00:52 (quit) martinhex: Remote host closed the connection
00:55 (join) martinhex
00:57 (join) rbarraud
01:30 (quit) rbarraud: Ping timeout: 276 seconds
01:45 (join) vu3rdd
02:00 (join) rbarraud
02:10 (quit) jonrafkind: Ping timeout: 265 seconds
04:16 (join) masm
06:46 (quit) rbarraud: Remote host closed the connection
06:47 (join) rbarraud
07:00 (join) xrymbos
07:01 xrymbos: hi
07:01 xrymbos: is there a concise way to get a list of numbers from 1 to 100?
08:01 Lajla: xrymbos, sure is.
08:01 Lajla: range
08:01 Lajla: rudybot: eval (range 1 101)
08:01 rudybot: Lajla: error: reference to undefined identifier: range
08:01 Lajla: I think
08:01 Lajla: Ahh, npt
08:01 Lajla: What is range called here man?
08:01 Lajla: It's always caleld range
08:01 Lajla: Oh well, then this
08:02 Lajla: rudybot: eval (list-build 100 (lambda (n) (+ 1 n))
08:02 rudybot: Lajla: error: eval:1:0: read: expected a `)' to close `('
08:02 Lajla: fff
08:02 Lajla: rudybot: eval (list-build 100 (lambda (n) (+ 1 n)))
08:02 rudybot: Lajla: error: reference to undefined identifier: list-build
08:02 Lajla: Ffff
08:02 Lajla: rudybot: eval (build-list 100 (lambda (n) (+ 1 n))
08:02 rudybot: Lajla: error: eval:1:0: read: expected a `)' to close `('
08:02 Lajla: ffffffffffffffff
08:02 Lajla: rudybot: eval (build-list 100 (lambda (n) (+ 1 n)))
08:02 rudybot: Lajla: ; Value: (1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100)
08:02 Lajla: Yeah
08:02 xrymbos: awesome
08:03 xrymbos: thanks :)
08:03 Lajla: xrymbos, I am actually better with this than this example might let you think
08:03 Lajla: xrymbos, anyeay
08:03 Lajla: way*
08:03 Lajla: What you can now do is (define (range start finish) (build-list finish (lambda (n) (+ n start)))) I guess
08:03 Lajla: It's not as good as the range that I have running here though
08:04 Lajla: Which also goes backawrds and can step
08:04 xrymbos: yup, cool
08:12 Lajla: xrymbos: http://pastebin.com/p3YT654s
08:12 Lajla: Haven't tested it though
08:12 xrymbos: oh man, you can do x.xs?
08:12 Lajla: Also, some would argue that my code is not idiomatic, but that is just how I roll.
08:12 xrymbos: awesome :)
08:13 Lajla: Well, it's not a pattern matching like in haskell
08:13 (join) hanDerPeder
08:13 Lajla: http://pastebin.com/bGcmk7bC this is indented slightly better
08:14 Lajla: NO
08:14 Lajla: It doesn't work with negative ranges.
08:14 Lajla: I failed
08:15 Lajla: Ahh
08:15 Lajla: I see the bug.
08:16 Lajla: Well actually let it be like that
08:16 Lajla: That's actually good
08:17 Lajla: if you want negative, use (range 10 4 -1)
08:18 Lajla: xrymbos, http://pastebin.com/mpfk8gC9
08:18 (quit) rbarraud: Read error: Connection reset by peer
08:18 Lajla: I stand by that version
08:44 (quit) xrymbos: Ping timeout: 276 seconds
08:58 (join) b-man_
09:30 Lajla: http://www.ccs.neu.edu/scheme/pubs/scheme2009-fb.pdf eli, why do you write PDF documents I randomly google?
09:30 Lajla: Stalker
09:43 (join) TOSTO
10:18 (join) MayDaniel
10:34 (join) confounds
10:38 (quit) samth_away: Quit: Ex-Chat
10:47 (quit) hanDerPeder: Quit: hanDerPeder
10:49 (join) hanDerPeder
10:49 eli shrugs
11:22 (quit) hanDerPeder: Quit: hanDerPeder
11:35 (quit) vu3rdd: Remote host closed the connection
11:42 (join) carleastlund
11:45 (join) danbrown
11:52 (join) sstrickl
11:52 (quit) b-man_: Remote host closed the connection
12:16 (join) jonrafkind
12:22 (quit) danbrown: Quit: danbrown
12:37 (join) tommc
12:41 (join) hanDerPeder
12:48 (quit) hanDerPeder: Ping timeout: 276 seconds
12:51 (quit) MayDaniel: 
12:52 (quit) confounds: Remote host closed the connection
13:09 (join) confounds
13:14 (join) hanDerPeder
13:16 (join) b-man_
13:18 jonrafkind: unstable/find is whacky.. it alternatively tests the cdr of the list and the car of the list
13:41 (quit) bremner: Quit: ZNC - http://znc.sourceforge.net
13:45 (join) bremner
13:47 (quit) bremner: Client Quit
13:48 (join) bremner
13:49 (join) danbrown
13:50 (quit) bremner: Client Quit
13:50 (join) bremner
13:53 (quit) bremner: Client Quit
13:55 (quit) tommc: Ping timeout: 265 seconds
13:57 (join) bremner
14:06 (quit) confounds: Remote host closed the connection
14:34 (join) mceier
14:37 carleastlund: jonrafkind, why is that behavior wacky?
