of leverage over the organization).

The Creative Commons licenses allow a great deal of freedom to modify, reuse, and redistribute content. Content creators can put a number of restrictions, or a combination on them, on their licenses. The restrictions are:

• Attribution (BY) – Any copies must cite the original author
• Share-Alike (SA) – Any copies must be shared under the same license as the original
• No Derivatives (ND) – No derivatives must be made; all redistributed copies must be verbatim copies of the original
• Non-Commercial (NC) – Derivatives or copies must not be made for commercial purposes

The entire number of combinations is 2^4 = 16; however, since the Share-Alike and No-Derivatives flags are mutually exclusive, four of those are invalid, and the last combination, no restrictions at all, is tantamount to releasing your content under the public domain, so is not a valid Creative Commons license. Also, licenses that do not have the Attribution clause used to exist, but have been phased out due to lack of demand.

Using these combinations, one can create the eleven Creative Commons licenses, which are named after their abbreviated restrictions: CC-BY-SA 3.0, for example. (3.0 is the version of the Creative Commons license; version 3.0 is the latest as of now.) Incidentally, this blog is licensed under the CC-BY-SA 3.0 license. That is, any derivatives or redistributions must be licensed under the same CC-BY-SA license, and must be attributed back to me.

The sheer power of the Creative Commons license comes partly from this — the ease of use with which content owners can license their content. You don’t need a law degree when licensing your content under these licenses; Creative Commons provides a clean, easy-to-use interface on their website that allows you to pick and match licenses according to your needs, and provides “human-readable” license deeds detailing, in very plain and simple language, the freedoms accorded by each Creative Commons license.

The Creative Commons licenses also have very high visibility on the Internet today. Not only are their icons professional-looking and recognizable from a distance, the sheer number of websites and services using their licenses is staggering. Some major users of Creative Commons include Flickr, the Wikimedia Commons, MIT OpenCourseWare, and soon Wikipedia. Furthermore, many service providers set Creative Commons as their default license (Blogger and even the Berkman Center’s blogging platform), coercing users who otherwise do not know or care about licenses to license their content under a Creative Commons license.

The advent of Creative Commons licenses is a welcome development for the open source community. More and more people are being introduced to the free licensing movement, and even large companies (including Microsoft with parts of its MSDN specifications and its MSDN Wiki) have begun to embrace this as well, further escalating this move. It’s reassuring to have one license framework that everybody can rely on and trust, and the stability and expertise of the Creative Commons organization can only help.

First, how exactly is Wikipedia content licensed? Here is the official license text from the Copyrights page:

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, with no Front-Cover Texts, and with no Back-Cover Texts.

As you can see, Wikipedia does not use the controversial Invariant Sections clause that the GFDL permits — that means there are no pages on Wikipedia detailing its philosophy that must be kept in derivatives (such as other websites that reuse Wikipedia content).

What does Wikipedia’s license mean for the average contributor? When you goto a Wikipedia page, hit “Edit this page”, make your changes, then finally hit “Save page”, you are licensing your contributions under the GFDL. This is because the edit page, similar in spirit to “clickwrap” licenses found on the Internet and in software, contains the text, “You agree to license your contributions under the GFDL” along with a footer with the text quoted above.

Now I expect that the average visitor to Wikipedia who makes a quick contribution (such as a spelling fix or changes some figures here or there) does not realize this fact. This might be a little problematic for Wikipedia, as there is no “I agree” button, or any other user interface element (such as a checkbox) that forces the user to agree to this license. In the past, the courts have found this to be a problem — in Specht v. Netscape Communications Corp. (2001), the court ruled that the Netscape license was not enforceable as there was no explicit “I agree” button for the user to click before downloading the software. The changes needed to ensure Wikipedia’s safety might be minimal — simply by preceding the text with ‘By clicking “save page” below, you…,’ for example. It would be interesting to see whether this is actually a problem legally.

What, then, does the GFDL mean for the established Wikipedian? Firstly, it means that there’s a slight learning curve — getting accustomed to the GFDL and the various restrictions and freedoms it poses is something that can only be learned through reading through the rather thorough documentation on Wikipedia itself. Second, it means that a lot of care must be taken to preserve edit histories.

Edit histories are the information you get when you click on the “History” tab of an article on Wikipedia; basic information about each revision to an article, such as the user name (or IP address, in case of anonymous contributors) and date and time of each revision, are presented. This list is necessary for copyright purposes; because Wikipedia articles are not copyrighted by the Wikimedia Foundation, or any other single entity, but by its contributors (due to standard Berne Convention rules), the list of contributors must be preserved throughout revisions in case somebody wishes to relicense an article (although this is unheard of as of yet) or more frequently, an article needs to be cited.

The preservation of edit histories is usually carried out automatically by the MediaWiki software; however, when merging two pages into one, for example, the origins of each portion of text must be noted on the attached discussion page of the resulting article.

Of course, this preservation of edit histories is not a burden associated exclusively with the GFDL; any Creative Commons license requiring attribution would also have the same burdens.

