Here's source and logs: http://www.imn.htwk-leipzig.de/~waldmann/draft/skpp11/subseqsum/Subseqsum.hs
The program is meant to show an application of the "third homomorphism theorem" approach (hom-based structural parallel programming). My observation (for this program) is that there is little speedup, indeed the threadscope picture shows a peak only at the very end, but the eventlog contains surprising (for me) stack and heap overflows. Would this work better with Data.Sequence instead of List? (Is there a really cheap way (O(1)) to split some Data.Sequence roughly in half?) PS: I keep telling my students that "structural parallel programming" is the right thing to do, but I find it surprisingly difficult to exhibit clear-cut examples that support the claim (and work with standard ghc, so students can reproduce it). I appreciate any comments. - J.W. _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
