It is fun to compare the histories of Clojure, Scala and F#. A pragmatic programmer created Clojure. A university professor created Scala. A research team from Microsoft created F#. Each language is characterized by its creator(s).
I find Scala a bit of an anomaly though, yes it was created by a university professor, but it was clearly designed for mass adoption - even down to the curly braces (although they are backtracking that decision now). I have lot of sympathy for this approach, but I think it has also been the underlying cause of many of their problems.
A university professor that created the Modula-2 compiler sold by Borland, and was one of the Pizza compiler designers, which contributed to Java generics.
Not your regular university professor, rather someone with industry experience, specially in the Java world.
You forgot to also mention that he was the original author of the Java 5 compiler which Sun took over. He is definitely not a guy in an ivory tower, almost the exact opposite in fact. But ivory towers are not necessarily bad. It was precisely an ivory table that gave us Lisp and Haskell. Lisp's Garbage Collector was a big influence on the original Java. And Java's chief architect has specifically mentioned Haskell as his influence in the future of Java.
Truth be told there are plenty of those folks as well.
In European universities you aren't necessarily required to do outside work, so there are plenty that do the path, student, Msc student, PhD student, Professor, department chair, retirement.
In fact, I had a couple of occasions where good professors told me to just taste the professional world between getting the degree, and trying to pursue further academic life.
My second attempt at it, while at CERN, kind of did it for me.
So I do understand where the stereotype comes from.
In the q&a with Guy Steele at the end he says one thing he would do differently is integrate transducers more deeply into the language from the beginning, but the concept emerged around the time of core async when they were wary of rewriting map filter reduce etc again for channels (rough paraphrase from memory, watched this a few days ago).
I wonder what that would look like, what core functions would use transducers that do not, and how.
