There's also http://redmonk.com/sogrady/2017/03/17/language-rankings-1-17/ http://redmonk.com/sogrady/2017/03/17/language-rankings-1-17... which uses # pull req

There's also http://redmonk.com/sogrady/2017/03/17/language-rankings-1-17/ http://redmonk.com/sogrady/2017/03/17/language-rankings-1-17... which uses # pull requests on github with # of tags on Stack Overflow. Another biased methodology for sure, but it paints a different picture.