oreilly.comSafari Books Online.Conferences.


Open Source and Free Documentation Licenses, Part 2: The Open Publication License

by Andrew M. St. Laurent, author of Understanding Open Source and Free Software Licensing

Open source licensing principles are almost always linked to software development, and for good reason. The idea of open source and free software licensing has met with great success in connection with software: the very name open source refers to the opening of software's source code so that it can be edited, copied, and modified.

As discussed in part one of this article, open source principles may be less readily applicable to documents and other non-software works. Nonetheless, there are a number of licenses and projects directed at applying open source and free software licenses to such works. In my book, Understanding Open Source and Free Software Licenses, I discuss the Creative Commons, an ambitious project intended to apply the power of open source development to a wide variety of works, including music, video, and the fine arts, as well as to documents. In part one, I examined the GNU Free Documentation License (FDL), a license that follows the GNU licensing philosophy by permitting free modification and distribution of a given document, upon the condition that the document so distributed or modified also be licensed under the GNU FDL.

The licenses described in this article, and in the upcoming third part of this series, takes an approach to licensing that is similar to open source software licenses like the BSD or MIT licenses, and does not impose such a "reciprocal" obligation on licensees. This is to say that these licenses do not require that licensees themselves distribute the work, or modified versions of the work, only under the license under which they themselves received the right to distribute the work or make modified works from it. The Open Publication license (OPL), which we'll cover today, was originally crafted for use with software manuals, but can be used in connection with any kind of document. The Open Gaming license, which I'll discuss in the forthcoming part three, was specifically designed for one set of documents, the System Reference Documents, version 3.5 (SRD v. 3.5). The SRD v. 3.5 consist of the mechanics underlying the third edition of the original Dungeons & Dragons gaming system created by TSR, Inc., and later bought by the Wizards of the Coast.

The Open Publication License

Related Reading

Understanding Open Source and Free Software Licensing
By Andrew M. St. Laurent

The Open Publication license was developed by Open Content. In addition to the basic license itself, the license provides some "Good-Practice Recommendations" and some options that may place additional limitations on the use of the rights granted by the license.

The Open Publication license in its "standard" form (that is to say, without any of the additional restrictions that may be added, as described in more detail below), operates much like the aforementioned "academic" software licenses such as the BSD or the MIT licenses. The copyright for the licensed work, as with these academic licenses, remains with the original author and publisher, although virtually every exercise of rights under copyright law is freely permitted to licensees and there is no requirement that derivative works be licensed under the Open Publication license.

Although the OPL has the benefits of being both short and comprehensible, it suffers from serious ambiguities that may weigh against its use. These potential problems are described in more detail below.

Section I of the license describes the basic and broad grant of rights in the license. In order to indicate that a particular work is licensed under the Open Publication license, the work should include such a notification immediately following the copyright notice. For example, a hypothetical work, "The Frog-15 Manifesto," would carry a notice like the following: "The Frog-15 Manifesto is copyright 2004 by Rick Jones. This material may be distributed only subject to the terms and conditions set forth in the Open Publication license, v. 1.0." The v. 1.0, which we are discussing here, is attached as Appendix A to this article.

In its basic form, the license permits the redistribution of the licensed work, including commercial redistribution, and also permits the distribution of the work in modified form, subject to certain additional restrictions. The work may be distributed in its original form so long as the Open Publication license itself or the incorporation by reference (such as that described in the previous paragraph) is displayed in the distributed form. If the work is distributed unmodified in electronic form, that is substantially the only restriction.

If the work is distributed in printed book form, the name of the author and the publisher as well as the title of the work must appear on all outer surfaces of the book, including, presumably, the front and back covers and spine. Works that are distributed electronically or otherwise in non-printed form do not need to comply with this requirement. Works may be aggregated freely with works under other licenses on the same media, so long as the aggregate work contains a notice that work licensed under the Open Publication license is included.

Section II of the license simply states the copyright of the licensed work remains with the copyright holder, and Section III contains some legal boilerplate that is described in more detail below.

Pursuant to Section IV, modified versions of the work may also be created and distributed, but these must comply with certain additional restrictions. The modified version must be identified as such, with the person making the modifications being identified and the date of the modifications given. Acknowledgements of the original author and publisher must be given pursuant to "normal academic citation practices:" a potentially ambiguous requirement. The location of the unmodified document may be given--another potentially ambiguous requirement. Finally, the names of the original author(s) may not be used to imply endorsement of the modified work without permission. Significantly, however, as already noted, the Open Publication license does not prohibit the distribution of modified works under more restrictive licenses, including a proprietary license. In other words, so long as the modified work complies with these relatively minimal restrictions, the modified work need not itself be open or available for legal modification.

In Section III, the license also contains two pieces of legal boilerplate, one disclaiming any warranties, including a warranty of non-infringement, and another asserting that the terms of the license are severable: in the event that a court finds that one or more terms of the license are unenforceable, those terms that are enforceable can be "severed;" that is, enforced without taking into consideration the unenforceable terms of the license.

That describes the basic functioning of the Open Publication license. Now let's look at the other sections.

Dissecting the Rest of the License

Section V of the license, "Good-Practice Recommendations," contains non-binding terms that are not required of licensees, but are "requested from and strongly recommended of redistributors." There are three "Good-Practice Recommendations:" first, that licensees give 30 days of notice to authors before redistributing work, to allow the author to provide updated versions of the work; second, that substantive modifications in modified works be clearly identified, either in the text or in an addendum; and third, that authors of works be given free hard copies from distributions of original or modified works. It is the licensee's choice whether or not to comply with these recommendations, as they are not required by the license.

