Last week, IBM released the Universal Information Management Architecture (UIMA) as open source under the Common Public License which provides for free and legal use in commercial and non-commercial pursuits. This move by Big Blue started a train of thought in my head about the the viability of Open Source projects started by large, publicly traded companies.
I have always wondered whether the inherent nature of large companies actually makes it impossible for them to run Open Source projects anywhere except the ground. It is one thing to buy the rights to an Open Source project, put it under new management and run it as you see fit, but it is quite another thing to start one. In my experiences working with and for large companies it seems as if their hierarchial makeup and centralized management styles would clash with the core tenets of Open Source.
Writing good software is not the only requirement for a successful Open Source project. If the community is not fully involved it defeats the purpose of releasing the code as open source. The more people inquire about aspects of the project, the more responses are posted in forums the better the project will be in the long haul. Now, I am in no way accussing IBM of doing this nor am I saying that is what is going to happen...I simply question whether large companies understand well enough the hidden dynamics of a successfull open source project. Dynamics of open collaboration, decentralized project governance and meritocracy.
For example, IBM has an established ecoystem of partners, ISV's and other practicioners who are involved with the UIMA. Which is a wonderful thing when you are trying to create buzz for a product. However, in terms of the open source development model, is it going to help the project itself? Once again I really don't know. Will these entities become valuable contributors to the project or will they be involved in name only?
Also, there is always the risk that a great number of potential contributors within the Open Source community might be turned off by the fact that a large (however open source friendly it might be) for-profit company. Will IBM make the necessary effort to stimulate the non-paid, non-IBM affiliated community into contributing to the project.
How will other companies who are interested in [contributing/partaking] take part if they are competing with Big Blue in some way. A couple of years back...HP has some interesting open source efforts that it sponsors (i.e. SmartFrog) but it never seems able to use its resources to energize a group of evangelists who are capable of "spreading the gospel". Granted, this may not be HP's strategy (I wouldn't know one way or the other), but I don't see very many reasons why they wouldn't want to have this happen. I see a similar thing happening with IBM unless they can find a way to slightly decouple their identity from their open source efforts so people can become interested in the projects itself, instead of the fact that the project is run/sponsored by IBM.
It seems like the best examples of successful Open Source are the ones who are able to galvanize a community around an idea that meets a need in the market and a well designed and well written codebase. The very strength of community is what has driven the acceptance of open source, not simply great code. When it is possible to gather talented people from across the world to contribute to a project, a great many things become possible. If IBM can do that, the sky is the limit.