My IFCLA banquet note about forking and IPR law
Below is my talk from the International Federation of Computer Law Associations conference banquet that took place in Helsinki last week. (It is post-edited to match what was actually said.)
I have to say I was quite honored to be asked to speak. I was preceded by Finlands Minister of Justice Tuija Brax and later in the evening followed by imho Finlands funniest magician Martti Vannas. The dinner was set in the old stock market building of Helsinki, an exquisite restaurant now. I'm happy to say the talk was well received and many of the lawyers came to thank me afterwards.
Also for myself it was an inspiring event to talk to an audience that was different from anything I've done before. By a coincidence, both mine and Mrs Brax talks touched on human rights topics so it made a nice continuation too.
International Federation of Computer Law Associations
Helsinki June 10-11th
Gala dinner, June 10th
Practicing key freedoms of free software - forking MySQL
What is the difference between an engineer and a lawyer?
Actually, it's not the beginning of a lame joke or anything. I would like to share some thoughts about what it means to be an engineer. So maybe afterwards in the bar, you can tell me if lawyers are different or similar.
We have a saying: If it isn't broken, don't fix it. But on the other hand, if something is broken, then it is my call as an engineer to fix it.
As an engineer, it is my goal in life to change the world, and to change it for the better. Take the dishwasher for example. So many hours of my life saved to do something else instead of doing the dishes. It was created by an engineer.
The invention of open source is like a grand slam from an engineering point of view, because it revolutionizes things on so many different levels. Most technical innovation - I claim - today happens within open source and it is considered by many a superior way to develop software. But that is just the tip of the iceberg. The concept of open source, or free software as it was originally called, is possibly a bigger legal and cultural innovation than the technology. This concept is now spreading into areas like open content, open data and open government. Even activities that don't strictly qualify as open source - like sharing a video on YouTube - are all about sharing. It is a powerful paradigm that sharing and giving away is more valuable than protecting and excluding. Talk about changing the world for the better!
The GPL license may have been written by a few lawyers, and looking at the newest version it seems the more lawyers that are included, the worse it gets. But it makes me proud to think that it's main author and original inventor is a software developer. A developer who wanted a better way to develop and use software, and took it upon himself to fix the system for doing that. Which in this case is not a computer system, but the copyright system.
I'm supposed to talk about my work at Monty Program. While I pretend to do that, I will first give you an exercise: What legal concepts or topics are relevant in the following story?
At Monty Program we run what is called a virtual company - just like we also did when we worked at MySQL. My morning routine for getting to work is essentially just to get out of bed and turn on the computer. This allows me to sleep about 1 hour more each night compared to having to dress up and commute to an office.
Starting my computer takes about one minute. After that I use various internet technologies to connect with my colleagues who also work from their home offices in Russia, Ukraine, Bulgaria and Germany, plus some Americans who really should be sleeping at this time. We will discuss MariaDB development with the engineers, and the sales pipeline with our sales guy and so forth.
Running a virtual company also allows us to hire the best candidate for the job, without being dependent on his location. You don't have to live in Silicon Valley to work on MariaDB.
When one of our architects graduated, he was offered a prestigious job at IBM Research labs, and a job at MySQL. The IBM job probably was a better option by all accounts. But because his wife wanted to live in Bulgaria, he chose MySQL. You might say that this is yet another kind of freedom enabled by the internet and the open working culture we have.
OK, so back to my legal quiz: What relevance does freedom of speech and free communication have with the above picture?
Well, it turns out many of our best engineers are from Russia or Eastern Europe. If we go back in time 20 years - just like the Minister of Justice also did... And we go back to the USSR, so to speak :-) it would have been unthinkable for an engineer in St Petersburg - or rather Leningrad - to connect on a daily basis with engineers in Western Europe and the USA. We simply could not work with him the way we do now because he would not be allowed to. He could probably go to jail simply for speaking English too well.
I'm here tonight as a lonely engineer being surrounded by lawyers, and apparently even some legislators. So I thought maybe it is appropriate to make the point that while the internet is the technical enabler that allows me to collaborate efficiently with my colleagues, it wouldn't be possible without some important basic rights, like freedom of speech and free communication. It is due to such legal freedoms that we have been able to create open source innovations like MySQL, which at Monty Program we now call MariaDB.
Which brings us to the next legal issue to discuss. So while MySQL's open source license allows us to take the code and continue working on it, we have to change the name because of trademark law, otherwise we'd get into trouble.
