Where open source hasn’t won
Open source has victories, but there are battles that still need to be fought.
A few years ago, we proclaimed that “open source won.” That may be true, but it’s not exactly a battle cry. What do you do after declaring victory? Go home and have a party, I suppose. Start up a few would-be unicorns, or get a job at one of them.
What we didn’t do was look seriously at what’s not open. Open source may have won, but it’s a victory that was limited to programming languages, databases, frameworks, and tools. That’s a fairly small part of the world. What else?
I’m not an open source purist; I don’t think that all code wants to be “free,” either as in beer or as in liberty. But I believe strongly that there are many areas where we need to struggle for increased openness; if we don’t, we’ll pay a very high price. So let’s try to “think different” (as Apple used to say). Open source has won, in at least some parts of the landscape, but what battles still need to be fought?
In the last few years, we’ve made a lot of progress with opening up data; but we need much more. Who is collecting your data, and what are they doing with it? What data do governments hold that should be released to the public? And, more important, what form is it in? Open source advocates would laugh at a company whose “open source” release consisted of a set of PDF files. In data, that’s all too common. We also need to ask what data should not be open: what data is legitimately private.
Open data models
Cathy O’Neil has frequently argued for the importance of open data models. As our future becomes ever-more intertwined with “big data,” we desperately need to know more about the models that are predicting our actions, driving our economy, and policing our streets.
Open UI and UX
I made this one up; I don’t really know what “open UI” means. But I’ve spent plenty of time with open source desktop tools that have horrendous user interfaces. Open source developers have never had a good handle on UI design. UI design is hard, the biggest users of open source software are typically other software developers, and software developers are notoriously tolerant of bad UI. How do we engage UI designers in the open source process? “Open” isn’t just about source code.
Open source biology
Who owns your genome? What does that even mean? A gene may be a piece of code (a sequence of letters) that could be subject to a software license. But an “open source gene” makes little sense without instructions about how to synthesize and deploy it. Open source biology may be the most important issue facing us; it’s also the most complex, with a tangle of ethical, legal, and practical issues.
In the past few years, we’ve seen plenty of interest in publishing models for 3D printers online. And that’s a small victory for open source. But there’s more. We won’t get far with open hardware if we don’t also have open protocols so our open hardware can communicate. A bigger and nightmarish issue facing the Internet of Things is billions of devices that are difficult, or impossible, to update. These devices are running out-of-date software, full of security vulnerabilities, and they will still be online in 10 years, 20 years, maybe even 30 years. Many of these devices are running an open source operating system, but few of them can (or will) be updated. The right to update is one of the more important (though less discussed) implications of open source. In the hardware world, that right is often forgotten.
Open source may have won. But there’s a lot more to “open” than source. It’s not time to start the party. And it’s certainly not time to rest on our collective laurels.