14:38 jonrafkind: i expected it to just look at each element of the list and run the predicate on it
14:38 jonrafkind: i eventually wrote my own function to do that
14:39 jonrafkind: because my predicate expects a certain type (the type of the element), when it receives a list of elements it barfs
14:40 carleastlund: What list?
14:40 carleastlund: The input to "find" from unstable/find does not have a list anywhere in it.
14:41 jonrafkind: find symbol? '((all work) and (no play)))
14:41 jonrafkind: im pretty sure the second argument is a list
14:41 jonrafkind: http://docs.racket-lang.org/unstable/find.html?q=unstable/find
14:41 carleastlund: Sure, and I can pass a list to print as well, but when describing what print does I don't say "the list".
14:41 jonrafkind: unless these docs have changed a whole lot
14:42 carleastlund: Which input to find is "the list"?
14:42 jonrafkind: the second argument
14:42 carleastlund: x : any/c
14:42 carleastlund: Doesn't look like a list to me.
14:48 (join) ttuttle
14:48 ttuttle: What exactly happens if I mutate the elements in a set?  (I'm writing a chat client, and trying to keep track of the people in a channel, when their nicknames might change.
14:49 jay-mccarthy: unexpected things
14:51 ttuttle: alright, I guess I'll just use a list.
14:51 jay-mccarthy: you can have a set that you keep in a box
14:52 jay-mccarthy: you just can't mutate the ELEMENTS of the set
14:52 ttuttle: oh, right
14:52 ttuttle: well
14:52 jay-mccarthy: (define s "Jay")
14:52 jay-mccarthy: (define ss (set s))
14:52 jay-mccarthy: (set-member? ss "Jay") => #t
14:52 ttuttle: I need to be able to mutate the items, since people's names can change.
14:52 ttuttle: So I guess I just can't use a set.
14:52 jay-mccarthy: (string-set! s 2 #\e)
14:52 jay-mccarthy: (set-member? ss "Jae") => unknown
14:53 jay-mccarthy: you could remove the old name and add the new name
14:54 (join) samth
14:54 ttuttle: ...hmm.
14:55 carleastlund: If you use an eq-based set, you're safe.  Mutation won't affect eq? or eq-hash-code.
14:55 ttuttle: so, as long as a user remains online, their box should never change, only the contents of the box
14:55 carleastlund: Or if you make a custom structure with a prop:equal+hash-code that is oblivious to its mutable fields, that'll work too.
14:56 ttuttle: I mean, the only reasonable identifier for an IRC user is their nickname.
14:56 ttuttle: and that can change.
14:56 carleastlund: In that case, it's not a very useful key for a set.
14:56 ttuttle: I mean, at any given point, it is.
14:56 carleastlund: If the only reasonable identifier is something that changes, it sounds like you don't have an identifier.  Or, you need to treat "changes" as removing one user and adding another.
14:56 ttuttle: But over time it's not.
14:57 ttuttle: I guess changes will need to be remove and readd.
14:57 carleastlund: Well, if your sets need to persist over time, they need element identity that does, too.
15:13 ttuttle: man, this is going to be full of O(n^2) operations >.<
15:13 ttuttle: or at least O(n_channels)
15:30 ttuttle: Man, I am remembering why I loved functional programming so much.
15:33 (join) MayDaniel
15:33 Lajla: ttuttle, as far as the client goes, it only needs to keep track of which nickname says what, and when a nickname changes I guess.
15:33 Lajla: Like, let's say I do /nick foo when you are not in this channel.
15:33 Lajla: Like you quit, I do /nick foo,and you come back.
15:34 ttuttle: Lajla: Oh, I know I can't track that.
15:34 Lajla: THere is no way for you to know then if I changed, or if I left and another joined.
15:34 Lajla: So you can't track of 'who is in the channel'
15:34 ttuttle: Lajla: I'm only tracking this for channels I'm in.
15:35 Lajla: I guess then you could just give them all a number.
15:38 (quit) MayDaniel: 
16:06 (quit) lisppaste: Ping timeout: 240 seconds
16:08 (quit) b-man_: *.net *.split
16:23 (quit) sstrickl: Quit: sstrickl
16:28 ttuttle: Is there any way to erase all the mappings in a (mutable) hash at once?
16:29 jay-mccarthy: why not just start using a new one
16:31 carleastlund: Alternately, use immutable hash tables.  Then using a new one is just as simple, and much more natural.  :)
16:33 (join) xrymbos
17:13 Lajla: ttuttle, is there any reason by the way why constructing a new set with the old set as a base won't work?