It turns out that the real restrictions of the GFDL (as used by Wikipedia, without the invariant sections clause) are only in the redistribution stage, the requirement to include the full text of the GNU Free Documentation License along with redistributed copies being the main one. Even if this is the only problem, it is a good enough reason to switch Wikipedia’s license — switching to a Creative Commons license will greatly increase the utility of Wikipedia content.

This doesn’t come as much of a surprise — Citizendium has some articles that are forks (copies that are subsequently modified) of their counterparts on Wikipedia, which is licensed under the GFDL. Since the Wikimedia Foundation has announced plans to work with the Free Software Foundation and the Creative Commons to make the GFDL compatible with the CC-BY-SA license, it’s only natural that Citizendium should go ahead and make their content available under the Creative Commons license. (For the time being, though, they are required to keep the originally Wikipedia-based articles licensed under the GFDL due to the viral provisions of the license.)

Apparently, however, there was a huge discussion among Citizendium editors on whether to use a non-commercial license (in this case it would be CC-BY-SA-NC) or not. The discussion (available in paraphrased form on Citizendium) seems to revolve around the practical reasons to choose one license over the other, such as distribution to those without Internet access.

What’s notable is that one of the major reasons against the GFDL/CC-BY-SA seems to be “we don’t want Wikipedia to reuse our articles,” along with a general feeling of hostility towards allowing other websites to redistribute Citizendium content. This seems to say a lot about Citizendium and its real thoughts on Open Source and free licensing — are they really out there to share a good encyclopedia, or are they just doing it to spite Wikipedia?

While I applaud Citizendium editors for ultimately choosing the CC-BY-SA approach, and thus allowing greater redistribution and reuse of its content, there seems to be a troubling sense of exclusivity among the Citizendium community. I can’t help but compare this with the Wikipedia community, whose foundations are built on selflessness and creating an encyclopedia for the (actual) better good.

The GFDL was originally designed for licensing software documentation for free software products. As the Preamble states, the license was developed “because free software needs free documentation: a free program should come with manuals providing the same freedoms that the software does.” However, the license goes on to state that the license can be used for other, more general documents (such as Wikipedia).

In a nutshell, the GFDL offers licensees the ability to “copy and distribute the [work] in any medium, either commercially or noncommercially,” and allows licensees to modify that work and redistribute it — provided that the derivative work be made available under the same terms as the original. The GFDL shares this viral quality with the GNU GPL, the other major license that the Free Software Foundation provides. This locks modifiers and distributors into the same licensing scheme; you can’t, for example, take a GFDL-licensed work, make a derivative, and release it for non-commercial use only.

One controversial aspect of the GFDL is that it provides provisions for “Secondary Sections” (basically, any sections in the work that describes the author or publisher’s connection to the work). These sections may not be modified in any resulting derivatives, and are called invariant sections. (If they are modified, the resulting copies must be redistributed under a different title). Secondary sections usually include documents detailing the philosophy of the original author (in the GFDL-licensed Emacs manual, for example, it is a lengthy section in the form of a Manifesto). Apparently this was written into the GFDL so that subsequent authors of documentation, who were sometimes in violation of the original philosophies as stipulated by the original author, would not be able to change the original manifestations thereof.

Another more practical problem of the GFDL is its full license text provision. Any distribution of GFDL-licensed material is required to include a copyright notice as well as the full text of the GNU Free Documentation License, which, when printed on US Letter size paper, is almost 8 pages long. Not only is this a problem, for example, when reprinting a GFDL work in a book, it’s an even bigger problem when all you need to do is redistribute one small image for inclusion in, say, a magazine.

Although it’s still rather rare for GFDL-licensed content to appear in print, the full license text provision of the GFDL has already led to some unwieldiness. For example, on Wikibooks, a Wikipedia sister project that collaboratively writes GFDL-licensed books instead of an encyclopedia, completed books are often distributed in PDF form. PDF such as these come with an 8-page appendix with the GFDL license text, complete with a paragraph on “How to use this License for your documents.”

While this is all well and good for spreading the GFDL, it leads to wasted paper and frankly, it looks rather unprofessional to have all this stuck to the back of a book. The GFDL still has a sense of the Unix geek ethos — the war they’re fighting is more guerilla than anything professional.

There are, however, alternatives to the GFDL that provide much of its liberties as well as its protections. A very good example of this is the Creative Commons licenses, specifically the Creative Commons Attribution Share-Alike (CC-BY-SA) license. In the next post I’ll cover Wikipedia and its relationship to the GFDL, and then go on to cover the Creative Commons licenses.

The blog post linked above is sort of inaccurate, calling it a “switch” to Creative Commons, but the real deal can be found in the official resolution from the Wikimedia Foundation.

This comes as a welcome development for Wikipedia – the GFDL has always been a dark, kept-under-the-rug feature of Wikipedia, as it has caused some well-known inconveniences to the Wikipedia community and those wanting to reuse Wikipedia content. (I’ll cover the details in a later post.)

So what exactly is the difference between the GFDL and the Creative Commons Attribution Share-Alike (CC-BY-SA) license? We’ll cover that next time.