00:10 (quit) Guest8272: Remote host closed the connection
00:13 (quit) offby1: Quit: time to reboot emacs; sayonara
00:14 (join) offby1
00:39 (quit) shkk: 
01:06 (join) jonrafkind
01:16 (join) blAckEn3d
01:18 (quit) blAckEn3d: Client Quit
02:06 (join) shimman
02:07 shimman: does racket provide a socket programming interface/lib?
02:08 jonrafkind: it gives you tcp and udp sockets
02:08 jonrafkind: http://docs.racket-lang.org/reference/tcp.html?q=network
02:08 jonrafkind: i think theres a more low level socket library on planet, if you need it
02:10 shimman: thanks jonrafkind -- i was looking for a low level socket library, any lead?
02:11 shimman: is where a warper for libpcap?
02:11 jonrafkind: http://planet.racket-lang.org/display.ss?package=socket.plt&owner=vyzo
02:12 jonrafkind: http://planet.racket-lang.org/display.ss?package=SPeaCAP.plt&owner=evanfarrer
02:12 (notice) rudybot: http://tinyurl.com/29nhj6e
02:14 shimman: jonrafkind, thank you!!
02:15 jonrafkind: no problem
02:20 (quit) klutometis: Ping timeout: 248 seconds
02:27 (quit) jonrafkind: Ping timeout: 248 seconds
02:30 (join) klutometis
02:31 (nick) klutometis -> Guest38273
02:44 (join) masm
03:10 (join) hellmage
03:38 (quit) shimman: Ping timeout: 252 seconds
03:47 (quit) hellmage: Ping timeout: 240 seconds
04:49 (join) mgummelt
04:50 mgummelt: Why do Racket runtime errors not provide source code line numbers?
04:51 mgummelt: For example this error: car: expects argument of type <pair>; given #<closure>
04:51 mgummelt: gives me no information about where the error occurred
04:51 mgummelt: all I get is an (incomplete) stack trace
05:03 (join) hanDerPeder
05:05 (quit) jay-mccarthy: Ping timeout: 258 seconds
05:19 (join) jay-mccarthy
05:42 (quit) mgummelt: Ping timeout: 252 seconds
05:43 (join) TimBrown
05:44 (quit) TimBrown: Client Quit
06:53 (quit) rbarraud: Ping timeout: 240 seconds
07:21 (join) spidermario
08:33 (join) mgummelt
08:36 mgummelt: So it seems that DrRacket shows me the line number of a runtime error, but it's not printed when I run the file under racket -t file.rkt.  Does anyone know why this is?
09:08 (join) b-man
09:08 (nick) b-man -> Guest95592
09:15 (quit) mario-goulart: Remote host closed the connection
09:15 (join) mario-goulart
09:46 (nick) samth_away -> samth
09:48 samth: mgummelt, the incomplete stack trace you're seeing is generated by this: http://docs.racket-lang.org/reference/contmarks.html?q=stack%20trace&q=errortrace#%28def._%28%28quote._~23~25kernel%29._continuation-mark-set-~3econtext%29%29
09:48 (notice) rudybot: http://tinyurl.com/342wgu3
09:48 samth: the better version in drracket is generated by http://docs.racket-lang.org/errortrace/index.html?q=errortrace
09:49 samth: in general, drracket provides many tools that aren't provided by the plain command line
09:49 mgummelt: thanks
09:50 mgummelt: Any reason for the difference?
09:53 mgummelt: I would think if DrRacket's error tracing capabilities are better, the same backend should be used with the command line
09:53 mgummelt: Obviously excluding any graphical only capabilities
09:54 (quit) masm: Ping timeout: 265 seconds
09:57 samth: mgummelt, using errortrace slows down the program
09:58 samth: you can use it on the command line, see the docs i linked
09:58 samth: Lajla, i'm really not the person to ask about compilation of lazy languages
09:58 samth: there are lots of papers on this, i recommend you read them
10:02 (join) masm
10:27 (quit) mgummelt: Quit: Page closed
11:39 samth: jay-mccarthy, ping
11:40 jay-mccarthy: pong
11:40 samth: if you go to the drdr page, and click on 20862 (the first result)
11:40 samth: the resulting github page has only 2 commits
11:40 samth: but john pushed 3
11:40 samth: so i think there's an off-by-1 error somewhere
11:41 jay-mccarthy: hmm. i can't click that push
11:41 samth: oh, now i can't any more
11:42 jay-mccarthy: you can only click pending pushes, not the currently building one
11:42 jay-mccarthy: and i guess it just switched
11:44 jay-mccarthy: (format "http://github.com/plt/racket/compare/~a...~a" start-commit end-commit)
11:44 jay-mccarthy: that's what it does
11:44 jay-mccarthy: maybe i don't know what ... means to github
11:52 samth: i think ... means [,)
11:52 samth: no, (,]
11:52 samth: so exclusive, inclusive
12:06 (join) jonrafkind
12:43 (quit) hanDerPeder: Read error: Connection reset by peer
13:01 Lajla: samth, not? =(
13:01 Lajla: Maybe I should ask Simon Peyton-Jones
13:07 samth: he certainly knows more about lazy languages than me
13:10 Lajla: samth, than I.
13:10 Lajla: samth, oh well..
13:10 Lajla: Seek and ye shall find.
13:24 (join) hanDerPeder
13:40 jay-mccarthy: samth: github doesn't seem to give you the ability to get a different compare view
13:41 samth: jay-mccarthy, does eli give you the previous sha?