17:22 (join) confounds
17:25 ttuttle: Lajla: that will work, but I've decided to use a hash and just remove and readd instead
17:25 Lajla: ttuttle, the inevitable quaestion arising: why scheme if you want to use destructive update?
17:25 ttuttle: Lajla: IRC is annoyingly stateful
17:26 ttuttle: Lajla: I am containing all the stateful stuff in one ugly class.
17:28 samth: Lajla, b/c Racket is a great language for writing imperative programs in
17:28 ttuttle: mutable hashes are lovely to work with
17:28 ttuttle: and this state is inherently mutable, so I see no reason to pretend it's not and then stick it in a box.
17:30 Lajla: samth, you are a very strange person.
17:31 ttuttle: samth is right though :P
17:38 samth: how is that strange?
17:38 samth: I would say the strange person is the one on #racket wondering why people are using Racket
17:39 ttuttle: I am using it because, even though my program has to have some imperative parts, it is a joy to implement the rest functionally, and Racket has things like events that free me from doing stupid gruntwork.
17:47 ttuttle: I have to admit it is confusing getting used to explicit currying.  (The last functional language I used was SML, for class.)
17:53 carleastlund: Racket is a functional, imperative, object-oriented, dynamic, typed, reactive language.  You can combine any or all of those features in one program, along with anything else you find in our libraries.  We don't believe in "our way or the highway".  We throw in everything, including the highway and the kitchen sink.
17:54 jonrafkind: you should put that on the racket-lang.org
17:54 ttuttle: it is awesome
17:54 carleastlund: And, ttutle: while mutation and mutable hashes are just fine, very little is "inherently mutable", depending on your programming model.
17:56 carleastlund: IRC identities inherently *change over time*, that is true.  Mutation is only one way of modeling change over time.
17:56 ttuttle: somewhat sadly, my event model is imperative.
17:56 ttuttle: (modules call an event-callback to propagate events.)
17:57 jonrafkind: isnt CML (store-passing style) a way to handle changes over time?
17:57 ttuttle: I would have used handle-evt, but that'd require that every low-level event (e.g., line read from the IRC server) mapped to one high-level event.
17:57 carleastlund: CML is not in any way related to store-passing style, as far as I know.
17:58 carleastlund: ttuttle: (a) that's probably not true, and (b) I don't see it as a problem anyway.
17:58 jonrafkind: oh maybe i confused terms, but the point is about store-passing style, not CML
17:58 ttuttle: carleastlund: okay, cool.
17:58 ttuttle: I don't see a particular problem with propagating events imperatively, as long as the flow is clear.
17:59 carleastlund: In any event, there is an advantage to sticking an immutable hash table in a box -- you get mutation at the top level, which can simplify your event model, but back-end libraries get to work with the data in a persistent manner, and "emptying" the hash table becomes really easy.
17:59 carleastlund: I'm a big fan of putting a thin layer of mutation at the top level of a program, and then programming purely functionally from there down to the bottom.  It's often a really convenient way to work -- you can interface with imperative APIs, but you can test your programs functionally and so forth.
18:01 ttuttle: hmm
18:02 ttuttle: so the interface between the "irc client" class and the "keep track of who's where" class is imperative, and I don't think I want to change that.
18:02 ttuttle: but I might be able to make the guts of the tracker class more functional.
18:03 (quit) jonrafkind: Ping timeout: 260 seconds
18:03 carleastlund: On the other hand, (define (reset-hash! h) (for ([k (in-hash-keys h)]) (hash-remove! h k))) is really easy to write.  ;)
18:11 (join) rmitt
18:22 (join) neilv
18:22 (quit) neilv: Changing host
18:22 (join) neilv
18:29 (join) Nanakhiel
18:33 (quit) Lajla: Ping timeout: 276 seconds
18:35 (quit) carleastlund: Quit: carleastlund
18:47 (join) rbarraud
19:01 (quit) danbrown: Quit: danbrown
19:17 (quit) rmitt: Quit: Page closed
19:27 (quit) mceier: Quit: leaving
19:52 (join) kingless
20:01 (quit) kingless: Quit: Colloquy for iPad - http://colloquy.mobi
20:32 (join) Fare
20:38 (quit) hanDerPeder: Quit: hanDerPeder
21:07 (quit) Fare: Quit: Leaving
21:13 (quit) confounds: Remote host closed the connection
21:16 (quit) xrymbos: Remote host closed the connection
21:17 (quit) neilv: Quit: Leaving
21:38 (join) masm1
21:39 (quit) masm: Ping timeout: 250 seconds
22:24 (quit) masm1: Quit: Leaving.
23:26 (join) neilv
23:26 (quit) neilv: Changing host
23:26 (join) neilv
23:38 (quit) rudybot: Remote host closed the connection
23:38 (join) rudybot