00:00 asumu: yoklov: I don't think the namespace thing solves referential transparency with macros.
00:00 asumu: But I could be wrong.
00:01 mithos28: What is the easiest way for patches? github pull requests?
00:01 asumu: mithos28: Yeah.
00:01 yoklov: asumu: there might be edge cases where it doesn't work, but it works extremely well in practice
00:02 asumu: I don't see why that's better than the well-understood hygiene algorithm from the 80s though.
00:02 yoklov: simplicity?
00:02 asumu: It's not actually any simpler.
00:02 asumu: Note that the hygiene algorithm is orthogonal to syntax-case or rules.
00:03 asumu: Or do you mean simple to implement?
00:04 yoklov: i'm not familiar with the algorithm
00:04 (quit) jyc: Ping timeout: 240 seconds
00:04 yoklov: i mean simple to understand
00:04 asumu: Well, you don't need to understand the algorithm to use it.
00:04 asumu: Since the macro system will just rename whatever needs to be renamed.
00:04 yoklov: using it limits what you can accomplish though, does it not?
00:05 sizz_: Fare: yes
00:05 Fare: cool
00:05 mithos28: asumu: what algorithm are you talking about?
00:05 asumu: mithos28: Dybvig algorithm.
00:06 asumu: (from the syntax-case paper)
00:06 sizz_: Fare: and there's a new edition of How to Design Programs in the works too.
00:06 yoklov: asamu: when you backquote something in clojure, it always namespace resolves the symbol, e.g. `foo becomes 'user/foo (if your namespace is user), there's no special behavior in macros
00:06 yoklov: the only difference is when it runs, which seems simpler
00:06 Fare: did they finish 2htdp or is that the new thing you're talking about?
00:06 yoklov: or at least, to me
00:06 mithos28: That is just the marking and how marks cancel, correct?
00:07 asumu: mithos28: Yeah, that's right. But the point is that works no matter how you decide to implement the transformers.
00:07 asumu: (as long as your syntax objects can store the mark information)
00:07 sizz_: Fare: no, that's what i'm talking about; its online here  http://www.ccs.neu.edu/home/matthias/HtDP2e/index.html
00:08 asumu: yoklov: Yeah, but in Scheme, identifiers in macros are renamed completely automatically. How is that harder to use?
00:08 mithos28: It seems weird to call that an algorithm though as that is sort of the definition of hygiene.
00:09 asumu: mithos28: That's true and I think everyone agrees on that. It's just that it's really hard to formally define hygiene.
00:09 asumu: (at least, that's what I hear from my colleagues who know way more about this than I do :))
00:10 mithos28: I have only read the papers, so you probably know more
00:10 yoklov: asumu: because a clojure macro is just a function which manipulates lists. there's much less magic going on
00:10 asumu: yoklov: a Racket macro is also just a function. It operates on syntax and produces syntax.
00:11 yoklov: if you can understand a function that rearranges lists, you can understand clojure/cl macros without much difficulty
00:11 yoklov: personally, i can't understand most of the racket macros i see, and have spent time trying to
00:11 asumu: rudybot: (define-syntax (foo stx) (if (free-identifier=? (stx-car stx) #'car) #'(displayln "hi") #'(displayln "hi")))
00:11 rudybot: asumu: your sandbox is ready
00:11 rudybot: asumu: Done.
00:12 asumu: rudybot: (foo car)
00:12 rudybot: asumu: error: reference to an identifier before its definition: stx-car in module: 'program phase: 1
00:13 asumu: rudybot: (require (for-syntax syntax/stx))
00:13 rudybot: asumu: Done.
00:14 asumu: rudybot: (define-syntax (foo stx) (if (free-identifier=? (stx-car stx) #'car) #'(displayln "hi") #'(displayln "nope")))
00:14 rudybot: asumu: Done.
00:14 asumu: rudybot: (foo car)
00:14 rudybot: asumu: ; stdout: "nope\n"
00:14 asumu: rudybot: (foo other)
00:14 rudybot: asumu: ; stdout: "nope\n"
00:14 asumu: Ah, I wanted the cadr instead.
00:15 asumu: But yeah, you can just deconstruct syntax objects kinda like lists if you want.
00:15 asumu: (this is obviously a stupid macro, but it shows you do not need to use syntax-* at all)
00:15 yoklov: i wasn't aware of that, but there's still a lot of baggage
00:16 mithos28: yoklov: and with it comes a lot of power
00:16 yoklov: as much power as being able to freely transform the source code as a list?
00:17 asumu: Yes, just like how Lisp macros are more powerful than C macros that can freely transform lexemes.
00:17 (join) jeapostrophe
00:17 (quit) jeapostrophe: Changing host
00:17 (join) jeapostrophe
00:17 mithos28: that isn't power that is just convinence
00:17 mithos28: how does clojure handle line numbers in macros?
00:18 yoklov: clojure debugging is messy in general, and a lot of that comes from java. the line numbers seem to be right almost always though (no column information)
00:19 yoklov: that's something that racket does well, but its a pain to use
00:19 yoklov: adding the srclocs and stuff is very repetitive
00:20 asumu: Shambles_: the DrRacket bug re: aborts has been fixed (since I don't think you got the bug closure e-mail)
00:20 asumu: (by Robby)
00:21 Shambles_: asumu: Wow!  That was really fast, at least compared to what I'm used to.  Heck, I'm used to bug reports being ignored indefinitely.  Thanks. :)
00:23 (quit) zyoung: Remote host closed the connection
00:24 yoklov: asumu: that doesn't really make sense, because C's preprocessor isn't turing complete (or even close, for that matter.), but i'm assembling the code in a language that is
00:28 mithos28: yoklov: what do you want, just that syntax-pairs work exactly like pairs with regards to car and cdr? Because that is very separate from the namespace of symbols versus marks.
00:30 yoklov: mithos28: yeah, you're right
00:30 asumu: yoklov: Yes, but a macro system that was turing complete and let you have control over all lexemes would also be pretty unfortunate.
00:30 asumu: Just like CPP is.
00:30 asumu: Point is: if your language is given too much freedom to do what it wants, you're not necessarily getting the most out of it.
00:31 asumu: Because as a programmer you cannot reason about it.
00:31 mithos28: I think you are arguing that syntax-pairs should work just like pairs, but racket is fairly strict on the standard operations not being generic and providing specialized generic versions
00:32 yoklov: mithos28: i'm in general arguing that scheme macros are overly complex
00:32 yoklov: i guess
00:32 (quit) jeapostrophe: Ping timeout: 260 seconds
00:33 yoklov: and that namespacing symbols solves that problem in a much simpler way
00:36 (join) Kaylin
00:55 (quit) yoklov: Quit: bye!
00:56 (join) jeapostrophe
00:56 (quit) jeapostrophe: Changing host
00:56 (join) jeapostrophe
01:06 (join) didi
01:06 (quit) axe_wielder: Ping timeout: 256 seconds
01:06 (quit) hash_table: Ping timeout: 256 seconds
01:06 mithos28: asumu: do you know anything about TR? I'd ask samth or stamourv but they are never on this late.
01:07 didi: Is it possible to flip the a canvas% vertically, so the bottom left becomes (0, 0)?
01:17 didi: Hum. Maybe applying a transformation to dc%?
01:17 didi: Nah, doesn't seem to be correct.
01:20 (join) jyc
01:24 (quit) surrounder: Ping timeout: 246 seconds
01:24 (join) surrounder_
01:33 (quit) didi: Quit: ERC Version 5.3 (IRC client for Emacs)
01:50 (join) veer
02:01 (quit) Fare: Quit: Leaving
02:09 (quit) mithos28: Quit: mithos28
02:11 veer: So when I use a lexer with input-string  :
02:11 veer: (a-lexer (open-input-string "\"\n\""))
02:11 veer: I get #(struct:token tok-string "\"\n\"")
02:11 veer: which is fine.
02:11 veer: But when I use lexer with a file containg only "\n" with quotes included:
02:12 veer: (call-with-input-file "somefile"   (lambda (in) (position-token-token (alexer in))))
02:12 veer: I get #(struct:token tok-string "\"\\n\"")
02:12 veer: note the double \\ in output.
02:12 veer: Any idea why different output assuming "\n" written in a file
02:13 veer: is equivalent to "\"\n\"" in repl.
02:13 jonrafkind: the "\n" in your string is a literal newline
02:13 jonrafkind: but in a file its the two characters \ and n
02:16 veer: so open-input-string interpret \n as newline ?
02:17 jonrafkind: no, "\n" in racket makes a newline
02:17 jonrafkind: you know what i mean? when you write a string in racket you are letting the racket parser interpret the string
02:17 jonrafkind: rudybot: (eval "\n")
02:17 rudybot: jonrafkind: your sandbox is ready
02:17 rudybot: jonrafkind: ; Value: "\n"
02:18 jonrafkind: rudybot: (eval (string-ref "\n" 0))
02:18 rudybot: jonrafkind: ; Value: #\newline
02:18 jonrafkind: if you want a literal slash followed by a newline *in racket* then you must escape it
02:18 jonrafkind: "\\n"
02:19 jonrafkind: and thats why you get an escaped backslash followed by an n when you parse a file, because racket is showing you how it knows to represent characters
02:24 veer: allright I got it , thanks jonrafkind
02:24 jonrafkind: ok
02:31 (quit) jeapostrophe: Ping timeout: 248 seconds
02:31 (join) plobzik
02:33 (quit) plobzik: Read error: Connection reset by peer
02:43 (quit) jonrafkind: Ping timeout: 246 seconds
02:46 jyc: are there any well-written open sourced racket applications I could study?
02:48 Kaylin: lots of drracket is written in racket I think http://racket-lang.org/download/
02:50 jyc: Kaylin, should have thought of that - thanks!
02:50 Kaylin: also prob simpler would be some of the little libraries on planet
02:55 (join) hkBst
02:55 (quit) hkBst: Changing host
02:55 (join) hkBst
03:04 (nick) surrounder_ -> surrounder
03:09 (join) rahul_
03:10 (join) __rahul__
03:14 (join) antithesis
03:26 (quit) Kaylin: Quit: Leaving.
03:26 (join) nilyaK
03:38 (quit) kvda: Quit: Computer has gone to sleep.
03:41 (quit) rahul_: Ping timeout: 245 seconds
03:42 (quit) __rahul__: Ping timeout: 260 seconds
03:42 (join) noelw
03:49 (join) surrounder
04:00 (join) kvda
04:03 (join) __rahul__
04:03 (join) rahul_
04:22 (quit) kvda: Quit: Computer has gone to sleep.
04:22 (quit) surrounder: Ping timeout: 246 seconds
04:23 (join) surrounder_
04:23 (quit) veer: Ping timeout: 260 seconds
04:31 (nick) surrounder_ -> surrounder
04:33 (quit) jyc: Read error: Connection reset by peer
04:53 (join) veer
04:59 (join) dzhus
05:00 (quit) dzhus: Remote host closed the connection
05:00 (join) dzhus
05:33 (join) kvda
05:46 (join) jesyspa
05:51 (quit) cdidd: Remote host closed the connection
05:53 (quit) kvda: Quit: Computer has gone to sleep.
05:55 (join) surrounder_
05:56 (nick) surrounder_ -> surrounder
06:04 (join) surrounder_
06:09 (join) mceier
06:19 (quit) surrounder_: Ping timeout: 246 seconds
06:20 (join) surrounder
06:24 (join) bitonic
06:31 (quit) nilyaK: Quit: Leaving.
07:03 (quit) dzhus: Remote host closed the connection
07:06 (quit) surrounder: Ping timeout: 246 seconds
07:06 (join) surrounder_
07:08 (quit) antithesis: Quit: antithesis
07:19 (quit) surrounder_: Ping timeout: 246 seconds
07:19 (join) surrounder
07:31 (quit) surrounder: Ping timeout: 246 seconds
07:31 (join) surrounder
07:38 (quit) jesyspa: Quit: Lost terminal
07:38 (join) jesyspa
07:39 (quit) SHODAN: Remote host closed the connection
07:40 (quit) surrounder: Ping timeout: 246 seconds
07:40 (join) SHODAN
07:40 (join) antithesis
07:56 (join) RacketCommitBot
07:56 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/peHcLA
07:56 RacketCommitBot: [racket/master] racket/gui: fix `canvas%' implementation `on-paint' - Matthew Flatt
07:56 (part) RacketCommitBot
08:05 (quit) Shviller: Ping timeout: 248 seconds
08:05 (quit) rahul_: Ping timeout: 245 seconds
08:06 (join) Shviller
08:06 (quit) __rahul__: Ping timeout: 240 seconds
08:09 (join) kanak
08:18 (join) rahul_
08:20 (join) __rahul__
08:26 (join) surrounder
08:37 (quit) surrounder: Ping timeout: 246 seconds
08:38 (join) surrounder_
08:46 (nick) surrounder_ -> surrounder
08:47 (join) getpwnam
08:48 (join) hash_table
08:49 (join) zyoung
08:53 (join) aidy
09:02 (join) snearch
09:02 (join) dnolen
09:13 (quit) antithesis: Ping timeout: 246 seconds
09:18 (join) surrounder_
09:25 (join) plobzik
09:26 (join) netrino
09:27 (quit) plobzik: Read error: Connection reset by peer
09:29 (quit) hash_table: Ping timeout: 246 seconds
09:29 (quit) getpwnam: Ping timeout: 246 seconds
09:29 (join) antithesis
09:47 (quit) surrounder_: Ping timeout: 246 seconds
09:48 (join) surrounder
09:52 (join) noam__
09:54 (quit) noam: Ping timeout: 260 seconds
09:54 (quit) noam__: Read error: Connection reset by peer
09:54 (join) noam__
09:55 (quit) noam__: Read error: Connection reset by peer
09:55 (join) noam__
09:58 (quit) noam__: Read error: Connection reset by peer
09:58 (join) noam__
09:59 (quit) noam__: Read error: Connection reset by peer
09:59 (part) aidy
09:59 (join) noam__
10:01 (quit) veer: Quit: Leaving
10:07 (quit) surrounder: Ping timeout: 246 seconds
10:07 (join) masm
10:08 (quit) acarrico: Ping timeout: 245 seconds
10:08 (join) surrounder
10:11 (quit) snearch: Quit: Verlassend
10:11 (join) jrslepak
10:12 (join) mithos28
10:13 samth: mithos28: merging your pull req now
10:15 mithos28: samth: thanks, I'll have another one for the polymorphic struct problem soon as well, just need to add test cases and test it
10:15 samth: mithos28: wonderful
10:16 (join) RacketCommitBot
10:16 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/8O1qLg
10:16 RacketCommitBot: [racket/master] Add support for multiple provides of the same identifier in TR. - Eric Dobson
10:16 (part) RacketCommitBot
10:16 (join) GrayMatterComput
10:16 mithos28: with regards to rename-transfromers, I originally thought it would be possible because you can examine the rename transformer and instead export a rename-transformer to a safe identifier
10:16 samth: mithos28: right, you could do that
10:17 samth: you'd have to treat the targets of rename-transformers as exporte
10:17 samth: d
10:17 samth: but there's a more general solution that i know how to do
10:17 samth: just haven't done it yet
10:17 samth: that will support arbitrary macros
10:18 mithos28: great, then I think that pr can be folded into that
10:18 GrayMatterComput: Hi all, does anyone know how to use _list or equivilent with ffi in Racket v5.1.3 or should I just upgrade my package to 5.2.1 (using Ubuntu 12.04)
10:20 (join) acarrico
10:21 (join) gridaphobe
10:24 (join) anRch
10:26 GrayMatterComput: Anyone? Basically, I finally realized that 5.2.1 has support for _list, _vector, etc but 5.1.3 does not. How can you send a list as a parameter to a C method in this version (without _list)?
10:27 mithos28: Upgrading seems like an easy method, is there a reason it is not?
10:28 GrayMatterComput: I would like to stick to the default packages from Ubuntu 12.04, which is currently Racket 5.1.3.
10:30 GrayMatterComput: This is for xpilot-ai (xpilot-ai.org) btw :)
10:30 mithos28: In that you don't want to install it manually or that you only want packages from the main repo? There is a ppa with newer versions that should make it painless.
10:32 mithos28: or are you trying to release something that the will work with the default packages?
10:33 GrayMatterComput: We were hoping the release would only be based on the main repo
10:34 GrayMatterComput: it isn't a huge deal to install the ppa, but I was curious as to how lists were passed before _list
10:34 samth: GrayMatterComput: you should ask on the mailing list
10:34 mithos28: _list is in 5.1.3, is there a problem with the version that is there?
10:34 mithos28: http://download.racket-lang.org/docs/5.1.3/html/foreign/foreign_procedures.html?q=ffi#(form._((lib._ffi/unsafe..rkt).__list))
10:34 rudybot: http://tinyurl.com/bmosltf
10:36 GrayMatterComput: I get _list: bad syntax in: _list
10:36 mithos28: are you using it inside a _fun
10:36 GrayMatterComput: all my other ffi stuff works great though
10:36 GrayMatterComput: yep
10:36 GrayMatterComput: is this right? (get-ffi-obj "start" ai (_fun _int _list -> _int)
10:37 mithos28: no
10:37 mithos28: you need (_list o _int) or something similar
10:38 mithos28: the o is for output saying that you are only passing in values with that argument, and the _int is saying what values are in the list
10:39 (join) getpwnam
10:39 (join) hash_table
10:40 GrayMatterComput: alright, then I still get _list: bad syntax in: (_list o _string) for (get-ffi-obj "start" ai (_fun _int (_list o _string) -> _int)
10:41 GrayMatterComput: with an i instead of o, it segfaults
10:41 mithos28: at that point, or at calling it point?
10:42 mithos28: I would expect when you call the function
10:42 mithos28: I messed up i and o, you want i
10:43 mithos28: and o requires the length which explains why it was bad syntax
10:43 mithos28: Are you sure you are calling the c function with the correct arguments?
10:45 GrayMatterComput: the arguments are correct (I wrote the function), it's basically argv and argc
10:46 mithos28: and it doesn't segfault when called from the command line?
10:46 GrayMatterComput: the segfault is from tying to use the value i get, because instead of reading "xpilot-ng-x11" (or w/e) when it gets to C, it reads ����
10:47 mithos28: you are returning an int but that sounds like you are returning a string
10:47 mithos28: can you paste the c and racket code somewhere
10:48 GrayMatterComput: the <?> symbols are from printf(argv[0]) in C, argv being the list that I passed from scheme
10:49 (quit) antithesis: Remote host closed the connection
10:49 GrayMatterComput: sure, I can upload the code in a minute
10:51 GrayMatterComput: http://xpilot-ai.org/downloads/RACKET/
10:52 (quit) asumu: Ping timeout: 248 seconds
10:52 GrayMatterComput: the method start is at the bottom of the .c file
10:52 (join) aidy
10:52 mithos28: where do you call the ffi function
10:53 GrayMatterComput: the user would call it in their bot, so at the end of Spinner.rkt
10:53 mithos28: found it, on another note you should try using modules instead of include
10:54 GrayMatterComput: yeah, that was my next thing to look into after I get this working ><;
10:54 (join) asumu
10:54 mithos28: why are you passing in the function loop to an int value?
10:54 GrayMatterComput: I'm not
10:55 mithos28: it looks like you are. you define the function loop (define (loop) (turnLeft 1)) and then call (start loop ...)
10:56 GrayMatterComput: if you look at rktAI.rkt, start (for racket) takes a function and a list, start (for c) takes an int and list
10:56 (quit) anRch: Read error: Connection reset by peer
10:56 mithos28: ah, missed that
10:56 (join) anRch
10:56 GrayMatterComput: np, the naming is a bit confusing
10:57 mithos28: and the first print statement is what is corrupted?
10:57 GrayMatterComput: yes
10:58 GrayMatterComput: ARGC: 3 ARGV[0]: ����
10:58 (join) tim-brown
10:59 GrayMatterComput: right at the beginning of start in C
10:59 mithos28: can you try casting the pointers to ints and printing those
11:00 mithos28: that would confirm that sane values are getting passed through
11:00 GrayMatterComput: so, like     printf("ARGC: %d\nARGV[0]: %s\d",argc,(int)argv[0]); ?
11:01 mithos28: need to change the %s to a %d but yes
11:01 mithos28: and can you add argv[1] and 2
11:01 mithos28: as the values should all look similar
11:01 GrayMatterComput: yep (and yeh, I changed n by accident
11:01 (quit) elliottcable: Remote host closed the connection
11:02 (join) elliottcable
11:03 GrayMatterComput: OH, WAIT. It might work now >.> I will send you cookies
11:03 mithos28: did you forget to recompile the C?
11:05 (join) jeapostrophe
11:05 (quit) jeapostrophe: Changing host
11:05 (join) jeapostrophe
11:05 GrayMatterComput: no, it must have been something in Racket I fouled up trying to fix it, but now I get ARGC: 3 ARGV[0]: xpilot-ng-x11 ARGV[1]: -name ARGV[2]: Spinner and the bot is named correctly
11:06 mithos28: ok, glad to help. C is annoying to debug
11:08 GrayMatterComput: honestly, I have the most difficult time with Racket ><; I typically program in C and Python (then Java and all those silly web languages and then Scheme/Racket)
11:08 GrayMatterComput: It's all those parentheses... anyway, many thanks, for what it's worth, I shall credit you in our final rkt file
11:14 (part) GrayMatterComput
11:21 (join) RacketCommitBot
11:21 RacketCommitBot: [racket] plt pushed 3 new commits to master: http://git.io/RRFmGg
11:21 RacketCommitBot: [racket/master] further correction to tests/gracket/blits prop - Matthew Flatt
11:21 RacketCommitBot: [racket/master] configure: use installed `libtool' by default - Matthew Flatt
11:21 RacketCommitBot: [racket/master] ffi/unsafe: fix arithmetic overflow and representation - Matthew Flatt
11:21 (part) RacketCommitBot
11:36 (quit) netrino: Ping timeout: 240 seconds
11:44 (quit) jeapostrophe: Ping timeout: 245 seconds
11:45 (join) aalix
11:46 (quit) acarrico: Ping timeout: 252 seconds
11:49 (join) netrino
11:49 (quit) sstrickl: Quit: sstrickl
11:51 mithos28: samth: pull request sent, tests are still running
11:51 samth: mithos28: great
11:52 (quit) dnolen: Ping timeout: 250 seconds
11:56 (join) RacketCommitBot
11:56 RacketCommitBot: [racket] plt pushed 2 new commits to master: http://git.io/4KoCBA
11:56 RacketCommitBot: [racket/master] adjust the timeouts to two drracket tests and add a little syncronization help in a third - Robby Findler
11:56 RacketCommitBot: [racket/master] adjust redex error message tests to match singular/plural fixes in - Robby Findler
11:56 (part) RacketCommitBot
11:58 (join) cdidd
11:59 (join) acarrico
12:08 (quit) anRch: Quit: anRch
12:21 (quit) noelw: Quit: noelw
12:25 (quit) noam__: Read error: Connection reset by peer
12:25 (join) noam__
12:25 mithos28: samth: tests passed
12:26 (quit) noam__: Read error: Connection reset by peer
12:26 (join) noam__
12:27 (quit) noam__: Read error: Connection reset by peer
12:27 (join) noam__
12:28 (quit) noam__: Read error: Connection reset by peer
12:28 (join) noam__
12:31 (quit) noam__: Read error: Connection reset by peer
12:32 (join) noam__
12:33 (quit) noam__: Read error: Connection reset by peer
12:33 (join) noam__
12:42 (quit) hkBst: Quit: Konversation terminated!
13:03 samth: mithos28: that was a silly oversight on my part, apparently
13:07 (quit) tim-brown: Remote host closed the connection
13:16 (quit) noam__: Read error: Connection reset by peer
13:16 (join) noam__
13:17 (quit) noam__: Read error: Connection reset by peer
13:17 (join) snearch
13:18 (join) noam__
13:18 (join) antithesis
13:18 (quit) noam__: Read error: Connection reset by peer
13:19 (join) noam__
13:20 (join) jonrafkind
13:25 (quit) surrounder: Ping timeout: 246 seconds
13:26 (join) surrounder
13:27 (join) RacketCommitBot
13:27 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/V0YDOw
13:27 RacketCommitBot: [racket/master] Attach struct info to constructor for polymorphic TR structs. - Eric Dobson
13:27 (part) RacketCommitBot
13:27 samth: eli: please close pull req 105 and 106
13:50 (join) jacius
14:09 (quit) rahul_: Ping timeout: 250 seconds
14:10 (quit) __rahul__: Ping timeout: 265 seconds
14:22 (join) anRch
14:22 (quit) snearch: Quit: Verlassend
14:22 (join) __rahul__
14:22 (join) rahul__
14:23 (quit) surrounder: Ping timeout: 246 seconds
14:24 (join) wtetzner
14:25 (join) surrounder
14:29 (quit) jrslepak: Quit: This computer has gone to sleep
14:54 (join) dnolen
15:10 (quit) netrino: Ping timeout: 250 seconds
15:16 (quit) anRch: Quit: anRch
15:18 (join) netrino
16:20 (join) yoklov
16:32 (join) ynniv
16:33 (quit) yoklov: Quit: bye!
16:34 (join) Fare
16:40 (join) jeapostrophe
16:40 (quit) jeapostrophe: Changing host
16:40 (join) jeapostrophe
16:43 (quit) acarrico: Ping timeout: 244 seconds
16:47 (join) acarrico
17:09 (join) zyoung_
17:11 (quit) zyoung: Read error: Operation timed out
17:14 (quit) hash_table: Ping timeout: 244 seconds
17:14 (quit) getpwnam: Ping timeout: 244 seconds
17:24 (quit) netrino: Quit: Ave!
17:29 (quit) antithesis: Quit: antithesis
17:35 (quit) Fare: Quit: Leaving
17:37 (quit) ynniv: Quit: ynniv
17:41 (join) yoklov
17:41 (join) yoklov_
17:42 (part) yoklov_
17:47 (quit) mceier: Quit: leaving
17:51 (quit) zyoung_: Remote host closed the connection
17:51 (join) kreol[Ukr]
17:52 (join) zyoung
17:54 (quit) dnolen: Ping timeout: 250 seconds
17:54 (quit) gridaphobe: Remote host closed the connection
17:56 (quit) rahul__: Ping timeout: 244 seconds
17:56 (quit) __rahul__: Ping timeout: 245 seconds
17:59 (join) nilyaK
18:09 (join) zyoung_
18:09 jeapostrophe: samth_away: stamourv: ping
18:12 (quit) zyoung: Ping timeout: 244 seconds
18:13 (quit) zyoung_: Ping timeout: 256 seconds
18:14 (quit) nilyaK: Ping timeout: 252 seconds
18:16 stamourv: jeapostrophe: pong
18:16 stamourv: (But leaving in ~15 mins)
18:16 jeapostrophe: I'm sending you an email... probably better form anyways :)
18:16 jeapostrophe: it can wait till tomorrow/whenever you can respond
18:16 jeapostrophe: thanks tho
18:16 stamourv: Ok.
18:17 (join) jyc
18:19 (join) getpwnam
18:20 (join) hash_table
18:24 (quit) jyc: Read error: Connection reset by peer
18:24 (join) Kaylin
18:35 (quit) Kaylin: Quit: Leaving.
18:38 (join) zyoung
19:02 friscosam: is there a stream match expander already in racket?
19:13 jeapostrophe: friscosam: not to my knowledge
19:18 (join) antono
19:23 (quit) masm: Quit: Leaving.
19:24 (join) masm
19:25 (quit) jonrafkind: Ping timeout: 260 seconds
20:04 (join) dnolen
20:05 (quit) jacius: Remote host closed the connection
20:18 (join) kvda
20:19 (quit) jeapostrophe: Ping timeout: 252 seconds
20:26 Shambles_: friscosam: What is a stream match expander?
20:28 Shambles_: I'm sort of worried that I seem to be in the extreme minority that /likes/ the very minimal and regular syntax in Lisp- and Forth-likes.  Also a bit surprised to see people complaining about the parens here.
20:29 (join) jrslepak
20:31 ChibaPet: No, that's what draws in most people I think.
20:32 ChibaPet: Simplicity, uniformity.
20:32 Shambles_: So far when they're brought up in this channel about all I hear are the complaints, not people saying they like it.
20:34 bremner: I think most people think saying I like lisp in a lisp channel is a bit pointless.
20:34 Shambles_: I absolutely adore knowing "this will happen, and then this will happen" without looking up some godforsaken table of operator priorities (for anything outside basic mathematics), or having to be very careful about how I write things, because the syntax is a minefield.  Heck, in Racket, and Common Lisp, even the order of evaluation of arguments is documented.  Can't say that for most languages, and might be bad taste to depen
20:35 ChibaPet: Saying I like Perl in a Lisp channel, now... :P
20:35 Shambles_: XD
20:35 bremner: I like perl.
20:36 ChibaPet: Same here, although some of the syntax for nested indirection makes me twitch.
20:36 Shambles_: I like that Perl looks somewhat less like opening a executable in a text editor than APL. :)
20:36 bremner: yeah, true enough.
20:36 bremner: (about the indirection)
20:36 (quit) Shvillr: Ping timeout: 244 seconds
20:36 (join) Shvillr
20:37 Shambles_: Thoug APL and J do at least have a nice, simple, defined evaluation order.
20:37 bremner: the J source code is kindof sad
20:38 Shambles_: bremner: Howso?
20:38 bremner: just awful coding style IMHO. and remember, I like perl.
20:38 Shambles_: :D
20:43 Shambles_: Is there much interest in other back-ends for Racket, like Whalesong?  I can think of a few applications that embed Perl (Pidgin) or Python (Blender, Gimp (though there is Guile as a option there), Inkscape, Scribus...) that it might be nice to run Racket in.
20:44 bremner: Shambles_: you can embed "real" racket, as well
20:45 Shambles_: bremner: I suppose so, if they'd accept the patches.  Pidgin has become hostile to embedding anything else. They want everybody to just use Perl or D-BUS.  Of course D-BUS doesn't exist on Windows.  I'm not sure if it exists/works on Mac OS X either.
20:46 jrslepak: Shambles_: download the J source code and take a look at it
20:47 bremner: jrslepak: heh, you've been scarred by it too?
20:47 bremner: I remember thinking oh, maybe I should package that for Debian
20:47 jrslepak: bremner: I was looking through it this morning... can't remember what I was looking for
20:48 jrslepak: not sure I _want_ to remember
20:49 Shambles_: I notice a intense hatred of whitespace and meaningful names, and a love of the C preprocessor.  None of this surprises me given the typical APL programming style.
20:50 Shambles_: I was more expecting all of the code to be formatted into different orgami animals or something else suitable for the IOCCC.
20:50 jrslepak: and the "hatred of … meaningful names" even carries into the file names
20:53 Shambles_: Yes.  Thou shalt not use names longer than 2 characters. :P
20:53 Shambles_: I'm quite impressed with some of the preprocessor abuse.  I didn't know some of this would even work.
20:53 Shambles_: And here I thought I was clever for knowing you could put things between the 'then' block and the 'else'.
20:54 Shambles_: Imagine what they could have done with C++ template metaprogramming!
20:55 ChibaPet: The worst source code I've ever seen was ... well, proprietary. The worst publically available source I've ever seen was procmail.
20:55 jrslepak: "typedef A                (*AF)();
20:57 Shambles_: When I last looked a lot of X Window source was ... really gross.
20:57 Shambles_: I last looked in the late 90's though.
20:57 jrslepak: and then, a few lines later, "typedef A                  X;"
20:58 Shambles_: Well, hey, if you're not going to use names longer than two characters, you have to reuse them eventually! ;)
20:58 jrslepak: typedef struct {I k,flag,m,t,c,n,r,s[1];}* A;
21:00 Shambles_: Funny thing is, some bad source isn't obvious.  I had never run into bad source code that /appeared/ to be well written until I worked on that last C++ project.  Everything had nice easy to read comments, and meaningful variable names.
21:01 Shambles_: On the downside, one class was over 8,500 lines, they decided to avoid multithreading by making everything one huge state machine, and it wasn't so much designed as random bits glued together until it more or less did what you wanted.
21:02 Shambles_: The FSM aspect made it super fun to work on.  The place where something was being fiddled or freed might be a *huge* distance (entire other module (i.e. directory of other files)) from where it was. allocated.
21:03 (join) jyc
21:03 Shambles_: It linked to both wxWindow and Qt, but didn't use either one for a GUI. It used its own GUI toolkit.  It also linked to DirectX, but used OpenGL for graphics.  It used FMOD for sound.
21:03 Shambles_: I can't tell you what it used those libraries for.  I never could figure it out, except that it wouldn't build without them.
21:03 jrslepak: oh... well, that's what I was about to ask
21:05 Shambles_: I'm not sure what situation could produce code like this.  My best guess was somebody in management had some brains and decided to make and enforce a coding style.  That could explain the surface seeming to be good, while it was all nasty inside.
21:05 Shambles_: Following the word, but not the spirit, of the law. :P
21:08 (quit) dnolen: Ping timeout: 246 seconds
21:49 (join) jacius
22:02 (quit) cky: Quit: WeeChat 0.3.6
22:02 (join) Fare
22:05 (join) cky
22:13 (quit) bitonic: Ping timeout: 252 seconds
22:13 (quit) cky: Quit: WeeChat 0.3.6
22:14 (join) cky
22:14 (join) bitonic
22:24 (quit) masm: Quit: Leaving.
22:27 (quit) jesyspa: Quit: leaving
22:30 (join) jonrafkind
22:37 (quit) bitonic: Ping timeout: 252 seconds
22:38 (join) bitonic
22:48 friscosam: Shambles_: for the match form you can make new 'matchers' using define-match-expander.  I was asking if someone had already written a 'matcher' for streams
22:48 friscosam: I'm converting some code that uses lists to uses streams, and it would be convenient
22:49 Shambles_: friscosam: I see.  You went from lists to streams to make it more general?
23:03 friscosam: I was actually trying to introduce some laziness
23:03 (part) aidy
23:08 (quit) Fare: Ping timeout: 246 seconds
23:15 (join) jeapostrophe
23:15 (quit) jeapostrophe: Changing host
23:15 (join) jeapostrophe
23:16 (join) antithesis
23:18 (quit) yoklov: Quit: computer sleeping
23:20 (join) vu3rdd
23:20 (quit) vu3rdd: Changing host
23:20 (join) vu3rdd
23:22 (join) Fare
23:30 (quit) jeapostrophe: Ping timeout: 252 seconds
23:33 (quit) aalix: 
23:39 (join) Kaylin
23:40 mithos28: Can anyone give me an example of in which TypedValue is needed in TR to maintain safety?
23:42 (quit) jyc: Read error: Connection reset by peer