Over the past three months, I've been communicating at length with several leaders in the Open Source Software (OSS) community about how best to license software patents in a way that supports the goal of OSS developers, users, and distributors. I've learned a great deal along the way about the uneasy relationship between OSS and software patents.
Briefly, OSS products:
…are systems whose human-readable ("source") code is always freely available to anyone who is interested in downloading it. This is in contrast to most commercial software, whose source code is considered intellectual property and a trade secret not to be disclosed. Advantages of open source include availability, extensibility, and the opportunity for peer review. Open source products are made available under a variety of licenses. …
Open source solutions have a number of advantages for a healthcare enterprise. The collaborative sharing of ideas and concepts practiced by users of open source software can create 'communities' of developers, partners, testers and users who interact with each other to further improve the software. This can speed up the development process, bringing in skills that a single software vendor would not be able to provide. And the community can also provide an alternative, though unconventional, avenue for technical support….At the data level, an open source software application does not strand critical health data in a proprietary format. [Reference Link]There are two types of OSS licenses:
- "Permissive" licenses—also called the "BSD style" license, come in many styles. They allow "derivative works" (different software programs in which some OSS code is embedded) to be used in proprietary (i.e., non-open source) software programs, which, I assume, makes BSD a type of "Commercial OSS (COSS)" license.
- The "Copyleft" license, which comes in only one style, and is promoted by the General Public License (GPL) published by the Free Software Foundation. Unlike the BSD licenses, the GPL requires the original source code and all derivative works to be free and open to anyone for any purpose, that is, they can never be used by proprietary programs. The GPL is one of about 50 different Free Open Source Software (FOSS) licenses.
OK, my conversations with leaders of the Free Software Foundation made very clear the deep-rooted resentment their community has against software patents. They claim that:
- ALL software patents (but no other types of patents) are "evil"
- ALL software patents are invalid or trivial, i.e., prior art already exists, but the patent office didn't realize it, or else the patent is for something that is obvious given the state of the art
- ALL people and companies who (might) sue for infringement of their patents are "predatory, unethical, greedy, freedom-destroying gangs of bandits" who don't care about hurting the world in order to make a quick buck at others' expense
- ALL software developers and distributors who use free software must be protected from lawsuits by any patent-holder, even if the developers and distributors make a profit by use of someone else's patented code/techniques.
As I understand it, there has been a rash of invalid/trivial software patents granted in recent years. Take Amazon's 1-click technology, for example. This patent claimed that running a common software process using a single mouse click is patentable compared to running that same process using several mouse clicks. How novel and unobvious is this? Not very! So, the free software community argues that it is justified in defending developers who use any type of one-click technique—or any other patented software techniques—against any infringement lawsuit. [See this link for more] I can certainly sympathize with them and do see their position as valid.
There is one aspect of this argument, however, that is invalid in my view. It has to do with the way they consider ALL software patents as trivial/invalid, and ALL patent-holders/inventors as evil, greedy, menacing predators. My psychological training has taught me that very few things are so black & white, and such gross generalizations are usually faulty. It would be much more rational to claim that (a) SOME software patents as trivial/invalid, or even MOST if you can prove it, and (b) SOME patent-holders/inventors rely on underhanded, unethical, predatory practices driven by solely greed. Failure to make any such distinctions weakens their case considerably! I've also learned that in adversarial situations, the greatest understanding and best resolution often comes from looking at things from both sides at the same time and then moving toward a "middle ground."
For example, imagine you are an inventor of a patented software technique that is neither trivial nor invalid, and it has great potential value. If you were ethical and truly want to contribute something beneficial to the world, you want to offer that patented process for free to nonprofit/noncommercial organizations, and maybe even to individuals for their personal use.
Further imagine that your family has sacrificed a great deal in the past two decades, and that family & friends have helped support your efforts by taking a big risk investing in your company. So, you figure it's reasonable to receive a small licensing fee only from companies and individuals who would make a profit by using your invention in their own software programs. After all, you reason, it's only fair since you've spent many thousands of hours and dollars developing your invention, and you would be helping others make money.
Now imagine you want to use an OSS license to offer your patented technique to nonprofits and for personal use, and you explain this plan to the FOSS folks. As soon as they hear you would be requiring profit-making companies to pay you a licensing fee, including profit-making FOSS companies, they threaten to defend those FOSS companies if you were to sue them for infringement, should they "happen to get in the way." That is, they would try to destroy you if you sued any profit-making FOSS company that uses your patented technique and refused to pay your royalty because those payments would eat into their profits. In other words, you, the inventor, would be consider unethical, greedy, and menacing because you have no right to restrict the profits others gain by using the invention you worked and sacrificed for years to develop! Maybe it's just me, but this all sounds very counterintuitive and outright unfair!
If you were that inventor, what would you do … how would you feel?
Well, that's precisely the situation I found myself in! I admit that it took me a while to figure out why I was feeling attacked, yet was being accused of being the attacker. But since I believe in the value of free open source, and since I do believe the FOSS folks with whom I've spoken are decent and honest people who have been abused by our patent system, I didn't want to become defensive, hostile … or give up. So, I offered several compromises.
Back in April, when I first started to consider open source, someone suggested the Trolltech "dual-licensing" model. It's an interesting approach, but not appropriate since as it is now written, it is for the dual licensing of copyrighted code and has nothing to do with patents. And then there's a Patent-Based Dual-Licensing Open Source Business Model, but that doesn't apply to me either since it distinguishes between software code alone and code used in a computer.
So, after lengthy consideration, I suggested that I would be willing to offer the patented technique for free to everyone in the FOSS community, as long as it was tied to a particular type of software program. That is, it could be used in a PHR, for example, but not in an EMR or CCR.
I was informed, however, that this constraint is called a "field of use" restriction, which is "completely antithetical to the purpose of the license, which is to make it both clear and legally binding that all recipients of the program have every right to copy, modify and redistribute the code for any purpose [italics added], so long as they don't limit anyone else's rights by e.g. changing the license terms on modified versions of the work. Such a modification of the license…would vitiate [i.e., trivialize] its purposes completely [which is why we] use the copyright on the license itself to prevent such modifications."
I then suggested that I strip out the code in my programs that run the patented technique and offer the remaining source code (which includes trade secrets) for free in a fully functional program under the liberal GNU GPLv3 FOSS license. I would concurrently offer the patented technique under a separate commercial license. Anyone who wanted who wanted to combine the code in the free programs with the patented technique, or use the patented technique in other programs, could do so by paying a small licensing fee via the commercial license. Unfortunately, they didn't see this as a good compromise because it precludes free software developers from using it, which "restricts their freedom."
Due to the lack of an adequate alternative, I concluded that the safest and fairest way forward is to start by publishing for free (via GNU GPLv3) at least one fully functional software program (including trade secrets); it will not, however, contain any code related to our patented techniques. Elsewhere, I will offer a commercial license for our patented technique to those who want to profit financially from its use. The patented technique will not have any "field of use" restrictions, but will require a small royalty fee. I expect this to be done in the next month or two. Depending on the interest and response from the OS communities, I will revisit this issue and decide what to do about our other programs.
I believe that by being open, honest and fair with the people in the OS communities, there will be no need for patent infringement litigation since developers and distributors will appreciate my position, realize that we're not "gangs of bandits," and act with integrity for a win-win relationship.
I welcome any comments & suggestions and will respond, although I'll be available only sporadically over the next two weeks.