Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 .. 5022
— 502,152 events total
2020-11-02 07:26:13 <danso> @type fmap
2020-11-02 07:26:14 <lambdabot> Functor f => (a -> b) -> f a -> f b
2020-11-02 07:26:43 <danso> is your `Null` defined like `data Null = Null` ?
2020-11-02 07:27:02 <danso> i think you most likely mean Maybe instead of Either
2020-11-02 07:27:05 → frankie2 joins (~frankie@185.104.184.43)
2020-11-02 07:27:30 <energizer> Maybe sounds like like it's in the right direction
2020-11-02 07:27:52 <danso> also fmap on Either maps a function over the right side, not the left one
2020-11-02 07:28:15 <danso> fmap on maybe is
2020-11-02 07:28:28 <danso> fmap f (Just a) = Just (f a)
2020-11-02 07:28:35 <danso> fmap f Nothing = Nothing
2020-11-02 07:28:44 <energizer> yeah that's what i mean ^
2020-11-02 07:29:19 <danso> if you want to get cute you could use <$> or <&>
2020-11-02 07:29:36 <dminuoso> In fact, <$> is very typical
2020-11-02 07:29:44 <danso> a <&> f is the same as `fmap f a`
2020-11-02 07:30:10 <dminuoso> With a bit of squinting <$> behaves a bit like ($), which is possibly why they both have a dollar in their name.
2020-11-02 07:32:16 → sQVe joins (~sQVe@unaffiliated/sqve)
2020-11-02 07:32:44 <energizer> how is <$> used here?
2020-11-02 07:33:10 <danso> f <$> a
2020-11-02 07:33:30 <danso> `<$>` is the same as `flip <&>`
2020-11-02 07:33:34 <danso> @type <$>
2020-11-02 07:33:36 <lambdabot> error: parse error on input ‘<$>’
2020-11-02 07:33:38 → christo joins (~chris@81.96.113.213)
2020-11-02 07:33:40 <danso> @type (<$>)
2020-11-02 07:33:42 <lambdabot> Functor f => (a -> b) -> f a -> f b
2020-11-02 07:35:12 <energizer> ok i get it
2020-11-02 07:35:45 <dminuoso> c.f.:
2020-11-02 07:35:48 <dminuoso> :t ($)
2020-11-02 07:35:49 <lambdabot> (a -> b) -> a -> b
2020-11-02 07:37:10 → dhouthoo joins (~dhouthoo@ptr-eiv6509pb4ifhdr9lsd.18120a2.ip6.access.telenet.be)
2020-11-02 07:37:17 <energizer> is <$> or <&> more used?
2020-11-02 07:37:27 <dminuoso> <$> by far
2020-11-02 07:37:45 <dminuoso> Depending on the project and situation you might either see fmap or <$> being used more
2020-11-02 07:37:51 <dminuoso> Sometimes its handy to use both at the same time
2020-11-02 07:37:59 <dminuoso> fmap f <$> g
2020-11-02 07:38:00 × Tario quits (~Tario@200.119.185.133) (Ping timeout: 256 seconds)
2020-11-02 07:38:42 → Tario joins (~Tario@201.191.91.236)
2020-11-02 07:39:03 <danso> haskellers are pretty inconsistent on this point, you can see
2020-11-02 07:39:29 <danso> <$> is preferred over <&> but >>= is preferred over =<<
2020-11-02 07:39:39 <dminuoso> Dunno about that
2020-11-02 07:39:43 <dminuoso> I prefer =<< by far :p
2020-11-02 07:40:03 <danso> and i prefer <&> !
2020-11-02 07:40:04 <dminuoso> In fact, =<< often arises when I write code with <$> first, just to realize it should have been bind instead.
2020-11-02 07:40:13 <danso> i don't think either of us are in the majority though :^)
2020-11-02 07:40:15 <dminuoso> I found <&> to be useful in writing decoders :)
2020-11-02 07:40:44 <dminuoso> f x 1 = getWord32 <&> Something
2020-11-02 07:40:48 <dminuoso> f x 2 = getWord32 <&> AnotherThing
2020-11-02 07:41:00 <dminuoso> For a sufficiently large f this aligns much more nicer
2020-11-02 07:41:08 <danso> one of the haskell survey questions should have been >>= vs =<<
2020-11-02 07:41:21 <danso> pity we just missed it
2020-11-02 07:41:29 <dminuoso> Didn't get the memo
2020-11-02 07:42:50 → toorevitimirp joins (~tooreviti@117.182.183.132)
2020-11-02 07:45:50 <energizer> why is it called <&> ?
