home2barns bookwriting

The Business Schools of Harvard and MIT consider Free/Open Source

(c) 2003 David Stutz

The Harvard Business School and MIT’s Sloan School combined to sponsor a Cambridge symposium last week that explored (and celebrated) the free and open source movements. Karim Lakhani and Eric Von Hippel left the door to the ivory tower ajar for me, and I found myself able to participate. This is a quick wrapup of some of my own observations, made after interacting with this very interesting group of people.

The Sword of Damocles

This has certainly been noted elsewhere, but patent infringement continues to be the most notable threat to the long term health of the free and open source movements. Two different threads from this conference combine to hammer this point home:

Obvious thread #1: The health of free/open source software projects is directly tied to their communities, whose true strength is conferred by unfettered access to source code. Sonali Shah said it succinctly and simply during her presentation on governance in open source and gated source projects: restrictions on the use, modification, or distribution of source prevent members of a project from investing heavily in the interests of the community. I have seen this phenomenon first-hand in the Rotor project, which is a gated source project that I have worked on for several years. People use Rotor enthusiastically for their own purposes, but very few have the desire to “give back” to the community beyond publishing their own work or helping on the mailing lists. Were Rotor code freely redistributable, I suspect that this would be a different story.

Obvious thread #2: All code, whether libre or proprietary, is subject to patent law. Unlimited liability, which can be triggered by contributions of source code that knowingly or unknowingly propagate patented material, and patent infringement, which can be triggered by the use of unlicensed patents, are important issues for any software developer that wishes to participate in free/open source communities. Many of these communities have spent countless hours dithering about the nuances of contractual relationships that characterize software licensing, while pretty much ignoring the equally important relationship that they have with the rules set down by their own and other governments. There are, of course, notable cassandras out there, but so far their prophecies have gone unheeded or have been met with a wall of powerful and passive denial. (The free software community has the strongest record here, calling for the revolutionary overthrow of copyright and patent laws. Many of the conference attendees, however, seemed to think that implementing this approach might be somewhat impractical.)

Corporations, universities, and even some governments now view patent portfolios as economic tools to be managed for risk/reward in the same way as other asset portfolios. Because of this, all software is going to be increasingly subject to patent actions. The impact that this has on open source projects will be much larger than the impact on proprietary software projects, since the strength of open communities is directly related to their intangible (but very real) sense of potential. Anything that interferes with perceived freedom will threaten the health of these communities, and patent actions could easily erode the motivation for developers to “give back,” resulting in either outright project death or else in the conversion of projects to subsidy-supported satellite status. I suspect that fragile open source communities are actually quite susceptible to criticalities that could be triggered by such actions.

Increased software patent litigation could easily curtail community participation by all but the most politically-motivated developers, could give the term “developing nation” a new meaning by isolating the production of free software to countries where patent law (or lack thereof) is conducive to this production, and could also trigger legal and audit-related expenses for corporations that provide indemnification to customers using open source software, such as IBM. In my mind, if patent litigation were to knock the wind out of open source development, many of its assets would become tightly bound to the corporations that could provide patent resources and know-how. This transformation would come at the cost of project freedom, self-determination, diversity, and would introduce the same kinds of legal and technical review processes that encumber private software development today. Needless to say, free and open source communities (especially those with large scale code assets) should be spending large amounts of effort on developing effective patent strategies right now.

Other Memorable Memes

The conference itself was short, but densely packed with information, and the papers make for interesting reading. Attendees that I met included economists, sociologists, anthropologists, lawyers, business school types, consultants, and a number of representative members of free and open source communities.

To most of the people attending, open source fits into the social fabric in completely familiar ways. This point was highlighted by the numerous questions along the lines of “is this open source thing actually any different than ‘open science’ or, alternatively, other movements driven by volunteerism?” These questions, and much of the discussion that ensued, made it clear that the success of the open source movement is no more astonishing to other attendees of this conference than it is to me. (So why is it so astonishing and vexing to certain large software companies? Some mysteries cannot easily be explained.)

The term “joining script” was bandied about with glee and abandon; it was used to represent the dance that wannabe members of a community must perform in order to become officially sanctioned members. I don't know whether this term comes from the sociology literature or whether it was just the product of an academic brain, fevered by some publication deadline, but whoever invented it deserves to be banished to the land of bad acronyms and overly abstract jargon where there is nothing to read but patent applications!

Finally, David McGowan, a law school professor who has tracked the open source movement for quite a while, wryly observed that free and open source licenses actually behave more like trademarked brands than anything else. A programmer may invoke the GPL or the OSI only having agreed to live by the rules of the tribe. (Or should I say “by following its joining script”?) Legal ramifications and formal two-party relationships are much less important than the overall status that comes from the brand’s image and the associated social norms that accompany it. This observation struck me as entirely accurate, and helps to explain many of the extra-legal trapping of open source licenses.

Who would have thought Richard Stallman had so much in common with Michael Jordan and other famous brand-identity pitchmen?

Last modified: 23 June 2003