WordPress Themes, GPL, and Copyright Case Law

Filed in Web DevelopmentTags: , , , , ,

Within the WordPress community, the question of GPL inheritance of WordPress themes erupts into contentious debate with the reliability - if not the frequency - of Old Faithful. While I understand that, according to the GPL interpretation of Matt Mullenweg, the Free Software Foundation (FSF), and the Software Freedom Law Center (SFLC), WordPress themes are derivative of WordPress and therefore must necessarily inherit WordPress' GPL, I would like to investigate the issue not in light of their interpretation but rather in light of copyright law and precedent case law.

Before I begin, let me add an important caveat: I have no qualms with the GPL. I have always released - and will continue to release - under GPL anything I develop related to WordPress. I do so because I choose to do so, as a means of making even a minor contribution to a project from which I believe I have personally benefited. I do have issues with how the GPL-inheritance question has been handled - but those issues are out-of-scope for this post.

Having (hopefully) made that point clear, let's begin!

What US Copyright Law Says

US Copyright law defines a "derivative work" as such:

A “derivative work” is a work based upon one or more preexisting works, such as a translation, musical arrangement, dramatization, fictionalization, motion picture version, sound recording, art reproduction, abridgment, condensation, or any other form in which a work may be recast, transformed, or adapted. A work consisting of editorial revisions, annotations, elaborations, or other modifications, which, as a whole, represent an original work of authorship, is a “derivative work”.

Note the key adjectives: recast, transformed, and adapted.

Consider also Section 102(b), which states:

In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.

This clause establishes the boundary around copyright between copyrightable expression, and non-copyrightable ideas.

Summarizing GPL Inheritance Requirements

To summarize GPL requirements regarding license inheritance for derivative works 1:

  1. The GPL only applies to distribution of a (modified or unmodified) GPL-licensed work, or a derivative work. Any activity involving use, modification, or creation of derivative works that does not involve distribution is outside of the scope of the GPL.
  2. Distribution of a (modified or unmodified) GPL-licensed work, or a derivative work, requires that such distribution be licensed under GPL.

The GPL is what is now referred to as a "copyleft" license: a modified public-domain license that takes advantage of the exclusive rights granted by copyright law to prevent derivative works from being restrictively licensed. Since the copyright owner has exclusive right to produce and to distribute derivative works based on the copyrighted work, the GPL intends to grant unlimited usage rights (to use, study, modify, etc.) to the end-user, while forcing follow-on developers of derivative works to release those works under the same license.

It is important to understand that, because the GPL explicitly defines any activity not involving distribution to be out of the scope of the license, and since right of distribution is solely derived from copyright law, that GPL derives its legal basis from copyright law alone. This distinction separates the GPL from most other traditional software licenses, which derive their basis for usage and modification restrictions not from copyright law, but from contract law.

Notes:

  1. WordPress is released under GPL version 2.0. I'll try to summarize below the parts of the license germane to derivative works.

    First, from the Preamble:

    The reason we have a separate public license for some libraries is that they blur the distinction we usually make between modifying or adding to a program and simply using it. Linking a program with a library, without changing the library, is in some sense simply using the library, and is analogous to running a utility program or application program. However, in a textual and legal sense, the linked executable is a combined work, a derivative of the original library, and the ordinary General Public License treats it as such.

    ...

    The precise terms and conditions for copying, distribution and modification follow. Pay close attention to the difference between a "work based on the library" and a "work that uses the library". The former contains code derived from the library, while the latter only works together with the library.

    Terms and Conditions, Clause 0:

    The "Library", below, refers to any such software library or work which has been distributed under these terms. A "work based on the Library" means either the Library or any derivative work under copyright law: that is to say, a work containing the Library or a portion of it, either verbatim or with modifications and/or translated straightforwardly into another language. (Hereinafter, translation is included without limitation in the term "modification".)

    Terms and Conditions, Clause 2:

    These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Library, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Library, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

    Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Library.

    Terms and Conditions, Clause 5:

    A program that contains no derivative of any portion of the Library, but is designed to work with the Library by being compiled or linked with it, is called a "work that uses the Library". Such a work, in isolation, is not a derivative work of the Library, and therefore falls outside the scope of this License.

    However, linking a "work that uses the Library" with the Library creates an executable that is a derivative of the Library (because it contains portions of the Library), rather than a "work that uses the library". The executable is therefore covered by this License. Section 6 states terms for distribution of such executables.

Feedback

Comments (Comments are closed)

77 Responses to “WordPress Themes, GPL, and Copyright Case Law”
  1. chip_bennett says:

    WordPress Themes, GPL, and Copyright Case Law – http://www.chipbennett.net/2010/07/wordp

  2. Like I said earlier. Excellent just excellent. or was it awesome? =)
    You are an excellent writer I must say.

    This really is how one must approach this subject. Find case law, talk about how things interact and why different cases is pro or against a certain interpretation.

    Now I’m awaiting an as detailed response from the themes/plugins are GPL camp. They really need to step up since none of the themes are GPL articles/posts comes even close to the depth of this article.
    From what I’ve understood the WP core crew are going to present their view on things on WP.org. I hope their pro-GPL article is as well researched as this one and includes references etc to cases that is in favour of their interpretation. Given that the only source of their arguments for at the moment is an biased response from SFLC the really need more unbiased sources.

  3. Mark Jaquith says:

    A work consisting of editorial revisions, annotations, elaborations, or other modifications, which, as a whole, represent an original work of authorship, is a “derivative work”.

    I’d absolute describe a WordPress code extensions (themes or plugins), as “elaborations, or other modifications.” Themes necessarily have WordPress code in them. They necessarily use code from the original (and GPL’d) WordPress themes. Plugins and themes absolutely “transform” and “recast” the WordPress application. That’s what makes them so great!

    In the end, neither of us are lawyers, and even if we were, this is new legal territory. Actually, it’s rather exciting to be involved in a passionate debate over something new like this that could have such broad-reaching implications, wouldn’t you say?

    I regard a WordPress extension, as an “elaboration” of WordPress code. You may consider it to be a separate work that is not at all “based upon” WordPress. It doesn’t sound like either of is coming closer to changing the other’s mind. We’ll see how it shakes out, I guess!

    P.S. how is this post shorter than the comments you left on mine? 😉 I kid, I kid.

    P.P.S. No website left because your OpenID functionality was just hanging when I tried.

  4. Mark Jaquith says:

    Totally missed the other pages. I take back my comment about brevity!

  5. Chip Bennett says:

    Mark,

    Thanks for stopping by and taking the time to read and comment. I agree that it is indeed exciting to be involved in such a passionate debate – though I appreciate that we can discuss such controversial and potentially incendiary topics as dispassionately as possible.

    As I said up front: I have no problems with the GPL, and appreciate the benefits it has provided to me. But, I also see a community benefit to discovering the truth, which I believe will cut through the emotional and sometimes propagandistic rhetoric (from both sides of the debate), so that it can cease being a source of division for the community.

    What I specifically wanted to look at with this post is not the logical or common-sense arguments, but what the courts have had to say on the matter. With something like copyright, in a common-law jurisdiction like the U.S., the law is written fairly open-ended, and the specifics are defined by court decisions. So, we can debate the logical and common-sense arguments into eternity, but what really matters is the reasoning behind the various court decisions involving copyright – and especially those involving software copyright.

