Based on the fact that CodeIgniter is up there, I think we can assume that this survey is oversampling PHP developers. The post doesn't say much about their methodology for distributing surveys, but I'm guessing not too many landed at the financial (web) software firm my friend works at - 500 developers strong, all working in .NET. My impression of bias is further strengthened by the numbers working in OS X. While I believe that more than 50% of PHP/Ruby on Rails developers are using Macs, I don't think that number will hold up when Java and .NET devs are included.

Or maybe it will. Who knows? Point is, unless they share their data collection methodology, this survey could just as likely be representative of the 1,400 folks who follow web directions on twitter.

Yeah I have to agree with spanktheuser, unless their is some transparency on how they are coming to these conclusions, it's hard to take these numbers seriously.