Center for American Progress

Encouraging the Production of Open Source Software

Encouraging the Production of Open Source Software

The National Open Source Fellowships and the National Open Source Prizes are a Commitment to Encourage the Growth of Our Collective Commons

Open source software is a contribution by corporations and by individual citizens to strengthening our Internet infrastructure, growing our economy, and making our society a better place. Open source software becomes a fundamental building block in commercial software. For example, Microsoft Windows incorporates many important open source components in their popular operating system. Open source software also answers needs that are not met by commercial software producers: the creation of solutions that benefit the community at large but may not be ready yet for the marketplace. For example, “Wiki” software was used to create the Wikipedia, a collective reference work that rivals the encyclopedias produced in prior eras. Open source software facilitates learning and the dissemination of knowledge. Open source software is a public good and deserves public support.

What can the U.S. government do to further encourage this important phenomenon? The government can take steps to encourage the production of new software and to reward the best software that has been produced using two simple mechanisms:

  1. A program for the awarding of National Open Source fellowships, available in equal numbers to students and to working professionals. Students who are awarded the fellowships would be able to cover the costs of their tuition and living expenses while producing a piece of valuable software that is available as open source. Likewise, a working programmer who was awarded such a fellowship could treat it much like a professor might treat a Guggenheim fellowship: an opportunity to take some time off from their job and devote full-time to the production of something they care deeply about and that benefits the community.
  2. A juried National Open Source Prize to reward the open source software products that have most benefited our collective commons. The fellowships encourage people to produce new software; the Prizes recognize real results that have already benefited society.

The Department of Education would be a logical home for such a program. However, it is important that the program not become a large government bureaucracy and instead leverage the resources of the community it is meant to help. This means that fellowship applications should go through a peer review process, the same kind of process that, for example, a National Science Foundation research proposal might go through. Selection of the winning fellowships would likewise be awarded through a selection committee drawn from the community. Selection of prize winners would similarly use a peer review system to filter initial applications and would be chosen by a distinguished jury of leading experts. This “open source government” approach means that the Department of Education program could be quite small, with perhaps a staff of five people.

The budget for the entire program would also be small by government standards. A $10 million fellowship fund would be enough to fund 200 National Open Source Fellows per year at a meaningful level. Likewise, a $5 million prize fund would be significant enough to make a real difference to the open source community. Even with government overhead and staffing, the total cost of the program would be well under $20 million per year.

How would one define what constitutes an open source project that would qualify? Again, we would leverage the existing work of the community. The Open Source Initiative is a 501(c)(3) non-profit that has developed an “OSI Certified” mark and has qualified many different license agreements that meet the community-developed definition of what constitutes open source. To qualify for a National Open Source fellowship or the National Open Source Prize, the software would have to bear a license agreement that bears the OSI Certified Mark.

We propose that the fellowship be available to individuals, though individuals might have the support of their employers. It is not inconceivable that a company might provide matching funds to any successful employees that win a fellowship, providing a sort of sabbatical program. However, it is important that the application process be based on individuals, not institutional affiliation, to encourage the broadest possible participation.

While fellowships would be awarded to individuals, the National Open Source Prize would be available to groups of people who develop software. The groups in some cases could be corporations, in some cases unincorporated associations of individuals. One of the requirements in applying for the prize would include an agreement among the individual contributors as to how the prize would be distributed. Some groups might choose to use the money to establish a foundation to govern further work; others might choose to distribute the proceeds among their members. Part of the due diligence of the selection process for the prizes would be to ensure that those applying for the prize do indeed properly represent the collective enterprise that produced the software.

In summary, a $20 million program by the U.S. government would provide a significant incentive to open source developers. fellowships would encourage the production of new software, and prizes would reward software already produced. Heavy involvement by the community in the peer review and selection process keeps the overhead of the program low and makes the program a form of “open source government.”

Carl Malamud is a Senior Fellow and Chief Technology Officer, and John S. Irons, Ph.D., is the Director of Tax and Budget Policy.

The positions of American Progress, and our policy experts, are independent, and the findings and conclusions presented are those of American Progress alone. A full list of supporters is available here. American Progress would like to acknowledge the many generous supporters who make our work possible.