Open source without open governance

(This is a posting that was stuck in my outbasket over the weekend. For obvious reasons I'm pushing it out of the out basket).

Last week I saw an intriguing headline in an InfoWorld e-mail blast:

Nokia: Open Source Developers Should Play By Our Rules

It's becoming clear that the phone maker thinks open source developers need to adapt to the ways of commercial software vendors, not vice versa. Read on:
http://cwflyris.computerworld.com/t/3299928/121450395/121092/0/

The article was intriguing but I had to get some work done, so I didn’t have a chance to investigate right away. The link pointed to a blog entry by Tom Sullivan which pointed to an InformationWeek article by Serdar Yegulalp which pointed to a Business Week article by David Meyer.

When it comes to open source mobile devices, I’ve been writing about Nokia’s Maemo efforts (and Intel’s derivative Moblin) almost since the blog started. I first heard about Maemo at HICSS-40 from friends of mine at ETH Zürich who have written a paper on Mameo. (I can only find the slides online).

Now not knowing anything else, one would expect that Intel would have a better hope of getting OSS, for two reasons. First, software is complementary to its main hardware business, while Nokia’s systems busines (like Apple’s or IBM’s) is based on software system integration.

Second, Intel has a long experience with open source as a founder (and the main funder) of OSDL — the late, great Linux promotion entity born back in 2000. It has also done other work in house to work with Linux (as part of its successful strategy to supplant RISC Unix boxes with x86 Linux boxes).

As part of my own research on open source, one of the things that I looked at was how firms interacted with open source communities. I chose as a co-author Siobhán O’Mahony, because no academic knows more about the dynamics open source communities than she does.

In April, we finally got our paper published, which we called “The Role of Participation Architecture in Growing Sponsored Open Source Communities.” It was in a special issue of Industry & Innovation (a European innovation journal), based on a track at last year’s EURAM conference on open source, user innovation and open innovation.

I could write a whole paper about Maemo and how it fits into other open source communities, but alas I’ve promised to write other papers in the next 60 days. So let me just quote from our conclusions:

However, our study showed that sponsored open source software communities are fundamentally different from autonomous communities in the potential for goal conflict between sponsor and community members. Although both sponsors and members seek widespread adoption, the primary goal of a corporate sponsor is profiting from its investment, while the goal of an open source community would be improving the capabilities of the shared technology.

To gain interest from a community of contributors, sponsors needed to at least provide transparency. The openness of sponsored communities differed most in terms of accessibility, with most sponsors retaining privileged (monolithic) rights for some portion of the community’s decisions. In a few open cases, the sponsor shared some control with the community—and when sponsors relinquished more control to the community, those sponsored communities were transformed into autonomous ones.

As a consequence, we also found a dramatic difference between most sponsored and autonomous communities in terms of design decision related to accessibility, particularly in terms of governance. Governance of autonomous projects was largely pluralistic, shared widely among community members, whereas the ultimate decisions of sponsored communities were (with rare exceptions) controlled by the sponsor.

So, to put it bluntly: companies want to have their cake and eat it too, but if they exert too much control, people will figure this out and individuals (or other companies) won’t bother to participate.

The problem for Nokia is, the cases where tight control is where you’re the only game in town (cf. MySQL, SugarCRM). There are many other initiatives building code based on Linux, so if Nokia tries to hold things too tight with Maemo, people will just join LiMo (if it ever opens its gate), Android (if it ever opens, and if it ever ships), or OpenMoko or (fill in the blank).