00:12 (join) jeapostrophe_
00:13 (quit) jeapostrophe_: Client Quit
01:51 (quit) jonrafkind: Ping timeout: 255 seconds
03:00 (join) lisppaste
05:30 (join) masm
05:42 (join) martinhex
05:43 (join) b-man_
06:05 (quit) vu3rdd: Quit: ERC Version 5.3 (IRC client for Emacs)
07:30 (join) jeapostrophe_
07:31 (quit) b-man_: Ping timeout: 255 seconds
08:17 (join) hanDerPeder
08:20 (join) MayDaniel
08:29 (quit) MayDaniel: 
08:52 (quit) jeapostrophe_: Quit: jeapostrophe_
08:57 (join) jeapostrophe_
09:02 (quit) hanDerPeder: Quit: hanDerPeder
09:04 (join) hanDerPeder
09:11 (quit) jeapostrophe_: Quit: jeapostrophe_
09:22 (join) evhan
09:28 (join) samth
09:29 samth: jeapostrophe, drdr still down
09:29 (join) danbrown
09:31 (quit) hanDerPeder: Quit: hanDerPeder
09:32 (join) jeapostrophe_
09:39 (quit) evhan: Ping timeout: 245 seconds
09:43 (join) evhan
09:47 (quit) jeapostrophe_: Quit: jeapostrophe_
09:54 (quit) evhan: Ping timeout: 255 seconds
10:06 (join) vu3rdd
10:07 (join) hanDerPeder
10:14 (join) pygospa
10:17 (quit) TheRealPygo: Ping timeout: 276 seconds
10:18 (quit) vu3rdd: Quit: ERC Version 5.3 (IRC client for Emacs)
10:23 (join) MayDaniel
10:30 (join) vu3rdd
10:32 (quit) hanDerPeder: Quit: hanDerPeder
10:32 (join) b-man_
10:46 (join) jeapostrophe_
10:46 samth: jeapostrophe_, please kick drdr
10:46 samth: it's been down for a while
10:47 jeapostrophe_: kicked
10:48 (join) Yann1
10:50 Yann1: why do I get "expand: unbound identifier in module in: for/vector"  when I try to use "for/vector"? According to the docs, its part of the language... http://docs.racket-lang.org/reference/for.html#%28form._%28%28lib._racket/private/base..rkt%29._for/vector%29%29
10:50 rudybot: http://tinyurl.com/33cdql7
10:50 samth: Yann1, almost certainly because you're either using an old version of Racket, or you are in the wrong language
10:52 Yann1: I'm using plt-5.0.1 and #lang racket
10:52 samth: i believe that `for/vector' is new in 5.0.2
10:52 samth: but i'm not sure
10:52 Yann1: is there a changelog somewhere?
10:54 samth: annoyingly, we seem to have forgotten `for/vector' in the release notes
10:54 (quit) jeapostrophe_: Quit: jeapostrophe_
10:55 Yann1: its not in doc/release-notes/racket/HISTORY.txt
10:55 samth: Yann1, that only has more significant changes
10:55 samth: but it was added after 5.0.1
10:56 Yann1: I will try plt-5.0.2, perhaps such things should be noted in the docs
10:56 samth: in which docs?
10:56 Yann1: http://docs.racket-lang.org/reference/
10:57 samth: that's the reference documentation for the current version, which it says at the top
10:57 samth: are you suggesting that every binding should list its version?
10:58 Yann1: yes
10:59 samth: i don't think that would be very useful, since the vast majority of binding have been there a while
11:08 (join) sstrickl
11:21 (join) jeapostrophe_
11:33 jeapostrophe: my build is broken
11:35 (quit) danbrown: Remote host closed the connection
11:36 samth: jeapostrophe, my build seems to work
11:37 jeapostrophe: I get: path->complete-path: expects argument of type <path (for any platform) or string>; given #<undefined>
11:38 jeapostrophe: There we go
11:40 (join) anRch
11:41 (quit) jeapostrophe_: Quit: jeapostrophe_
11:42 samth: jeapostrophe, we really ought to be able to warn about that bug
11:42 (join) fjl
11:42 jeapostrophe: i agree
11:43 (join) hanDerPeder
11:50 (join) jeapostrophe_
12:09 Yann1: Question: is "for" iterative or recursive (internally)?
