00:00 (join) jeapostrophe
00:11 rekahsoft: how is the documentation on racket-lang.org and planet racket (the reference documentation) generated? i assume something to do with embedding scribble into my code?
00:12 rekahsoft: nevermind..found some good documentation on http://docs.racket-lang.org/scribble/how-to-doc.html
00:20 (quit) lucian_: Ping timeout: 246 seconds
00:38 (quit) jeapostrophe: Quit: jeapostrophe
01:15 (quit) realitygrill: Quit: realitygrill
01:19 eli: lewis1711: There's a hiding policy thing that hides library macros by default.  You'll need to disable hiding for the `for' (or completely to see the whole thing.)
01:34 (join) CoolgyFurlough
01:37 CoolgyFurlough: sorry to ask such a noob question, but how do you install racket on Ubuntu 11.04
01:38 CoolgyFurlough: ?
01:41 (quit) CoolgyFurlough: 
01:41 jonrafkind: fail
01:41 (join) CoolgyFurlough
01:50 jonrafkind: sup
01:50 jonrafkind: you can compile from source or use a binary installer
01:50 jonrafkind: or use a PPA for ubuntu
01:54 CoolgyFurlough: sorry, i was having trouble with the .sh file
01:54 CoolgyFurlough: then i figured out about chmod
01:54 jonrafkind: or just 'sh blah.sh'
02:13 (quit) jonrafkind: Ping timeout: 240 seconds
02:54 (quit) CoolgyFurlough: 
03:42 lewis1711: eli: oh yes so it does. thanks
03:44 (join) masm
03:47 (join) Fulax
04:22 (join) mceier
04:29 (join) tfb
05:07 (join) noelw
06:35 (join) MayDaniel
06:58 (join) jeapostrophe
07:14 (quit) MayDaniel: Read error: Connection reset by peer
07:24 (join) ChibaPet
07:41 (quit) jeapostrophe: Quit: jeapostrophe
07:42 (join) kPb_in
07:53 (join) jeapostrophe
07:55 (quit) DGASAU: Ping timeout: 260 seconds
08:00 (join) DGASAU
08:22 lewis1711: https://gist.github.com/1153948 seems like a bug - this compiles fine if I comment out line 3, which should have no effect on type checking.
08:40 (quit) weirdo: Quit: WeeChat 0.3.6-dev
08:45 (join) weirdo
08:48 (part) lewis1711
08:49 (quit) weirdo: Client Quit
09:31 (quit) jeapostrophe: Quit: jeapostrophe
09:32 (join) jeapostrophe
09:32 (quit) rekahsoft: Ping timeout: 240 seconds
09:32 (join) rekahsoft
09:41 (quit) tfb: Quit: gone
09:41 (join) tfb
10:09 (quit) jeapostrophe: Quit: jeapostrophe
10:36 (join) hussaibi
10:40 (join) dnolen
10:43 (quit) dnolen: Client Quit
11:03 (quit) kPb_in: Quit: kPb_in
11:07 (join) jeapostrophe
11:11 (quit) noelw: Quit: noelw
11:37 (join) dnolen
11:37 (join) RacketCommitBot
11:37 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://bit.ly/oKMFqx
11:37 RacketCommitBot: [racket/master] fix `regexp-match' performance for short matches on long strings - Matthew Flatt
11:37 (part) RacketCommitBot
12:02 tauntaun: Suppose that struct B is a subtype of struct A and that function f takes one argument x, expected to be of type struct A, and performs a functional update on it, returning a value of struct A.  Suppose f is actually passed an instance y of struct B.  Function f will presumably perform the functional update on the "A part" of y, but I'd like it also to preserve the "B part" of y, so that it returns a B instance. Is this possible in
12:02 tauntaun: Racket?
12:02 tauntaun: (struct-copy seems to say 'no')
12:06 (join) MayDaniel
12:15 (quit) MayDaniel: Read error: Connection reset by peer
12:16 (nick) samth_away -> samth
12:20 samth: tauntaun: that's not in general possible without really gross hacks
12:20 samth: i think this is a limitation of the current struct system that we need to address, but structs are already pretty complicated
12:21 (quit) hussaibi: Ping timeout: 250 seconds
12:22 (join) hussaibi
12:25 tauntaun: samth: I was afraid so (and I agree that structs have gotten a bit complicated).  A corollary question: does this mean that I should switch from structs to classes?  I'm hoping not, because I'd really like the code to stay referentially transparent as it grows bigger and bigger, and I don't see how that's possible with dynamic method dispatch.