Ok, so the American lawyer is nodding his head there. I've noticed a big cultural difference here between US and Europe. When we go to conferences in the US, everyone is always coming to us saying: Oracle will sue you. They always sue everyone. Larry Ellison even sued his yacht club. In Europe people don't think about that at all, they just encourage us to continue forking MySQL. Well, we'll see. So far they haven't done anything. If you ask me, it would be great PR for us if they did sue.
So that brings up the next question: Is trademark law a good or a bad thing? And is it a good thing that copyright legislation enables us to fork MySQL, as it is called? And in general, what kind of laws and legal principles would be beneficial to open source, or just innovation in general?
Open source is collaboration among individuals. Those individuals often are employed by some company, but fundamentally, the powerful open source communities consists of individuals. So just like with the above example of me communicating with my colleagues in Russia, anything that empowers the individual, like freedom of speech and other basic human rights, are usually good for open source.
So far so good, we are all for human rights. I hope. [fingers crossed]
So let's look at intellectual property laws with a similar mindset.
Copyright in my opinion is a very successful concept. The crucial success factor of copyright legislation is that it gives the creator of a work the copyright. The software engineer doesn't need to hire a lawyer to somehow register his copyright, it is automatically his. If someone wants to use his software or acquire rights to it, it is up to that other party to come to the creator and negotiate with him and the developer can choose the terms he licenses his work under.
Copyright has low administrative overhead, it empowers the individual who actually does the creative work, and it scales well. These properties are fundamental to why open source works so well, because it is possible to form large communities of software developers who can choose to share their work with each other.
Unfortunately recent additions to copyright law, and I mean legislation that try to address the issue of Digital Restrictions Management show an opposite trend and have strayed away from this successful principle. Quite apparently that legislation has had zero effect on piracy during the last decade, but at the same time it has allowed businesses to use it for centralizing and monopolizing their customer base: You can buy music from Apple's music store that will only play on a device manufactured by Apple, and it is illegal for you as an individual to do anything about that. In this power play there are also casualties, for instance blind people are now unable to have their e-books read out loud, because it is illegal for an engineer to create the software that could do that. If you ask me, there should not be laws to protect that kind of technology, there should be laws against it instead.
The existence of open source is actually a testimony to how brilliant and versatile the copyright system is. While open source licenses originally were seen to be somehow anti-copyright, they actually rely on copyright legislation themselves and are able to use it to their advantage.
The reason why MySQL chose to be open source and why they chose the GPL, which is the most popular open source license, was simple. There was a demand for a good open source database and MySQL decided to pursue that demand. From a business point of view it was a choice to maximize user adoption and popularity and to get embedded as a component in the Linux stack known as LAMP.
The copyright system enabled and supported this kind of strategy. It is designed in a way that it is more or less neutral with respect to the strategy chosen by the rights holder. It could easily have been the opposite, we could have a copyright system that was specifically designed to protect the interests of proprietary software companies, large movie studios against indepents and book publishers against the authors. But it is not. The system allows the rights holder to choose their own strategy, and will protect any strategy they choose.
And so MySQL became popular. So popular that it was last year acquired by its competitor. I and many others resigned and continued to work on the code in a new company.
Now, if you are an IPR lawyer, it is probably your job to think of ways to stop that from happening. If on the other hand you work on the software procurement side, this is good news for you. Your database provider was just bought by its main competitor, but you don't need to worry. Whatever the new owner does, the open source code will live on. You can continue to ask for offers from other vendors. You are in control of your destiny. Like I said, this was a key reason for many early MySQL adopters to choose MySQL.
Our developers that now work on MariaDB have spent from 10 to 25 years working on MySQL. For some of them, including Monty himself, it is essentially their entire career as a software developer. The result of their work - the code they have created with their own bare hands is what keeps a majority of the world wide web today running. Needless to say there is a big emotional attachment and pride in that work.
Now imagine if they had been working for a proprietary software company. If you get fired, or leave the company, or the product line is discontinued, then you don't have access to your life's work anymore. In an acquisition the destiny of your life's work is in the hands of the highest bidder. But since the MySQL code is open source, we can continue to work on it and enhance it.
In business terms I would say that open source is all about risk management. As long as there is interest from developers and customers, open source minimizes the risk that the project would fail due to poor management or hostile takeovers. ...both of which, by the way, apply for Sun and Oracle respectively. The risk is minimized by not allowing managers and owners to control the project.
