On the Revival of Dynamic Languages

On the Revival of Dynamic Languages by Nierstrasz O, Bergel A, Denker M, Ducasse S, Gälli M, Wuyts R.

This paper reads a little like propaganda for dynamic languages. The discussion on Lambda the Ultimate (linked below) raises some interesting points and points out some of the inconsistencies in their argument (static languages can't prove everything, so let's use dynamic languages which prove nothing).

That aside, the authors do discuss some interesting language features. First-class namespaces can be used, as they describe, to implement horizontal sub-classing (a compelling use case has been discussed on LtU). Similarly, their desire for pluggable type systems seems a little reminiscent of the approach that the PyPy project has taken with object spaces. On the other hand, I don't really like their suggestion that image-based languages are the way to go (I much prefer to write code in my favourite editor, manage it using my favourite version control system, etc.). All up, it's an interesting read (if only for the descriptions of the various language features the authors like).

DOI | CiteULike | Del.icio.us | LtU