    And, as I said in the post: I don’t see copyright case law supporting the argument that themes are derivative of WordPress because they are “elaborations” or “modifications” of WordPress. I’d love to hear another interpretation of this or other copyright case law, that would support such an argument.

    I agree that neither of us is a lawyer, and I would certainly welcome a lawyer explaining how I am misinterpreting the cases I’ve cited (or, for that matter, backing me up on that interpretation!). Regardless, I find it interesting – and instructive – for us to discuss these decisions, even as laymen.

    (And I do apologize for getting overly wordy in my post and in my comments.) 😉

    p.s. thanks for the OpenID heads-up. I’ve not checked in on it for a while. I’ll go do some investigation, and figure out what’s wrong.

  6. Chip Bennett says:

    @Mark:

    Totally missed the other pages.

    Are the footnotes obscuring the pagination? I’m still experimenting with both.

  7. Ryan says:

    Hell of an article Chip!

    Can’t wait to see if there’s an outcome on this topic from the potentially upcoming Matt vs Chris legal battle.

  8. Doug Stewart says:

    Chip:
    As I said on Twitter — well done.

    While I, too, am sympathetic to the GPL side of things, the more I read, the less convinced I am that the Foundation, Matt, et al. would prevail if they take Chris up on his request to be sued. There are simply too many factors, including the necessity of a technically-competent jury, to declare a potential suit as an easy win and the substantive downsides to a loss are indeed severe.

  9. Mark Jaquith says:

    Trying my comment again, now that I’ve seen pages 2 through 4.

    A work consisting of editorial revisions, annotations, elaborations, or other modifications, which, as a whole, represent an original work of authorship, is a “derivative work”.

    I’d absolute describe a WordPress code extensions (themes or plugins), as “elaborations, or other modifications.” Themes necessarily have WordPress code in them. They necessarily use code from the original (and GPL’d) WordPress themes. Plugins and themes absolutely “transform” and “recast” the WordPress application. That’s what makes them so great!

    WordPress is non-compiled code; therefore by definition, no static linking exists between WordPress core and Themes.

    Why the distinction on static linking? PHP is compiled into bytecode on the fly (or, not on the fly in the case of using a PHP opcode cache). PHP code can also be compiled. Why should that make a difference? I don’t think the compiled or non-compiled nature of the code, or the file in which code is contained, makes it any more or less separate, as they all are all one as one application. Not WordPress and a Theme communicating or interoperating, but running together as one functional unit (etc, you’ve read my post).

    The Nintendo court distinguished Midway by pointing out that the circuit board distributed in Midway substantially copied and replaced a portion of the original video game circuitry, whereas the Nintendo add-on did not actually incorporate any of Nintendo’s protected expression.

    The add-on was separate hardware that allowed the user to change three 6- or 8-byte values. It’s hardly on the level of a theme or a plugin which can change entire swaths of code, add new functionality, or disable existing functionality. In terms of modification, the Game Genie was about the equivalent of an air horn and “timing codes” so social conservatives can bleep out or distort swears in songs. And it’d only work three times a song, so it’d be useless against Kid Rock (ZING).

    Also, Galoob appears to refute the argument that Themes are derivative of WordPress due to their dependence upon WordPress core, their design to interact with WordPress core in the same way that WordPress interacts with itself

    It was a 24 byte one-way filter, completely external to the game. It didn’t even have the ability to add or remove code — only flip one of 192 light switches into a different position. It’s a far more subtle form of modification than WordPress extensions have.

    The Sega court held that a copyright owner could not erect barriers intended to prevent interoperability with follow-on works, and that copying of code required to accomplish that interoperability, in the expression of a separate, copyrightable expression, did not constitute copyright infringement.

    From what I can tell, Accolade did not substantially incorporate that code into their work. Sega was saying that Accolade privately converting decompiling their code was infringement, even though they never distributed it (at least, not substantially). That’s not what we’re saying.

    This incorporation must involve the actual creative expression of the protected work, not merely the mechanisms of interaction, which are non-copyrightable ideas.

    To be clear, expression means code, in this case. The implementation of an idea. So the idea of filters is not copyrightable. WordPress’ filter implementation is.

    In order to be derivative, a work must bear substantial similarity to a protected work, based on an Abstraction-Filtration-Comparison test.

    I think Thesis would fail this test (that is, it would bear substantial similarity). There may be a gray area where themes that don’t do much in their code and just display a list of posts have less similarity than themes which integrate themselves more tightly into WordPress code. We’d have to see how it went in court.

    Further, as Sega and Sony demonstrate, even a Theme’s use of WordPress core function/global variable calls and filter/action hooks merely constitutes incorporation of non-copyrightable mechanisms of interoperability.

    This is the crux of our disagreement. I do not consider them mechanisms of interoperability. They are mechanisms of changing WordPress. They are WordPress itself! WordPress has external APIs for interoperability. RSS, Atom, AtomPub, XML-RPC. If you want to look at the Sega case, having someone look at WordPress source code in order to figure how to interoperate with those APIs wouldn’t violate the WordPress copyright.

    Theme PHP files generally describe a semantic HTML markup. This semantic markup on its own is likely not copyrightable, as it falls under merger and/or scenes a faire principles.

    The HTML is a distraction. That’s not really the issue. It’s the PHP code.

    Even the basic PHP in Themes very possibly constitutes non-copyrightable expression of ideas, processes, or methods of operation – and as such is not covered by copyright law.

    Again, there may be gray area here where “basic PHP” is evaluated by courts differently than something like Thesis. Coincidentally, and tangentially, the integration with WordPress is roughly proportional to the benefit that would come from it being freely licensed, and as well proportional to the extent to which an anti-GPL theme developer might feel inclined to give a restrictive user license. No one would sue because of a foreach loop. 🙂

    Anyway, thanks for writing this. It seems very much in the spirit of my post (the difference being that my post being more of a technical analysis, this being more of a legal exploration). It’ll take a court case to look at both technical and legal aspects and give us a conclusive answer. It was interesting reading all these other cases, but I think what they make most clear is that copyright in the digital age is complicated, and there is still a lot of legal ground yet to plow.

    I really do have to disengage from this whole debate. I just got in trouble with the Mrs. for spending so much time on it! I’m going to go back to coding with WordPress and releasing everything I distribute under a non-restrictive license. It’s easy to forget that beneath the bickering and the mountain of legal and technical minutia, there’s an amazing and fun publishing platform that was downloaded 10 million times in the last month! 🙂

  10. Sean Kaye says:

    Great piece! Incredibly well researched and thought out.

    I agree with you that Themes are not derivatives and are effectively a working example of “fair use”.

    One thing I am struggling to get a grip on and your opinion would be interesting is, what exactly is WordPress protecting under the GPL? What is the copyrightable material?

    For example, the functions and things to make database calls or do screen layouts are not necessarily able to be covered and the output content belongs to the users. Most of what WordPress does is to shunt data to and from a database or organise the user’s data for presentation.

    So I ask, what exactly is covered?

    I also think your analysis on “loss” is excellent. As I’ve said on Twitter, themes are additive in their impact to the product so there really can’t be “loss” and therefore no damages. Would you agree?

    Excellent stuff.

  11. Chip Bennett says:

    @Doug Stewart:

    I think, even more so now, that Thesis is a special case, due to the extent of copying of core WordPress code. (Note: I don’t know what that extent is alleged to be, fully.) Interestingly, the Altai decision certainly implies that Thesis can be held accountable for infringement for version X.Y, even while version X.Y+1, which presumably has all core WordPress “copy pasta” removed, is held not to be infringing.

    But, I don’t think even Thesis would be a “slam dunk” case, because I don’t know what result the court would get from an Abstraction-Filtration-Comparison test. It is entirely conceivable that the court could rule either way (again: not knowing the full extent of what was copied from core into Thesis).

    If Thesis were ruled to be infringing, I don’t think the ruling would have that much reach into other Themes. But, if Thesis were ruled not to be infringing, the results would be devastating for advocates of GPL inheritance.

  12. Chip Bennett says:

    @Mark Jaquith:

    I appreciate your (as usual) thoughtful reply. I’ll incorporate your thoughts into the response it deserves, as free time, and its appropriate use, allow.

  13. Chip Bennett says:

    @Sean Kaye:

    Great piece! Incredibly well researched and thought out.

    I agree with you that Themes are not derivatives and are effectively a working example of “fair use”.

    Note: I believe themes are not inherently derivative. Unless and until a court actually considers a specific Theme, it will be difficult to say anything more definitively.

    Thesis is a good example: a court will have to determine if the “copy pasta” code constitutes infringement of copyrightable expression from WordPress core. I couldn’t even hazard a guess as to how that would turn out, because such a suit would be unprecedented, and because I’ve not seen the exact details.

    One thing I am struggling to get a grip on and your opinion would be interesting is, what exactly is WordPress protecting under the GPL? What is the copyrightable material?

    For example, the functions and things to make database calls or do screen layouts are not necessarily able to be covered and the output content belongs to the users. Most of what WordPress does is to shunt data to and from a database or organise the user’s data for presentation.

    So I ask, what exactly is covered?

    Here’s one quick-and-dirty example that I thought up: Widgets.

    The idea of Widgets isn’t copyrightable, and likely, WordPress’ specific means of implementation of Widgets isn’t, either. However, WordPress’ expression of Widgets – i.e. the specific Widgets defined in core – would likely be copyrightable.

    Another example: Comments.

    The idea of Comments isn’t copyrightable, and, likely, WordPress’ specific means of implementation of Comments isn’t, either. However, WordPress’ expression of Comments – i.e. the code defined by wp_list_comments() function call – would likely be copyrightable.

    So, a court probably wouldn’t consider using WordPress’ defined mechanism of filtering wp_list_comments() to constitute a derivative work; however, a court very likely would consider the wholesale incorporation of wp_list_comments() to be derivative.

    I also think your analysis on “loss” is excellent. As I’ve said on Twitter, themes are additive in their impact to the product so there really can’t be “loss” and therefore no damages. Would you agree?

    In the post, I didn’t get into any of the relief specified by copyright law. It is likely a topic for its own post (if necessary). But, in brief: relief in a copyright infringement case isn’t necessarily dependent upon loss or damage incurred by the owner of the infringed work. The Act specifies statutory damages based solely upon the severity of the infringement, and that can be awarded at the court’s discretion. So, Thesis could be subjected to statutory fines, regardless of any loss incurred (or not) by WPF.

    Excellent stuff.

    And thanks for reading!

  14. chip_bennett says:

    @MikeWas I referenced your WordPress Theme/GPL derivative analysis post here: http://www.chipbennett.net/2010/07/20/wo… (would love your input)

  15. RJack says:

    Your “not a contract” claim is erroneous:

    “It is important to understand that, because the GPL explicitly
    defines any activity not involving distribution to be out of the
    scope of the license, and since right of distribution is solely
    derived from copyright law, that GPL derives its legal basis from
    copyright law alone. This distinction separates the GPL from most
    other traditional software licenses, which derive their basis for
    usage and modification restrictions not from copyright law, but
    from contract law.”

    The principle that a license is a contract was established by the
    Supreme Court in DE FOREST RADIO TEL. CO. V. UNITED STATES, 273 U.S.
    236 (1927). http://supreme.justia.com/us/273/236/case.html

    “No formal granting of a license is necessary in order to give it
    effect. Any language used by the owner of the patent or any conduct
    on his part exhibited to another, from which that other may properly
    infer that the owner consents to his use of the patent in making or
    using it, or selling it, upon which the other acts, constitutes a
    license, and a defense to an action for a tort. Whether this constitutes
    a gratuitous license or one for a reasonable compensation must, of
    course, depend upon the circumstances; but the relation between the
    parties thereafter in respect of any suit brought must be held to be
    contractual, and not an unlawful invasion of the rights of the owner”

    This contract principle applies specifically to copyright licenses:

    “Although the United States Copyright Act, 17 U.S.C. 101 – 1332,
    grants exclusive jurisdiction for infringement claims to the
    federal courts, those courts construe copyrights as contracts and
    turn to the relevant state law to interpret them.”; Automation by
    Design, Inc. v. Raybestos Products Co., 463 F.3d 749, (United
    States Court of Appeals for the Seventh Circuit 2006).
    http://ftp.resource.org/courts.gov/c/F3/463/463.F3d.749.05-1172.html

    No federal court in the past eighty-three years has ever ruled that a
    copyright license is anything other than a contract. Eben Moglen’s
    assertion that “licenses are not contracts” is simply untrue under U.S
    law. A copyright license is either an enforceable contract or it’s not
    a valid license.

  16. Chip Bennett says:

    @RJack,

    I actually never said that the GPL isn’t a contract. To claim I made a “not a contract” claim is a mis-characterization of my point, which was that, in granting an implied license of unrestricted use (modifying, copying, et al) except in the case of distribution, the GPL eschews all protections granted by contract law outside of copyright law.

    The GPL is basically a modified Public Domain license, that grants full usage rights to the end user, and only imposes the restriction that any distribution be released under the same license. Explicitly stating everything other than distribution as out of scope causes the GPL to derive its enforceability solely from copyright law, rather than contract law in general.

    Hopefully that clears up the mis-conception.

  17. I’d absolute describe a WordPress code extensions (themes or plugins), as “elaborations, or other modifications.”

    WP extension does not elaborate or modify WordPress it self. Only the information created by the end user. No code is ever added, altered or removed from WP core.

    Themes necessarily have WordPress code in them. They necessarily use code from the original (and GPL’d) WordPress themes.

    Using code would mean for loops, if statements etc. They don’t necessarily use actual code from original themes. Then comes the question if traversing an array is copyrightable. Also ain’t there fair use principle. If the application is meant to be extended and you have to study its code to extend it sounds odd that use of the codelayout and function names would be enough to inherit GPL.

    Plugins and themes absolutely “transform” and “recast” the WordPress application. That’s what makes them so great!

    Thing here is that they don’t recast and transform WordPress application. They transform the information inputted by the end user. Themes/Plugins are also loaded by WP. Not the other way around.

    Why the distinction on static linking?

    Dynamic linking is considered by many to be allowed. Not by the biased FSF how ever.
    GPL#Linking and derived works

    PHP is compiled into bytecode on the fly (or, not on the fly in the case of using a PHP opcode cache). PHP code can also be compiled. Why should that make a difference?

    It is a matter of when and how it happens. It is done by the end user, not by the developer/distributor. The end user is the one that creates a derivative work by your own admission but it is not distributed so the GPL does not come into effect.

    The add-on was separate hardware that allowed the user to change three 6- or 8-byte values. It’s hardly on the level of a theme or a plugin which can change entire swaths of code, add new functionality, or disable existing functionality.

