Douglas E. Phillips is author of The Software License Unveiled: How Legislation by License Controls Software Access, which reframes the debate between proprietary and free software to ask whether “legislation by license” should control either kind of software access. In the article below he looks at Google’s new Chrome operating system.
Google recently announced plans for a new PC operating system, the Chrome OS. Within minutes, the media reported an OS war and began telling us who loses and who wins. The Chrome OS will smash Windows! Or else Chrome is actually good for Microsoft and (despite running on a Linux kernel) threatens Linux! Who knows? We might actually have to wait until the new OS comes out, which Google says won’t happen for another year.
Whatever becomes of the Chrome OS, though, Google’s decision about how to license it tells us something about where the software license itself is going. And that’s worth looking at now, given that how software is licensed shapes how it’s created and used.
Google promises open source licensing, which points to an approach the company already uses for the Chrome browser. This approach combines a royalty-free source code license to developers with proprietary end user terms. Expect to see more of this hybrid model across the software industry in the years to come.
Source code for the browser is available under the BSD license, which is short, sweet, and (in licensing jargon) permissive. You can use BSD-licensed code in a new work and distribute the new work any way you choose.
The separate terms of service for the browser’s executable version don’t impose a fee, but they do authorize Google to deliver targeted advertising when you use it. So the browser – and most of Google’s other services – are free, but only in the sense that broadcast television is free (or was free, before cable and satellite found a way to charge for it). You pay (some would say dearly) for broadcast TV, not by sending money to ABC, CBS, NBC, or Fox, but by tolerating a high-fat diet of ads we hate.
You pay Google in more or less the same way, except that the advertisements usually tend to be more relevant. In return, you get access to content that compares to network TV, in scope and depth, as the Pacific Ocean compares to a puddle (taking into account that the Pacific Ocean also has its shallow parts and sharks). Most people seem to like this bargain. It works, however, only if you agree to let Google keep track of what you’re looking for online and serve ads in response. Without this proprietary element, there would be no Google as we know it and, among other things, no Chrome browser or Chrome OS. The hybrid license lets all this happen.
The main objection to permissive and hybrid licensing comes not from the Microsofts of the world, but from advocates of the GNU General Public License. You can borrow source code from a GPL-licensed work and include the code, royalty-free, in a new work. But if you want to share the new work with others, then the GPL requires that you also license the new work – including the parts that weren’t borrowed – under the GPL. In the GPL philosophy, protecting software freedom means denying the freedom to mix free code with any code that is subject to proprietary licensing terms.
According to Google, the Chrome OS will consist of the Chrome browser “running within a new windowing system on top of a Linux kernel.” Because the Linux kernel is GPL-licensed, Google will have to provide that part under the GPL. The GPL’s “copyleft” condition applies, however, only if the new work is a “derivative work” as a matter of copyright law. The Chrome browser clearly is a separate work, and the new windowing system apparently will be as well (as is the existing X11 windowing system, which is often used with Linux but not licensed under the GPL).
Google’s not likely to license the non-Linux parts of the Chrome OS under the GPL, and for good reason. Consider the Web itself. Tim Berners-Lee, who invented the Web, at first wanted the Web software to be licensed under the GPL. He changed his mind after people told him that the strings attached by the GPL could severely limit the Web’s reach. So he arranged instead to have the Web software dedicated to the public domain. The Mosaic graphical browser included parts of the original Web software, which meant that avoiding the GPL was key to making Mosaic something on which for-profit companies could build. Without the freedom to monetize Web technologies, the private investment that helped the World Wide Web live up to its name almost certainly never would have occurred. That’s why a combined open source and proprietary licensing model probably makes sense for the Chrome OS. It’s also why your next software license, like your next car, may well be a hybrid.