Or I was going titled this blog ‘The GPL is Not a Solution for World Peace’ but I didn’t.
Last week’s big news was that the Nokia is going to buy outright Symbian and then open source the Symbian operating system under the Eclipse Public License (EPL) and start the not-for-profit Symbian Foundation. Luckily we have analysts like Stephen O’Grady from Redmonk that provides a detailed and insightful analysis of the announcement. However, I was surprised how Nat Torkington’s post regarding Symbian’s license decision misses the mark on the importance of licenses in the growth of platform ecosystems.
Nat’s basic argument is Linux under GPL has:
Essentially, the Linux kernel developer community is the Christmas Truce for the Unix platform developers–a place where they cooperate rather than compete … because the license dictates that they do it.
and that because Symbian under the EPL and Google Android (under APL) allow for proprietary extensions so:
I think this will slow down the success of their platforms and means neither will unlock the true potential of an open mobile platform. I believe true demilitarized openness is a necessary, but not sufficient, condition for open mobile platform success.
I think Nat gives far too much credit to the selection of the GPL to the success of Linux. The choice of license is just one component to a successful platform; more important is the governance of the platform. Linux is successful and a place to cooperate because it is a vendor neutral community. No one vendor has control over the Linux kernel, so it is a safe choice for multiple vendors and developers to participate. If Linux was controlled by a single vendor, say Red Hat, then I highly doubt IBM, Novell, Canonical, Monta Vista, etc would participate in the evolution of the kernel.
A successful platform needs a vibrant ecosystem of individuals and vendors that provide products and services that add value to the platform. The key to enabling this ecosystem is an ‘Architecture of Participation‘ where vendors or individuals, can innovate, contribute and profit from providing value to the platform. To enable this level of participation there needs to be a well defined component model that allows for extensions by the community. If this is the case, then commercial products are just one way to participate and can in fact accelerate the success of a platform. Profitable vendors have a reason to invest in the success of the base platform.
Nat is correct that for Symbian to be successful, they need to reach out to a wider community of developers and vendors to support their platform. However, the GPL is not going to force vendors to embrace Symbian or create ‘a truce’. What Symbian needs to enable is:
1) a vendor neutral governance model that allows for multiple competing vendors to collaborate
2) an Architecture of Participation that includes a common component model that is used for extensions.
These factors are what will drive the cooperation and collaboration on the Symian platform. The formation of the Symbian Foundation is one important step in that direction but the execution of the governance will be important to watch. As Stephan O’Grady mentions:
what is the governance model? Who will have the right to commit code? How will that list be managed and grown? How will the project manage to rapidly innovate without compromising the stability necessary to handset devices?
Symbian’s selection of the EPL provides them an opportunity to reach the greatest number of ecosystem participants. Lets see how they do with the execution on the other factors.