    Show me one piece of theme/plugin code that alters the actual WP core code and not what the functions return? Theme/plugins uses the turn on/off switches that is already in WP core code. They does not alter any code to make it work. If I remodel a keyboard to be an arcade control and then use that with a game meant for keyboards have I altered the game? If I change which keys the game should listen to through the interface the game provides have I altered the core code of the game?

    It was a 24 byte one-way filter, completely external to the game. It didn’t even have the ability to add or remove code — only flip one of 192 light switches into a different position. It’s a far more subtle form of modification than WordPress extensions have.

    Which theme/plugin adds and removes WP core code? By your own admission themes/plugins can turn off/on functions through the interface WP it self provides. No alteration of core code required.

    This is the crux of our disagreement. I do not consider them mechanisms of interoperability. They are mechanisms of changing WordPress. They are WordPress itself! WordPress has external APIs for interoperability. RSS, Atom, AtomPub, XML-RPC. If you want to look at the Sega case, having someone look at WordPress source code in order to figure how to interoperate with those APIs wouldn’t violate the WordPress copyright.

    Problem is the vague definition of an API and that you really have different definitions of it depending of the situation. In essence WP itself is a framework meant to be interacted with or a platform as Matt has called it.
    If its a platform then the systems library exception could come into effect.

  18. RJack says:

    You’re attempting to create new copyright law. The “distribution” of a derivative work as a whole is a matter of contract law. 17 USC § 106(2) governs only “preparation” of derivative works. The Copyright Act is silent concerning the distribution of derivative works as a whole.

    There are two distinct copyright owners in a newly created derivative work:

    17 USC § 103. Subject matter of copyright: Compilations and derivative works

    (b) The copyright in a compilation or derivative work extends only to the material contributed by the author of such work, as distinguished from the preexisting material employed in the work, and does not imply any exclusive right in the preexisting material. The copyright in such work is independent of, and does not affect or enlarge the scope, duration, ownership, or subsistence of, any copyright protection in the preexisting material.

    Absent a valid contractual agreement between the two distinct owners (preexisting and modifying) of their respective copyrights, no derivative work as a whole may be distributed.

    The GPL cannot create a new right or scope restriction concerning distribution of derivative works as a whole — that is the sole province of Congress.

  19. Chip Bennett says:

    @RJack

    Absent a valid contractual agreement between the two distinct owners (preexisting and modifying) of their respective copyrights, no derivative work as a whole may be distributed.

    I’m not disputing that GPL gains its validity under contract law. Clearly, it does.

    Rather, I’m saying that the exclusive right to make and distribute a work, or a derivative of that work, derives from copyright law, and that any other protections otherwise afforded by contract law have been eschewed by the implied license granted by virtue of the GPL. (In fact, most of the exclusive rights protected by copyright law have been eschewed by this same implied license.)

    Again: IANAL. I’m not understanding how we are saying vastly different things, or how the differences in what we are saying in any way impact the overall analysis?

  20. Ken Newman says:

    “Show me one piece of theme/plugin code that alters the actual WP core code and not what the functions return?” ~ Andreas Nurbo

    So code must modify the original program’s source code at runtime to be derivative? So I can copy code and modify it in my program as long as I don’t have the program write the changes back to the original source? Ridiculous. (And plugins have used preg replace for that)

    By the way, WordPress “provides “the interface” “for turning off/on functions” for it’s own use. Derivatives are by definition WordPress. WordPress and it’s licensed derivatives are the only programs which can legally use the internal APIs and interfaces in a distribution.

  21. Chip Bennett says:

    @Ken Newman:

    So code must modify the original program’s source code at runtime to be derivative? So I can copy code and modify it in my program as long as I don’t have the program write the changes back to the original source? Ridiculous. (And plugins have used preg replace for that)

    Clearly, if a Theme copies/pastes core WordPress code into itself, then it has incorporated code from WordPress. I don’t think Andreas is disputing that point at all. (But even then, the question for the courts would be one of whether or not the copied code constituted copyrightable expression.)

    But, as for modifying “the original program’s source code at runtime”: WordPress Themes (and Plugins) do not modify WordPress core code. Do a checksum of core WordPress files, then intsall, activate, use, deactive, and uninstall a Theme – if you once again performed a checksum of core WordPress files, it would be identical to the previous checksum. Themes, by their very design, do not modify core WordPress code.

    By the way, WordPress “provides “the interface” “for turning off/on functions” for it’s own use.

    Agreed. And, as per the case law cited here, mere use of that interface – whether it is external, internal, or even if core needs to be decompiled to determine the mechanism – does not cause a work to be considered derivative. See Sega and especially Sony.

    Derivatives are by definition WordPress.

    I’m not sure what you mean by this statement. Can you clarify? (And I’m not trying to be obtuse; I want to make sure I’m understanding what you’re trying to say.)

    WordPress and it’s licensed derivatives are the only programs which can legally use the internal APIs and interfaces in a distribution.

    Not true. In fact, according to the cases cited, the courts will not permit, under copyright protection, artificial hurdles to be erected to prevent follow-on works from making use of another work’s mechanisms for interoperability.

    I welcome any references to copyright case law that indicates otherwise; I couldn’t find it, though.

  22. It is very refreshing to see the GPL Theme issue debated on an intellectual level without all the emotion. This article and Mark Jaquith’s are the best I’ve read for either side. I was strongly for the “themes are a derivative” side before reading this but now I am more on the fence and able to look at the whole issue with a broader perspective.

    Whatever the outcome, it will be very exciting to see how this whole thing plays out. My only concern is that if WordPress and Thesis go to court we might not get a ruling on the derivative work issue since so much Thesis code was taken directly from WordPress core.

  23. Ken Newman says:

    @Chip Bennett

    I’m saying that the exclusive right to make and distribute a work, or a derivative of that work, derives from copyright law, and that any other protections otherwise afforded by contract law have been eschewed by the implied license granted by virtue of the GPL.

    I Think what you are saying is:

    Copyright law says the rights to copy, modify, and distribute exclusively belong to the creator. You must have contracted permission to copy or create a derivative for distribution. The GPL grants contracted permission if you accept the terms.

    I’m only confused by you writing “implied.” The default position (Copyright Law) is that you have no right to copy, modify, sublicense, or distribute the code. You can only gain permission through the license (Contract Law), so in order to gain that permission you must agree. It’s pretty explicit (fully and clearly expressed).

  24. Chip Bennett says:

    @Ken Newman

    Copyright law says the rights to copy, modify, and distribute exclusively belong to the creator. You must have contracted permission to copy or create a derivative for distribution. The GPL grants contracted permission if you accept the terms.

    I think that’s a fair statement – with the qualifier that even copies do not fall under the scope of GPL unless distributed.

    I’m only confused by you writing “implied.” The default position (Copyright Law) is that you have no right to copy, modify, sublicense, or distribute the code. You can only gain permission through the license (Contract Law), so in order to gain that permission you must agree. It’s pretty explicit (fully and clearly expressed).

    The implied license involves any activity outside of distribution (which both implies and includes copying) of the work or a derivative. In other words, the implied license involves any type of end-use activity. The license for end-use activity is implied, rather than express, because GPL explicitly states that such activity is outside its scope.

  25. Chip Bennett says:

    @Chris Olbekson

    I lament the injection of emotional appeal into the discussion, as it doesn’t help the community find the truth, and come to a conclusion or compromise. So, I try to avoid emotional appeal whenever possible.

    As for a potential lawsuit against Thesis: it depends on the approach of the suit. I see many similarities to Altai here, since presumably any plagiarized code in Thesis 1.7 (or whichever) will be removed in Thesis 1.8. So, the lawsuit could seek in infringement claim on the two versions individually – just as Computer Associates did against Altai.

    In that case, the court would (hopefully) decide on the merits of each version (with plagiarized code, and without) separately – giving us some resolution on the question of whether or not Themes are inherently derivative of WordPress.

  26. Ken Newman says:

    Agreed. And, as per the case law cited here, mere use of that interface – whether it is external, internal, or even if core needs to be decompiled to determine the mechanism – does not cause a work to be considered derivative. See Sega and especially Sony.

    That’s a misread of the case law.

    The Sega issue was, first, making a copy for testing, learning (fair use) and then second, whether a password was copyrightable or available for distribution. Doesn’t really aply well when you understand that a password file deliberately set up to block interoperability and an internal API aren’t even close with regard to this case.

    As for Sony, “The court determined that the BIOS represented non-copyrightable tools necessary for implementation of creative expression, and that they cannot be used as artificial barriers or hurdles to fair use.” Again, it’s about fair use. (Creative expression is fair use) How is WordPress’s internal workings (like Sega’s password file) “artificial barrier to fair use”?
    (A commercially distributed theme isn’t fair use. Using internal APIs outside of the license isn’t an example of fair use, because the license is permissive to fair use.)

    Each of these cases deal with fair use, and intentional artificial barriers to fair use but Every fair use scenario is covered and allowed by the GPL. There is no barrier to fair use with GPL.

    The case law you cite actually supports the opposite of what you presented.

  27. Chip Bennett says:

    @Ken Newman:

    As for Sony, “The court determined that the BIOS represented non-copyrightable tools necessary for implementation of creative expression, and that they cannot be used as artificial barriers or hurdles to fair use.” Again, it’s about fair use. (Creative expression is fair use) How is WordPress’s internal workings (like Sega’s password file) “artificial barrier to fair use”?
    (A commercially distributed theme isn’t fair use. Using internal APIs outside of the license isn’t an example of fair use, because the license is permissive to fair use.)

    Each of these cases deal with fair use, and intentional artificial barriers to fair use but Every fair use scenario is covered and allowed by the GPL. There is no barrier to fair use with GPL.

    The case law you cite actually supports the opposite of what you presented.

    This conclusion isn’t supported by the facts – and decision – of the Sega case. Quoting again from Douglas Hass:

    In its preamble, the GPL also essentially argues that its license does not erect an artificial hurdle. The GPL carries no fees and expressly makes source code freely available for the public. At first glance, this free software approach would support “growth in creative expression, based on the dissemination of other creative works and the unprotected ideas contained in those works, that the Copyright Act was intended to promote.” However, the Sega court rejected this favorable license argument. The court extended the right of developers to create compatible modules free from the control of the original copyright holder even if the original copyright holder was willing to license that right under other terms. Sega had offered Accolade a license agreement that would have allowed Accolade to create Sega-compatible games with the condition that Sega manufacture those games. Despite this license offer, the court declined to find copyright infringement in Accolade’s creation of Sega-compatible games without a license… Like Sega’s license, the GPL would impermissibly apply the Act’s principle of growth to only some works (those carrying its license).

    The parallels here should be clear. The application of copyright law here isn’t a matter of fair use, but of the right of developers to create compatible, follow-on works that interoperate with a protected work. It’s not a matter of fair use, because those compatible, follow-on works aren’t derivative.

  28. Chip Bennett says:

    @Ken Newman:

    Look also at the wording of the decision itself. To wit (at 1523):

    Public benefit need not be direct or tangible, but may arise because the challenged use serves a public interest. Id. In the case before us, Accolade’s identification of the functional requirements for Genesis compatibility has led to an increase in the number of independently designed video game programs offered for use with the Genesis console. It is precisely this growth in creative expression, based on the dissemination of other creative works and the unprotected ideas contained in those works, that the Copyright Act was intended to promote.

    Here, the alleged infringement was the decompiling of the video game object code – which would not be an infringement under GPL. Nevertheless, consider the court’s reasoning: the “growth in creative expression” resulting from additional Sega-compatible video games, is “precisely” what “the Copyright Act was intended to promote.” The “creative expression” of additional video games is directly analogous to the “creative expression” of additional WordPress themes.

    Further, with respect particularly to the PHP required to render a WordPress theme, consider this part of the decision (at 1524, citations omitted):

    Computer programs pose unique problems for the application of the “idea/expression distinction” that determines the extent of copyright protection. To the extent that there are many possible ways of accomplishing a given task or fulfilling a particular market demand, the programmer’s choice of program structure and design may be highly creative and idiosyncratic. However, computer programs are, in essence, utilitarian articles — articles that accomplish tasks. As such, they contain many logical, structural, and visual display elements that are dictated by the function to be performed, by considerations of efficiency, or by external factors such as compatibility requirements and industry demands. In some circumstances, even the exact set of commands used by the programmer is deemed functional rather than creative for purposes of copyright. “[W]hen specific instructions, even though previously copyrighted, are the only and essential means of accomplishing a given task, their later use by another will not amount to infringement.”

    This reasoning, by my reading, refutes Mark Jaquith’s contention that the code itself is the “expression”. The code itself does not exist as a means unto itself, but rather as a means to carry out some function, purpose, or other creative expression. In order to determine what parts of WordPress’ core code are copyrightable as creative expression, it will first be necessary to define, at varying abstraction levels, the purpose of WordPress.

    (I gave some examples above. See: Widgets, Comments.)

    And finally (at 1527-1528):

    We conclude that where disassembly is the only way to gain access to the ideas and functional elements embodied in a copyrighted computer program and where there is a legitimate reason for seeking such access, disassembly is a fair use of the copyrighted work, as a matter of law. Our conclusion does not, of course, insulate Accolade from a claim of copyright infringement with respect to its finished products. Sega has reserved the right to raise such a claim, and it may do so on remand.

    If the “finished products” – i.e. Accolade’s Sega-compatible video games were derivative works, and therefore infringing on Sega’s copyright, why was this point not argued and decided upon the District Court’s remand of the decision? if Accolade’s Sega-compatible video games, that incorporated the functional requirements of Sega’s video games in order to achieve interoperability with the Sega console, were not derivative, then clearly, neither are WordPress themes.

  29. RJack says:

    Your initial premise,
    “Since the copyright owner has [the] exclusive right to produce and to distribute derivative works based on the copyrighted work, the GPL intends to grant unlimited usage rights (to use, study, modify, etc.) to the end-user, while forcing follow-on developers of derivative works to release those works under the same license”
    is simply untenable under U.S. law. As I previously stated, the GPL is a contract between a preexisting author and a modifying author. The “follow-on developers of derivative works” to which you refer are the “… all third
    parties under the terms of this License” referred to in the GPL’s sec. 2(b).

    You can’t “force” those third party developers to accept the GPL’s terms.
    There is a fundamental principle of contract law prohibiting the parties to a contract from binding nonparties. See, e.g., EEOC v. Waffle House, Inc., 534 U.S. 279, 294 (2002) (“It goes without saying that a contract cannot bind a nonparty.”).

