00:28 (join) acquo
00:30 (quit) acquo: Client Quit
02:17 (quit) jao: Ping timeout: 240 seconds
02:27 (join) aLeSD
02:27 aLeSD: hi all
02:42 jonrafkind: hi
02:50 aLeSD: someone could explain me what's lambda ?
02:54 (quit) mheld: Quit: mheld
02:59 jonrafkind: it makes a function
03:00 jonrafkind: like 2 is a number, (lambda (x) x) is a function that accepts one argument and returns the same argument
03:03 (join) mheld
03:08 (quit) mheld: Ping timeout: 264 seconds
03:18 aLeSD: jonrafkind: in racket you use small functions ?
03:18 aLeSD: I mean you try to divide everything in atoms
03:20 jonrafkind: its usually good programming practice to write small functions
03:23 Lajla: aLeSD,
03:24 Lajla: (define (square x) (* x x))
03:24 aLeSD: hi Lajla
03:24 Lajla: Is actually just sugar for (define square (lambda (x) (* x x))
03:24 Lajla: In Scheme, you store a function in a variable like any other value.
03:24 Lajla: aLeSD, remember set! ?
03:24 Lajla: You can also do (define x 3)
03:25 Lajla: And later (set! x (lambda (x) (* x x))
03:25 Lajla: )
03:25 aLeSD: Lajla, if i can use (define (square x) (* x x))
03:25 aLeSD: why use (define square (lambda (x) (* x x)) ?
03:26 Lajla: aLeSD, no real reason.
03:26 Lajla: But it does the same.
03:26 aLeSD: you mean that lambda is unusefull ?
03:26 Lajla: aLeSD, it's not where you would usually use lambda.
03:26 Lajla: aLeSD, say you want a function to return another function.
03:26 aLeSD: ok let's remove lambda from racket
03:27 aLeSD: :)
03:27 aLeSD: I don't belive that is the same than define
03:28 Lajla: For instance (define (square-or-double which) (lambda (x) ((cond ((equal? which "square") *) ((equal? which "double") *)) x x)))
03:28 Lajla: If you do (square-or-double "square") it returns a squaring function
03:29 Lajla: If you do (square-or-double "double") it returns a doubling function
03:29 Lajla: aLeSD, the entire language would break apart.
03:29 Lajla: For instance, (let ((x 3) (y 1)) (+ x y)) is traditionally just sugar for ((lambda (x y) (+ x y) 3 1)
03:30 jonrafkind: you wrote * when you meant +
03:30 Lajla: (define (f x) ...) is just rewritten by the compiler to (define f (lambda (x) ...))
03:30 Lajla: jonrafkind, you are quite correct.
03:30 Lajla: Where did you muster this astute correctness?
03:31 jonrafkind: elementary
03:31 Lajla: I don't get a 'dear lajla'?
03:32 jonrafkind: that will cost $3.50
03:32 aLeSD: ok
03:34 Lajla: I thought Sherlock holmes worked for free.
03:34 Lajla: aLeSD, lambda is such a often used and fundamental part of the language that many people think the name is far to long for such a commonly used thing, and I would agree.
03:35 Lajla: But aLeSD remeber that I told you that for-each was a function and not syntax?
03:35 aLeSD: ok
03:35 aLeSD: i am going slowly
03:35 aLeSD: give me time to learn let set etc
03:35 aLeSD: lol
03:38 (quit) jonrafkind: Ping timeout: 255 seconds
03:40 Lajla: aLeSD: http://codepad.org/cEOQAVyh
04:17 aLeSD: ok
04:17 aLeSD: ?I will use lambda when I want to create a function that doesn't need a name
04:19 aLeSD: Lajla
04:19 Lajla: aLeSD, you will use lambda when you create a function.
04:20 Lajla: Like I said (define (f x) ...) is just sugar for (define f (lambda (x) ...))
04:20 Lajla: The compiler rewrites it to the latter form before proceeding
04:21 aLeSD: Lajla, I will create a calculator
04:21 aLeSD: mmm
04:22 aLeSD: no
04:22 aLeSD: ok I will create the first stuff with fluxus
04:23 aLeSD: I think that you have to be very smart to program in racket
04:23 Lajla: yeah
04:23 aLeSD: how you decide to program the stuff is very important
04:23 Lajla: we're all pro.
04:23 Lajla: Pretty much
04:23 Lajla: It's not really a RAD language.
04:44 (quit) aLeSD: Ping timeout: 240 seconds
06:06 (quit) cataska: Quit: leaving
06:36 (join) masm
07:03 (join) mceier
08:02 (join) MayDaniel
08:46 (quit) masm: Quit: Leaving.