12:11 jeapostrophe: irrelevant to the user, but recursive, but tail recursive, so iterative again
12:11 Yann1: thanks :)
12:13 (quit) b-man_: Remote host closed the connection
12:14 (join) carleastlund
12:26 (quit) anRch: Quit: anRch
12:31 (join) mceier
12:35 (join) anRch
12:35 (join) jonrafkind
12:43 (quit) jeapostrophe_: Quit: jeapostrophe_
12:52 (quit) vu3rdd: Ping timeout: 252 seconds
13:00 (quit) anRch: Quit: anRch
13:03 (quit) masm: Quit: Leaving.
13:14 (join) Alex_Gaynor
13:16 stamourv: jeapostrophe: did you fix that path error when building?
13:16 stamourv: I'm getting it too
13:16 Alex_Gaynor: Hi, I'm working on a polynomial time implementation of longest common subsequence (for class clearly), and I've got myself a nice little recursive function, except without memoization it's definitely not polynomial time, is there a way to implement memoization without using functions ending in ! (requirement of the assignment), or do I need to rethink my algorithm?
13:16 stamourv: I tried nuking the build dir + make clean, but no luck
13:16 stamourv: building from a fresh clone works, though
13:19 samth: stamourv, he did fix it
13:20 samth: Alex_Gaynor, you can pass around the memo table
13:20 Alex_Gaynor: samth: right, but how can I set a value in it, without using hash-table-set! or something like that
13:21 samth: you can't, you'll have to construct a new table
13:21 samth: using `hash-set' (no !) for example
13:21 Alex_Gaynor: oh, nifty missed that, thanks
13:21 stamourv: samth: do you know how?
13:22 samth: stamourv, see https://github.com/plt/racket/commit/f8ca8f36770a3a726131ecc0552fbb267ce4d2d7
13:22 rudybot: http://tinyurl.com/32nod7d
13:23 stamourv: oops, missed that commit
13:23 stamourv: thanks
13:24 Alex_Gaynor: samth: are hash tables available by default, or do I need to require something?
13:24 Alex_Gaynor: r5rs FWIW
13:24 samth: in `racket/base' and `racket', they're available
13:24 samth: do you have to use r5rs?
13:25 Alex_Gaynor: Yup
13:30 samth: then your instructor is unlikely to want you to require various bits of Racket
13:30 samth: I'd use a list and add to it with `cons'
13:31 Alex_Gaynor: Yeah, I'll write it as-is with a hashtable, follow up with the professor, and replace with a list implementation if necessary
13:31 Alex_Gaynor: thanks for the pointer
13:43 eli: jonrafkind: I tried that, inside and outside of vnc, and it did't miss any enter.
13:43 jonrafkind: it doesnt happen all the time for me either
13:44 eli: Once in how many?
13:44 jonrafkind: once every few seconds I guess
13:44 jonrafkind: let me try something scientific..
13:45 eli hands jonrafkind an amulet of bug reporting
13:45 jonrafkind: ok it just happened, i typed a lot of stuff and now enter doesnt work. but all other kjeys do
13:45 eli: long lines?
13:45 jonrafkind: not really, 30 characters
13:46 eli: Doesn't happen to me.
13:47 jonrafkind: oh it could be my vi thing screwing up
13:48 eli coughs
13:48 eli: A minor detail...
13:48 jonrafkind: o wait
13:48 jonrafkind: no wait
13:48 jonrafkind: it just happened in a non-vi mode text editor
13:48 eli: "text editor"?
13:49 jonrafkind: text% editor<%>
13:49 (quit) mikeX: Ping timeout: 240 seconds
13:49 (part) Alex_Gaynor: "Leaving..."
13:50 eli: You should of course disable anything you have completely and then try it again.