Section VI, entitled "License Options," lists options that can be added to the basic form of the Open Publication license by including additional text after the license reference or to the copy of the license. There are two such additional options. First, the licensor (usually the author of the work) can restrict distribution of the work in modified form by adding the sentence "Distribution of substantively modified versions of this document is prohibited unless prior permission is obtained from the copyright holder" to the license or the license reference. Second, the licensor can limit the commercial exploitation of the licensed work by adding the sentence "Distribution of the work or derivative of the work in any standard (paper) book form is prohibited unless prior permission is obtained from the copyright holder" to the license or the license reference.

Finally, a "Policy Appendix" that is not "considered part of the license" gives instructions on how to subscribe and post to the Open Publication Authors' List, and a contact for questions about the license. The "Policy Appendix" also states that authors may impose their own license terms on Open Publication works so long as these license terms are no more restrictive than the Open Publication license.

License Issues

The Open Publication license is fairly comprehensible and its intentions are certainly good. Nonetheless, there are serious problems with the license that may encourage authors to strongly consider other licenses.

The first, and most serious, problem is in the "License Options" section. It seems simple enough for an additional section to be added to a license reference contained in the work. However, this is may well create confusion if both the license reference and a copy of the license are included in a particular work. For example, the author of "The Frog-15 Manifesto" may not wish to permit the creation of derivative works from the manifesto, and accordingly includes the following license reference in the body of his work: "The Frog-15 Manifesto is copyright 2004 by Rick Jones. This material may be distributed only subject to the terms and conditions set forth in the Open Publication license, v. 1.0. Distribution of substantively modified versions of this document is prohibited unless prior permission is obtained from the copyright holder." However, Jones also includes the entire text of the license as an appendix to the manifesto, including Section IV, which explicitly permits the creation of modified works. A future licensee may not read the license reference carefully, or at all, and may assume that he has the right to freely modify the work, contrary to the intentions of the author.

The solution to this problem is the crafting of substantially similar but distinct licenses, each of which is specifically tailored to the rights that a given author wants to grant. The Creative Commons system of licenses is exactly that. By attempting to reach this result without drafting separate licenses, the Open Publication license creates a very real chance of confusion.

The two options described in Section VI also both contain ambiguities or limitations that don't encourage their use. The first option, the one just discussed, limiting the ability of licensees to make derivative works, prohibits only "substantive modifications" to the licensed work, which are themselves defined as changes to the "semantic content of the document" and not "mere changes in format or typographical corrections." Although apparently clear at first look, this "substantive modifications" definition is dangerously ambiguous. For example, does the use of only a section of the work constitute a "semantic change"? Similar problems are presented by potential derivative works that reorganize the licensed work or incorporate other text, albeit without changing any of the text from the original work. Such works, at least arguably, don't change the "semantic content," but undoubtedly would change the work.

The second option in Section VI, prohibiting commercial publication of the work or derivative works in standard book format, also presents ambiguities. An author may want to prevent any commercial exploitation of the work; but even if this option is included, a licensee could, arguably at least, reprint the work in a magazine. Moreover, the licensee could unquestionably reproduce and distribute such a work in an electronic format. An author using the Open Publication license and looking to head off such results could, in theory, add another clause prohibiting such uses of the work. This option, however, would take such a "revised license" outside of the scope of Open Publication licenses, as the "Policy Appendix" seems to prohibit the application of terms to Open Publication licenses that are more restrictive than the license itself, which these terms certainly would be. Such an author would almost certainly be better served by using another license.

The inclusion of the "Policy Appendix" and the "Good Practice Recommendations" in the body of the license itself also raises significant problems: neither is part of the license, but the terms included in them may confuse licensees. The "Policy Appendix" does not contain much of substance except the ambiguous permission granted to Open Publication authors to include their own license terms that do not add further restrictions to the Open Publication license. This permission in itself creates a real ambiguity as to whether, if authors do add their own license terms that are more restrictive, such terms would be enforceable. The "Good Practice Recommendations" are further written so that they may seem mandatory to at least some licensees, who may decide not to use the licensed work based on their (false) belief that these recommendations are mandatory. Such recommendations might appropriately be made in a FAQ or other explanatory document, but should not be incorporated into the license itself.

Finally, the Open Publication license doesn't adequately describe how to address generations of modified works or the inclusion of multiple works licensed under the Open Publication license in one medium, such as an anthology of articles. The GNU FDL and the Creative Commons licenses address these situations much more clearly. Indeed, the Open Publication license is particularly ambiguous on the question of anthologies. Section III permits the inclusion of licensed works, or parts of licensed works, on a single medium along with other works, including works licensed under other types of licenses. The larger work, including the licensed work, sounds like an anthology. However, Section IV, includes among the list of works described as "derivative works" "anthologies." The ambiguity thus arises as to whether such a use of the licensed work has to comply with the relatively few restrictions of Section I or with the somewhat greater restrictions of Section III.


In sum, while the Open Publication license is unambiguous when applied at least to some licensing situations, it has many ambiguous parts that may weigh against using it in open source document projects.

In the conclusion to this three-part series on open source and Free Documentation Licenses, Andrew will break down the Open Gaming license. Stay tuned.

Andrew M. St. Laurent is an experienced lawyer with a long-time interest in intellectual property, particularly software licensing.

In August 2004, O'Reilly Media, Inc., released Understanding Open Source and Free Software Licensing.

Return to

Sponsored by: