Wednesday, February 18, 2009

From User to Contributor

As open source software becomes more widely used, the flow of contributions back to open source projects become more important. Contributions back to an open source project are not only an indicator of the health of a particular project's community, but they are critical to ensure these projects are able to grow and realize the benefits of community involvement. Matt Asay noted in a recent Open Road Blog posting, a lack of contributions from enterprises "may have serious, negative consequences for the long-term health of the open-source ecosystem."

A recent article by Dan Woods in Forbes explains that this is the result of a gap between the "enlightened self-interest" of individuals who contribute for their own benefit, and institutional collaboration, which is "much lower than expected and hoped for, based on patterns of individual participation." While the Forbes article provides some excellent background on the problem and how we reached this point, I would like to dig a little deeper into the reasons why enterprises don't contribute, and what changes are needed to encourage more contribution.

Enterprises (primarily for-profit corporations, but also government entities and foundations) often have few incentives to contribute to open source projects, and many incentives to not contribute. At a fundamental level, the traditional thinking of enterprises has been that ownership of intellectual property is an important part of preserving corporate assets. Though the open source movement is changing these views, ownership and tight control has also traditionally been seen as crucial to preserving a competitive advantage. In addition, enterprises often suffer from the failings of bureaucratic organizational structures, which make the process of obtaining the multiple layers of necessary approvals for release of intellectual property rights difficult if not impossible. Finally, another basic concern of for-profit enterprises is that any time spent on open source projects is time taken away from core, for-profit activities.

No doubt the above described bias against open source contribution is the product of short-sighted thinking. While we should not expect that enterprises would change their position on openness in a short time frame, we can propose short-term, high-impact fixes that quickly and strongly demonstrate the benefits of open source and encourage more openness. Some specific examples include:

  • Appeal to employees of enterprises to voice their desire to participate in open source projects, and use resulting the groundswell of employee requests as a way to incent enterprises to develop a well-reasoned participation policy rather than perpetuating ad-hoc or unknown participation.
  • Create model policies that enterprises can easily adopt to permit employee participation and facilitate enterprise contribution, with an emphasis on the benefits of participation over the traditional focus on ownership and control of IP rights.
  • Clearly demonstrate cases in which contributions create a win-win situation for enterprises and communities, possibly in a public white paper format.
  • Begin with the low-hanging fruit: contributions related to internal IT infrastructure components rather than contributions that are perceived to compromise competitive advantage; tackle the issue of contributions related to competitive-advantage at a later time after the benefits of contribution are clear.
  • Acknowledge that strategic and competitive situations might exist that warrant withholding or delaying making contributions, while emphasizing that it is a rare case when choosing not to contribute is the best course.

Some of these are aspirational, but the underlying themes are important: (1) Every technology company has a massive base of employees who are interested, if not heavily involved, in open source, and these employees can be the agents of change within enterprises. (2) Enterprises need to better understand the economic and other benefits of contribution, and the lessening importance of ownership and control of IP rights. (3) This movement should occur incrementally by starting with the types of projects that are least likely to encounter the traditional organizational resistance.

These changes will happen on their own over time, but we can establish an environment that encourages enterprises to speed progress. At stake is the ability for open source projects to survive and reach the critical mass necessary for all community members to realize significant benefits of the open source movement.