Re: Basically tells you all you need to know about Stack Overflow users...
According to the article: Those taking the survey are not a particularly diverse group
There is more than one kind of 'diversity' in my view, particularly when doing a sample for surveys, which is actually more along the lines of "representative sample" vs "whatever they had". If the sample represents Stack Overflow, then I suppose it MIGHT be a good survey.
But as for programming languages in general, I tend to believe TIOBE, if for no other reason than they've been doing this a while and have a pretty scientific way [it seems] of getting programming lingo popularity out of whatever data set they use. interestingly, C jumped over Java a bit this month (Java was neck/neck #1 vs C #2 for quite a while again) and so Java and C run close to one another in popularity and continue to do so.
In any case when you see survey results that are "this different" it makes you wonder how accurate the survey is.