13:51 (join) mikeX
14:08 (join) anRch
14:12 (join) b-man_
14:15 (join) danbrown
14:32 (quit) fjl: Remote host closed the connection
14:39 jonrafkind: whats the difference between `set-pen' and `set-brush' ?
14:39 jonrafkind: on a dc
14:57 jeapostrophe: i thought a brush was a pen with a color and a pen is just a shape?
14:58 jeapostrophe: i don't really know though
14:59 jonrafkind: ok ill try to figure it out and clarify the docs
15:00 jonrafkind: on another topic, jeapostrophe and eli, do you like the fact that eli-tester's `test' function calls `error' if any tests fail? id rather it just printed the errors and moved on
15:00 jeapostrophe: i like it, it cooperates with embedded test calls
15:00 jonrafkind: meaning if one test fails the rest dont occur?
15:01 jeapostrophe: naw (test a (test b) c)
15:01 jonrafkind: i thought there was one global error counting thing
15:01 jeapostrophe: ya that's what i mean
15:04 jonrafkind: it still works without the `error' call
15:04 jonrafkind: calling `error' doesn't make the global error increase.. right?
15:04 jeapostrophe: i like a good explosion i guess
15:06 eli: jonrafkind: Yeah, it could work without throwing an error --
15:07 eli: and incidentally, your `#:on-pass' thing made me do a particular extension that will make it easy to just print errors.
15:07 eli: But for the default I think that throwing an error makes more sense.
15:07 (join) Yann3
15:07 (quit) Yann1: Ping timeout: 252 seconds
15:07 jonrafkind: it usually confuses me, i have some code under the (test) thing that I wanted to run and I forget why it doesnt run
15:08 eli: Um...
15:08 eli: Clearly it should run with some expected result that is not a test failure.
15:08 jonrafkind: (test) (something)
15:08 jonrafkind: `something' wont run if `test' throws an error, thats all im saying
15:09 eli: (define (my-tests) (test stuff)) (something) (my-tests)
15:09 jonrafkind: (printf "hello world!\n") (+ 1 2)
15:09 eli: No, (define (my-tests) (test stuff)) (something) (my-tests)
15:32 (quit) MayDaniel: 
15:42 (quit) sstrickl: Quit: sstrickl
15:57 (quit) anRch: Quit: anRch
16:36 (join) Fare
17:10 (quit) b-man_: Remote host closed the connection
17:46 (join) Eren
17:46 Eren: hello
17:47 Eren: although I selected advance student and added image.ss and universe.ss teachpacks, I cannot use "crop"
17:47 Eren: reference to an identifier before its definition: crop
17:47 Eren: does anyone have idea?
17:47 Eren: and yes, I use DrRacket, doing my one of projects for uni :)
17:48 carleastlund: Eren, did you select htdp/image or 2htdp/image?  The crop function is only in the latter teachpack.
17:49 Eren: carleastlund: I am using preinstalled teachpacks which, I think, are by-default
17:49 carleastlund: Both come with DrRacket.
17:50 carleastlund: Ah, but I see both are not on the teachpacks list.
17:50 Eren: I can only see "image.ss" under "Language->Add Teachpack"
17:51 Eren: btw, Advanced Student is selected for language
17:51 carleastlund: I see what's going on.  The new "2htdp" stuff is phasing out the Teachpack menu.  To get the new image teachpack, just write "(require 2htdp/image)" at the top of your program, after removing the old image teachpack.
17:52 carleastlund: Sorry for the growing pains here -- the 2htdp libraries and the new teachpack process are a work in progress.