    This lack of privity renders the GPL unenforceable. My point concerning the unenforceability of the GPL was pointed out to Eben Moglen by Professor Robert P. Merges of the Berkeley Law School in June of 2000:

    ********************************************************************
    “Professor Robert P. Merges of the Berkeley Law School noted some of
    the problems in his “The End of Friction? Property Rights and
    Contract in the ‘Newtonian’ World of On-Line Commerce” (12 Berkeley
    Tech. L.J. 115), in which he describes the GPL as “informal (i.e.,
    not legally enforceable) restrictions on digital content.”

    Not legally enforceable?

    “By its own terms, the copyleft agreement is an unusual license; at
    the most basic level consider the problem of determining damages
    when the licensee frustrates the licensor’s expectation of zero
    profits under the contract,” writes Merges. “But what is most
    significant about the agreement is that it purports to restrict
    subsequent transferees who receive software from a licensee,
    presumably even if the licensee fails to attach a copy of the
    agreement. As this new transferee is not in privity with the
    original copyleft licensor, the stipulation seems unenforceable.”

    This isn’t as complicated as it sounds. “Privity” simply means party
    to the contract, and it works both ways: I can sign a contract with
    you whereby you give me some money and I send flowers to your
    mother. If I fail to send the flowers, you can sue me, but your
    mother can’t, because she’s not a party to the contract. Likewise, I
    might sign an agreement with you whereby I give you some money and
    in return you agree that your mother will come to my house and, I
    don’t know, bake a pie some Sunday afternoon. Your mother is not
    obligated to live up to the terms of the agreement, because she’s
    not party to it. If she doesn’t bake the pie, I can’t sue her but I
    can sue you.) Merges’s point is that the GPL isn’t an enforceable
    contract because it obligates parties who did not agree to it.”
    ******************************************************************
    http://www.linuxplanet.com/linuxplanet/reports/2000/1/

    Moglen’s answer was that a license was *not* a contract.

  30. Sean Kaye says:

    There’s an interesting paper on the Queensland University of Technology site about Legal Issues for the use of FOSS and in particular it goes into some of the legal risk elements of the GPL. Obviously this is based on Australian law, but it is an interesting read, especially page 15.

    The paper also touches briefly (without going into too much depth) about how the GPL v2 may run afoul of the Trade Practices Act here in Australia. The paper talks about warranties a bit, but the TPA is much broader than that. If something like the GPL were to run aground in Australia, it would almost certainly be with the Trade Practices Act as opposed to even Copyright Law.

    There’s also a decent explanation of derivative works (as it relates to software) in the paper which is worth reading.

  31. Fill says:

    Something that has confused me is that the GPL FAQ says you can take GPLed code and use it for private use (to modify/study/etc.) BUT says that code running a publicly available website is hardly ‘private use’ and therefore the code should be made public as well.

  32. Chip Bennett says:

    @Fill,

    Simply running WordPress on a publicly accessible web site is not distribution. If anything, it is most analogous to public performance. Public performance is use, not distribution, as defined by the Copyright Act. Thus, by the wording of the GPL, it is outside the scope of the license.

    If the FAQ says otherwise, it’s simply wrong.

    I believe you’re referring to this part of the FAQ (emphasis added):

    A company is running a modified version of a GPL’ed program on a web site. Does the GPL say they must release their modified sources?

    The GPL permits anyone to make a modified version and use it without ever distributing it to others. What this company is doing is a special case of that. Therefore, the company does not have to release the modified sources.

    It is essential for people to have the freedom to make modifications and use them privately, without ever publishing those modifications. However, putting the program on a server machine for the public to talk to is hardly “private” use, so it would be legitimate to require release of the source code in that special case. Developers who wish to address this might want to use the GNU Affero GPL for programs designed for network server use.

    This statement contradicts the clear wording of the Copyright Act. Distribution (covered by the GPL) and public performance (outside the scope of GPL) are clearly defined as two separate actions in the Act. Such an argument, if used in court, would not withstand even the most basic scrutiny. Public performance is not distribution, and the GPL clearly states that anything outside of distributions of copies of modified or unmodified code or its derivatives is not covered by the license.

  33. Yuhong Bao says:

    Yep, even the FAQ itself in the next sentence suggests the Affero GPL if you want that.

  34. saulgoode says:

    Kudos on a well-written conspectus of various applicable cases in law. Nonetheless, I respectfully disagree with your ultimate conclusions. In particular, the following claim:

    Extant case law involving copyright infringement – and especially, software copyright infringement – does not view linking to and interaction with a protected work to constitute a derivative work, …

    The problem is with the conflation of “derivative work” with “infringing derivative work”. There are actually several ways a work could be categorized:

    It is not derivative.
    It is derivative but not infringing (original material unprotected by copyright).
    It is derivative and would be infringing, but licensing provides authorization.
    It is derivative and unauthorized, but the Fair Use doctrine applies.
    It is derivative and unauthorized, and it is infringing.

    Two of the three rulings which you cited as supporting your conclusion (Sony and Sega) were each based upon the Fair Use defense being successfully applied to the derivative work, not that the work was deemed non-derivative. In the third (Galoob), at issue was not whether the Game Genie device itself was a derivative work, but whether its output constituted a derivative work. Galoob, in particular, seems to have no bearing on any GPL-related questions of derivative work — though of all the cases cited, Galoob seems to provide the most comprehensive general analysis of derived works.

    At best, the Sony and Sega rulings suggest that the fourth category (Fair Use) might apply to linking and interaction of plug-ins or modules. That, however, is quite distinct from not being a derivative work — in fact, to reach the point of Fair Use being addressed, the work has to be considered derivative.

    But Fair Use is always a fact-based analysis of a particular situation and a Fair Use defense may well be viewed in a much different light if applied to the facts surrounding a GPL-licensed program than to more traditionally licensed copyrighted works. In the Sega case, the courts maintained that allowing Fair Use derivatives provided a public benefit, a growth in creative expression, and dissemination of other creative works. Likewise, in the Sony case it was decided that NOT permitting Fair Use derivative works would make it impossible for others to compete in the marketplace.

    Considering that GPL licensing directly has as its goal to promote public benefit and the sharing of creative works, while still enabling others to compete, one wonders whether the courts would ever be able to justify a Fair Use defense against it. Quite frankly, the GPL does a better job of “promoting the Arts and Sciences” than does copyright law itself; and as Fair Use is a defense against copyright holders overextending their monopoly to the point of suppressing creativity and sharing, it is a fairly weak proposition when confronting the General Public License.

    GPL licensing has encouraged millions of creative and inventive minds to share their expressive works freely with society. I for one would be extremely disappointed if the courts were to ever place the commercial interests of a single individual or company above that advancement.

    Nonetheless, thank you for your well-written article and for providing a background to this discussion.

    Regards.

  35. Considering that GPL licensing directly has as its goal to promote public benefit and the sharing of creative works, while still enabling others to compete, one wonders whether the courts would ever be able to justify a Fair Use defense against it.

    It depends on who you ask actually. One goal is that software should be free as in beer. Another is to make all software free as in freedom. Etc. Its an ideologial license really. RMS calls it a social movement.
    When it comes to copyright etc there seems to be monetary aspect also.

    GPL licensing has encouraged millions of creative and inventive minds to share their expressive works freely with society. I for one would be extremely disappointed if the courts were to ever place the commercial interests of a single individual or company above that advancement.

    GPL will still be doing just fine even if it doesnt extend to plugins/themes. WP and other software will still be developed etc. They just cant benefit from the code of propriety extensions and those extensions can’t use the code of the GPL software. The GPL software it self is not hurt by propriety stuff and I don’t think people would stop licensing their stuff under the GPL just because they don’t have too.

