I roughly interpret the graph as anything being below the line leaning towards the "getting things done" camp and anything above leaning towards "academics". Th

I roughly interpret the graph as anything being below the line leaning towards the "getting things done" camp and anything above leaning towards "academics". This is a bit distorted because FLOSS projects are favored (there's lots of "getting stuff done" code that is not on Github) and some languages have great communities outside Stackoverflow (Rust). Prolog seems like a good example. It is used in AI classes and "different enough" to leave students with lots of questions for assignments (imo). At the same time it's not used in many industry projects and if it is companies tend to use proprietary Prologs (like Sicstus) and the code is unlikely to be on Github. I suppose one could also argue that good setup/tooling and documentation helps to push a language below the line. I'd say Elixir is a good example of this. So by and large I'd lean towards languages below the line when trying something new. Am I wrong? Alas the next language I want to play around with (F#) is above the line. But the domain modeling with F# talk/website is very intriguing.