00:24 (join) mithos28_
00:24 (quit) mithos28: Read error: Connection reset by peer
00:24 (nick) mithos28_ -> mithos28
00:25 (quit) Fare: Ping timeout: 246 seconds
00:38 (join) Fare
00:57 (quit) Fare: Ping timeout: 246 seconds
01:32 (join) jeapostrophe
01:57 (quit) m4burns: Ping timeout: 240 seconds
02:01 (quit) jeapostrophe: Ping timeout: 265 seconds
02:11 (quit) jonrafkind: Ping timeout: 245 seconds
02:16 (quit) yoklov: Quit: bye!
02:30 (join) hkBst
02:30 (quit) hkBst: Changing host
02:30 (join) hkBst
02:35 (quit) djcb: Read error: Connection reset by peer
02:45 (join) hkBst_
02:45 (quit) hkBst: Read error: Connection reset by peer
02:45 (quit) hkBst_: Changing host
02:45 (join) hkBst_
03:02 (join) dnolen
03:05 (quit) mithos28: Quit: mithos28
03:13 (quit) dnolen: Remote host closed the connection
03:27 (join) m4burns
03:32 (join) ahinki
03:32 (quit) ahinki: Client Quit
03:33 (join) ahinki
03:33 (quit) ahinki: Client Quit
03:34 (join) ahinki
03:39 (join) mmajchrzak
03:53 (part) Gertm: "WeeChat 0.3.7"
03:56 (quit) m4burns: Ping timeout: 260 seconds
04:14 (quit) mmajchrzak: Remote host closed the connection
04:18 (join) mceier
04:23 (join) noelw
04:27 noelw: tommc: my advice is to be very careful where you use continuations. If you use them everywhere and don't have a clear model for reclaiming them your memory consumption will get out of hand.
04:33 (join) antithesis
04:53 (join) masm
05:03 (join) tim-brown
05:14 (quit) Kaylin: Quit: Leaving.
05:19 (quit) cataska_: Ping timeout: 244 seconds
05:21 (join) gciolli
05:28 (join) cataska
05:33 (join) fftb
05:40 (quit) bitonic: Quit: WeeChat 0.3.5
05:40 (join) bitonic
05:53 (quit) antithesis: Quit: yes leaving
06:27 (join) antithesis
06:40 tim-brown: does anyone know how places compare with Erlangs processes, in terms of performance, footprint, etc.
06:43 noelw: Places are OS threads
06:43 noelw: Erlang processes are not IIRC
06:43 noelw: The Erlang VM (in SMP builds) maps processes to OS threads
06:43 noelw: Again, IIRC
06:44 noelw: Places will be more heavy-weight than Erlang processes
06:44 tim-brown: i'll look more at erlang (see if it has its own process scheduler)
06:44 noelw: Racket threads would be the natural comparison
06:44 bitonic: tim-brown: erlang has its own runtime yes. It manages its threads internally distributing them among few OS threads
06:44 tim-brown: racket threads aren't SMP scalable, though?
06:45 bitonic: (I don't know anything about racket threads)
06:45 bitonic: tim-brown: the main advantages is that you can spawn an insane number of them, and switching is really fast
06:46 tim-brown: i'm just wondering if racket could be (relatively easily) made as scalable as erlang likes to boast
06:46 bitonic: well I don't know about "relatively easily", but you definitely can
06:47 bitonic: but you'd have to add that facility to the VM (if it doesn't have it already)
06:48 tim-brown: racket VM can handle futures -- but i've shied away from them, given the caveats in the documentation about what I'm allowed to do with them
06:48 tim-brown: or am i being too timid?
06:48 bitonic: tim-brown: I don't know what those are. But I use Haskell a lot and Erlang at work, and their lighweight threads are really nice
06:49 bitonic: (GHC, the main Haskell compiler, has similar facilities)
06:50 bitonic: In GHC cases each program is linked to a 50k lines C runtime that handles all that stuff, so I suspect that doing that is not "relatively easy", given its size :P
06:51 tim-brown: how limiting are futures folks, in your varied experiences?