08:53 (join) sunnyps
09:39 (join) jeapostrophe
09:42 (quit) jeapostrophe: Client Quit
09:45 (join) jeapostrophe
09:45 (quit) jeapostrophe: Client Quit
10:05 (join) jeapostrophe
10:05 (quit) Lajla: Read error: Connection reset by peer
10:06 (join) Lajla
10:09 (quit) jeapostrophe: Client Quit
10:28 cky: rudybot: (define (square-or-double type) (let ((func (case type ((square) *) ((double) +)))) (lambda (x) (func x x))))
10:28 rudybot: cky: your r5rs sandbox is ready
10:28 rudybot: cky: Done.
10:29 cky: rudybot: ((square-or-double 'square) 5)
10:29 rudybot: cky: ; Value: 25
10:29 cky: rudybot: ((square-or-double 'double) 5)
10:29 rudybot: cky: ; Value: 10
10:29 cky: :-D
10:29 cky: (Just my reformulation of Lajla's function.)
10:31 Lajla: cky, yeah, I didn't want to introduce my padawan to symbols just yet.
10:31 cky: :-)
10:31 Lajla: I will do so when he is ready.
10:31 Lajla: And can control his powers better.
10:31 cky: rudybot: (define (square-or-double type) (let ((func (case type (("square") *) (("double") +)))) (lambda (x) (func x x))))
10:31 rudybot: cky: Done.
10:31 cky: rudybot: ((square-or-double "double") 5)
10:31 rudybot: cky: error: procedure application: expected procedure, given: #<void>; arguments were: 5 5
10:31 Lajla: I sense much fear in him, I fear that his childhood experiences may have clouded his judgement
10:31 cky: Hmpf. Oh well.
10:31 Lajla: lolololololo
10:32 Lajla: I think
10:32 Lajla: that case cannot compare strings
10:32 Lajla: case uses eqv?
10:32 cky: Fail. :-P
10:32 (join) mheld
10:32 cky: Ah.
10:32 Lajla: THis is a part I don't like though.
10:32 Lajla: like that whole member, memq, memv and all.
10:33 Lajla: I would rather use somethinf like (foo (lambda (k) (eqv? k the-value-I-search)) some-list)
10:33 Lajla: Just using a unary predicate.
10:34 cky: Isn't that called find (from SRFI 1)?
10:36 Lajla: Seriously?
10:36 Lajla: Then I totally missed that when reading 1.
10:36 Lajla: I use that, but call it give-first
10:37 Lajla: So that I can name filter...
10:37 Lajla: give-all
10:37 Lajla: I am a sucker for consistency
10:37 cky: Hahahahaha.
10:38 Lajla: Though
10:38 Lajla: I am renaming it to
10:38 Lajla: first-when
10:38 Lajla: En all-when
10:38 Lajla: So that I can be consistent
10:38 Lajla: and have first-while
10:38 Lajla: and all-while
10:39 Lajla: I never liked equal? to be honest.
10:39 Lajla: I mean, what if you invent your own super duper data structure.
10:39 Lajla: Should you update equal? for that?
10:40 Lajla: I also think that code that uses equal? instead of vector-equal? or string-equal?, id est, code that is ignorant about whether it are vectors or strings in that case is generally silly.
10:40 Lajla: Or lists, whatever
10:43 cky: *nods*
10:43 cky: Yeah, I don't know what to do about equal? either.
10:43 cky: I wish there's a formal system for extending equal?
10:43 cky: Kind of like setf-handlers in CL.
10:45 cky: But, kind of pointless, like you say.
10:45 Lajla: cky, yeah, I think it should ust be limited to string-equal, vector-equal, et cetera.
10:45 Lajla: And if you invent some nice fancy datastructure.
10:45 Lajla: Like a Lajla.
10:45 Lajla: Just make a lajla-equal
10:46 cky: That reminds me.
10:46 cky: Making a generic function for equal? could be nice.
10:46 Lajla: And don't go like (define equal? (let ((old equal?)) (lambda xs (if (lajla? (car xs)) (apply lajla-equal? xs) (apply old xs))))))
10:46 Lajla: Or something silly like that
10:46 Lajla: what do you mean?
10:46 cky: I'm thinking of CLOS-style generic functions.
10:47 cky: So you can specify multiple variants of equal?, each with a different type signature.
10:47 cky: So you could define a (equal? <string> <string>), (equal? <character> <character>), (equal? <number> <number>), etc.
10:48 cky: And after all that, the most generic equal? could just be defined to be eq?.
10:48 Lajla: No, I think that is silly too.
10:48 Lajla: I think that people should just know what they are comparing and use string-equal? vector-equal? et cetera accordingly.
10:49 cky: But that makes it hard to write generic algorithms. :-(
10:49 Lajla: And code that is written so that you do not know that statically is silly most of the time.
10:49 Lajla: It is hard enough as it is.
10:49 Lajla: Because vectors do not cons.