So what about trademarks? Trademark law kind of fares well in this comparison too. When he started his new company, my boss Monty neglected to register the domain name montyprogram.com. So within 24 hours it had been registered by a Chinese man. Even if we had not registered Monty Program as a trademark, we were able to argue the case for a so called common law trademark, and eventually the domain was transfered back to us. But what we also learned in the process was that if we had gone through the effort to make the company name a registered trademark, we would have gotten our domain name back much faster. Of course, if Monty had thought of such issues to begin with, then he would have also registered the domain name and not gotten into this trouble at all.
So trademark law kind of combines the best of both worlds: You can get a trademark without the legal overhead of registration, but you can make your trademark stronger and more secure by actually registering it.
Which then brings us to patents.
Usually when people criticize software patents, they focus on issues like the poor quality of granted patents, evidenced by facts like so and so many patents are invalidated at trial. And yes, that is of course a problem, but to me that is more like a symptom of the real problem.
From my perspective, the whole patent system is a problem, because the system is designed in a centralized way with high overhead of legal costs. An individual open source developer cannot realistically apply for even a single patent, let alone actually enforce it against anyone. So the system is beneficial to those medium and large companies that have a powerful legal department rather than a powerful engineering department. This is the exact opposite to the copyright system that empowered each individual engineer in a distributed, low overhead fashion.
The other obvious problem is the opaqueness and arbitrariness of the patent system. For copyright, I can always know that the code I created with my own hands is owned by me. If I start to worry about software patents, then there is no way of knowing whether somewhere out there there would be a patent that could be used against me, against some work I've created completely by myself. For this reason, from an engineer's point of view patents are often seen as purely a playground for lawyers, with very little connection to the actual innovative work done by engineers.
And the poor quality of granted patents is of course just a symptom of the centralized way patents are granted. The collective intelligence of the attorneys that file patent applications will always outsmart those poor overloaded examiners at the patent office. It is basically an unfair competition. In engineering we call such a centralized system a single point of failure. If you are building a cluster, you want to avoid single point of failures. If you have one, it means it is just a matter of time before something goes wrong.
So why am I speaking about this tonight? I just thought, with so many legal experts present, maybe one day some of you will be in a position to influence legislation, or case law, or just contracts and common practices.
In fact we just spoke about it at my table, that for a lot of things we come up with, there is no legislation and we just make up the practices as we go. Google Street View, Google Books, YouTube, are they legal or illegal? The idea that someone would go out and photograph all of Finland is so crazy, of course there is no legislation that covers it.
So if you are ever in a position to change our legal world for the better, I wanted to give you some hints what properties to go after. I'd wish you could design legal systems that
* empower individual engineers,
* that have low legal overhead and scale well
* that make it easy to collaborate and share
* and that are as neutral as possible with respect to the technology or business strategy being used. [this last point was added after the talk]
If you do that, I can promise you we will continue to innovate like you've never seen before!
On this note I'd like to end with a citation of Michael Tiemann, Vice President of Open Source Affairs for Red Hat. This is from his testimony in a recent trial where Red Hat and Novell invalidated a patent that had been used by a patent troll to attack them:
-- in the case of proprietary software, it's typically developed by a small number of people working in secret on the program. And they -- the total number of people who are working on that is necessarily limited by the total number of people within that company.
By contrast, in the world of open source, the entire world, or at least all those who are connected to the internet, can potentially be developers. And that was one of the things that really sparked my imagination about open source.
I've been teased by my counsel about being a smart kid, but in 1987, I knew immediately I was not the smartest kid in the whole world. And so one of the exciting things about open source was I always had a chance to work with people smarter than me. And when I talk with open-source developers, every single one of them has the experience that there is somebody smarter than them teaching them something. So we have a motto in the world of open source, which is nobody is as smart as everybody. That's been my experience, and that's been the experience of all the people I've worked with in open source.
PS: It was interesting to note at which points the lawyers would laugh:
- The note about GPL getting worse the more lawyers are involved.
- My account of Monty not registering montyprogram.com and related trademark issues seemed to be funny all throughout.
- Btw, the "American lawyer" nodding his head was Lawrence Rosen.
- Something in the section about open source being a form of risk management was funny, but I cannot know if this was calling Sun management incompetent or just the point of developers not allowing managers and owners to control their project. Or it may have just been late laughter about the trademark story.
- The concept of a single point of failure was for some reason very amusing.