When I hear things like Business Open Source Use Up 26% in One Year, I get a little interested.  Excited even.  I like hearing good news about good things, like most people.  But it doesn't override the skeptic in me that forces me to ask, "Says who?"

Follow me if you can.  I followed the link from Slashdot to Computer Business Review -- a publication I'd never heard of before but which I found linked from a trusted news source, so it must be all right, right?  Then on to the main page of the company responsible for the original survey and a short jump to their press page where there was a release that summarized the survey results.  (If you get this far and wonder why the press release looks so familiar, it's because the cub reporter working for CBR basically reworded it and published it as an article.)  There was, however, no information about the methodology used, number of surveys sent out, response rate, respondent demographics, or anything else that would give me a little insight into the meaning of it all.  Just think, before the invention of hypertext I could have just read the blurb on Slashdot, taken it as gospel, and gone to bed like people do.

I also found out that:

  • The 26% increase they report is on the basis of the number of open source packages used by companies responding to the survey over similar figures collected in 2006.  As far as I'm concerned, this tells me about as much about real OSS penetration into major companies as the nonsense notifications and invitations I have waiting for me on Facebook.
  • The most common license in use was found to be the Apache license which was used in 62% of the cases.  GPL was way down the list with 27% coverage.  This shouldn't come as a surprise to anyone reading a little further because...
  • The list of the top open source packages was slanted heavily toward open source libraries and development tools (with a particular bias toward Java-based products) with Hibernate and Struts claiming the two top spots and somewhere in the vicinity of 80% of the top 25 most popular packages being exclusively or predominantly used for Java application development and deployment.

It's this last bit that leaves me feeling a little used.  The rankings are sufficiently screwy -- i.e. Eclipse beating Apache, Jasper Reports making the list while Linux does not, etc. -- that it really makes someone who knows enough about IT and has two brain cells to rub together want to ask whether or not anything useful can be inferred based on a shallow survey of a very limited subset of the technology sector where the results lack any kind of transparency.

Is open source on the rise in the enterprise?  Possibly.  Seems like it.  I can't say for sure based on what I see here, and the echo chamber that is news on the Internet sure isn't helping.