06:51 tim-brown: parallelism and "relatively easy" rarely share a sentence
06:51 bitonic: yeah :)
06:52 noelw: Futures are for parallelism, not concurrency
06:52 noelw: Threads are for concurrency, not parallelism
06:52 noelw: Places are for parallelism and concurrency
06:52 noelw: Does that help?
06:52 tim-brown: may well do!
06:59 (join) jao
07:00 (quit) jao: Changing host
07:00 (join) jao
07:01 (join) MayDaniel
07:08 (join) m4burns
07:12 (quit) jao: Remote host closed the connection
07:15 aidy: has anyone written a fast prime sieve in racket?
07:15 (quit) Skola: Quit: Lost terminal
07:16 bremner`: aidy: isn't that one of the examples in the programming languages shootout?
07:17 aidy: probably isn't a real sieve
07:19 aidy: oh, that actually looks like one
07:39 mrcarrot: what is a prime sieve?
07:39 bremner`: Erostrathenes?
07:39 (join) samth
07:40 bremner`: http://en.wikipedia.org/wiki/Sieve_of_Eratosthenes
07:40 bremner`: apparently I can't spell in Greek.
07:41 bremner`: there are fancier (and faster) versions
07:43 aidy: I implemented one which can produce infinitely many primes rather fast http://www.cs.hmc.edu/~oneill/papers/Sieve-JFP.pdf
07:45 bremner`: I doubt that
07:45 bremner`: ;)
07:45 aidy: read the paper :P
07:45 aidy: it's not that hard to do
07:46 bremner`: this word infinite... I do not think it means what you think it means...
07:46 bremner`: or maybe we disagree about quickly...
07:46 aidy: el oh el
07:47 aidy: I meant as a stream rather than flagging a fixed size vector
07:47 bremner`: yeah, I figured as much, just teasing ;).
07:49 aidy: :p
07:53 (quit) gciolli: Quit: Leaving.
08:04 (quit) MayDaniel: Read error: Connection reset by peer
08:18 (join) kanak
08:20 (quit) jrslepak: Quit: This computer has gone to sleep
08:24 (join) jeapostrophe
08:31 (join) Karli
08:57 (join) anonymous
08:57 (nick) anonymous -> Guest96534
09:00 Guest96534: Hi, I've got a problem with measuring runtime of sorting with time-apply. I sort list 10 times but sometimes it returns 0, sometimes normal time. Please help me, it's for my lab :)
09:02 chandler: Are you calling `time-apply' 10 times or calling `time-apply' on a procedure that sorts the list 10 times in a loop?
09:02 (join) dzhus
09:02 Guest96534: a call time-apply on a procedure that sorts the list 10 times in a loop
09:03 chandler: Do the results get more reliable if you sort the list more times in the loop?
09:03 chandler: Say 100 or 1000 times?
09:04 chandler: Generally when running small benchmarks you'll want to run it enough times that the overall run time is more than a second.
09:05 Guest96534: yep, but sometimes it returns normal results
09:06 Guest96534: 100 times is extremely slow, insertion sort is slow
09:07 chandler: Are you runing this in DrRacket?
09:09 chandler: If so, can you try at the command-line racket REPL?
09:09 Guest96534: yep, I'm running in DrRacket
09:12 (quit) jeapostrophe: Ping timeout: 272 seconds
09:14 Guest96534: how to include source file in command-line? :) I've never used
09:14 (join) noelw_
09:15 hkBst_: Guest96534: just do: racket file
09:17 (quit) noelw: Ping timeout: 246 seconds
09:17 (nick) noelw_ -> noelw
09:28 (join) Guest__
09:29 (quit) Guest96534: Ping timeout: 245 seconds
09:34 (quit) samth: Ping timeout: 246 seconds
09:34 (join) jao
09:35 (quit) jao: Changing host
09:35 (join) jao
09:44 Karli: Hi, is there anyone here who considers himself to be sort of an expert on the 'load' function in racket?