  36. Chip Bennett says:

    @saulgoode:

    Two of the three rulings which you cited as supporting your conclusion (Sony and Sega) were each based upon the Fair Use defense being successfully applied to the derivative work, not that the work was deemed non-derivative. In the third (Galoob), at issue was not whether the Game Genie device itself was a derivative work, but whether its output constituted a derivative work. Galoob, in particular, seems to have no bearing on any GPL-related questions of derivative work — though of all the cases cited, Galoob seems to provide the most comprehensive general analysis of derived works.

    At best, the Sony and Sega rulings suggest that the fourth category (Fair Use) might apply to linking and interaction of plug-ins or modules. That, however, is quite distinct from not being a derivative work — in fact, to reach the point of Fair Use being addressed, the work has to be considered derivative.

    I disagree with your assessment of Galoob. It is directly applicable to WordPress core and Themes. The Theme interjects itself between WordPress core and the person viewing the content managed/output by WordPress, in order to modify the appearance of that content. Just as the Game Genie did so in a manner that was determined by the court not to be derivative, so does a WordPress Theme.

    Bear in mind that, with Sega, the alleged infringing took place in the intermediary copying of object code, not in the development of Sega-compatible games. It was this intermediary copying (of the entire work, both copyrightable and non-copyrightable content), in order to reverse-engineer the Sega game in order to define the methods of interoperability, that was deemed to be a fair use of a protected work.

    In fact, the court differentiated between the copyrightable content (the elements of the video game itself) from the non-copyrightable content (the functional requirements of interoperability) of the protected work. The Sega court expressly stated that infringement occurred due to the copying of the copyrightable content of the object code, but that such copying was fair use since it was the only means of accessing the non-copyrightable content that was necessary for development of follow-on (and by definition, non-derivative) works.

    The Sony court used much the same reasoning.

    The most important point from both decisions, with respect to WordPress Themes, is that the courts determined that any part of the protected work incorporated into the follow-on work for the purpose of achieving interoperability was non-copyrightable method-of-operation content. Therefore, since such content was non-copyrightable, it by definition did not cause the follow-on work to be considered a derivative of the protected work.

  37. Dennis Steed says:

    This is a fascinating discussion.

    However, I’m afraid that any Matt vs Chip court battle would be disappointing.

    As I understand it, Chip’s themes incorporate code from GPL’d WordPress themes. As such, I suspect any court will rely on that simple and clear fact to rule against Chip.

    Unfortunately, the court does not have to say anything in its decision about whether hooking a non GPL’d theme to the WordPress core is problematic.

    What is really needed is a case where the WordPress theme is written entirely from scratch and incorporates no GPL’d code.

    Dennis

  38. Chip Bennett says:

    @Dennis Steed:

    However, I’m afraid that any Matt vs Chip court battle would be disappointing.

    As I understand it, Chip’s themes incorporate code from GPL’d WordPress themes. As such, I suspect any court will rely on that simple and clear fact to rule against Chip.

    I believe you mean Chris – as in Chris Pearson – not Chip – as in, me. 🙂

  39. Dennis Steed says:

    You are correct, I did mean to say “Chris” rather than “Chip”. Don’t know how to edit the original post though.

    Dennis

  40. Mike Wasylik says:

    Chip,

    An excellent and thorough review of the case law and arguments on both sides. I’m going to bookmark this for reference.

    Mike

  41. saulgoode says:

    The Sega court expressly stated that infringement occurred due to the copying of the copyrightable content of the object code, but that such copying was fair use since it was the only means of accessing the non-copyrightable content that was necessary for development of follow-on (and by definition, non-derivative) works.

    I disagree with your definition; basing a work upon non-copyrightable content does not make it non-derivative, it would be derivative but non-infringing. If I produce a photo album of paintings by Leonardo daVinci, it is still a derivative work even though it doesn’t infringe upon any copyrights.

    Semantics aside, I would point out that while the court in the Sega trial did address accessing non-copyrightable content of the object code, they did not actually delineate what that obtained, or why. As such, it is rather a stretch to conclude from that case that modules or plug-ins which inter-operate with programs don’t access copyrightable content; especially so since the “object code” under consideration was machine statements burned into ROM.

    Given the construct of the machine and the development approach taken at the time, there was no function prototyping, no symbolic referencing, and no runtime linking information that could be gleaned from reverse engineering the Sega games’ object code. Aren’t these aspects of a program indicative of the creative and intellectual elements that the programmer has fixated in the software which make it copyrightable?

    Isn’t that precisely what copyright is supposed to protect: the creative and intellectual choices made by an author, artist, or programmer that are distinct from the choices that would be made by somebody else? Certainly Ian Fleming shouldn’t be able to prevent others from writing stories about heroic government agents battling evil villains who are members of global terrorist organizations, but when those stories call their hero James Bond and he works for MI6 and is fighting Ernst Blofield of S.P.E.C.T.R.E. then it pretty well starts to fall under consideration as being a derivative work.

    Likewise, I would argue, that copyright law as applied to computer programs does not preclude people from employing the algorithms and functionality from other programs, but once your program (or plug-in/module) starts calling a particular set of functions with a particular set of parameters in a particular sequence, you have entered the realm of creating a derivative work based on that other computer program*. If the creative and intellectual aspects of a computer program aren’t protectable by copyright then what exactly makes a program copyrightable?

    * But not necessarily infringing.

  42. Chip Bennett says:

    @saulgoode:

    I disagree with your definition; basing a work upon non-copyrightable content does not make it non-derivative, it would be derivative but non-infringing.

    Copyright only extends to that part of a work that is copyrightable. Content can be non-copyrightable for myriad reasons (public domain, copyrighted by others, merger, scenes a faire, etc.); but regardless of the reason particular content is deemed non-copyrightable, if content is determined to be non-copyrightable, then it by definition isn’t copyrighted. If it isn’t copyrighted, then it (again, by definition) isn’t derivative.

    If I produce a photo album of paintings by Leonardo daVinci, it is still a derivative work even though it doesn’t infringe upon any copyrights.

    Assuming the paintings (or the photos thereof) are in the public domain, then you have created a compilation. Your copyright doesn’t extend to the photos themselves, but rather to the creative manner in which you compiled them. Such compilations have a very narrow scope in terms of copyright protection.

    I could take those same photos and do pretty much anything with them – except for re-create your exact compilation – without violating your copyright.

    Semantics aside, I would point out that while the court in the Sega trial did address accessing non-copyrightable content of the object code, they did not actually delineate what that obtained, or why.

    I would argue that such differentiation is exactly what the Sega court performed. See Sega at 1523:

    The declarations of Accolade’s employees indicate, and the district court found, that Accolade copied Sega’s software solely in order to discover the functional requirements for compatibility with the Genesis console — aspects of Sega’s programs that are not protected by copyright. With respect to the video game programs contained in Accolade’s game cartridges, there is no evidence in the record that Accolade sought to avoid performing its own creative work.

    Clearly, the Sega court considered Accolade’s games themselves to be original, creative work (i.e. non-derivative), and considered the “functional requirements for compatibility” to be “not protected by copyright”. This differentiation is directly applicable to a WordPress Theme: the creative expression of the Theme is original, creative work, and the “functional requirements for compatibility” with WordPress (i.e. function calls, filter/action hooks, etc.) are not protected by copyright.

