#haskell - Thu 3 May 2007 between 00:00 and 00:26

NY Lost Funds



narain> let test 1 = 1; test n = 3*(n-1) in test 10
lambdabot27
kpreidchowmeined: ghci accepts only expressions; that's a definition. you can either put it in a file and load the file in ghci, or write "let test = ... in test"
I recommend having a file in which you define stuff and reload in ghci
chowmeinedok
kpreid(you can reload with just ":r")
narainsorear: have you send the mail to the buggy blog guy yet? can you @tell me when he responds?
actionddarius hacks PDF by hand.
ddariushacks PDF by hand.
dfrankewhat's new in HEAD?
CaleIn fact, just : on its own is enough to reload the file in ghci
dolioHmm... Wadler google talk.
hpasteLoganCapaldo pasted "Generalizing ShowS, sort of, what's wrong with this picture?" at http://hpaste.org/1682
faxsnoouou
oops
Caleheh
Too bad unicode characters aren't allowed in nicks.
dolioLoganCapaldo: I think, if anything, 'type M m a = forall a m. ...' would have a and m in '...' unrelated to the a and m in 'M m a'.
faxCale: yeah!
dolioYou'd be introducing new type variables that shadow the outer ones.
faxI think its really due time someone revise the IRC protocols
CaleThat would be nice.
donsPseudonym: thanks for raising that issue on the ml.
actionnarain wonders if every latin alphabet has an upside-down version in unicode
narainwonders if every latin alphabet has an upside-down version in unicode
LoganCapaldodolio: oh is that whats happening? that kind of makes sense
donsPseudonym: I'd written something of a reply yesterday, then gave up in frustration
Pseudonym: the ml lists are fragmenting a bit
dolioLoganCapaldo: That'd be one problem. I'm not sure if you can have constrained type variables in a synonym.
LoganCapaldodolio: well it only complains really when I give the functions explicit signatures
dolioLoganCapaldo: Note: 'type M = forall a m. (MonadPlus m) => m a -> m a' is also accepted.
*Main> :t (undefined :: M)
(undefined :: M) :: forall (m :: * -> *) a. (MonadPlus m) => m a -> m a
LoganCapaldoHmm maybe that's what I want
dolioI suspect not...
Because I think the type 'm a -> M' is essentially 'm a -> (forall b n. MonadPlus n => n b -> n b)'
Where you might be able to lift out the 'forall...', but it still ammounts to 'm a -> n b -> n b' not 'm a -> m a -> m a'.
hpasteLoganCapaldo annotated "Generalizing ShowS, sort of, what's wrong with this picture?" with "Progress, maybe?" at http://hpaste.org/1682#a1
LoganCapaldounitMonadPlusS I still can't give the type I want...
Or I can, just not with the names I want to use
hpaste ate the b
or my clipboard ate the b rather

Page: 2 9 16 23 30 37 44 51 58 65 72 79 86 93 100 107 114 

IrcArchive

NY Lost Funds