09:48 Karli: I'm trying to load a file in the middle of the program, to get the same effects of 'include'-ing it. When running the command in the repl prompt it works fine, but as soon as i put it inside a module it goes crazy and throws the following error:
09:48 Karli: maps\Awesomeroom.txt:1:0: compile: unbound identifier (and no #%app syntax transformer is bound) in: #%top-interaction
09:48 Karli: racket* goes crazy
09:49 (join) dous
09:50 noelw: load + module = sad times
09:50 noelw: Don't do it
09:54 Karli: but i wanna :/
09:54 (join) jrslepak
09:54 bremner`: I'm no expert, but maybe dynamic-require ?
09:55 Karli: hm, will try
09:56 Karli: but the code in the file is simply a code snippet, not a module
09:57 Karli: the most weird thing of all is that the exact same usage works fine in a different part of the program, where (load filename) returns the exact contents of the file as if it were written inline in the definitions
10:05 (join) dnolen
10:05 asumu_: Karli: Why do you need to load a code snippet? Why not make the other thing a module?
10:13 (join) jeapostrophe
10:14 (nick) samth_away -> samth
10:15 Karli: good question, just the way we made it i guess
10:27 asumu_: Karli: well, if you can make the other thing a module that's probably easier.
10:27 asumu_: `load` does odd things and it's pretty hopeless.
10:28 (join) djcb
10:41 Karli: hm, i suppose you're right
10:41 Karli: thanks
10:41 (quit) Karli: Quit: Page closed
10:45 (quit) ahinki: Quit: ChatZilla 0.9.88.1 [Firefox 12.0/20120328051619]
10:46 (quit) fftb: Remote host closed the connection
10:49 (join) shadgregory
10:50 (quit) antithesis: Quit: yes leaving
10:50 (join) antithesis
10:58 (quit) hkBst_: Quit: Konversation terminated!
11:21 (join) MayDaniel
11:32 (quit) MayDaniel: 
11:33 (quit) ssbr_: Ping timeout: 248 seconds
11:35 (quit) jrslepak: Quit: This computer has gone to sleep
11:40 (quit) bitonic: Quit: WeeChat 0.3.5
11:40 (join) bitonic
12:09 (join) jonrafkind
12:18 (join) jtpercon
12:21 (join) anRch
12:24 (quit) dnolen: Ping timeout: 276 seconds
12:28 (nick) asumu_ -> asumu
12:35 jonrafkind: is there a convenient way to abstract over the #lang for scribble documents? so I can easily change between scribble/jfp and scribble/sigplan for example
12:38 asumu: jonrafkind: maybe you could have two submodules (one for each) and then include the contents of another scribble document in both.
12:38 asumu: Or just have two modules in general.
12:40 (quit) abbe: Excess Flood
12:42 (quit) dous: Remote host closed the connection
12:43 jonrafkind: well it only seems to matter what the #lang is in my main.scrbl file, the rest can use whatever
12:44 (join) ashish
12:46 (quit) anRch: Quit: anRch
12:46 asumu: I just meant if you used submodules you could have DrRacket (or raco presumably) just choose which one to run.
12:46 asumu: Which effectively parameterizes it.
12:46 asumu: Don't know if that'll actually work though.
12:47 jonrafkind: right
12:47 jonrafkind: anyway whats a good way to typeset non-text/code stuff
12:47 jonrafkind: not necessarily racket code though
12:48 jonrafkind: i was using verbatim but its sort of ugly
12:59 (join) RacketCommitBot
12:59 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/SVZNoA
12:59 RacketCommitBot: [racket/master] Rewrite hash-keys so that only keys, not values, are accessed. - Stevie Strickland
12:59 (part) RacketCommitBot
13:01 samth: jonrafkind: if verbatim is ugly, use a better font
13:01 jonrafkind: how do you do that
13:07 (quit) tim-brown: Remote host closed the connection
13:08 (join) dnolen
13:11 (quit) dented42: Ping timeout: 252 seconds
13:14 samth: jonrafkind: if you just take out the courier option, it helps
13:14 samth: personally, i like \usepackage{inconsolata} in latex
13:14 jonrafkind: oh so you convert scrbl files to latex and then run latex commands?
13:15 jonrafkind: i like scribble --pdf main.scrbl
13:20 samth: jonrafkind: no, you include that declaration in the style.tex
13:20 jonrafkind: and use ++style style.tex ?