17:52 Eren: carleastlund: thank you for your help, it seems to work now, let me try the function
17:53 Eren: functional programming is really a must for a computer science student, most of the universities here do not relize its importance though :)
17:58 (quit) danbrown: Remote host closed the connection
18:48 (join) jeapostrophe_
18:58 (quit) jeapostrophe_: Quit: jeapostrophe_
19:00 (quit) mceier: Quit: leaving
19:09 (quit) hanDerPeder: Quit: hanDerPeder
19:40 (join) b-man_
20:11 (quit) b-man_: Ping timeout: 245 seconds
20:15 (join) jeapostrophe_
20:20 (quit) jeapostrophe_: Quit: jeapostrophe_
20:31 (join) b-man_
20:35 (quit) jonrafkind: Ping timeout: 250 seconds
21:11 (quit) b-man_: Ping timeout: 252 seconds
21:12 (join) b-man_
21:24 (join) samth_
21:24 samth_: jeapostrophe, drdr needs some libraries to be installed
21:27 (quit) b-man_: Ping timeout: 255 seconds
21:29 (join) jeapostrophe_
21:40 askhader: Are there no racket libraries for audio-playback operations on Linux?
21:41 (quit) jeapostrophe_: Quit: jeapostrophe_
21:51 (join) RyanRN
22:06 samth_: askhader, see rsound on planet
22:07 askhader: brilliant
22:09 askhader: But only reads .wav ohsad
22:09 eli: askhader: If you just want to listen to music, see collects/ffi/examples/xmmsctrl.rkt
22:10 askhader: Yes, I won't be manipulating the music.
22:10 eli: And if you want to play sounds in mp3s, look for the allegro library.  (It's something like sdl.)
22:10 samth_: askhader, write a gstreamber binding and put it on planet!
22:10 eli: askhader: In this case you can do the xmms thing, it's probably the easiest.
22:11 eli: gstreamer is for more than just listening; and besides, it's designed to be easy to interface through sub-processes.
22:11 askhader: it, being the xmms wrapper?
22:11 (quit) carleastlund: Quit: carleastlund
22:12 askhader: I would consider writing mpd bindings if mpd didn't suck bullocks.
22:12 samth_: yes, i understand that gstreamer does lots of things - that's why a planet library would be so useful!
22:14 eli: askhader: Yes -- if you just want to play some music, then controlling xmms or running some `play' subprocess would be very easy.  Shouldn't take more than a few minutes.
22:14 (join) jeapostrophe_
22:15 askhader: I was thinking of just writing an mplayer wrapper since mplayer plays pretty much everything already and the actual playback of media files is irrelevant to the computation I'm trying to achieve with racket. I could alternatively write .m3u playlist files.
22:19 eli: askhader: That's probably easy too, if it's practical.
22:19 eli: (to use mplayer, that is.)
22:20 eli: I think that VLC can also be controlled easily via a listening port.
22:20 askhader: Ample options.
22:31 eli: askhader: BTW, yet another option is Windows -- the MCI interface is *extremely* easy to use, and it abstracts nicely over all the media types the system knows about.
22:32 askhader: Ja winbl0wz >.>
22:34 eli: askhader: I'm as much up for flaming it as you'd expect, but the bottom line wrt to a sound interface is that it is the most convenient one.
22:34 eli: Linux suffers from 2^n interfaces, all incompatible in various ways,
22:35 eli: and OSX forces you to write 2 tons of boilerplate code just to do anything.
22:35 eli: (OSX was probably the most PITA-ish platform for the rsound interface that was mentioned earlier.)
22:36 askhader: I'm in total agreeance.
22:44 samth_: agreeance - word of the day
22:46 eli: Advantage of being a non-native speaker: easily invent words according to the rules, and easily parse such words regardless of being valid words or not.
22:47 eli: samth is canadian, so he's borderline.
22:53 Fare: eli: so Windows, the only good platform?
22:54 eli: ...as far as sound support goes.
22:54 Fare: dammit, I was hoping to be able to quote you out of context
22:55 eli: :)
22:55 eli: The Windows and Linux sides are not surprising here, it's OSX that was a surprise in just how horribly it sucks.
23:01 samth_: eli, you'll be punished for this
23:03 Fare: for revealing your canadianness?
23:04 offby1: HE'S CANADIAN?!?
23:04 offby1 flees
23:06 eli: It could be fun to see how far I could go with that joke...
