Re: [Haskell-cafe] MTL vs Free-monads, what are your experiences

With constraint kinds, you can make type synonyms that expand to constraints, solving your syntactic objection, but I don't want to discuss those further. Far as error handling goes, I use MonadCatch/MonadError: https://hackage.haskell.org/package/exceptions-0.8.3/docs/Control-Monad-Catc... Here catch is: catch :: (Exception e, MonadCatch m) => m a -> (e -> m a) -> m a Catch implies Throw, so. This part,

Another advantage is that the Purescipt example uses a concrete type

I'm not sure I understand. What are you saying is concrete here? You still have a row of effects which could be any valid handler. I don't think I understand the distinction you're making. Eff itself is parameterized by the row of effects (# !) and the return type (*), so there's nothing especially "concrete" about that to my mind as contrasted with something instantiable as any Monad known to have a Console API and a means of furnishing random values. On Wed, Oct 19, 2016 at 1:41 PM, Will Yager wrote:

It seems that there are several advantages to the Purescript approach.

For example, this

catchException :: forall a e . (Error -> Eff e a) -> Eff (err :: EXCEPTION | e) a -> Eff e a

would be unwieldy to express using typeclasses, requiring at least three constraints. I also find this style easier to read than constraints, as it requires no mental substitution. E.g. if I see

(Foo m, Bar m, Bar n) => Baz -> -> m a

When I get to the end of the type, I have to go back to the beginning to figure out what m is. I can't read left-to-right. This happens a lot with constraint-based monad composition.

Another advantage is that the Purescipt example uses a concrete type, which is often easier to reason about than "ad-hoc" typeclass abstractions like MonadRandom. However, it looks like you still get the flexibility of ad-hoc typeclasses, because you get to pick any function that discharges the effect type in the given effect monad.

Like I said, I have not used it, but these are what I've noticed from topical observation.

Apologies for the formatting; copying that code example appears to have confused the iOS mail app.

Cheers,

Will

On Oct 19, 2016, at 12:26, Christopher Allen wrote:

It's not really more direct. It's an unordered collection of effects you can use. IME it's a less efficient mtl-style, but YMMV.

Taking an example from a PureScript tutorial:

func :: Eff (console :: CONSOLE, random :: RANDOM) Unit

Can just as easily be:

func :: (MonadConsole m, MonadGimmeRandom m) => m ()

(mangled name so it doesn't overlap with a real class)

There are other differences, but they haven't amounted to much for me yet.

Kmett's Quine has a good example of some homespun mtl-style: https://github.com/ekmett/quine

On Wed, Oct 19, 2016 at 12:17 PM, Will Yager wrote:

Can anyone comment on the use of Purescript-style effect monads as compared to MTL and Free? While I have not used them in practice, they seem to express the "intent" of monad composition a bit more directly than the approaches we use in Haskell.

Cheers,

Will

_______________________________________________

Haskell-Cafe mailing list

To (un)subscribe, modify options or view archives go to:

http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe

Only members subscribed via the mailman list are allowed to post.

-- Chris Allen Currently working on http://haskellbook.com

-- Chris Allen Currently working on http://haskellbook.com