2020-11-02 07:46:28 <dminuoso> energizer: Presmuably because (&) is a flipped ($)
2020-11-02 07:46:39 <dminuoso> % :t ($)
2020-11-02 07:46:40 <yahb> dminuoso: (a -> b) -> a -> b
2020-11-02 07:46:43 <dminuoso> % :t (<$>)
2020-11-02 07:46:43 <yahb> dminuoso: Functor f => (a -> b) -> f a -> f b
2020-11-02 07:46:46 <dminuoso> % :t ($)
2020-11-02 07:46:46 <yahb> dminuoso: (a -> b) -> a -> b
2020-11-02 07:46:49 <dminuoso> % :t (<$>)
2020-11-02 07:46:49 <yahb> dminuoso: Functor f => (a -> b) -> f a -> f b
2020-11-02 07:47:08 <dminuoso> % :t (<&>)
2020-11-02 07:47:09 <yahb> dminuoso: Functor f => f a -> (a -> b) -> f b
2020-11-02 07:47:10 <dminuoso> Sorry
2020-11-02 07:47:12 <dminuoso> :)
2020-11-02 07:47:17 <dminuoso> % :t (&)
2020-11-02 07:47:17 <yahb> dminuoso: a -> (a -> b) -> b
2020-11-02 07:47:27 × christo quits (~chris@81.96.113.213) (Remote host closed the connection)
2020-11-02 07:48:26 → Sanchayan joins (~Sanchayan@106.200.218.30)
2020-11-02 07:48:48 → christo joins (~chris@81.96.113.213)
2020-11-02 07:48:51 <energizer> alright, so why are they called & and $ ?
2020-11-02 07:49:05 <danso> dminuoso, the survey is definitely still open
2020-11-02 07:49:08 <danso> i believe it was posted today
2020-11-02 07:49:28 <danso> energizer, because they're common things to do, so they should be short
2020-11-02 07:50:27 <energizer> $ kinda looks like flipping, yknow
2020-11-02 07:50:57 → mimi1vx joins (~mimi@tulipan.habr.nat.praha12.net)
2020-11-02 07:51:11 <energizer> & is like 'and' but i'm not sure if fmap is related to 'and'
2020-11-02 07:51:15 <dminuoso> Well you have to call them something..
2020-11-02 07:51:33 <dminuoso> There's very limited ASCII characters available
2020-11-02 07:52:34 <energizer> yeah sometimes notation is just "gotta pick something" and sometimes there's a hidden connection to some concept
2020-11-02 07:52:38 <danso> i think you have misunderstood
2020-11-02 07:52:44 <danso> $ is apply
2020-11-02 07:52:48 <danso> as in function application
2020-11-02 07:52:53 <danso> f(x) is f $ x
2020-11-02 07:52:54 <dminuoso> energizer: I dont think there's a particular reason for those choices on function application.
2020-11-02 07:53:14 <danso> & is apply with its arguments flipped, so x & f is f(x)
2020-11-02 07:53:36 <energizer> mhmm
2020-11-02 07:53:47 <danso> the parallel is that <$> is fmap and <&> is fmap with its args flipped
2020-11-02 07:54:00 <dminuoso> energizer: You could ask on haskell-cafe though, if you're curious.
2020-11-02 07:54:02 × mimi_vx quits (~mimi@2a01:490:16:1026:8cbe:b3f3:f284:209c) (Ping timeout: 264 seconds)
2020-11-02 07:54:10 <energizer> ok that does make it clearer. thanks.
2020-11-02 07:54:14 <dminuoso> It's possible someone there knows of a connection, perhaps they were burrowed from another language
2020-11-02 07:54:30 <dminuoso> Or maybe from some field in mathematics
2020-11-02 07:54:54 <dminuoso> Though I think the primary reason would be
2020-11-02 07:55:18 <dminuoso> There aren't many other characters ASCII characters left that suggest different things
2020-11-02 07:55:28 → cosimone joins (~cosimone@2001:b07:ae5:db26:d849:743b:370b:b3cd)
2020-11-02 07:55:35 → gehmehgeh joins (~ircuser1@gateway/tor-sasl/gehmehgeh)
2020-11-02 07:55:44 <dminuoso> -+*^ are already used for arithmatic
2020-11-02 07:55:51 <danso> i don't know of languages that predate haskell that use an operator for fn application
2020-11-02 07:56:04 <dminuoso> @ cant be used as a standalone operator because its used for as-paterns
←Prev  Next→
Page 1 .. 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 .. 5022

All times are in UTC.