> Were you using Haskell as the pseudocode? Actually, in the ICFP paper mentionned above, we describe the most important parts of our Arrow implementation and then wrote a complete tutorial on 2.5 pages. So, that was not pseudocode but real, executable code. However, we did not use the Arrow syntax, to simplify the translation in other languages.
Using Haskell was motivated (on a suggestion of Conal Elliott) to prove the purity of our solution and that our constructs actually instanciate Arrow and its friends. Btw, for those of you interested in the Arrow instance, including Mealy, Conal Elliott blog [http://conal.net/blog/] is a must-read. -- Pierre-Evariste DAGAND http://perso.eleves.bretagne.ens-cachan.fr/~dagand/ _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
