I sympathize! > While we're here, I'd love to get your opinion on polysemy -- it looks like the best thing since sliced bread for free(r) monads, the combinati

I sympathize! > While we're here, I'd love to get your opinion on polysemy -- it looks like the best thing since sliced bread for free(r) monads, the combination of the usability of freer-simple and the efficiency gains of fused-effects. > I've long secretly held the believe that mtl-style and free-monads were roughly equivalent in power and efficiency equality was only a matter of time, but I am really looking to the free approach because effect algebras seem to be a fantastic tool -- they compose more cleanly than mtl does. I've not tried polysemy and I don't think I will unless it comes up on the job. I'm inured to effect systems after the fifth or sixth flavor. I've heard it's nice from people who've wanted to use it though. I've written my share of production Haskell and I haven't found mtl to be as bad about composition as people tell me it is.