23:11 (quit) jeapostrophe_: Quit: jeapostrophe_
23:13 offby1: Don't joke about that kinda thing
23:14 (join) danbrown
23:15 samth_ tries to think of nationalities to associate with eli
23:16 samth_: i think american is probably the most offensive
23:16 eli: Ha!
23:17 eli: I thought that you'd try similarly with a neighboring country, which will be ... difficult.
23:23 (join) jonrafkind
23:24 Fare: samth: roman or turkish citizen.
23:29 jonrafkind: samth, did you really just use LoL in an email
23:29 jonrafkind: please turn in your computer badge. you are suspended for a month
23:32 eli: jonrafkind: http://landoflisp.com/
23:32 jonrafkind: yea I know..
23:33 jonrafkind: so now I can use lol with impunity right? since it has a valid meaning?
23:33 jonrafkind: LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL LOL v
23:33 eli: Only as LoL.
23:34 askhader: So I've written a basic function to produce a simple nested vector structure that represents a directory tree. The following paste contains a simplified version of the function that computes the structure. I'd like to know what you folks think of the style of the code? (a 'track' is a flat data-type) http://paste.lisp.org/display/116430
23:34 jonrafkind: lOl
23:35 eli: askhader: Why vectors?
23:35 (join) vu3rdd
23:35 askhader: o(1) reference time, I think was the deciding factor.
23:36 eli: What's the sizes that you'll be dealing with?
23:37 Fare: next, Land-of-Lisp, the racket edition?
23:37 Fare: Region-of-Racket ?
23:37 Fare: Reign-of-Racket ?
23:37 askhader: Aproximately 58 gigs worth of music on my private library.
23:37 askhader: Length in the order of tens of thousands, I think.
23:37 askhader: Maybe more.
23:38 eli: Fare: That acronyms much worsly.
23:38 eli: askhader: So it's not more than a few thousands for the file entries, right?
23:38 askhader: In production, yes.
23:39 eli: ...and whatever you're doing with it, it's intended for operations that the user is doing wrt to music.
23:39 Fare: Racket's Outer Territory, Foreign Land ?
23:39 askhader: wrt?
23:39 eli: WRT: with respect to
23:39 (join) emma
23:39 eli: In that case, I think that the difference between vectors and lists is neglegible,
23:40 askhader: When /wouldn't/ the difference be neglegible and how do you figure?
23:40 eli: and if it's all for human-driven operations, then it's optimizing in a place that won't see any difference.
23:40 askhader: Well the operations with this data certainly isn't all human driven.
23:41 eli: Probably around one or two orders of magnitude more files,
23:41 askhader: I'm going to be using these structures to attempt to implement some sort of machine learning technique for playlist generation.
23:41 eli: and possibly subject to extensive *cheap* and automated operations.
23:41 eli: So, if you're dealing with tags a lot, the cost of reading them will dominate the result anyway.
23:42 askhader: So you think I ought to use mutable lists instead.
23:42 Fare: Racket's Outstandingly Typical Fun Land ?
23:42 askhader: Yes, I was thinking the same thing in terms of tags.
23:42 eli: Reading a tag from a single file takes more time than copying a 10k list...
23:43 eli: And there's not even any need for mutable lists -- just use plain lists.
23:44 askhader: Is the decision between lists and vectors usually one of performance?
23:45 eli: It's more of a question of what fits your needs more naturally, performance is only a factor in the above cases -- where there's a *lot* of data, and/or when it's used very heavily.
23:47 askhader: Well the 'listed' version of that program certainly looks cleaner. One call to map got rid of a complete (let ..) nest
23:47 Fare: what about the decision between (unary) church numerals and builtin integers?
23:48 Fare: Speed has always been important otherwise one wouldn't need the computer. —Seymour Cray
23:49 (quit) samth_: Ping timeout: 252 seconds
23:50 eli: Fare: That's an awful quote.
23:51 eli: askhader: See also `fold-files'.
23:57 (join) jeapostrophe_
23:58 askhader: eli: Oh, very nice.