> >> As an aside, I think strictFoldl is not automatically paralellizable, but >> myFoldl may be. [...] > >If dreams of automatic parallelisation of lazy functional programs were >pennies, I'd have $2.35. > >Alastair ;-) But those who are still dreaming might want to do some reading. One reference that springs to my mind is Jonathan Hill's PhD thesis: Data-parallel lazy functional programming September 1994 http://www.dcs.qmw.ac.uk/publications/report_abstracts/1995/701.html ... Development of the higher order parallel map, fold, and scan is presented, a trio of functions that is fundamental to a data-parallel paradigm. Unlike other parallel implementations of these functions, particular attention is given to their non-strict nature. Claus
