Software

OPINION

Is the GPL OK for DNA?

Duct tape is the world’s most versatile invention. It makes ducts airtight, holds car parts together, patches holes in canoes and keeps rock musicians from tripping over wires onstage. It rips easily, holds firm even underwater and provides a quick fix for almost any emergency. But sometimes, duct tape fanatics go too far.

You can buy wallets made entirely of duct tape. Some even make clothes out of it. The same is true with open-source licensing. It has many useful applications, but it is easy to get carried away. Take, for example, the increasing interest in applying open-source licensing to biotechnology.

Opening the Software Bazaar

Open-source development allows for software “bazaars” where contributors come together in less structured groups to contribute to a software project. Without the right licensing, this is not possible. Unless contributors have access to the source code for a project, they cannot effectively contribute. More importantly, if the contributors don’t get what they want out of the project, they won’t put in the effort, even if the source code is available.

Open-source contributors want the freedom to use the existing code and their own contributions as they desire, as well as the ability to retain credit for their contribution. They also want to know that downstream, users can’t take control of the code and restrictively license modifications of it (although sometimes that is less of a factor, so a licensor might use a license without those restrictions). These aspects of open-source software licenses are made possible by the nature of the intellectual property that protects them.

A critical aspect of open-source licensing is that the license terms follow the creation, i.e., those with access to the creation are bound by the license. It is also important that those without access to the creation not be bound by any rights; otherwise it would be possible for unknown third parties to assert rights to the creation.

Open-Sourcing Biotechnology

There is momentum growing to apply open-source licensing to biotechnology. One organization pursuing this goal is an Australian agricultural life science institute, Cambia. It proposes licensing technologies such that the techniques and components used to create products under its model are available to all.

It is unclear what intellectual property rights it would use to bring life to its model or how it would free up the uses of the licensed technology while preventing others from constraining it.

Intellectual property rights are generally negative rights. They exclude others from doing something. Copyrights grant their holders the right to exclude others from copying. Once you have that right, you can negotiate terms of a license to obtain other rights in exchange for permission to copy. The exclusive copying right includes copying, making derivative works, distributing and other rights. Therefore, someone with a copyright can control the conditions under which derivative works are made and distributed. Most software modification is the creation of a derivative work.

Copyright Law and GPL’ed Works

Creating software independently, even if it is identical to other copyrighted code, is not copying of the other code, and therefore unknown third-parties would not have rights to the software. Furthermore, extracting the functionality of copyrighted code and reusing it elsewhere is not considered copying, as ideas, functionality, etc. are not considered copyrightable.

These aspects of copyright law are significant when considering open-source development. If someone independently creates software, duplicating at most the functionality of someone else’s code, the independent creator has no obligations under copyright law and can license that code under any suitable terms, such as those provided by the GPL.

When downstream users modify the code, they would most likely be bound by those terms as they create derivative works. In essence, the copyright holder can strike a bargain along the lines of “if you see my code and use it, you have to do so on my terms.” This is an ideal use for the GPL because its restrictions on “proprietization” of code travel with the code and attach to each recipient of the code.

Patent Law and GPL-ed Works

Patents also give their owners exclusive rights, and licenses give permission to exercise those rights under desired terms. However, patent rights are different in several ways. First, there is no independent creation as in copyright — if someone later develops a patented item without being aware of the patent rights of another, it is still patent infringement. Second, there are no rights to derivative works that fall outside the patent grant.

Each patent contains claims clearly specifying what the patent rights cover. Even after reading a patent, if another person develops something that is derived from knowledge learned in the patent but outside the scope of the claims, there is no patent infringement. As a result, there is no way to obtain a bargain wherein those who see your work and build on it have more obligations than those who do not see your work and build the same thing.

Thus, in terms of breadth of intellectual property protection, patent rights are narrower and broader, making it harder for GPL-style licensing. Patent rights are narrower in that anyone can build on what can be learned from the patent, so long as they avoid the claims. Patent rights are broader in that they can be asserted against anyone.

Open-Sourcing and Biotechnology

Open-source models could work for biotech developments that are protectable under copyright, but not much in biology is copyrightable. Gene sequences are not copyrightable because they are functional and are not works of authorship. A new method of transferring genes may be patentable, but it is not typically copyrightable. Thus, developing a new method and then freely licensing it does not assure licensees that other patent rights would not block them.

