Comments on: Lazier functional programming, part 1 http://conal.net/blog/posts/lazier-functional-programming-part-1 Inspirations & experiments, mainly about denotative/functional programming in Haskell Sat, 26 Sep 2020 21:06:12 +0000 hourly 1 http://wordpress.org/?v=4.1.17 By: Conal Elliott » Blog Archive » Adding numbers http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-754 Mon, 25 Oct 2010 22:44:21 +0000 http://conal.net/blog/?p=190#comment-754 […] Lazier functional programming, part 1 […]

]]> By: Conal Elliott » Blog Archive » Lazier functional programming, part 2 http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-753 Sun, 19 Sep 2010 19:36:30 +0000 http://conal.net/blog/?p=190#comment-753 […] About « Lazier functional programming, part 1 […]

]]> By: conal http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-752 Sat, 18 Sep 2010 20:55:26 +0000 http://conal.net/blog/?p=190#comment-752

You don’t say what we have at our disposal to solve the puzzle. If we are restricted to lambda definable functions there isn’t much to do. If we have something like por it’s a different matter.

augustss: I didn’t want to bias people’s explorations.

]]> By: David Sankel http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-751 Thu, 16 Sep 2010 14:34:28 +0000 http://conal.net/blog/?p=190#comment-751 @dimtar: Awesome!

]]> By: dimitar http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-750 Wed, 15 Sep 2010 15:00:46 +0000 http://conal.net/blog/?p=190#comment-750 I was wrong again: f ⊥ ≡ g ⊥ is too strong. From Albert Lai’s explanation the minimal strictness is achieved when: either f g ⊥ = f⊥ ⊓ g⊥. In Haskell:

either :: (HasLub c, HasGlb c) => (a -> c) -> (b -> c) -> Either a b -> c
either f g x = (f undefined) `glb` (g undefined) `lub` strictEither f g x

When c is flat domain it is essentially David Sankel’s solution.

]]> By: Dave http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-749 Wed, 15 Sep 2010 13:29:06 +0000 http://conal.net/blog/?p=190#comment-749 The opposite of ‘strict’ is ‘lax’, so ‘less strict’ would be ‘laxer’.

]]> By: dimitar http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-748 Tue, 14 Sep 2010 18:42:59 +0000 http://conal.net/blog/?p=190#comment-748 Yeah, my guess is broken. I was naive hoping that the restriction f ⊥ ≡ g ⊥ can be relaxed.

]]> By: David Sankel http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-747 Tue, 14 Sep 2010 13:34:40 +0000 http://conal.net/blog/?p=190#comment-747 using lub instead of unamb would make my version a bit lazier.

]]> By: David Sankel http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-746 Tue, 14 Sep 2010 03:02:59 +0000 http://conal.net/blog/?p=190#comment-746 Making my version more efficient.

either :: (Eq c, Lub c) ⇒ (a → c) → (b → c) → Either a b → c
either f g e = t1 `unamb` t2
  where t1 = let a = f undefined
                 b = g undefined
             in if a == b then (a `lub` b) else undefined
        t2 = case e of Left a → f a
                       Right b → g b

]]> By: David Sankel http://conal.net/blog/posts/lazier-functional-programming-part-1#comment-745 Tue, 14 Sep 2010 02:55:37 +0000 http://conal.net/blog/?p=190#comment-745 lazierEither does not meet the consistancy requirements for lub. consider:

lazierEither (const True) (const False)

⇓

(const True . fromLeft) `lub` (const False . fromRight)

]]>