12:26 (join) jonrafkind
12:34 tauntaun: (When I say 'switch', I don't mean change all structs to classes everywhere in the code.  I mean only in particular cases.)
12:44 samth: tauntaun: first, i don't see a conflict between referential transparency and classes
12:44 samth: second, i don't see how classes would fix the problem
12:50 (join) lucian
12:50 (quit) hussaibi: Ping timeout: 252 seconds
12:51 samth: rudybot: tell lewis1711 the problem is that `ffi' provides a different binding for `->'
12:51 rudybot: samth: ehrm... sorry for coming back on the issue but...
12:51 samth: rudybot: help
12:51 rudybot: samth: help [<what>], version, quote, source, seen <nick>, uptime, t8 <from> <to> <text> ..., init [<lang>], eval <expr> ..., give <to> <expr> ..., apropos <str> ..., desc <id>, doc <id>, later "tell" <nick> <something> ...
12:51 samth: rudybot: later tell lewis1711 the problem is that `ffi' provides a different binding for `->'
12:51 rudybot: minion: memo for lewis1711: samth told me to tell you: the problem is that `ffi' provides a different binding for `->'
12:57 (quit) tfb: Quit: sleeping
12:57 (join) sstrickl
13:07 tauntaun: samth: my goal is to be able to change a field value in a supertype.  I prefer functional updates, but this implies that each subtype will need its own function to update that same field.  Classes prevent the problem because, as you know, the subtypes would need only to inherit one supertype method.
13:07 (join) chturne
13:08 tauntaun: As for referential transparency, I understand the term to mean that the behavior of the program is determined entirely by its text, and it seems to me that this gets broken by method dispatch, which, as you know, is determined only at runtime.
13:10 asumu: tauntaun: How about (if (complicated-but-deterministic-computation) f1 f2)? Is that referentially transparent?
13:12 asumu: I think what method dispatch does do is potentially break modularity, but you don't need methods for that.
13:13 tauntaun: asumu: if complicated-but-deterministic-computation is RT, then it seems to me the answer is yes.
13:13 tauntaun: If I'm wrong, please correct me.
13:14 asumu: Not saying you're wrong. RT isn't rigorously defined so I don't know what it means either.
13:15 tauntaun: asumu: I'm surprised.
13:15 tauntaun: I thought RT means that any function invocation in the program text can be lexically replaced by the function's body.
13:16 tauntaun: So that the program text becomes mathematics.
13:23 tauntaun: asumu: If you know of another definition of RT, please do let me know about it.
13:27 (join) soegaard
13:38 (quit) lucian: Read error: Operation timed out
13:38 (quit) soegaard: Quit: Page closed
13:40 (join) soegaard
14:05 samth: tauntaun: dynamic dispatch is equivalent to refactoring to add a bunch of conditionals -- it's referentially transparent IFF the rest of the language is
14:07 tauntaun: Ah.
14:07 tauntaun: Thanks!
14:08 samth: tauntaun: you need to play some tricks for classes to fix the problem
14:09 tauntaun: samth: Do you mean tricks to perform functional updates?
14:09 samth: ie, you'd have to basically write out all the code duplication that you'd need to implement an updater for (your particular) structs
14:10 tauntaun: OK, so are we converging on the conclusion that it might be better to scrap RT and allow objects to be mutable?
14:10 tauntaun: samth: ^
14:10 tauntaun: s/better/easier
14:13 samth: tauntaun: no, i'm just saying that either way, there's probably a bunch of boilerplate to write
14:13 samth: avoiding mutation is still usually a good idea
14:14 tauntaun: samth: but mutation obviates boilerplate
14:14 tauntaun: (say that three times real fast)
14:14 samth: yes
14:14 samth: but it leads to other problems
14:15 tauntaun sighs
14:16 tauntaun: samth: The context, btw, is a game.  The structs in question represent game agents.  I recall your saying once that that is in fact a classic use case for mutation (e.g., agent moves from one room to another, or agent picks up an object).  Any further thoughts?
14:17 samth: tauntaun: i don't have much really enlightening to say on the balance between mutation and not
14:17 tauntaun: Actually, at that time you referred me to a pretty good explanation in HtDP.
