My colleague Dan Woods just posted one of his typically thought-provoking questions on his Blog at O’ReillyNet. Referencing Jared Diamond’s superlative (and Pulitzer-winning) book “Guns, Germs and Steel: The Fates of Human Societies”, Dan asks which one of the two software creation models will come to dominate the other — Open Source or Commercial?
While the clash of these titans is tempting to consider, my first thought was that a different book than Diamond’s was more appropriate, one that didn’t suggest such a huge difference in originating circumstances between OSS and COTS.
Geography
In Diamond’s thesis, physical conditions of the geography and the distribution of high-protein grasses for farming were the root causes of the ultimate dominance of European societies. But the differences between Open Source and Proprietary Software don’t spring from different origin conditions that affected their rate of development on a linear scale.
Economics
Instead, a different book came to mind, The Wealth and Poverty of Nations by David S. Landes. Landes’ fantastic work strikes me as a great companion to Diamond’s, where Landes takes an economic historian’s perspective on the same fundamental question of why some nations came to dominate others, with entertaining and insightful lessons from around the globe over several thousand years of history. Landes pays much more attention to cultural preconditions, and the power of ideas and organizations within a society. It’s particularly fascinating to learn that Imperial China routinely produced technical innovations but a variety of cultural factors suppressed their development, and didn’t value what they offered. Meanwhile, radically different cultural conditions in Northern Europe produced a nearly opposite effect.
History
I see OSS vs. Commercial as less about indigenous peoples versus gun-toting and germ-bearing white guys than it is about England versus Spain, or some other case where the resources may vary and the basic knowledge is similarly rooted, but the cultural, philosophical and religious differences took two different groups who lived very close to each other in dramatically different directions. And in many cases this happened in a short period of time. I don’t say England vs. China because the origin differences are too great compared to software. I’m even uncomfortable with the analogy of Open Source to a nation-states or commercial groups in history. Is Apache Software Foundation the East India Company? Or if that’s Microsoft, then is ASF the British privateers?
Softwarity
Setting aside analogies and trying to take them on their own terms, both Open Source and Commercial Software each have a lot going in their favor. Software runs the world more and more every day, and commercial software (with all its flaws and failures) has played a major role in that growth. Core technologies of the Internet may have come out of DARPA and .edu but commercialization has been a driving engine of expansion and evolution throughout my professional lifetime. It seems to me that there’s a far more symbiotic relationship between the two than most people acknowledge. Not so black and white.
Open Source has influenced me greatly over the years, and I know I’m not alone among my peers and profession. I think it’s unlikely that we’ll see one of these models dominate the other simply because they tend to serve different goals and do different things well. There are undeniable economic factors at play — I believe the backlash against the high margins of PC software in the 90′s fed Open Source a great deal, and IBM may soon enough be declared visionary for their embrace of it as a foundation for growing services revenue — but I’m not ready to see a world without commercial software vendors, and that’s not just wishful thinking. As Dan points out, there are some problems that just don’t make it onto the OSS radar, but have significant value to businesses and consumers.
Merge
There is a different path to consider, in which the two begin to blend. I hear a lot of commercials on public radio these days, and major supporters of Apache with three-letter corporate names. Huge patent holders like IBM are sharing their intellectual property, and even Microsoft is exploring shared source initiatives. Some of my favorite commercial software companies have an open source license but you pay for the use of their intellectual property (at a great low price). Already the rise in value of Open Source software means that money is being poured into it by those who expect to get some back. So software providers of any stripe will continue to be shaped by financial support that makes creation possible, and the demands of the market for what’s created.
WYSIWYG
In the end I want my software to be innovative, cheap, fast, stable, supported, transparent, extensible and secure. I find a lot in Open Source that does well in many of these criteria, and the next generation of commercial software is going to have to compete with what’s in this new model. I expect that it will, and it’s going to do that by learning from open source, and building on it. In many ways, Open Source is pointing the way to and stimulating the next generation of software. But that doesn’t mean we can get the business value we need from it without commercial ventures that make it possible to do so.