
Also don't forget that the "popularity" as estimated by TIOBE is just based on statistics of how often people talk about a given language online, and it doesn't look like their algorithm is particularly sophisticated. So, as we have probably already said numerous times, while it'll be reasonably correlated with how much people use a given language worldwide compared to others, the reasons people talk about it are undoubtedly varied, and a language that has many quirks is more likely to be talked about online more often than others, as people overall tend to talk more about their problems/look for solutions and the negative side of things, rather than the positive one.
If I was the author/maintainer of a programming language, I would be happier seeing how many projects there are with it, and I'd be more concerned seeing how much it's talked about online, knowing that 80% of what's being talked about would probably show how badly designed it is. Well designed tools should require minimal assistance.