13:21 jonrafkind: ah yes that worked. neat
13:21 (quit) bitonic: Quit: WeeChat 0.3.5
13:32 (join) fftb
13:39 (join) epic
13:41 epic: I wonder, is racket faster than CLisp? I'm wondering if one could write a CAD program with this!
13:41 jonrafkind: if you do the graphics in opengl its probably ok
13:43 samth: epic: there are many common lisp implementations
13:45 epic: thank you all. I read Paul Graham's 100 year language but commercial implementations are even more than the MS Studio suite, and I watched a video where an author said that with C-Lisp, wait overnight for his AI world to populate, or two minutes with Franz Lisp...
13:46 epic: ...so PG uses CLisp but his bottleneck is the server, right? I need something fast (but will learn Racket either way!)
13:46 samth: racket is plenty fast for most purposes
13:47 bremner`: except the typed racket compiler
13:47 bremner`: ;)
13:47 epic: Oh, I thought the typed would be faster!
13:47 epic: The compiler already knows the types!
13:47 bremner`: the resulting code is often faster
13:48 bremner`: I am just teasing samth because he is (one of?) the main typed racket author(s)
13:48 epic: Oh! Thanks!
13:49 samth: bremner`: at least i'm the author of the slow parts :P
13:49 bremner`: epic: it's pretty much a universal tradeoff between compilation speed and runtime speed.
13:49 epic: And it's compiled on the fly when running the program?
13:50 epic: (the application)
13:51 bremner`: when developing, yes. Later, not necessarily.
13:51 samth: epic: racket does some compilation at runtime, and other compilation ahead of time
13:51 bremner`: the model is pretty similar to java (with a JIT), I think
13:54 samth: bremner`: that's less true than you would think
13:55 samth: in most jvms, javac is really simple, and the jit performs real optimization
13:55 bremner`: right
13:55 samth: roughly the reverse is true in racket
13:55 samth: 'raco make' performs real optimization, and the jit is just a code generator
13:55 bremner`: ah, ok.
13:55 samth: that's become a little less true over time, but the smarts are mostly in the AOT compiler
13:59 epic: What problems would one have developing a cross-platform graphics-intensive program like CAD with Racket?
14:00 (join) dented42
14:01 samth: epic: well, we build drracket, a cross-platform gui-intensive program in racket
14:02 samth: and most of the problems are the usual problems -- cross-platform guis are hard, programming is hard, that sort of thing
14:02 bremner`: 3D graphics is a different set of challenges, I guess.
14:04 epic: How about coding logic in Racket, and then interfacing with .Net, which can't do macros? (To tie into AutoCAD.)
14:05 bremner`: interfacing with C will be easier
14:06 (join) anRch
14:07 epic: My problem is everything is C++ and names are mangled differently with each release. Only the .Net interface remains reasonably constant. (Plus, I have abandoned Windows for Linux, except for AutoCAD at work.)
14:10 (quit) dnolen: Ping timeout: 245 seconds
14:10 epic: That is too deep and too far along!
14:11 (quit) dented42: Ping timeout: 245 seconds
14:11 epic: Thank you so much for the help, bremner` and samth!
14:13 epic: ..and jonrafkind for the opengl advice.
14:14 epic: ...back to Racket for me, then!
14:16 (quit) epic: Quit: Leaving
14:22 (join) dented42
14:24 (join) Kaylin
14:25 (join) esiwmas
14:44 (join) jhemann
14:46 (join) Shvillr_
14:46 (quit) Shviller: Disconnected by services
14:46 (nick) Shvillr_ -> Shviller
14:48 (quit) Kaylin: Quit: Leaving.
14:49 (join) bitonic
14:51 (quit) rudybot: Ping timeout: 240 seconds
14:54 (quit) anRch: Quit: anRch
14:54 (join) rudybot
15:00 (join) noam_
15:03 (quit) noam: Ping timeout: 245 seconds
15:22 (quit) jhemann: Ping timeout: 250 seconds
15:23 (part) snorble_
15:27 (join) snorble_
15:28 (join) jhemann
15:39 (quit) djcb: Remote host closed the connection
15:48 (join) djcb
15:52 (join) dyoo
15:58 (quit) dented42: Quit: Computer has gone to sleep.