14:17 tauntaun: I should stop agonizing and bite one bullet or the other.
14:18 samth: yeah, there's only one way to find out
14:19 tauntaun: Anyway, thanks again for the help.
14:19 tauntaun goes away in search of a coin to toss.
14:19 (part) ChibaPet
14:21 (join) bluezenix
14:23 (join) hussaibi
14:23 (quit) chturne: Ping timeout: 240 seconds
14:27 (join) chturne
14:40 (quit) sstrickl: Quit: sstrickl
15:07 (join) MayDaniel
15:22 (join) RacketCommitBot
15:22 RacketCommitBot: [racket] plt pushed 5 new commits to master: http://bit.ly/pGYEHm
15:22 RacketCommitBot: [racket/master] fix optimizer problems with mutability checking, inlining - Matthew Flatt
15:22 RacketCommitBot: [racket/master] configure: make --disable-jit imply --disable-futures - Matthew Flatt
15:22 RacketCommitBot: [racket/master] fix reader graph-notation equivalence - Matthew Flatt
15:22 (part) RacketCommitBot
15:36 (quit) dnolen: Ping timeout: 252 seconds
15:52 (join) dnolen
16:23 (join) bluezenix1
16:26 (quit) bluezenix: Ping timeout: 250 seconds
16:33 jonrafkind: whats the syntax for heredocs in racket
16:34 (quit) rekahsoft: Ping timeout: 246 seconds
16:51 (quit) chturne: Quit: Leaving
16:56 soegaard: jonrafkind: http://docs.racket-lang.org/reference/reader.html?q=here#(idx._(gentag._188._(lib._scribblings/reference/reference..scrbl)))
17:09 (part) shofetim: "ERC Version 5.3 (IRC client for Emacs)"
17:21 (join) lucian
17:29 (join) anRch
17:32 jonrafkind: (system "racket") works but (process "racket") doesn't.. whats the deal?
17:33 jonrafkind: err (process "racket foo") fails
17:36 jonrafkind: ah it does work.. user error
17:36 (quit) jeapostrophe: Quit: jeapostrophe
17:39 stamourv: Lubarsky's second law of cybernetic entomology: I
17:39 stamourv: t's _never_ a compiler bug.
17:39 jonrafkind: what are his other laws
17:39 stamourv: Exception to Lubarsky's second law of cybernetic entomology: Except when you wrote the compiler.
17:39 stamourv: The first one is: There's always _one_ more bug.
17:40 jonrafkind: the first law is definately true
17:41 (join) sstrickl
17:49 (quit) lucian: Read error: Operation timed out
18:11 (join) RacketCommitBot
18:11 RacketCommitBot: [racket] plt pushed 3 new commits to master: http://bit.ly/qzwMLQ
18:11 RacketCommitBot: [racket/master] typo in the description - Eli Barzilay
18:11 RacketCommitBot: [racket/master] Added explicit copyleft page to the website. - Eli Barzilay
18:11 RacketCommitBot: [racket/master] `purify-http-port' returns a new port, so make it close the old one. - Eli Barzilay
18:11 (part) RacketCommitBot
18:13 (quit) soegaard: Ping timeout: 252 seconds
18:20 (quit) anRch: Quit: anRch
18:26 (join) RacketCommitBot
18:26 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://bit.ly/r3bfd5
18:26 RacketCommitBot: [racket/master] allow left-aligned figures - Matthias Felleisen
18:26 (part) RacketCommitBot
18:27 (join) cb`
18:35 ohwow_: If I am in REPL (Geiser currently), is there a way to run all top-level expressions?
18:38 (join) jeapostrophe
18:40 ohwow_: hm nevermind actually
18:41 ohwow_: i have another question actually. I have a module which exports a frame% object. I want to use it as a component in my main module and put it on another frame
18:41 ohwow_: how can I do that?
18:41 ohwow_: I didn't find any set-parent option
18:42 (quit) jeapostrophe: Client Quit
18:43 jonrafkind: actually I think there is some set-parent thing now, i asked for it a while ago
18:45 ohwow_: send: no such method: set-parent for class: frame%
18:45 ohwow_: I read in documenation that the container might be a good choice tho
18:45 ohwow_: :)
18:46 ohwow_: ah, but frame is a top-level-window% while container accepts only sub-window% as a child
18:51 ohwow_: How can I find all classes which implements specific interface?
18:54 jonrafkind: yea haha.. grep..
18:55 ohwow_: o
18:56 ohwow_: Hm.. I though that this can be quite important feature. It's kinda strange that on a class page you can see which interfaces it is implementing, but not vice versa. I might be wrong tho.
18:56 ohwow_: I asked that on the mailing list
18:56 jonrafkind: yea, java has that on some of its pages (implementing classes x, y, z), i like that feature
19:00 ohwow_: Hm, okay. Basically what I want to do is to write a library (module) which exports a panel/window which can be embedded/used in other programms. However, if I use the frame% class for this (which is preferable, since it can be shown without embeding) I have no way of specifing a parent for it nor I can add it as a child to a panel. When I use pane% or panel% I *must* specify parent when exporting (which I obviously can not do).
19:01 ohwow_: Am I doing something wrong here? I want to basically write what might be called a "GUI component".
19:02 jonrafkind: you could export a function that takes a parent and returns the frame
19:02 jonrafkind: or you could write a mixin
19:02 ohwow_: Hm, that's an interesting solution.
19:02 jonrafkind: so that someone provides a frame and you attach your stuff to it
19:02 jonrafkind: (ohwows-thing my-frame)
19:02 ohwow_: Writing my own class can be a solution to but I really want to export an *object*
19:02 jonrafkind: mixins are quite nice
19:03 jonrafkind: well you know an object isn't really a gui component..
19:03 ohwow_: ah true
19:03 jonrafkind: is a mixin what you want?
19:03 (join) jeapostrophe
19:04 ohwow_: Well. May I explain my general problem if you have time?
19:04 jonrafkind: i suppose
19:06 ohwow_: Are you familiar with LOGO language/turtles language? Racket has it's own implementation, but I want to write my own just for fan. I want to implement it as a component so I could embedd it and add aditional buttons, for example, for controlling turtles. Here's what I have so far: https://gist.github.com/1134811
19:07 ohwow_: If I would have a mixin I would have to encapsulate (current-turtle) too I guess?
19:07 jonrafkind: you could make hte current turtle a member of the mixin
19:08 ohwow_: Okay, right now I can do (current-turtle (turtle ...)) which would change the parameter
19:08 ohwow_: how that would look if I use a class?
19:09 ohwow_: (I am sorry for being so clueless, I have only basic understanding of Racket's class system. I promise to read it deeply later, but you guys just wrote soooo much code and made lots of awesome features I have to learn all of them)
19:11 jonrafkind: you would just use the turtle member of the class, its just a plain variable
19:20 (quit) MayDaniel: Read error: Connection reset by peer
19:25 ohwow_: ah i see, thanks
19:25 ohwow_: i would look into mixins then
19:29 (quit) dnolen: Ping timeout: 252 seconds
19:40 (quit) DGASAU: Remote host closed the connection
19:55 (join) DGASAU
20:04 (quit) mceier: Quit: leaving
20:06 (join) dnolen
20:17 (quit) jeapostrophe: Quit: jeapostrophe
20:23 (nick) samth -> samth_away
20:33 (quit) jonrafkind: Remote host closed the connection
20:35 (join) SeanDeNigris
20:43 (quit) masm: Ping timeout: 268 seconds
20:54 (join) jeapostrophe
21:12 (nick) elliottcable -> fuuuuuuuuuck
21:15 (nick) fuuuuuuuuuck -> elliottcable
21:22 (quit) jeapostrophe: Quit: jeapostrophe
21:24 (quit) bluezenix1: Quit: Leaving.
21:26 (quit) hussaibi: Ping timeout: 240 seconds
21:27 (join) bluezenix
21:28 (quit) bluezenix: Client Quit
21:30 (join) hussaibi
21:31 (join) hussaibi_
21:32 (join) kevin01123
21:32 (quit) kevin01123: Remote host closed the connection
21:35 (quit) hussaibi: Ping timeout: 258 seconds
21:40 (quit) martinhex: Ping timeout: 240 seconds
21:43 (quit) hussaibi_: Ping timeout: 276 seconds
21:47 (join) hussaibi
21:56 (join) shofetim
22:11 (join) martinhex
23:44 (join) jonrafkind
23:46 em: This is kind of a strange question, but when you guys make a racket program, sometimes you probably don't want to just run it in drracket. How do you make a racket program that has some sort of standalone interface?