4 ms·
A couple years back I read a blog post with someone complaining about how hard it was to implement the traditional Lisp "map" / "mapcar" function. Specifically
by klodolph 16y ago
A couple years back I read a blog post with someone complaining about how hard it was to implement the traditional Lisp "map" / "mapcar" function. Specifically, one that took N lists and a single function with N arguments. The author was totally right, too. I could implement this generalized "map" function in Haskell, but it required defining an extra type class which is a little extreme.
Of course, the real problem is that people who are new to Haskell expect to sit down and solve problems in the same way they've been solving problems all along. This author was used to the "map" function from Lisp, and wasn't thinking about whether it was necessary, or whether Haskell had a different way of doing things.
Here's what the "mapN" function does:
> mapN (+1) [1,2]
[2,3]
> mapN (+) [1,2] [5,8]
[6,10]
The reason that we don't want to use "mapN" in Haskell is because of the following problem:
> mapN (+) [1,2]
???
The result is either "[Int] -> [Int]" or "[Int -> Int]", with no clue as to which is correct. This is why Haskell has "zipWith", "zipWith3", ... Not as pedantry, but because the alternative is ambiguous. (You can use Control.Applicative + ZipList, but that's a tad verbose).
My conclusion is that Haskell's differences are what throw people off -- programmers have to relearn how to do things they thought they had figured out for good years ago. (In this case, "everything is curried" means "varargs requires type classes"). It's FINE if you don't want to learn to do things differently and I won't think less of you, I respect your choices for what you do with your precious free time, and I understand that you might want stuff just to work now without learning anything, just please stop stereotyping the Haskellers as a bunch of ivory tower PhDs. I'm not even done with my Bachelor's.
- llimllib 16y agoIf you want to suggest that the main problem is people "trying to solve problems in the same way they've been solving problems all along", then you should have an answer for how you implement memoization "the haskell way". If not, your "conclusion that Haskell's differences are what throw people off" is incorrect. I don't think the author of the email is "stereotyping the Haskellers as a bunch of ivory tower PhDs", he's relating a story that's actually true of finding out that a basic technique he wanted to implement was most easily implemented by following the technique in a PhD thesis.
- astine 16y agoNow, I could be very wrong, but it is my understanding that Lazy Evaluation will provide the main benefits of memoization, without any action on the part of the programmer.
- jongraehl 16y agoThat's not generally true, although you can make it work out that way in some cases (I think threads mean that some work may be duplicated).
- klodolph 16y agoBe careful lest thou misunderstand the GHC runtime, which is fairly subtle. In Haskell, the semantics are unchanged if the runtime evaluates a particular thunk multiple times. Haskell is fairly unique in this regard. For exactly this reason, Haskell threads do not need to synchronize when evaluating thunks in multicore environments. Synchronization has a cost, and that cost can exceed the cost of duplicate evaluation. It's a trade off.
- klodolph 16y agoTrue, that. For some functions, memoization in Haskell is dead easy. Like "fibs", which is generally a poor example function but it illustrates a technique you can use for dynamic programming in Haskell that doesn't work in strict languages: fibs = 0 : 1 : zipWith (+) fibs (tail fibs) Now, I don't know if this technique works for the function you're interested in. If you want something more general, you can do soemthing like: memoize :: Ord a => (a -> b) -> IO (a -> b) memoize f = do ref <- newIORef Map.empty return $ \x -> unsafePerformIO $ do m <- liftM (Map.lookup x) (readIORef ref) case m of Just y -> return y Nothing -> let y = f x in do modifyIORef ref (M.insert x y) return y That's just off the top of my head (probably has syntax errors, never mind the type checker), and it has a number of drawbacks. For one, the map will get larger and larger -- you could use a LRU purging scheme but that's more work than I'm doing today. It doesn't support polymorphic functions, this could be fixed with some extra care but because of the way "memoize" would have to work you'll need to know a decent amount about the Haskell type system. There, I responded to llimllib's ultimatum that somehow my conclusion was incorrect if I didn't personally know how to implement a memoization function. Can I have a cookie? That rant at the end of my post above was me venting frustrations, on a lot of Haskell forums I see people posting about how much they think Haskell is not the language for them and I think, "If it's not for you, can you go to a forum that's not about Haskell?"