    As such, it is rather a stretch to conclude from that case that modules or plug-ins which inter-operate with programs don’t access copyrightable content; especially so since the “object code” under consideration was machine statements burned into ROM.

    Even if your conclusions regarding the Sega decision are correct (and as I argue, they aren’t), this follow-on conclusion appears not to take Galoob into consideration. The Galoob Game Genie is exactly what you describe: a “[module] or [plug-in] which interoperate[s] with [a] program” – and the Galoob court determined that the Game Genie was not a derivative work.

    Given the construct of the machine and the development approach taken at the time, there was no function prototyping, no symbolic referencing, and no runtime linking information that could be gleaned from reverse engineering the Sega games’ object code. Aren’t these aspects of a program indicative of the creative and intellectual elements that the programmer has fixated in the software which make it copyrightable?

    No. Those are primarily “functional requirements for compatibility” and scenes a faire (see also: Altai).

    Isn’t that precisely what copyright is supposed to protect: the creative and intellectual choices made by an author, artist, or programmer that are distinct from the choices that would be made by somebody else? Certainly Ian Fleming shouldn’t be able to prevent others from writing stories about heroic government agents battling evil villains who are members of global terrorist organizations, but when those stories call their hero James Bond and he works for MI6 and is fighting Ernst Blofield of S.P.E.C.T.R.E. then it pretty well starts to fall under consideration as being a derivative work.

    And thus, you have nicely described scenes a faire – but, it seems, you have mis-applied the concept to WordPress. WordPress is a content management system used primarily as a blogging platform. To repeat two aspects I’ve referenced above: WordPress cannot claim copyright for Widgets and Comments (both are scenes a faire for a CMS/blogging platform). However, the creative expression contained in the code behind wp_list_comments() is copyrightable. Likewise, the actual Widgets included in core WordPress are copyrightable.

    What is not copyrightable, however, is the use of the wp_list_comments() function call. It is merely a method of compatibility. Likewise, extending the WP_Widgets class to incorporate new Widgets is not copyrightable (nor are the new Widgets inherently derivative of WordPress).

    Likewise, I would argue, that copyright law as applied to computer programs does not preclude people from employing the algorithms and functionality from other programs, but once your program (or plug-in/module) starts calling a particular set of functions with a particular set of parameters in a particular sequence, you have entered the realm of creating a derivative work based on that other computer program*.

    Merely making function calls is not derivative, because doing so in no way incorporates copyrightable content of a protected work in a follow-on work. Making those function calls with a particular set of parameters in a particular sequence – if such parameters and sequence are required for interoperability with the program – is likewise not derivative, because such parameters and sequence are “methods of operation” and therefore non-copyrightable.

    If the creative and intellectual aspects of a computer program aren’t protectable by copyright then what exactly makes a program copyrightable?

    That depends entirely on the purpose of the program – and that is a question that, as the Altai court indicated, needs to be answered using a process of Abstraction, Filtration, and Comparison.

    It seems that you are arguing for a Wheelan-level abstraction, which no court in the past 18 years has applied.

  43. RJack says:

    As happens with too many discussions concerning the GPL license,
    there seems to be an automatic assumption that the GPL is an
    enforceable license. Before launching into an analysis of what
    constitutes a derivative work or what constitutes an infringing
    derivative work, why not ask “is section 2 of the GPL license
    legally enforceable?”

    The Software Freedom Law Center has forever claimed that section 2(b)
    is a “condition” on the permissions to copy, modify and distribute GPL
    code. See for example the Best Buy Inc. litigation:

    “23. Under the License, Mr. Andersen grants certain permissions to other
    parties to copy, modify and redistribute BusyBox so long as those
    parties satisfy certain conditions. In particular, Section 2(b) of the
    License, addressing each Licensee, states: You must cause any work that
    you distribute or publish, that in whole or in part contains or is
    derived from the Program or any part thereof, to be licensed as a whole
    at no charge to all third parties under the terms of this License. Thus,
    if a Licensee redistributes a version of BusyBox, it may do so only
    under the terms of the License. Distribution of BusyBox under any other
    terms or conditions is not allowed.”
    http://www.softwarefreedom.org/resources/2009/busybox-complaint-2009-12-14.pdf

    The term “condition” in a legal complaint means a “condition precedent”.
    We can rest assured that Judge Scheindlin who is hearing the Best Buy
    Inc. case will interpret the term “condition” this way. See e.g.:

    “n52 “Plaintiffs bring claims for “Contract Failure of Condition”
    against each defendant. The Court is not familiar with this term. I
    assume “Contract Failure of Condition” is a claim for breach of a
    condition precedent.” Abu Dhabi Commercial Bank, et al. v. Morgan
    Stanley & Co., et al., 1:2008cv07508, SDNY, (2008).–Judge Shira A.
    Scheindlin. http://amlawdaily.typepad.com/AbuDhabi.pdf

    So… what is a “condition precedent” in a contract?

    “A condition is an event, not certain to occur, which must occur, unless
    its non-occurrence is excused, before performance under a contract
    becomes due.” Restatement (Second) of Contracts § 224 (1981).
    “Conditions precedent are disfavored and will not be read into a
    contract unless required by plain, unambiguous language.” Effects
    Associates, 908 F.2d at 559 n.7.”; Jacob Maxwell, Inc. v. Veeck, 110
    F.3d 749 (11th Cir. 1997).
    http://ftp.resource.org/courts.gov/c/F3/110/110.F3d.749.96-2636.html

    In the case of the GPL, the performance that is subject to the
    conditioning events is in section 2: “You may modify your copy or copies
    of the Program or any portion of it, thus forming a work based on the
    Program, and copy and distribute such modifications…” One event
    (condition precedent) which must be satisfied is section 2(b):

    “You must cause any work that you distribute or publish, that in whole
    or in part contains or is derived from the Program or any part thereof,
    to be licensed as a whole at no charge to all third parties under the
    terms of this License.”

    The obvious question is how you do gain copyright permission to copy,
    modify and redistribute the GPL’d code? You can’t *copy*, *modify* or
    distribute any code unless you *first* satisfy the condition (event)
    that any derivative work “you distribute or publish, that in whole or
    in part contains or is derived from the Program or any part thereof,
    to be licensed as a whole at no charge to all third parties under the
    terms of this License”.

    So… how do you distribute and license code that you have not yet
    received permission to create (copy and modify)? In contract law this is
    known as an impossible condition — the conditioning event depends on
    performance of the contract and performance of the contract depends on
    the conditioning event.

    Impossible conditions occurring in a contract are strictly construed
    against the drafter. The GPL is truly an unenforceable license.

  44. Chip Bennett says:

    @RJack,

    I’m tracking you now. Thanks for clarifying and elaborating.

    You bring up an interesting question regarding GPL – one that I’d not considered. As such, I’m not really in a position to comment on it. I think the point you bring up is fascinating, but largely irrelevant for the WordPress community at-large, which has generally accepted the validity of the GPL.

    I’ll have to look into it some more.

  45. Alexander Terekhov says:

    Nice article.

    Supported by: Dangerous Liaisons—Software Combinations as Derivative Works? Distribution, Installation, and Execution of Linked Programs Under Copyright Law, Commercial Licenses, and the GPL, By Lothar Determann

    Keep up the good work of disspeling the GPL myths.

    regards,
    alexander.

Trackbacks

  1. Twitted by ericlbarnes