10:49 cky: See, I come from C++, where generic algorithms are used a lot.
10:49 cky: I'm not talking about cons.
10:49 cky: I'm talking about conceptually higher-level operations.
10:49 Lajla: cky, I would disagree that they are more generic than dynamic typing.
10:50 Lajla: Usually it involdes not having to rewrite a function that sorts an array of floats when you already have one that sorts an array of integers.
10:50 Lajla: cky, do you know Haskell, or Ocaml, or any of that type?
10:51 cky: I understand the basic ideas about type inference, if that's what you mean.
10:51 Lajla: Well, not that.
10:51 Lajla: I mean, type variables.
10:51 Lajla: How it shows that C++ templates are silly.
10:51 cky: I vaguely remember those.
10:52 Lajla: I mean, there is no real reason for a function that sorts an array with quicksort to care about what the type of the elements is really.
10:52 cky: Right, as long as the ordering function works correctly, that's all that's required.
10:52 Lajla: YEah
10:53 Lajla: so, this nice Hindly Milner type system leaves that type as a type variable.
10:53 Lajla: So the type of this function is (a -> a -> bool) -> [a] -> [a]
10:53 Lajla: Basically, a is foralled.
10:53 Lajla: Universal quantificationationahaskash
10:54 cky: Heh.
10:54 Lajla: So the sort function is correctly typed, provided that its comparison operator is of type a -> a-> Bool
10:54 Lajla: And its input list of of type [a]
10:54 Lajla: And it also outputs a list of type [a] based on that.
10:55 Lajla: But its greatest asset is
10:55 Lajla: THat it shows just how silly C++ is. :')
10:59 cky: Heh.
11:00 (join) jeapostrophe
11:05 Lajla: cky, but another part is yeah that most (all?) expressions can have some type signature be inferred by the compiler.
11:05 Lajla: THough, the point is that it's not always the one you want.
11:05 Lajla: Like, if you go factorial n = prod [1 .. n]
11:05 Lajla: It infers its type as Int -> Int
11:05 Lajla: But maybe you wanted float -> float
11:06 Lajla: Because prod also works on floats.
11:06 Lajla: And as far as I know, hoaskell does not have a numerical tower like good old scheme
11:10 cky: *nods*
11:33 (quit) Lajla: Ping timeout: 260 seconds
11:35 (quit) jeapostrophe: Quit: jeapostrophe
11:48 (join) Lajla
12:22 (join) anRch
12:48 (quit) MayDaniel: Read error: Connection reset by peer
12:54 (quit) anRch: Quit: anRch
12:56 (quit) sunnyps: Quit: Leaving
13:00 (join) jeapostrophe
13:04 (join) anRch
13:17 (quit) jeapostrophe: Quit: jeapostrophe
13:30 (join) cinch
13:44 (quit) anRch: Quit: anRch
14:08 (join) jonrafkind
14:15 (join) MayDaniel
14:26 (join) jeapostrophe
14:26 (quit) jeapostrophe: Client Quit
14:45 (join) Fare
15:07 (quit) Fare: Quit: Leaving
15:57 (join) mwolfe
16:06 (quit) MayDaniel: Read error: Connection reset by peer
16:45 (quit) em: Read error: Connection reset by peer
16:46 (join) emma
16:48 (nick) emma -> em
16:57 (quit) offby1: Read error: Connection reset by peer
16:58 (join) offby1
17:01 (quit) tv|z: Ping timeout: 240 seconds
17:03 (join) tv|z
17:08 (quit) offby1: Read error: Connection reset by peer
17:08 (join) offby1
17:16 (join) anRch
17:40 (quit) anRch: Quit: anRch
18:13 (join) MayDaniel
18:19 (quit) offby1: Ping timeout: 260 seconds
18:27 (quit) pattern: Ping timeout: 250 seconds
18:28 (join) elly
18:29 (quit) MayDaniel: Read error: Connection reset by peer
18:32 (join) offby1
18:35 elly: is there a good example of a nontrivial program written in Racket that makes use of racket's contract system somewhere?
18:53 jonrafkind: typed racket
19:28 (quit) offby1: Read error: Connection reset by peer
19:28 (quit) mwolfe: Remote host closed the connection
19:29 (join) offby1
20:44 (quit) jonrafkind: Ping timeout: 240 seconds
21:00 (quit) mceier: Quit: leaving
21:09 (join) bremner
21:26 (quit) offby1: Read error: Connection reset by peer
21:43 (join) offby1
21:49 askhader has never used racket's contract system. Just informal contracts found within comments
22:07 (join) jeapostrophe
22:23 (join) jonrafkind
23:09 (quit) jeapostrophe: Quit: jeapostrophe
23:59 (join) jeapostrophe
23:59 (quit) jeapostrophe: Client Quit