Recently, there were two enormous acquisitions in the world of Open Source Software. Back in January, Sun Microsystems announced it was purchasing MySQL for $1 billion. About two weeks later, Nokia announced it was acquiring Trolltech for $153 million, about a seventh of what was paid for MySQL. Both purchases have huge implications for Open Source – MySQL, despite its shortcomings, is the most popular database for web application stacks, and QT, the major product from Trolltech is the foundation for KDE and a variety of other cross platform graphical programs. In both cases, the new parent companies have pledged that little will change with regards to the Open Source offerings. Nokia has made it clear that they will continue to support Maemo, which is based off GTK+ and GNOME on their “n-series” of internet tablets, while moving to a QT based user interface for their mobile phones.
Whenever acquisitions like this happen in the Open Source world, it naturally raises some concerns, despite pledges from the companies. It wasn’t long after Novell purchased SuSE that people realized that something was at odds with their prior purchase of Ximian – the two desktop environments were diametrically opposed to each other, having standardized on different widget toolkits and development toolsets. Eventually, the influence of the Ximian people won out, and SuSE changed to have more of a GNOME focus, although KDE still seems to be the dominant environment on the platform. Community members also start to fear a loss of voice within the community and may start to feel alienated from the new corporate project overlords. But is this really a concern?
For more insight, lets look at the seven largest Open Source acquisitions (original source):
What is the common thread between all of these topics? They’re all predominantly commercial type projects to begin with. Sure, there were always hobbyists who used MySQL, Cygwin, Xen, SuSE and QT, but the majority of development was still done in the cathedral. Zimbra and JBoss are almost exclusively designed for enterprise class deployments, with miniscule community outside of the main development. The fact, is that despite their appearances and the availability of source source code, most of these projects never were the type of Open Source that most people participate in for fun or skills.
Rather, these projects are a different class, business open source. They’re fundamentally different from projects such as GNOME, KDE, or even Mozilla. In most cases, these projects are backed by outside investors who seek to build a business around the project, the community is just a side effect of the process. While they have community forums, mailing lists, publicly accessible source code repositories, and bug trackers, who really calls the shots when it comes to development? In most cases, while the decisions are influenced by the community, the primary sponsor of the project makes all the critical decisions. Leaving little ability for the community to experiment with new implementations or ideas – for fear their work will be rendered obsolete in the next release.
This has an even more insidious effect – it robs the community of a skilled development pool. Rather than providing volunteers the ability to experiment on these ideas, by managing the project in a hierarchical environment, these individuals are given little chance to learn on the code base, possibly stripping the sponsor of a valuable talent pool. It also reduces the affective attachment to the project of the community, reducing most of the participants to a utilitarian perspective, willing to leave if another project provides enough benefit to overcome the switching costs.
The traditional view of open source has been that the community owns the software, and they make the decisions around it. However, this clearly is not always the case. When considering participation beyond the utilitarian level, be sure to ask yourself if your voice will be heard and your contributions will be valued. Otherwise, you may just be contributing to someone else’s bottom line.