Not having much time to produce original content, it’s a pleasure to find a post on a subject I was considering to cover, and actually being treated better than I could have hoped to do myself: a post about HPCS on Bartosz Milewski’s blog. Other than the fact that Fortress is not under development but has currently “wound down” (as Guy Steele said in the keynote speech of the JVMLS 2013), the article is up to date. All three approaches resonate with the theme in my post on array processing. I just want to add that both Guy Steele (Fortress) and Vijay Saraswat (X10) have had great influence on me, the former via Common Lisp and Scheme (cf. TinyScheme), the latter via the “Concurrent Constraint Programming” book (cf. my thesis on Concurrent Constraint Logic Programming [Technical Report CERMICS 92-04]). And Fortress intended to tackle the problem of properly dimensioned calculations, which had also been one my favorite subjects.
Even mainstream Java has been oriented towards how to do efficient and practical computing (cf. “Arrays 2.0“).