Ryan Ingram wrote:
On 10/24/07, apfelmus [EMAIL PROTECTED] wrote:
So, instead of storing a list [∃a. Show a = a], you may as well
store a
list of strings [String].
I've seen this before, and to some extent I agree with it, but it
breaks down for bigger examples due to sharing.
In most
In a sense, that's also the reason why stream fusion à la Duncan +
Roman + Don uses an existential type
data Stream a where
Stream :: ∀s. s - (s - Step a s) - Stream a
data Step a s = Done
| Yield a s
| Skip s
I thought there was a deeper reason for
TJ wrote:
data Showable = forall a. Show a = Showable a
stuff = [Showable 42, Showable hello, Showable 'w']
Which is exactly the kind of 2nd-rate treatment I dislike.
I am saying that Haskell's type system forces me to write boilerplate.
Nice :) I mean, the already powerful
Thanks for posting this, I finally understand existentials!
This bit was specially helpful:
So, how to compute a value b from an existential type ∃a.(a - a)? ...
Could you give a specific example of computing existential types?
I think this shows why digested tutorials, as opposed to
Thanks for the concise explanation. I do have one minor question though.
-+- A more useful example is
∃a. Show a = a i.e. ∃a.(a - String, a)
So, given a value (f,x) :: ∃a.(a - String, a), we can do
f x :: String
but that's pretty much all we can do. The type is isomorphic to
