Tuesday, March 21, 2006

Essential Language Support for Generic Programming

Essential Language Support for Generic Programming by Jeremy Siek and Andrew Lumsdaine.

This paper presents an extension of System F called System FG. The motivation for the work is in the construction of generic algorithms and, more importantly, the verification of their use. It looks to me as though the authors have taken a little bit too much of C++ on board.

The concepts of the paper are very similar to Haskell's type classes (according to section two, the main difference is that type classes are designed to allow Hindley-Milner type inference) and the associated types seem to me to be filling a similar role as functional dependancies. I found it rather interesting and it's the first type system paper I've read since I achieved a vague understanding of type systems.

ACM | Google | Del.icio.us | CiteULike | LtU

No comments: