00:10 (quit) francisl: Quit: francisl
00:13 (quit) snorble_: Ping timeout: 260 seconds
00:34 (quit) lewis1711: Ping timeout: 255 seconds
00:34 (join) lewis1711
00:39 (join) sw2wolf
00:43 (quit) kofno: Remote host closed the connection
00:47 (join) jonrafkind
00:47 (quit) jonrafkind: Changing host
00:47 (join) jonrafkind
00:53 (quit) mceier: Quit: leaving
01:13 (quit) jackhammer2022: Quit: Textual IRC Client: http://www.textualapp.com/
01:14 (join) kofno
01:22 (quit) kofno: Ping timeout: 245 seconds
01:22 sw2wolf: how to understand this code snippet: ((or/c string? input-port?) . -> . list?) ?
01:23 sw2wolf: especially . -> .
01:23 (quit) jeapostrophe: Ping timeout: 265 seconds
01:23 lewis1711: ugh
01:23 lewis1711: oh
01:23 lewis1711: it's a function that takes either a string or an input-port, and outputs a list
01:23 lewis1711: I really hate infix ->
01:23 lewis1711: (-> (or/c string? input-port?) list?)
01:24 sw2wolf: thx
01:24 lewis1711: typed racket forces you to use infix -> which is just wrong.
01:25 sw2wolf: infix -> seems strange
01:27 sw2wolf: is the space between . and -> essential in ". -> ." ?
01:28 lewis1711: think so. it works with any racket procedure IIRC
01:28 lewis1711: rudybot: (1 .+. 2)
01:28 rudybot: lewis1711: error: reference to an identifier before its definition: .+. in module: 'program
01:28 lewis1711: rudybot: (1 . + . 2)
01:28 rudybot: lewis1711: ; Value: 3
01:30 sw2wolf: thx
01:31 sw2wolf: rudybot: (cons 'a 'b)
01:31 rudybot: sw2wolf: your sandbox is ready
01:31 rudybot: sw2wolf: ; Value: (a . b)
01:31 sw2wolf: wonderful rudybot !
01:32 lewis1711: I always found it odd that this didn't work
01:32 lewis1711: rudybot: (rest (cons 'a 'b))
01:32 rudybot: lewis1711: error: rest: expected argument of type <non-empty list>; given: '(a . b)
01:32 lewis1711: rudybot: (cdr (cons 'a 'b))
01:32 rudybot: lewis1711: ; Value: b
01:32 (quit) dca: Ping timeout: 244 seconds
01:33 sw2wolf: maybe rudybot is still not a complete REPL ?
01:35 lewis1711: doesn't work on 5.3.1 either
01:36 sw2wolf: hope racketter makes it work
01:38 asumu: `rest` is a list procedure. Intentionally doesn't work on pairs.
01:39 sw2wolf: rudybot: (rest '(1 2 3))
01:39 rudybot: sw2wolf: ; Value: (2 3)
01:39 sw2wolf: asumu: what does rudybot mean ?
01:40 asumu: I'm not sure what you mean.
01:40 asumu: rudybot's just showing you the value that computes.
01:40 lewis1711: asumu: I guessed it might be intentional. IIRC racket tries to de-emphasies the cons cell stuff
01:41 sw2wolf: i mean what does rudybot stand for ?
01:41 asumu: The name? You should ask offby1.
01:41 sw2wolf: offby1: are you here ?
01:42 asumu: lewis1711: I don't really think it de-emphasizes cons cells. It's just that cons cells have perfectly good accessors already (car, cdr).
01:42 asumu: Although it's not always consistent, so that's not really great.
01:42 asumu: For example, `list-ref` technically works on non-lists...
01:42 asumu: rudybot: (list-ref (cons 1 #f) 1)
01:42 rudybot: asumu: your sandbox is ready
01:42 rudybot: asumu: error: list-ref: index 1 too large for list (not a proper list): '(1 . #f)
01:42 asumu: rudybot: (list-ref (cons 1 #f) 0)
01:42 rudybot: asumu: ; Value: 1
01:43 lewis1711: but first is exactly the same as cons isn't it?
01:43 lewis1711: sorry, as car
01:43 sw2wolf: rudybot: (first (cons 'a 'b))
01:43 rudybot: sw2wolf: error: first: expected argument of type <non-empty list>; given: '(a . b)
01:44 asumu: It's not. `first` means "get the first element of this list". `car` means "get the first thing in the pair".
01:44 lewis1711: I always thought first/rest car/cdr was just choosing saner names, that's why I was surprised
01:44 lewis1711: (car '(1 2 3))
01:44 asumu: Just like you wouldn't expect `head` and `fst` to do the same thing in Haskell.
01:44 lewis1711: rudybot: (car '(1 2 3))
01:44 rudybot: lewis1711: ; Value: 1
01:44 lewis1711: yeah but haskell has tuples
01:45 lewis1711: I never use car or cdr except for pairs. I'm 2 nu skool
01:45 asumu: Pairs are tuples.
01:46 asumu: The only real difference is that in Haskell, lists are not represented with tuples. At least not in a way you can observe.
01:47 sw2wolf: rudybot: (doc 'open-input-string)
01:47 rudybot: sw2wolf: error: reference to an identifier before its definition: doc in module: 'program
01:47 asumu: rudybot: doc open-input-string
01:47 rudybot: asumu: http://docs.racket-lang.org/reference/stringport.html#(def._((quote._~23~25kernel)._open-input-string))
01:47 asumu: sw2wolf: ^ that's how
01:48 sw2wolf: thx
01:49 sw2wolf: are there any guide to use rudybot ?
01:51 asumu: sw2wolf: private message it and say "help"
01:51 lewis1711: huh, I didn't actually know you could use "pairs" as 3 tuples
01:51 lewis1711: then you have to use the godawful cadr stuff
01:52 sw2wolf: asumu: thx
01:57 lewis1711: woah, this lib uses stuff from "scheme/foreign"
01:58 lewis1711: positively ancient
02:11 (join) mceier
02:21 (join) Kaylin
02:37 (quit) mithos28: Quit: mithos28
02:38 (join) mithos28
02:48 (quit) ambrosebs: Ping timeout: 245 seconds
02:55 (join) dca
02:58 asumu: It appears I just spent several hours trying to figure out why my code wasn't working. Did lots of printf debugging.
02:59 asumu: Then I invoked the nuclear option and just did 'raco setup -c' and rebuilt.
02:59 asumu: Now it works. :\
02:59 asumu is sometimes very unhappy about how 'raco' chooses to compile things...
03:00 (join) cdidd
03:01 (join) hkBst
03:06 (quit) karswell: Remote host closed the connection
03:10 (quit) ormaaj: Ping timeout: 240 seconds
03:14 (quit) mithos28: Quit: mithos28
03:16 (join) karswell
03:27 jonrafkind: asumu, were you writing a #lang ?
03:27 (join) ormaaj
03:28 jonrafkind: ive found that if you have a lang and you do 'raco setup foo' the edit main.rkt and provide more stuff, the new stuff isn't seen by a user of that lang
03:28 jonrafkind: for honu i never allow it to make the 'compiled' directory while im doing development
03:29 (join) mithos28
03:30 jonrafkind: http://this-plt-life.tumblr.com/
03:33 (join) ambrosebs
03:42 (join) tilde`
03:45 mithos28: If I have an object, is there a way to get its pointer address?
03:46 (quit) jonrafkind: Ping timeout: 260 seconds
04:02 (topic) -: Racket: http://racket-lang.org -- RacketCon videos: https://www.youtube.com/playlist?list=PLQC7pQ_gNllaY_NFQTwHFZTpjeR_PwKmb -- logs at http://racket-lang.org/irc-logs/
04:02 (names) -: gabot rapacity didi ozzloy tilde` ambrosebs mithos28 ormaaj hkBst cdidd dca Kaylin mceier sw2wolf lewis1711 mye cataska Shviller spiderweb Shvillr Raynes Nisstyre acarrico elliottcable noelw chrxn jaimef BeLucid em Twey cky m4burns ivan\ Demosthenex bremner sid0 Haffe Shambles_ SeanTAllen noam stchang errstr dsantiago otterdam sizz sirdancealot7 parcha sethalves samth_away jschuster ChibaPet roderic tobi offby1 JStoker SHODAN Draggor antono fasta ohama xian
04:02 (names) -: friscosam majoh asumu epsil pono lusory rudybot jrslepak_neu hyko greghendershott bartbes chandler stamourv mario-goulart danking dsp @ChanServ ivan` gf3 eMBee Cryovat mattmight jamessan bartbes_ blubberdiblub felipe
04:02 (notice) calvino.freenode.net: [freenode-info] channel trolls and no channel staff around to help? please check with freenode support: http://freenode.net/faq.shtml#gettinghelp
04:03 (join) jrslepak
04:04 (quit) tilde`: Ping timeout: 250 seconds
04:09 (quit) ormaaj: Ping timeout: 276 seconds
04:09 (join) karswell
04:16 (part) sw2wolf: "ERC Version 5.3 (IRC client for Emacs)"
04:23 (join) ormaaj
04:27 (join) MightyFoo
04:27 (nick) MightyFoo -> tim-brown
04:28 tim-brown: rudybot: init typed/racket
04:28 rudybot: tim-brown: your typed/racket sandbox is ready
04:29 (join) tilde`
04:29 tim-brown: What more hinting can I provide the type checker with:
04:29 tim-brown: rudybot:(for/first: : Natural ([n : Natural (in-naturals 1)]) n)
04:29 rudybot: tim-brown: error: Type Checker: insufficient type information to typecheck. please add more type annotations in: #%top-interaction
04:30 (quit) ambrosebs: Ping timeout: 244 seconds
04:30 tim-brown: OK, so this works:
04:30 tim-brown: scratch that
04:31 tim-brown: this works:
04:32 tim-brown: rudybot: (cast (for/first: : Natural ([n (in-naturals 1)]) n) Natural)
04:32 rudybot: tim-brown: error: #<undefined>:1:53: type-check: type name used out of context in: Natural
04:32 tim-brown: rudybot: init typed/racket
04:32 rudybot: tim-brown: your typed/racket sandbox is ready
04:32 tim-brown: rudybot: (cast (for/first: : Natural ([n (in-naturals 1)]) n) Natural)
04:32 rudybot: tim-brown: error: #<undefined>:1:53: type-check: type name used out of context in: Natural
04:33 tim-brown: rudybot: that works in my drracket -- what's your problem?
04:33 rudybot: tim-brown: yeah I use DrRacket on my mac...
04:38 (join) bitonic
04:46 didi: How do I prevent `response/xexpr' from escaping legit html code? I have a string with html what I would like to be rendered as such.
04:50 (quit) tilde`: Read error: Connection reset by peer
04:51 (join) tilde`
04:59 didi: OK, the trick is to use `make-cdata'.
05:05 (quit) mceier: Ping timeout: 260 seconds
05:06 didi: Aha! `string->xexpr' works too.
05:06 didi: The mailing list never fails me.
05:09 didi: Hum, although `string->xexpr' doesn't seem to like sloppy html that I don't control. Oh well, `make-cdata' will do.
05:26 (join) soegaard
05:30 (quit) hkBst: Ping timeout: 250 seconds
05:31 (quit) soegaard: Quit: soegaard
05:32 (join) hkBst
05:48 (join) tilde`_
05:51 (quit) tilde`: Ping timeout: 252 seconds
05:58 (quit) karswell: Remote host closed the connection
05:59 (quit) hkBst: Ping timeout: 255 seconds
06:00 (join) hkBst
06:00 (quit) hkBst: Changing host
06:00 (join) hkBst
06:08 (join) karswell
06:11 (join) masm
06:13 (nick) tilde`_ -> tilde`
06:14 (join) tilde`_
06:18 (join) sw2wolf
06:21 sw2wolf: rudybot, ap regexp
06:21 rudybot: sw2wolf: you want to collect both the evens from the sublist (car l) and from the sublist (cdr l), so you have to call (cons al dl), you have to multiply (* ap dp), and sum (+ as ds). and pass all those values as argument to col.
06:21 sw2wolf: rudybot, doc regexp
06:21 rudybot: sw2wolf: http://docs.racket-lang.org/reference/regexp.html#(def._((quote._~23~25kernel)._regexp))
06:22 sw2wolf: rudybot, (regexp-match #rx"a|b" "cat")
06:22 rudybot: sw2wolf: ; Value: ("a")
06:24 didi: My algorithm: (unless (regex-work? regex) (type-gibberish! regex))
06:24 (join) Aune
06:25 (nick) tilde`_ -> tilde`
06:27 (quit) Kaylin: Read error: Connection reset by peer
06:36 (quit) didi: Remote host closed the connection
06:37 (join) didi
06:45 (quit) sw2wolf: Quit: Leaving
06:49 (join) sw2wolf
06:50 (quit) bitonic: Ping timeout: 250 seconds
06:54 lewis1711: dungeon before: '#hash((#(struct:posn 48 48) . #<procedure:object>))
06:54 lewis1711: dungeon after: (hash (posn 49 48) #<procedure:object>)
06:54 lewis1711: why are these two printing differently? :/
06:58 (join) nathanpc
07:02 lewis1711: they're both done with printf
07:07 (quit) sw2wolf: Remote host closed the connection
07:23 (join) bitonic
07:23 (join) tilde`
07:24 (part) lewis1711
07:30 (join) kofno
07:30 (quit) karswell: Remote host closed the connection
07:40 (join) karswell
07:45 (quit) bitonic: Ping timeout: 255 seconds
07:57 (join) dnolen
08:09 (quit) tilde`: Read error: Connection reset by peer
08:09 (join) tilde`
08:13 (quit) hkBst: Ping timeout: 245 seconds
08:13 (quit) tilde`: Read error: Connection reset by peer
08:14 (join) hkBst
08:14 (quit) hkBst: Changing host
08:14 (join) hkBst
08:14 (join) MayDaniel
08:15 (join) tilde`
08:15 (quit) tilde`: Read error: No route to host
08:20 (join) tilde`
08:26 (quit) tilde`: Read error: Connection reset by peer
08:26 (join) tilde`
08:27 (join) bitonic
08:41 (join) ambrosebs
08:44 (quit) karswell: Remote host closed the connection
08:47 (join) jeapostrophe
08:47 (quit) jeapostrophe: Changing host
08:47 (join) jeapostrophe
08:49 (join) zyoung
08:54 (join) karswell
08:55 (join) soegaard
08:57 (quit) masm: Ping timeout: 264 seconds
08:57 (join) masm1
09:11 (quit) zyoung: 
09:12 (join) zyoung
09:12 (quit) zyoung: Remote host closed the connection
09:12 (join) zyoung
09:15 (quit) karswell: Remote host closed the connection
09:17 (quit) bitonic: Ping timeout: 255 seconds
09:18 (join) mizu_no_oto
09:24 (join) bitonic
09:25 (join) karswell
09:27 (join) francisl
09:34 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
09:44 (quit) ormaaj: Read error: Connection reset by peer
09:46 (join) ormaaj
09:52 (join) RacketCommitBot
09:52 RacketCommitBot: [racket] plt pushed 2 new commits to master: http://git.io/rJBOrQ
09:52 RacketCommitBot: racket/master 4cc287f Chenxiao: Improve ftp client. Add upload, progress monitor and something else.
09:52 RacketCommitBot: racket/master e0de33a Matthew Flatt: net/ftp: make progress proc keyword, change progress protocol...
09:52 (part) RacketCommitBot
09:53 (quit) bitonic: Ping timeout: 244 seconds
09:54 (join) mizu_no_oto
10:00 (join) bitonic
10:04 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
10:07 (join) anRch
10:09 (join) myx
10:10 (join) mceier
10:11 (join) francisl_
10:13 (quit) francisl_: Client Quit
10:14 (quit) soegaard: Quit: soegaard
10:18 (quit) karswell: Remote host closed the connection
10:23 (quit) jrslepak: Quit: What happened to Systems A through E?
10:28 (join) karswell
10:31 (quit) myx: Ping timeout: 255 seconds
10:35 (quit) dnolen: Ping timeout: 246 seconds
10:36 (join) hkBst_
10:36 (quit) hkBst_: Changing host
10:36 (join) hkBst_
10:37 (quit) hkBst: Ping timeout: 248 seconds
10:39 (nick) samth_away -> samth
10:43 (join) mizu_no_oto
10:45 (quit) mizu_no_oto: Client Quit
10:47 (quit) jeapostrophe: Ping timeout: 264 seconds
11:00 (join) mizu_no_oto
11:02 (quit) mizu_no_oto: Client Quit
11:05 (join) masm
11:07 (quit) masm1: Ping timeout: 250 seconds
11:09 (quit) anRch: Quit: anRch
11:15 (join) mizu_no_oto
11:19 (quit) ambrosebs: Ping timeout: 265 seconds
11:29 (join) jrslepak
11:43 (quit) jaimef: Excess Flood
11:44 (join) myx
11:45 (quit) bitonic: Ping timeout: 265 seconds
11:46 (join) jaimef
11:47 (quit) hkBst_: Quit: Konversation terminated!
11:48 (quit) karswell: Remote host closed the connection
11:49 (quit) masm: Quit: Leaving.
11:51 mithos28: My vim is now much more stable, and not crashing. Seems like the tagged part of scheme_malloc_tagged is important
11:53 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
11:55 (quit) kofno: Read error: Connection reset by peer
11:55 (join) kofno_
11:58 (join) karswell
11:59 (join) mizu_no_oto
12:08 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
12:08 (join) bitonic
12:15 (join) tsion
12:16 (quit) tsion: Changing host
12:16 (join) tsion
12:16 stamourv: tim-brown: Rudybot uses an old version of Racket, that doesn't have `cast'.
12:17 stamourv: tim-brown: Also, `ann' should work there, instead of `cast'. `ann' doesn't have a runtime cost, while `cast' does.
12:17 stamourv: lewis1711: Did you change the printing options in DrRacket? Or change the printing parameters (I forget which one is relevant here.)?
12:21 (quit) bitonic: Ping timeout: 264 seconds
12:26 (join) jeapostrophe
12:26 (quit) jeapostrophe: Changing host
12:26 (join) jeapostrophe
12:26 (join) bitonic
12:29 (join) mizu_no_oto
12:31 (quit) tilde`: Quit: kthxbai
12:38 (quit) bitonic: Ping timeout: 255 seconds
12:42 (quit) francisl: Quit: francisl
12:43 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
12:44 (join) bitonic
12:52 (join) francisl
12:54 (quit) karswell: Remote host closed the connection
13:03 (quit) bitonic: Ping timeout: 248 seconds
13:04 (join) karswell
13:05 (quit) Nisstyre: Ping timeout: 265 seconds
13:09 (quit) Aune: Remote host closed the connection
13:11 (join) mizu_no_oto
13:14 (join) nejucomo
13:19 (join) Nisstyre
13:21 (join) jonrafkind
13:23 (quit) francisl: Quit: francisl
13:28 (join) Kaylin
13:33 (join) bitonic
13:39 (quit) bitonic: Ping timeout: 244 seconds
13:40 (join) dnolen
13:41 (quit) Demosthenex: Ping timeout: 250 seconds
13:43 (join) francisl
13:43 (join) Demosthenex
13:49 (join) dyoo
13:52 (quit) karswell: Remote host closed the connection
13:59 (join) RacketCommitBot
13:59 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/5cucJw
13:59 RacketCommitBot: racket/master 4ff4c67 Danny Yoo: Modernize cfg-parser from mzscheme to Racket libraries.
13:59 (part) RacketCommitBot
14:02 (join) karswell
14:08 (quit) mye: Quit: mye
14:09 (join) mithos28_
14:11 (quit) karswell: Remote host closed the connection
14:11 (quit) mithos28: Ping timeout: 265 seconds
14:11 (nick) mithos28_ -> mithos28
14:20 (join) soegaard
14:21 (join) karswell
14:22 (join) tilde`
14:43 jonrafkind: jeapostrophe, fist and sound? pretty wierd..
14:43 jeapostrophe: heh
14:43 (quit) didi: Remote host closed the connection
14:43 jeapostrophe: 1st and 2nd would be cuter
14:43 jonrafkind: hm true
14:44 (join) Blkt
14:44 jonrafkind: i wonder if the tokenizer can be changed to allow numbers to start identifiers
14:45 jonrafkind: oh wait.. you can already do that
14:45 jonrafkind: so use 1st and 2nd!
14:46 jeapostrophe: hit reply :)
14:47 jonrafkind: bikeshedding debates are for irc :p
14:51 (join) RacketCommitBot
14:51 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/XLJrcQ
14:51 RacketCommitBot: racket/master db280d0 Danny Yoo: Add scribble/doclang2 with keyword support for customization, and documentation.
14:51 (part) RacketCommitBot
15:00 (join) RacketCommitBot
15:00 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/beKX5g
15:00 RacketCommitBot: racket/master f374cb1 Danny Yoo: Add separate section to avoid conflict at scribble time.
15:00 (part) RacketCommitBot
15:04 (quit) tilde`: Ping timeout: 276 seconds
15:09 (join) tilde`
15:11 (quit) mizu_no_oto: Quit: ["Textual IRC Client: www.textualapp.com"]
15:17 (quit) dyoo: Quit: dyoo
15:23 samth: jeapostrophe: any sense for why this occasionally times out: http://drdr.racket-lang.org/25892/collects/tests/racket/mzq.rktl ?
15:25 jeapostrophe: it looks like it dies in Optimization
15:26 (join) mithos28_
15:26 jeapostrophe: maybe there's a non-deterministic jit error?
15:34 (quit) karswell: *.net *.split
15:34 (quit) tsion: *.net *.split
15:34 (quit) majoh: *.net *.split
15:34 (nick) mithos28_ -> mithos28
15:35 (join) karswell_
15:37 greghendershott: Lately I find myself avoiding all of the above (car, cdr, first, rest) in favor of using match.
15:38 greghendershott: It usually seems so much clearer.
15:38 greghendershott: Am I being lazy and is that horribly slow or something?
15:41 (join) tsion
15:41 (quit) tsion: Changing host
15:41 (join) tsion
15:42 (join) dsevilla
15:43 (join) majoh
15:44 (join) friscosa1
15:45 samth: match is usually fast
15:46 (join) ivan`_
15:46 (quit) tilde`: Read error: No route to host
15:46 (join) tilde`
15:47 (quit) Twey: *.net *.split
15:47 (quit) ChibaPet: *.net *.split
15:47 (quit) friscosam: *.net *.split
15:47 (quit) ivan`: *.net *.split
15:48 (join) ChibaPet
15:48 (quit) ChibaPet: Changing host
15:48 (join) ChibaPet
15:49 (nick) ivan`_ -> ivan`
15:50 (quit) karswell_: Remote host closed the connection
15:51 (join) Twey
15:57 greghendershott: if loving match is wrong, I don't want to be right
15:58 (quit) tsion: Quit: Leaving
15:59 (quit) dnolen: Remote host closed the connection
16:01 (join) karswell_
16:05 asumu also always uses match when possible
16:17 (join) lewis1711
16:22 samth: greghendershott: more seriously, I put a lot of work into having match generate fast code
16:22 (join) mizu_no_oto
16:22 samth: most likely, it's faster than the code you would hand-write
16:23 greghendershott: maybe match should go in racket/base
16:28 (quit) karswell_: Remote host closed the connection
16:32 samth: greghendershott: `match` is a lot of code, that's why it's not there
16:37 greghendershott: ah. well maybe someday the "loader granularity" (can't think of the right term) could be finer than whole modules
16:38 (join) karswell_
16:40 samth: greghendershott: i don't understand what you're suggesting
16:42 greghendershott: not loading unused/unreachable code
16:42 greghendershott: "I see you've required racket/string, but you're only using these 3 items out of 100. I will only load the 3."
16:46 greghendershott: I'm sure that's waaaaay easier said than done. although naively it seems like the GC kinda already does something very similar
16:54 samth: greghendershott: it all depends what you mean by "load"
17:00 greghendershott: hmm. In the re-provide idiom (core.rkt requires some modules, but only re-provides a portion of those modules), what happens with the un-re-provided stuff?
17:01 greghendershott: Is it loaded on startup and sitting there in memory, just not accessible?
17:01 greghendershott: Or not loaded (or kept very long — same difference??) at all?
17:06 (quit) karswell_: Remote host closed the connection
17:06 (quit) Demosthenex: Ping timeout: 265 seconds
17:08 (join) Demosthenex
17:10 (join) didi
17:10 (quit) kofno_: Remote host closed the connection
17:10 (join) dyoo
17:13 dyoo: 17:01: it should also be loaded, because you can do things like http://docs.racket-lang.org/reference/Namespaces.html#(def._((quote._~23~25kernel)._module-~3enamespace)) to jump into the context of the module, provided or not
17:13 (join) masm
17:13 dyoo: greghendershott: sorry, forgot to add your name to the above reply
17:16 (join) karswell_
17:17 greghendershott: dyoo: Ah I see. So require must support some dynamic reflective features like this. Everything in the module is potentially available, could be used.
17:17 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
17:17 (quit) myx: Ping timeout: 265 seconds
17:17 dyoo: greghendershott; yeah.  The "demodularizer" project that one of Jay's students is doing is meant to let you strip out unused parts of modules.
17:18 (join) mizu_no_oto
17:18 dyoo: I think Blake Johnson's the person who's spearheading it?
17:18 jeapostrophe: dyoo: "doing" should really be "done", yes it's Blake
17:18 jeapostrophe: but it will be even more useful in 1-2 weeks
17:18 dyoo: Ah, thanks!
17:19 jeapostrophe: because he's about to commit something that can RE-optimize the code once the module boundaries are done
17:19 greghendershott: dyoo: I was vaguely aware of that. That eliminates unneeded requires?
17:19 jeapostrophe: to discover more x-module inlining, etc
17:19 greghendershott: Or you said "unused parts of modules"
17:19 greghendershott: Ah
17:19 (join) francisl_
17:19 greghendershott: With a tool like that, does racket/base serve a purpose anymore?
17:20 (quit) tilde`: Quit: kthxbai
17:20 jeapostrophe: greghendershott: yes
17:20 greghendershott: *avoided typing "is racket/base required anymore", to avoid a pun*
17:20 greghendershott: oh why?
17:20 dyoo: greghendershott: if I understand it right, it "demodularizes", so it's a whole-program transformation.
17:21 jeapostrophe: because racket uses racket/class, which has top-level code that initializes the class system. the demodularizer can't remove code that is at the top level, because you could depend on its effects
17:21 jeapostrophe: in racket/class's case, you are unlikely to depend on the effects, but demod doesn't know
17:21 (join) jao
17:21 (quit) jao: Changing host
17:21 (join) jao
17:21 jeapostrophe: i consider this an error in racket/class, it should initialize lazily
17:21 jeapostrophe: but there's not much of a reason to fix that now
17:21 jeapostrophe: who knows how many more instances of this there are
17:22 jeapostrophe: but i'd like to fix them as demod/reopt become more used
17:22 dyoo: unrelated question: is there a reason why net/uri-codec's uri-encode doesn't encode paren characters with %28 and %29?  I know parens are legal and all, but they play havok with a lot of tools online that don't have the most robust URL parsers.
17:22 dyoo: and %28 and %29 should play more nicely with the outside world.
17:23 jeapostrophe: dyoo: i presume it because uri-codec implements the spec, and others don't
17:23 (quit) francisl: Ping timeout: 260 seconds
17:23 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
17:23 (quit) francisl_: Ping timeout: 260 seconds
17:24 greghendershott: Racket net/uri-codec does RFC 2396 but I needed 3986 for my AWS lib because of some chars in SQL-like queries
17:24 greghendershott: 3986 encodes a few more, like ! ' ( ) *
17:25 jeapostrophe: the two libraries could probably share code
17:25 greghendershott: https://github.com/greghendershott/aws/blob/master/util.rkt#124
17:26 greghendershott: I forgot about this, at the time I meant to suggest that net-uri/codec be enhanced (or add a flag) for 3986
17:26 greghendershott: I could do this up and make a pull request, if you want
17:26 (join) anRch
17:27 dyoo: jeapostrophe: sure, but there are a lot of important apps and sites that are known to be a bit lax.  (The IRC client I'm using Colloque, for example), Stack OverFlow, Wikipedia, etc.  They aren't going to change for us, so might as well compensate for spec failure.  Basically, http://www.codinghorror.com/blog/2008/10/the-problem-with-urls.html
17:29 dyoo: greghendershott: yeah, would be helpful.  I keep hitting this pain point whenever including Racket doc links in web sites.
17:32 greghendershott: dyoo: Prefer uri-encode with a new optional flag for 3986, or leave that as-is and add a new function named something like "uri-encode/rfc-3986" — preference either way?
17:33 greghendershott: *tries to get bikeshedding done in the requirements phase*
17:33 greghendershott: :)
17:33 dyoo: greghendershott: looking at the rest of the library, making it a parameter seems the most consistent, similar to how (current-alist-separator-mode) works
17:34 (quit) didi: Remote host closed the connection
17:34 greghendershott: dyoo: Ah, good point.
17:34 dyoo: wait, reading the documentation in http://docs.racket-lang.org/net/uri-codec.html suggests that they may be doing something already with parens.  Reading...
17:35 dyoo: Hm.
17:36 dyoo: form-urlencoded-encode may be what I was looking for.
17:36 dyoo: rudybot: (require net/uri-codec)
17:36 rudybot: dyoo: your sandbox is ready
17:36 rudybot: dyoo: Done.
17:36 dyoo: rudybot: (form-urlencoded-encode "(hello world)")
17:36 rudybot: dyoo: ; Value: "%28hello+world%29"
17:36 dyoo: Ha.
17:38 dyoo: Ok, so maybe I should be asking for Scribble to use form-urlencoded-encode rather than uri-encode
17:42 (join) neilv
17:42 (join) bitonic
17:46 (join) RacketCommitBot
17:46 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/JdQK_g
17:46 RacketCommitBot: racket/master 6eef00a Matthew Flatt: raco setup: fix problem with doc index database...
17:46 (part) RacketCommitBot
17:51 (join) dnolen
17:57 dyoo: greghendershott: never mind: it looks like Scribble does its own thing with regards to escaping: it doesn't even use net/uri-codec.  I'll propose this patch and see what people think.  https://github.com/dyoo/racket/commit/f36c64ed1667cc44fb870d6be37b0b7fcc0418fc
17:57 rudybot: http://tinyurl.com/aj9kelw
18:03 (join) RacketCommitBot
18:03 RacketCommitBot: [racket] plt pushed 7 new commits to master: http://git.io/Qlm04A
18:03 RacketCommitBot: racket/master 5286e50 Asumu Takikawa: Minor doc fix
18:03 RacketCommitBot: racket/master 5c80743 Asumu Takikawa: Refactor to use syntax/id-table
18:03 RacketCommitBot: racket/master 2810888 Asumu Takikawa: Fix tests where type variable scoping is wrong
18:03 (part) RacketCommitBot
18:04 (quit) zyoung: Ping timeout: 250 seconds
18:07 (quit) karswell_: Read error: Connection reset by peer
18:09 (quit) jeapostrophe: Ping timeout: 276 seconds
18:15 (quit) spiderweb: Quit: ERC Version 5.3 (IRC client for Emacs)
18:17 (join) karswell_
18:22 (join) zyoung
18:26 (join) RacketCommitBot
18:26 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/-SnVXQ
18:26 RacketCommitBot: racket/master 0c11601 Asumu Takikawa: Fix documented name of `unsafe-cons-list`
18:26 (part) RacketCommitBot
18:31 (quit) anRch: Quit: anRch
18:32 (quit) jrslepak: Quit: What happened to Systems A through E?
18:34 (join) kofno
18:34 (quit) soegaard: Quit: soegaard
18:41 lewis1711: crud, I have to resort to C for speed. time brush up on the ffi stuff
18:42 dyoo: lewis1711: what process is expensive?
18:42 (quit) kofno: Ping timeout: 260 seconds
18:42 lewis1711: dyoo: calling a C function multiple times in quick succession :)
18:43 dyoo: Huh. :)
18:43 dyoo: lewis1711: possible to batch the calls together?
18:43 lewis1711: how do you mean? can this be done from racket side?
18:44 lewis1711: the situation is a C lib, that allocates a large 2d structure, then provides a function to mutate it at each index. which means I have to loop over the structure in racket and call "mutate at x, y" loads of times
18:44 lewis1711: and each call of that has some FFI overhead
18:45 lewis1711 wonders if there is a way to declare C arrays from racket
18:47 (quit) MayDaniel: Read error: Connection reset by peer
18:47 dyoo: lewis1711: I believe so.
18:47 dyoo: checking...
18:47 lewis1711: yeah there is
18:48 dyoo: http://docs.racket-lang.org/foreign/C_Array_Types.html
18:48 lewis1711: so I should just be able to write a little helper function in C, that takes the C map, and a C flat array of values it needs, and loops through it all. maybe
18:49 dyoo: I could have sworn I saw something like this on the mailing list a few weeks ago, and that mflatt had patches in devel to make the ffi faster in simple cases like this
18:53 dyoo: potentially commit 42a0342109ae0f55f55ddfd0e505926f92d8ee1e was what I was thinking of.  http://git.racket-lang.org/plt/commit/42a0342109ae0f55f55ddfd0e505926f92d8ee1e
18:53 rudybot: http://tinyurl.com/as4ean4
18:53 dyoo: lewis1711: so it might be worthwhile to see if Racket git head is faster for your application, if you're not already using it.
18:56 lewis1711: dyoo: hmm, do you have any idea when this patched racket version will be released? I mean it's not *prohibitively* slow right now, but I can see the writing on the wall, so to speak
18:56 dyoo: dunno
18:56 dyoo: You could always use the pre-release: I believe it's kept up to date with git head
18:56 dyoo: lewis1711: http://pre.racket-lang.org/
18:57 dyoo: it's a little bit of the cutting edge, but it might be worthwhile just to see if it helps with the slowness
19:01 (join) jeapostrophe
19:01 (quit) jeapostrophe: Changing host
19:01 (join) jeapostrophe
19:03 lewis1711: hmm actually 3 * speedup won't be enough really
19:03 lewis1711: I need order of magnitude - and I know form experience it's super fast in straight C. so I'll have to crack out the ol' K&R
19:09 (quit) zyoung: Remote host closed the connection
19:11 lewis1711: what does it mean for a hash table to "hold its keys weakly?"
19:14 lewis1711: oh it's just about jc
19:16 lewis1711: *gc
19:20 dyoo: lewis1711: yes, it's a gc thing.  "Weak references"
19:27 dyoo: lewis1711: the array initialization thing: might want to check on the mailing list too.  I think that Neil Toronto (of the math library) ran into something like this as well.  I could be wrong, but couldn't hurt to bring up the problem.
19:33 lewis1711: dyoo: I will look into it, I think it may be easier just to do it in C though
19:33 lewis1711: also, is there anyway to return a value from a hash table, and not a reference?
19:34 dyoo: lewis1711: don't understand the question yet
19:34 dyoo: can you explain in more detail?
19:35 lewis1711: dyoo: well, if you get a value from a struct, and mutate it, the struct recognises it
19:35 lewis1711: if you do the same with a hashtable, it does not
19:35 dyoo: ?!
19:35 dyoo: Let me check.
19:35 dyoo: rudybot: (define ht (make-hash))
19:35 rudybot: dyoo: Done.
19:35 dyoo: rudybot: (define b (box "foo"))
19:35 rudybot: dyoo: Done.
19:36 dyoo: (hash-set! ht 'a-box b)
19:36 lewis1711: forgot to tell rudybot :)
19:36 dyoo: rudybot: (hash-set! ht 'a-box b)
19:36 rudybot: dyoo: Done.
19:36 dyoo: rudybot: (box-set! (hash-ref ht 'a-box) "bar")
19:36 rudybot: dyoo: error: reference to an identifier before its definition: box-set! in module: 'program
19:36 dyoo: dah...
19:36 dyoo: rudybot: (set-box! (hash-ref ht 'a-box) "bar")
19:36 rudybot: dyoo: Done.
19:37 dyoo: rudybot: (unbox b)
19:37 rudybot: dyoo: ; Value: "bar"
19:37 lewis1711: huh. so that's what boxes are for..
19:37 dyoo: lewis1711: so I still don't quite understand what you're seeing yet.  :)
19:37 dyoo: do you have an example?
19:37 lewis1711: I am having an issue when my object system, based around a hashtable. you send it a message, get the value back, mutate it, and the object doesn't care because it just returned a copy
19:38 lewis1711: dyoo: well I didn't know about the box thing :/
19:38 dyoo: Weird.  It shouldn't be returning a copy unless you've done something to explicitly copy
19:38 dyoo: If you can gist.github.com it somewhere, I could take a look
19:39 dyoo: Are you inadvertanly just set!-ing a variable?
19:39 lewis1711: dyoo: https://gist.github.com/4281456
19:39 dyoo: (Rather than mutating structure)
19:40 lewis1711: line 51 is how you get a value out from a message
19:40 lewis1711: I think your box thing will solve it
19:40 dyoo: ok, looking at it now
19:41 (join) mizu_no_oto
19:42 lewis1711: I think I have to make all the values in the hashtable boxed
19:43 dyoo: Alternatively, provide a separate setter! for your slots.
19:43 dyoo: So I see that your object supports several messages, and the last one looks like the lookup one.
19:44 dyoo: It acts as your "getter", so I think similarly you want a "setter", if I'm reading the code right.
19:44 nejucomo: Are boxes the primary way to have mutability in data structures?
19:44 dyoo: normally, structures are marked mutable or not to allow data-structure mutability
19:44 dyoo: example:
19:45 dyoo: rudybot: (struct person (name age) #:mutable)
19:45 rudybot: dyoo: Done.
19:45 lewis1711: dyoo: there is 91, that works. but my issue is not that I can't change what a slot refers to. more that I can't mutate the value that it refers to. if you get what I mean
19:45 dyoo: rudybot: (define p (person "danny" 1))
19:45 rudybot: dyoo: Done.
19:45 nejucomo looks up (box)
19:45 dyoo: rudybot: (set-person-age! p 33)
19:45 rudybot: dyoo: Done.
19:45 dyoo: rudybot: p
19:45 rudybot: dyoo: ; Value: #(struct:person "danny" 33)
19:46 dyoo: So mutability's off in data structures by default: you turn it on either globally, or you can mark individual fields as mutable or not.
19:46 nejucomo: How can I change the browser executable used by (help) in xrepl?
19:47 dyoo: nejucomo: checking… it's the one mentioned in http://docs.racket-lang.org/net/sendurl.html if I'm not mistaken
19:48 dyoo: nejucomo: you might need to first set the browser in DrRacket preferences, after which xrepl will follow the same preference toward which browser it uses.
19:50 (join) sw2wolf
19:51 dyoo: lewis1711:  you may be confusing set! ing a variable binding vs. mutating a value, so I'm trying to make sure that's not what's happening
19:51 (join) zyoung
19:53 lewis1711: dyoo: yes! I want to do the latter
19:54 lewis1711: I am not sure what objects are supposed to do when you ask them about their slots.
19:54 lewis1711: whether they should be boxed..
19:56 nejucomo: I don't understand the need for boxes.  What's your goal lewis1711?
19:57 nejucomo rereads backlog...
19:59 (quit) Kaylin: Read error: Connection reset by peer
19:59 nejucomo: lewis1711: When you say you cannot mutate the *value* a slot refers to, what do you intend?
20:00 lewis1711: if a return a value from the object, it's just a copy
20:00 nejucomo: Do you want to do something like: (let ([x (my-object 'x)]) (mutate! x))  -and have my-object be mutated?
20:00 lewis1711: mutating it does nothing
20:00 nejucomo: It's a reference, I would assume.
20:00 nejucomo: Can you give an example of mutating it?
20:00 lewis1711: something like (set! (myobject 'want-to-change) 3)
20:00 lewis1711: (my-object 'want-to-change) => 3
20:01 nejucomo: Ah, I see.
20:01 lewis1711: right now (my-object 'want-to-change) just returns a fresh value
20:01 nejucomo: You'd have to override set! in that case.
20:01 nejucomo: Can you actually do (set! (myobject 'x) 3)  ?
20:01 nejucomo: Doesn't that trigger an exception?  (I'm new to racket, so I could be wrong.)
20:02 nejucomo: I assumed set! was a special form whose second parameter is a symbol, and it rebinds the given binding.
20:02 nejucomo: -but now that I think about it, I remember reading about set! macros.
20:02 nejucomo: If you want that to work with set!, I think you need a set! macro.
20:02 dyoo: lewis1711: how about adding to your objects a different message:
20:02 dyoo: something like:
20:02 dyoo: (my-object 'set! 'want-to-change 3)
20:03 lewis1711: yes but how would I do that?
20:03 lewis1711: when it's just getting uboxed values from a hashtable
20:03 lewis1711: you can't mutate a value in a hashtable that I know of
20:04 dyoo: give me a sec
20:04 dyoo: rudybot: (define ht (make-hash))
20:04 rudybot: dyoo: Done.
20:04 dyoo: rudybot: (hash-set! ht 'name "mal")
20:04 rudybot: dyoo: Done.
20:04 dyoo: rudybot: ht
20:04 rudybot: dyoo: ; Value: #hash((name . "mal"))
20:04 dyoo: rudybot: (hash-set! ht 'name "Castle")
20:04 rudybot: dyoo: Done.
20:04 dyoo: rudybot: ht
20:04 rudybot: dyoo: ; Value: #hash((name . "Castle"))
20:05 (quit) zyoung: Remote host closed the connection
20:05 lewis1711: rudybot: that's just replacing one string with another
20:05 rudybot: lewis1711: If you have no double quotes "" in your bbdb file, then I have at least some certainty that I won't break stuff when replacing all "" with nil ;)
20:06 lewis1711: ugh
20:06 lewis1711: I mean dyoo :D
20:06 dyoo: rudybot: (define s (string-copy "Rei"))
20:06 rudybot: dyoo: Done.
20:06 nejucomo: dyoo's suggestion would be one of the cases on the match in line 47, something like: [(list 'set name value) (hash-set! slots name value)]
20:06 dyoo: (hash-set! ht 'pilot s)
20:06 dyoo: rudybot: (hash-set! ht 'pilot s)
20:06 rudybot: dyoo: Done.
20:06 dyoo: rudybot: ht
20:06 rudybot: dyoo: ; Value: #hash((pilot . "Rei") (name . "Castle"))
20:07 lewis1711: what I mean is you can't do this
20:07 dyoo: rudybot: (begin (string-set! s 0 "e") (string-set! s 1 "v") (string-set! s 1 "a"))
20:07 rudybot: dyoo: error: string-set!: expects type <character> as 3rd argument, given: "e"; other arguments were: "Rei" 0
20:07 dyoo: doh
20:07 lewis1711: rudybot: (define ht (make-hash '((v . (vector 0))))
20:07 rudybot: lewis1711: if you're talking about logical cpus (HT) not physical cores, you can do 32 cores for 400$ (4x AMD Opteron Quad-core @ 100$ per). Not sure they make quad-CPU boards for those, though.
20:07 dyoo: rudybot: (begin (string-set! s 0 #\e) (string-set! s 1 #\v) (string-set! s 2 #\a))
20:07 rudybot: dyoo: Done.
20:07 dyoo: rudybot: ht
20:07 rudybot: dyoo: ; Value: #hash((pilot . "eva") (name . "Castle"))
20:08 dyoo: mutable values are mutable
20:08 lewis1711: yeah but you have defined s first
20:08 asumu: Lies. "Eva" was never an Eva pilot! ;)
20:08 lewis1711: I don't know my way around that
20:09 dyoo: asumu: I know, but asuka doesn't fit in three characters
20:09 dyoo: lewis1711: let me try another example:
20:09 lewis1711: rudybot: (define ht (make-hash `((a ,(vector 0))))
20:09 rudybot: lewis1711: And yeah, the argument that you can warp in a ht with enough energy for a storm immediately after the upgrade while other races have to wait longer is potent, but this is applies to like any protoss unit due to the warpin mechanic.
20:09 lewis1711: why does it keep doing that
20:09 lewis1711: rudybot: (define ht (make-hash `((a . ,(vector 0))))
20:09 rudybot: lewis1711: i use ht in elisp
20:09 dyoo: no idea, and why the heck is it talking about starcraft?!
20:10 lewis1711: rudybot: (define ht (make-hash `((a . ,(vector 0)))))
20:10 rudybot: lewis1711: your sandbox is ready
20:10 rudybot: lewis1711: Done.
20:10 nejucomo: rudybot: (define ht (make-hash (list (cons 'a (vector 0)))))
20:10 rudybot: nejucomo: your sandbox is ready
20:10 rudybot: nejucomo: Done.
20:10 lewis1711: ok
20:10 lewis1711: rudybot: (vector-set! (hash-ref 'a) 0 42)
20:10 rudybot: lewis1711: error: hash-ref: expects 2 to 3 arguments, given 1: 'a
20:10 nejucomo: rudybot: (define ht (make-hash `((a . ,(vector 0)))))
20:10 rudybot: nejucomo: Done.
20:10 lewis1711: rudybot: (vector-set! (hash-ref ht 'a) 0 42)
20:10 rudybot: lewis1711: Done.
20:10 lewis1711: (hash-ref ht 'a)
20:10 lewis1711: rudybot: (hash-ref ht 'a)
20:10 rudybot: lewis1711: ; Value: #(42)
20:11 lewis1711: what?
20:11 lewis1711: oh I give up
20:11 nejucomo: Weird, the same expression works for me.
20:11 lewis1711: if that works then..
20:11 nejucomo: lewis1711: Your lookup returned a vector.
20:11 lewis1711: maybe it's because I was trying to mutate C values
20:11 nejucomo: Did you expect that?
20:11 lewis1711: yes
20:11 lewis1711: before I did something like that
20:12 lewis1711: the moral of the story is 1. thanks 2. I have a headache 3. I need to mutate C data in C
20:12 dyoo: Oh..  Ok, context is _very_ important here. The ffi might be doing funny things: right, it's meant to bridge Racket values with C values, so it may be copying things for you.
20:12 nejucomo: But you realize why (vector-set! (hash-ref ht 'key) 0 32) is quite different from (set! (hash-ref 'key) 7) right?
20:12 (join) francisl
20:13 lewis1711: nejucomo: yeah. i was surprised it works
20:13 nejucomo: Yikes...  implicit copying of C values sounds scary.  ;-)
20:13 lewis1711: that would explain why it's so slow as well..
20:13 lewis1711: if that's indeed what it's doing
20:14 nejucomo: vector-set! can be a function, and you're passing a vector as the first argument.
20:14 dyoo: lewis1711: in that case, maybe you want *pointers* to C data
20:14 nejucomo: set! cannot be a function.
20:14 dyoo: that might exactly be the problem here.
20:14 nejucomo doesn't know anything about ffi yet.
20:14 (join) Kaylin
20:17 dyoo: lewis1711: I don't have time right now to look at FFI bindings, but maybe someone else here may be able to help.
20:18 (quit) nejucomo: Quit: leaving
20:18 lewis1711: it's the exact bit i need to do in C anyway
20:18 lewis1711: the simpler the FFI bindings are, the better I reckon
20:19 dyoo: lewis1711: yup.  Sorry about not knowing FFI so well; I haven't played with that library in a while.  Good luck!
20:19 (quit) dyoo: Quit: dyoo
20:22 neilv: the best ffi is often avoiding ffi altogether
20:23 neilv: i finally got all the misc. c code out of one racket app, and it was a big win for performance, as well as eliminating a potential source of impossible-to-debug bizarre bugs
20:24 lewis1711: neilv: yeah that's another option. I could code this myself in racket.
20:24 lewis1711: why not, i already did a 3d bresenham
20:25 neilv: when i want to have c code, i put it in a separate process
20:27 neilv: or, if you do it without c at all, in racket, and get good performance with the "unsafe" numerical procedures (or with typed racket), the core developers will be pleased
20:41 (join) kofno_
20:41 (quit) jonrafkind: Ping timeout: 250 seconds
20:45 (join) ambrosebs
20:51 (join) spiderweb
20:53 (join) dyoo
20:53 dyoo: asumu: ah, thought about this.  "Yui" would be the _perfect_ replacement for "Rei".  Or is that the other way around?  :)
20:54 (quit) dyoo: Client Quit
20:54 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
20:55 (join) mizu_no_oto
20:57 (join) jonrafkind
20:59 (quit) mizu_no_oto: Client Quit
21:04 neilv: for calling a thunk as a racket gui/base app exits... i see exit;insert-on-callback, but i don't want to use the framework.  set the executable-yield-handler value to insert my callback and then call the gui/base one?
21:06 (join) RacketCommitBot
21:06 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/phnfaQ
21:06 RacketCommitBot: racket/master be49c67 Danny Yoo: Documentation example for string-copy.
21:06 (part) RacketCommitBot
21:09 (join) mizu_no_oto
21:12 sw2wolf: rudybot: (set "Foo" "Bar")
21:13 rudybot: sw2wolf: ; Value: #<set: "Bar" "Foo">
21:13 (quit) francisl: Quit: francisl
21:14 sw2wolf: rudybot: (string->lowercased-words "hello, world!")
21:14 rudybot: sw2wolf: 
VERSION  Hello from outside ze channel

21:19 (quit) cdidd: Ping timeout: 255 seconds
21:25 (join) jrslepak
21:30 (quit) karswell_: Remote host closed the connection
21:35 (quit) spiderweb: Quit: ERC Version 5.3 (IRC client for Emacs)
21:37 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
21:37 lewis1711: is gensym suitable for a UUID?
21:38 jonrafkind: i suppose you could use the gensym'd value as a starting point to generate a uuid
21:38 jonrafkind: but then you might as well use (random) too
21:38 (join) gcr`
21:38 jonrafkind: actaully gensym might be a bad idea because its deterministic
21:39 gcr`: Quick question: If I (define-cstruct _AVFormatContext (...)), and then I have an _AVFormatContext-pointer, does this behave like a _pointer or a _gcpointer? Or is the distinction even necessary?
21:40 (join) karswell_
21:41 lewis1711: my other idea was just a function that kept state, and produced numbers starting at 0.
21:41 jonrafkind: gcr`, it depends if you allocate the object from racket or not
21:42 jonrafkind: if you allocate it with (malloc) and don't use 'raw then its allocated by the gc, iirc
21:42 jonrafkind: if its allocated by some library then its a _pointer, not managed by the gc
21:43 gcr`: Hm. That makes sense. Thanks, jonrafkind
21:45 sw2wolf: rudybot: (current-error-port)
21:45 rudybot: sw2wolf: ; Value: #<output-port:string>
21:45 neilv: you can /msg rudybot
21:45 asumu: lewis1711: if you want UUIDs, probably using a library made for that is best: http://planet.racket-lang.org/display.ss?package=uuid-v4.plt&owner=zitterbewegung
21:45 rudybot: http://tinyurl.com/6r5o57g
21:46 (join) RacketCommitBot
21:46 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/MvcRXg
21:46 RacketCommitBot: racket/master 5baf63f Matthew Flatt: syntax-local-get-shadower: fix interaction with marked bindings...
21:46 (part) RacketCommitBot
21:47 sw2wolf: neilv: /msg is good, then it canot disturb others
21:47 (quit) lusory: Quit: leaving
21:50 neilv: that uuid library code could be improved
21:52 neilv: but the rfc itself is sloppy with its language
21:54 (quit) dnolen: Ping timeout: 265 seconds
22:00 (quit) bitonic: Ping timeout: 244 seconds
22:03 (quit) kofno_: Remote host closed the connection
22:05 (join) kofno
22:13 gcr`: So another question: If the FFI is responsible for all allocation and deallocation of a struct, and if all communication of this struct is done through pointers, can I have a (define-c-struct (...)) that doesn't list all of the struct's fields? The reason is this library declares some fields at the end of the C struct as "private" and "subject to change", but it's just a convention in the header file :/
22:13 gcr`: sorry, by "FFI is responsible for all allocation..." I really mean the remote library handles that with its allocate_foo() functions
22:14 gcr`: that return pointers*
22:14 (quit) neilv: Quit: Leaving
22:16 (quit) ambrosebs: Ping timeout: 265 seconds
22:22 (quit) karswell_: Remote host closed the connection
22:29 (join) mizu_no_oto
22:29 (quit) nathanpc: Quit: Computer has gone to sleep.
22:30 (join) francisl
22:32 (join) karswell_
22:38 lewis1711: damn it I need better stack traces
22:44 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
22:50 (join) dnolen
22:51 (join) mizu_no_oto
22:55 (quit) Shviller: Ping timeout: 260 seconds
22:55 (join) Shviller
22:56 jonrafkind: lewis1711, have you tried errortrace?
22:57 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
22:59 (join) spiderweb
23:03 (quit) masm: Quit: Leaving.
23:04 (quit) francisl: Quit: francisl
23:09 (join) mizu_no_oto
23:09 (quit) mizu_no_oto: Client Quit
23:53 (quit) kofno: Read error: Connection reset by peer
23:55 offby1: sw2wolf: the name "rudybot" came from "rudebot", which came from "fartbot".
23:55 jonrafkind: really?
23:55 jonrafkind: rudybot, are you the grandson of fartbot?
23:55 rudybot: jonrafkind: you made that rule for great grandson which is pretty equivalent to my (grandson) rule
23:58 rudybot: I take no credit for my own name.