August 2, 2011

Developing country developing open source

Over the weekend, I flew to Johannesburg, South Africa for a conference on open source software hosted by the School of Business and Economic Sciences at the University of the Witwatersrand (WITS). I was invited to give the opening keynote, and spoke about my current study integrating prior work on open IT strategies. (More on that later).

The conference featured both local and international speakers; the former were a mixture of faculty, few graduate students and one practitioner.

One interesting panel had an argument (between two friends) over conflicting IP goals of South African government policy. IT consultant Derek Keats complained that a 2008 Intellectual Property Rights from Publicly Financed Research and Development Act (a more centralized twist on Bayh-Dole) would make it difficult or impossible for SA academic researchers to give away IP to open source projects.

Meanwhile, Prof. Robert Vivian argued that a government mandated preference for OSS was arrived at via unconstitutional means. (I noted my conclusion based on a decade of OSS research: the best procurement policy is to mandate a fair price comparison between OSS and proprietary software — one that tends to allow buyers to use the threat of switching to reduce lock-in rents but avoid the disruption of actually switching.)

Among the international speakers, Sebastian von Engelhardt of Friedrich-Schiller-University talked about his 2009 survey of 6,000 German software developers and IT consultants. He worried that those that used OSS might have a harder time getting capital or growing than those that developed proprietary software, but generally found the two types were similar except that the OSS firms tended to be smaller and younger.

Not all the papers were about open source software. Alessandro Rossi of University of Trento summarized a series of papers on how Wikipedia activity changes after an article is tagged as having a major problem. For example, if a WikiSimple article is tagged as having readability problems, the readability quickly goes up — but the rate of other changes goes down.

Marcel Bogers (@bogers) summarized antecedents to open innovation in prior research such as evolutionary economics, transaction cost economics, social network theory and the resource-based view. My personal favorite was a paper by Maradona Gatara on mobile crowdsourcing in Kenya.

The workshop closed with the big gun — economist Richard Langlois of U. Conn. Summarizing a forthcoming book chapter, he noted the dramatic parallels between early radio and early PCs: assembly was cheap while the components were valuable, hobbyists played an important role, the business quickly shifted to the importance of software.

The main difference is that during World War I, the US nationalized radio patents to facilitate ship-to-shore communication, and then spun them off after the war to create the Radio Corporation of America.

Disputing The Electronic Century by the late Alfred Chandler, Langlois concluded that the large RCA R&D labs were not a spur to innovation but a drag on innovation — because RCA wasted its resources trying to find new areas rather than improve existing technologies. Vertical integration and control by RCA (and Columbia) prevented the sort of modular innovation that became commonplace in the PC industry.

This was the first of what (funding permitting) may become an annual conference. Meanwhile, the local representative of Nine Sigma was handing out flyers promoting the “4th SA Innovation Summit” on 30 August. The summit is dedicated to “change, revolution, transformation, metamorphosis, breakthrough” and proudly proclaims that “we are anti-red tape.”

No comments: