I think I found the source survey that is the grist for the initial blog entry by Dana Blankenhorn, "The war is over and Linux won." The survey is available from www.ibm.com/linux/opensource/ (right column link).
It seems the survey population would certainly have some bias. Participants were people who had signed up for IBM Linux/open source newsletters. To IBM's or the survey company's credit, they clearly the describe the survey parameters in the 1st paragraph of the article. To quote from the article:
"In September 2006, IBM sponsored a survey of readers of its Open Source & LinuxLine newsletter to gauge the levels
of adoption, planning, and platforms for Linux and other open-source software. The survey was conducted by Unisphere
Research via e-mail notification, which directed participants to a Web-based survey instrument. A total of 434 anonymous
responses were collected by the survey deadline."
The survey itself contains some interesting bits, but I would be reluctant to directly extrapolate from the survey population to the overall market. That said, I think the up-front disclosure of methods in the actual survey negates any FUD charge. I chalk it up to Linux/open source exuberence by the blog author.
Slashdot Mirror
I think I found the source survey that is the grist for the initial blog entry by Dana Blankenhorn, "The war is over and Linux won." The survey is available from www.ibm.com/linux/opensource/ (right column link).
It seems the survey population would certainly have some bias. Participants were people who had signed up for IBM Linux/open source newsletters. To IBM's or the survey company's credit, they clearly the describe the survey parameters in the 1st paragraph of the article. To quote from the article:
"In September 2006, IBM sponsored a survey of readers of its Open Source & LinuxLine newsletter to gauge the levels of adoption, planning, and platforms for Linux and other open-source software. The survey was conducted by Unisphere Research via e-mail notification, which directed participants to a Web-based survey instrument. A total of 434 anonymous responses were collected by the survey deadline."
The survey itself contains some interesting bits, but I would be reluctant to directly extrapolate from the survey population to the overall market. That said, I think the up-front disclosure of methods in the actual survey negates any FUD charge. I chalk it up to Linux/open source exuberence by the blog author.