16:04 (quit) sajith: Ping timeout: 260 seconds
16:05 (quit) bluephoenix47: Ping timeout: 276 seconds
16:07 (join) Name_
16:08 (quit) Guest__: Ping timeout: 245 seconds
16:08 Name_: Hello. I've got a problem with measuring sorting time using time-apply procedure. I sort 25 times one list and sometimes I get 0 seconds as a result. What can be the problem?
16:09 danking: How can I override the default printer/writer for syntax objects?
16:10 danking: I want it to print as an sexp instead of #<sourceloc contents>, and I would like to drop sourceloc entirely if possible. I need to easily read it.
16:13 asumu: Name_: how are you using it?
16:16 (join) dented42
16:23 Name_: pass a procedure that sorts list 25 times
16:24 jonrafkind: why is 0 seconds a problem
16:24 Name_: cause I need to make graphics
16:24 Name_: plots
16:25 jonrafkind: what other timings do you get
16:25 Name_: 0.64, 1.24 and something like that
16:26 Name_: 2594.0 ^)
16:26 jonrafkind: for the same set of lists each time?
16:26 jonrafkind: or the same list I guess
16:26 Name_: well nope
16:26 jonrafkind: so then likely the lists that take 0 seconds are probably mostly sorted already
16:27 Name_: well I use insertion sort
16:28 Name_: so it doesn't mind sorted it or not
16:28 Name_: matter
16:28 Name_: bad english :(
16:29 (join) Fare
16:30 (part) Nisstyre: "Leaving"
16:32 dyoo: danking: the current-print parameter controls how objects are "printed"
16:32 dyoo: would you be able to override that?  For example:
16:33 dyoo: (define original-print (current-print)) (current-print (lambda (x)                  (if (number? x)                      (original-print "NUMBER")                      (original-print x))))
16:34 dyoo: danking: the write and display functions won't be affected, but regular toplevel print would, as would the "print" function.
16:35 (quit) antithesis: Remote host closed the connection
16:35 dyoo: Name_: how large is the list that you're sorting?
16:42 (quit) kanak: Ping timeout: 245 seconds
16:44 (quit) fftb: Remote host closed the connection
16:44 (quit) dyoo: Ping timeout: 245 seconds
16:49 (quit) dented42: Quit: Computer has gone to sleep.
16:54 asumu: Name_: can you post your code snippet too? (gist.github.com or something)
16:57 (join) dented42
16:57 (quit) mceier: Quit: leaving
16:58 Name_: It's about 16 thousands elements
16:59 Name_: http://pastebin.com/KprjbGPb
16:59 Name_: my coву
16:59 Name_: code
17:05 asumu: Name_: maybe some are empty somehow
17:05 asumu: due to the regex?
17:07 Name_: Hm, in files numbers, so I don't think that some files are empty
17:07 Name_: or what do you mean?
17:18 Name_: even if I sort 1 list I sometimes get normal time and sometimes I get zero
17:20 (join) Name__
17:23 (quit) Name_: Ping timeout: 245 seconds
17:24 Name__: Any idea?
17:26 asumu: Name__: So you're sorting a constant list?
17:26 asumu: And it's still not working?
17:27 asumu: Name__: BTW, using set! in your loop there is a really un-Rackety way of doing it.
17:27 asumu: Better would be to use for/list and get a list of times, then sum.