13:44 jay-mccarthy: no because the command-line tool has more options than github
13:44 samth: i'll figure out the git command to get the previous sha
13:44 jay-mccarthy: ~a..~a is what we really want
13:45 jay-mccarthy: there is a feature request in the github stuff
13:47 samth: if you just do 'git log --format=format:"%P" <start-sha> -1'
13:47 samth: that will give you the previous sha
13:51 (nick) tewk_ -> tewk
13:56 samth: tewk, thanks for the research on that benchmark
14:01 tewk: samth: if you have more questions, other things you want to try or look at, I'm happy to try to expose more info.
14:03 samth: tewk, why do you think it's spending so long marking?
14:04 samth: is it repeatedly traversing that vector?
14:19 jonrafkind: planet needs user comments
14:22 samth: how so?
14:22 samth: it has bug reporting
14:22 jonrafkind: so you can view what the community thinks about a package on the package description page
14:22 samth: ah
14:25 (join) b-man
14:26 (nick) b-man -> Guest2187
14:26 (quit) Guest95592: Ping timeout: 276 seconds
15:08 (nick) Guest2187 -> b-man_
15:27 (quit) b-man_: Remote host closed the connection
15:34 (quit) asumu: Read error: Connection reset by peer
15:36 Lajla: (define ((adder x) y) (+ x y)), I didn't know that was possible in Scheme
15:36 Lajla: I always used (define (adder x) (lambda (y) (+ x y)))
15:38 chandler: It's a Racket extension.
15:38 clklein: It's way more fun to use the unicode lambda letter than to spell the letter's name. I recommend that.
15:40 Lajla: chandler, it seems to work thi #lang r5rs
15:40 Lajla: Also, SICM uses it
15:40 Lajla: clklein, types harder
15:41 Lajla: Και συ, τεκνον clklein?
15:43 chandler: I don't know why it works in the R5RS language, but it's definitely not part of R5RS (or R6RS).
15:44 Lajla: chandler, I didn't find it in the specs either nop
15:44 Lajla: But SICM does use it
15:45 chandler: OK, maybe this extension originated in some other implementation.
15:47 Lajla: clklein, συ παιδεραστος.
15:50 clklein: See, wasn't that fun?
15:52 Lajla: clklein, using greek letters?
15:52 clklein: unicode in general
15:53 tewk: samth: Matthew seem to think you guys were optimizing a vector of complexes into two flvectors, is that not the case?
15:54 tewk: If I had to guess I would say marking vectors of complexes.
15:56 tewk: samth: If I get some time this weekend, I'll try to run callgrind and see if complex_obj_MARK is the hotspot.
15:56 Lajla: clklein, Ты педераст.
15:56 Lajla: Damn, that took me as long as possible
15:57 Lajla: This Russian keyboard layout is incomprehensionable
16:05 samth: tewk, we're not doing that
16:05 samth: since that would involve a lot of copying
16:06 samth: and since we'd have to keep the vector alive in the case we're concerned about, since it's an in-place fft
16:29 tewk: samth: running callgrind now, but I'm I bet its the complex vectors.
16:36 samth: tewk, i expect that's the case
16:36 samth: but once we know that, does it help us make it faster?
16:37 (quit) lisppaste: Quit: Want lisppaste in your channel? Email lisppaste-requests AT common-lisp.net.
16:47 (join) lisppaste
16:56 tewk: samth: we need something like typed-primitive-value-vectors or better analysis and optimization.
17:00 samth: it seems like the gc could avoid retraversing large objects that hadn't been mutated
17:15 (quit) hanDerPeder: Quit: hanDerPeder
17:17 (join) b-man
17:17 (nick) b-man -> Guest52282
17:18 (join) hanDerPeder
17:29 (quit) spidermario: Remote host closed the connection
18:31 (join) asumu
18:42 (quit) Lajla: Read error: Connection reset by peer
19:14 (join) Lajla
19:15 Lajla: samth, http://codepad.org/AxpDiPmU are you proud at me despite my weirdness? u.u
19:23 (quit) masm: Quit: Leaving.
19:23 jonrafkind: now write a cps converter from direct-style to cps
19:23 jonrafkind: and try to debug the results!!
19:26 Lajla: jonrafkind, one thing at a time, STarCraft II just game out.
19:26 Lajla: My compiler is currently on hold and secondary to totally pwning terran turtle players
19:26 Lajla: I just love turtles, they overdefend their front, you warp in some dark templars at the back, assasinate all their workers and go for their critical buildings thereafter, often they don't haven have detection
19:27 jonrafkind: once i beat nethack maybe ill get to starcraft
19:27 Lajla: Point about a turtler is that you already hint 'I'm not going for an early rush' so I can basically advance to dark templars and phase prisms easily without fear fo being overrun
19:29 jonrafkind: <Lajla> Point about FOOBARBAZBUZ is that WAMAKAKFOOLFUPOP so I can basically HUTOROWNFPEI without JOOJOFNUOT
19:29 jonrafkind gives the thumbs up
19:30 Lajla: jonrafkind, starcraft is master strategy and jargon, you know that
19:38 chandler: Really? And here I thought it was about being humiliated by Koreans.
19:40 (join) rbarraud
19:56 Lajla: chandler, they have superior strategy and jargon, you know that
19:56 Lajla: Especially the jargon
21:16 (quit) jonrafkind: Ping timeout: 260 seconds
22:02 emma: Who is bzlib? That guy is prolific.
22:02 (quit) Guest52282: Remote host closed the connection
22:21 (join) jonrafkind
22:49 (join) RyanRN
23:00 (quit) hanDerPeder: Quit: hanDerPeder
23:43 (quit) asumu: Quit: Leaving