I worry about firms capturing the open source movement. The most spectacular examples of open source emerged from the collaborative efforts of small groups joined in non-rival common interest. All those groups have suffered from very natural growing pains.
All institutions encounter two problems as the grow: coordination challenges and concentration of power. Those that steward the institutions respond to these problems by importing organizational devices from their personal experience. One of the reasons to study civics in school is to have a more enlighten less superstitious understanding of what you get from from the application of any given organizational device. It’s a shame that teaching civics isn’t a more serious enterprise.
There are standard ways to tackle the organizational problem, some amount of command and control are inevitable. You have lots of choices for example: laws, tools, standards, hierarchical organization, town meeting, representative democracy, etc. In designing solutions to the coordination problem it’s typical to pull something from bunch of different bags and combine them complex ways. For example my town as a representative town meeting, a board of selectmen, a town manager, a school board, and a school superintendent. All these, along with a large pool of professional practice and national and state regulation play a hand in coordinating various aspects of how we run the public schools.
Governance design isn’t just about solving the coordination problems. It is also about tempering the concentration of power that natural emerges in all institutions. What if the superintendent becomes too powerful? What if the town meeting decides to mandate home schooling for all?
The stewards of institutions pull solutions from their experience to solve these problems as well. For example progressive taxation is one way of tempering self reenforcing concentration of power from the distribution of income. For example in the US we have, until recently, feared concentration of power in the banking industry – so we had potent regulatory barriers to prevent banks from consolidating into a handful of national banks as is common in other countries. Or in another example when we build the interstate highway system we diffuse some the power of cities. The electoral college and the seats in the house of representatives hand power to weak or small states and in doing so disenfranchise citizens in large powerful states. That was intentional. It was a design to temper the power and coordination advantage of the elite. Apparently the founding fathers worried about the elite.
In the open source community we have not, I think, been sufficiently aware of these problems. Some people have, of course, but most people cling to the presumption that we can run all open source on a collegial town meeting model. Where coordination problems are talked out and where all parties are peers of similar power. Sadly, for some projects, that doesn’t scale. More of concern is that it is becoming clear that that model isn’t robust in the face the arrival at the table of a powerful focused individual or organization.
These concerns are one of the reasons that towns often shift from an open town meeting to a representative town meeting. An open town meeting is all well and good until a handful of over enthusiastic nut cases start showing up at every meeting and advocating some silly agenda – say mandatory taser ownership. Those who care at that point force the long tail of disinterested people into the voting booth to elect representatives. By design they temper the power/enthusiasm of the small group with the long tail.
How you govern open source projects is getting increasingly subtle. Open source is now a key part of the supply chain for huge firms and for the economy at large. This creates incentives for the customers of open source to come to the table, or worse to attempt to capture it. Not because of any evil agenda, but as a means to temper the risk of a supplier who’s nature is hard to understand. Mystery is risk.
One way we have tried to tackle this problem is licenses. These play a two sided role. One the one hand they act like the banking regulations as a way of limiting the ability of a single organization to aggregate a majority share of the open source commons. On the other hand they act to lower barriers to accessing the commons in a way not unlike the interstate highway system they energize the long tail.
In-spite of that we have a handful of very large aggregated points in on the landscape of the open source commons – FSF, Apache, the standards bodies like W3, Oasis, and the Java community. You can’t avoid the emergence of cities on a new landscape – you always get the power-law curve. But you do need to look closely at how they are governed.
You have to solve both problems. These institutions need to be well coordinated. These institutions need to be guarded against capture by the powerful players around them.
We need to solve these problems professionally and not by superstition and magical thinking.
The chart at right has a dot for