17:28 Name__: I know that code isn't good
17:28 Name__: ok, I'll try, but I don't think it will help
17:30 (quit) jhemann: Ping timeout: 246 seconds
17:32 (join) yoklov
17:35 (join) jhemann
17:43 (join) jhemann_
17:46 (quit) jhemann: Ping timeout: 260 seconds
18:03 (quit) jeapostrophe: Ping timeout: 276 seconds
18:05 (quit) dzhus: Remote host closed the connection
18:08 (join) RacketCommitBot
18:08 RacketCommitBot: [racket] plt pushed 2 new commits to master: http://git.io/-CSz6g
18:08 RacketCommitBot: [racket/master] avoid allocation for 'paren' structs - Robby Findler
18:08 RacketCommitBot: [racket/master] rackety - Robby Findler
18:08 (part) RacketCommitBot
18:11 (quit) jhemann_: Read error: Connection reset by peer
18:13 (join) jhemann_
18:14 (part) jtpercon
18:21 (quit) jhemann_: Ping timeout: 250 seconds
18:22 (join) jrslepak
18:27 (quit) jonrafkind: Ping timeout: 246 seconds
18:30 (quit) noam_: Read error: Connection reset by peer
18:31 (join) noam
18:42 (nick) samth -> samth_away
18:43 (quit) bitonic: Quit: WeeChat 0.3.5
18:43 (quit) noam: Read error: Connection reset by peer
18:45 (join) noam
18:46 (quit) jrslepak: Quit: This computer has gone to sleep
18:47 (join) mmajchrzak
18:49 (join) jrslepak
19:03 (quit) Fare: Ping timeout: 260 seconds
19:06 (quit) Name__: Ping timeout: 245 seconds
19:20 (quit) mmajchrzak: Ping timeout: 272 seconds
19:25 (join) jhemann
19:39 (quit) bmp: Quit: Bye!
19:43 (join) Kaylin
19:47 (quit) jhemann: Ping timeout: 250 seconds
19:55 (join) Fare
19:58 (join) jhemann
19:58 (quit) masm: Quit: Leaving.
20:03 (join) jonrafkind
20:42 (join) bmp
20:44 (quit) jonrafkind: Ping timeout: 260 seconds
20:46 (join) jeapostrophe
20:48 (join) RacketCommitBot
20:48 RacketCommitBot: [racket] plt pushed 3 new commits to master: http://git.io/TRKEKw
20:48 RacketCommitBot: [racket/master] streamline `vectorof' contact implementation - Matthew Flatt
20:48 RacketCommitBot: [racket/master] ffi/unsafe/com: `com-object->eq?' implies `equal?' - Matthew Flatt
20:48 RacketCommitBot: [racket/master] racket/gui: fix `reparent' with a `pane%' target - Matthew Flatt
20:48 (part) RacketCommitBot
20:49 (join) dyoo
20:50 dyoo: Someone needs to let Name_ know that the code is not measure what he or she thinks it's measuring
20:50 dyoo: The code is time-applying the system sort function, rather than the insertion-sort written in the code
20:51 dyoo: any way to let the person know this besides hoping they see it in the irc logs?
21:02 (quit) dyoo: Ping timeout: 245 seconds
21:02 (quit) noelw: Ping timeout: 252 seconds
21:07 (join) dalaing
21:12 (quit) jhemann: Ping timeout: 246 seconds
21:17 (quit) bmp: Remote host closed the connection
21:17 (join) bmp
21:19 (quit) dalaing: Ping timeout: 246 seconds
21:21 (join) dalaing
21:25 (quit) dalaing: Ping timeout: 264 seconds
21:27 (join) samth
21:30 (join) jhemann
21:33 (join) asurai
21:38 (quit) bmp: Ping timeout: 260 seconds
21:43 (quit) samth: Ping timeout: 246 seconds
21:50 (quit) jeapostrophe: Ping timeout: 265 seconds
22:05 esiwmas: I'm thinking about really getting into racket and following htdp, should I try emacs/geiser or just drracket?
22:07 bremner`: esiwmas: do you already know emacs well?
22:08 bremner`: I think you will have to use drracket for some things anyway. For example graphics and the debugger.
22:20 (join) dalaing
22:24 (join) dented42_
22:24 (quit) dented42: Read error: Connection reset by peer
22:24 (nick) dented42_ -> dented42
22:37 (quit) em: Ping timeout: 246 seconds
22:40 (join) emma
22:44 (quit) dented42: Ping timeout: 244 seconds
23:01 (join) veer
23:09 (join) jonrafkind
23:15 (join) jeapostrophe
23:16 (join) dented42
23:26 (join) zerbitx
23:33 (quit) jeapostrophe: Ping timeout: 244 seconds
23:35 (quit) esiwmas: Ping timeout: 244 seconds
23:39 (quit) jao: Ping timeout: 246 seconds
23:57 (quit) zerbitx: Ping timeout: 245 seconds