Yesterday, Kristin provided some good analysis of a new research report from The 451 Group. The report concludes that open source is less of a business model than a business tactic. Within the post about the report, there is an executive summary (PDF) which points to some interesting trends for open source, outside the constraints of the question The 451 Group was trying to answer.
One of the big findings in the 451 Group's study was this:
"The line between closed and open source has blurred as FOSS is embedded in proprietary products and commercial extensions have been added to FOSS."
The 451 Group's research included 114 open source vendors, and companies who qualified included Red Hat, Alfresco, IBM and Oracle. It's the IBMs and the Oracles included in this study that skew the results toward blends of open source and commercial software.
There are still a lot of open source projects that are completely divorced from commercial goals and proprietary components. These often tend to be smaller projects, but many small projects result in outstanding applications.
Recently, we interviewed Javier Paniza, the creator and lead developer of OpenXava. He made the following comment when asked how he was monetizing his project:
"OpenXava has no business model. Indeed, a true open source project cannot have a business model because an open source project is not from a company or individual, but a creation shared by a community. On the other hand, it's possible for a company to have a business model around an open source project."
In response, a reader disagreed with Javier:
"I would rephrase it to 'in order for a true open source project to thrive it HAS TO HAVE a business model.'"
I happen to think all of these conclusions ring of narrowcasting. Open source is the wild kingdom. Beasts of all sizes and shapes contribute to the ecosystem, ranging from totally non-commercial efforts to commercial efforts that wrap in purely open source components. That's part of the beauty of it.
Once I remember the great writer Nabokov citing a problem he perceived psychiatrists to have: "They attempt to put complicated people in simple boxes." I'm not endorsing his opinion there, but it is an argument against reductionism. Reductionism isn't what we need when we evalutate the open source landscape.