That was an interesting read, and something that I largely agree with.
Although I don't write much Scala, I love developing in JRuby (and Ruby in general) and Clojure, but in a lot of situations Plain Old Java (POJ) really makes the most sense. The issues are speed of experimental programming in alternative JVM languages vs. overhead to lack of tooling and deployment issues.
I think that the future "most used language of the JVM" will likely not be Java but I don't know what that language will be (and it has maybe not been created yet).
I say this as someone who's coding experience is essentially all web stack, and wouldn't consider myself a computer scientist, but that was one of the best reads this year. Gives me so much insight in why choosing a popular platform cough LAMP cough can be a good idea even when it feels so broken.
Although I don't write much Scala, I love developing in JRuby (and Ruby in general) and Clojure, but in a lot of situations Plain Old Java (POJ) really makes the most sense. The issues are speed of experimental programming in alternative JVM languages vs. overhead to lack of tooling and deployment issues.
I think that the future "most used language of the JVM" will likely not be Java but I don't know what that language will be (and it has maybe not been created yet).