Yesterday's piece on phony open source drew pretty good traffic for a holiday.
It also drew a thoughtful e-mail from Tim Yeaton, president of Black Duck Software. He interrupted his efforts to counter OpenLogic's new open source scanner to say that open source is bound to be assimilated into the software mainstream.
(Tim is pictured here on casual Friday at the company's blog, commenting on the Oracle-mySQL situation.)
Some assimilation is inevitable, he noted. After all, 1 in 200 of us are now related to Genghis Khan. It's how DNA rolls. (This explains why those who dream of past lives don't imagine themselves as medieval scullery maids.)
The key to understanding open source assimilation, and protecting yourself from it, he added, is to understand the difference between projects and products:
- open source "projects" are usually community-driven efforts made available under an open process and/or open source license. We track over 220,000 such projects licensed under 1800 different open source licenses in our Knowledgebase.
- open source "products" typically take some form of freely available open content: Linux packages, open source-licensed code, web or other content, etc. and make it useful in some way: via support, via packaging & refinement, etc.
In the process of making it useful, they are able to commercialize some aspect of it and generate a return to continue to invest (support contracts, commercial licenses for value-added components or capabilities, advertising, etc.).
So from our perspective, it's not so much about pegging a certain percentage of code or content that allows a "product" to be considered open source, it's more about if the product uses open source in a meaningful way to make it more useful to customers, and if the company building the product acts as a good citizen in open source (e.g., honoring the authors intentions, community process participation, contributions back, etc.).
There is a question of degree to be debated on these dimensions to be sure. In our work with customers actively utilizing open source in multi-source development (including ISVs and embedded solutions providers), we found that, on average, 22% of their code was open source.
Since these are pretty progressive companies when it comes to effective use of open source, it may be a useful threshold to consider when one thinks about what an open source "product" is.
Yeaton's view is not unusual. The Great Recession has turned most open source executives into pragmatists. They want to balance the visibility of their code with their ability to retain customer cash flow.
Personally I find this to be a reversible trend. Good times can breed idealism, and profitable idealism in open source can become a selling point that pushes companies back toward principle in the name of profit.
But that's just me. What say you?