The burgeoning open-biology movement cannot assure those who create new developments that the creations can be freely used and still prevent proprietization of the technology. But proponents of open-source biology might be able to create a commons where information can be presented without restrictions. They might also create owned pools of patents licensed under GPL-type terms, with the understanding that not all contributions can be controlled and not all of the rights can be obtained.

The Science Commons is a new project of the Creative Commons (the organization that developed a number of models for copyrighted works licensing) for promoting the use and access of scientific data. The Public Library of Science was formed in response to the restrictive licensing of scientific journals that limited availability of that information. These movements are ideal for dissemination of information, but merely having information is not enough. Free use requires that recipients be able to use what they learn.

Open-source biology might work by just distributing developments and using community pressure to encourage everyone to follow the spirit of sharing. With dual licensing, those who restrict have to pay for that right, raising money to fund free development. If the patents were truly blocking, then they could be pooled and only licensed to those who agreed to open standards.

Duct-Taping a Solution

Even the most devoted fans of duct tape know that for all its versatility, it is only a temporary solution. It might hold the hoses in your engine together for awhile, but eventually you’ll need to replace them.

Maybe some combination of open-source licensing and intellectual property protection can create a viable biotechnology movement. But to be safe, don’t discard your leather wallet and cotton clothes just yet.


Phil Albert, a LinuxInsider columnist, is a patent attorney and partner with the San Francisco office of the intellectual property law firm Townsend and Townsend and Crew LLP.


2 Comments

  • Anything which is patented and has a standard unique representation in a text file must also be automatically copyrighted. DNA has such a form as a 4 letter language, A C T G, written in text files which feed into a DNA printer which outputs them in that same order. Since DNA is both a language and uses quantum properties to define behaviors statistically, it is a kind of quantum software, and any precedents about software apply to it, including that its a form of communication and can be licensed open source GPL or any other software license. That means any evolutions containing it, even if there are many other parts, must open source the whole life form, which is why they call it a viral license. GPL is a contract, but effectively we make law by using it or writing our own when it spreads. We build on precedents like the recent one that says Human genes can not be patented because they are discovered in existing strings of DNA, and removing parts of those strings is not an invention. If I were inventing new strings of DNA, I’d upload their text form to sourceforge, for example, with a GPL copyright statement, and let de facto standards handle it as it spreads into other open source.

  • One hopes that Mr Albert shows more attention to substance and research on behalf of his clients than he’s showing in this column. The institute in question in Australia is not Cabria, but CAMBIA. The statement that something ‘is not clear’ should state ‘it is not clear to me because I have not made the slightest effort to find out’. Anyone who wants to see the substance behind loose rhetoric, please go to http://www.bios.net at which not only are the license concepts and community concepts spelled out, but the formative documents of the activity (called the BIOS Initiative) and its community innovation space, BioForge are described in some detail. Also, on this site – as a fundamental tool to understand patents and their implications – is the largest, free, full-text database of patents in the world, including US, European, Australian and PCT (international applications) in OCR’s full text form, as well as INPADOC, the world patent status database. And the Patent Lens, CAMBIA’s IP transparency efforts, includes comprehensive ‘claims’ landscape analyses of key technology areas. This will be expanded most likely to make analyses of blocking in software that could hijack or torpedo the OS movement.
    Also on our http://www.bios.net site, are disclosures of ‘kernel’s of patented technology that will seed the development of Open Source toolkits as tangible examples of using IP not to exclude but to impose covenants and grantbacks onto licit, inclusive community activities.
    Its troubling that intellectual property professionals shoot from the lip without doing relevant research. Googling is not hard, nor is reading. Understanding, however, does take some intellectual generosity of spirit.
    I encourage informed discussion about mechanisms to craft innovative communities and the elaboration of ‘protected technology commons’ spaces. The concepts and strategies we’re ‘duct-taping’ together with serious effort by skilled and informed IP professionals, scientists, policy and innovation specialists and good hearted citizens are critical not ony for life sciences to have more humane and productive impacts, but will be similarly critical in the software community.

Leave a Comment

Please sign in to post or reply to a comment. New users create a free account.

LinuxInsider Channels