00:00 (quit) Fare: Ping timeout: 260 seconds
00:05 (quit) dented42: Ping timeout: 248 seconds
00:06 (quit) nejucomo: Ping timeout: 276 seconds
00:11 (join) Fare
00:12 (join) dented42
00:23 (quit) ormaaj: Ping timeout: 256 seconds
00:29 (quit) Kaylin: Ping timeout: 252 seconds
00:31 (join) Kaylin
00:37 (join) ormaaj
00:44 (join) cdidd
00:45 (join) spiderwe`
00:46 (quit) spiderweb: Ping timeout: 255 seconds
00:46 (quit) spiderwe`: Client Quit
00:47 (join) spiderweb
00:55 Sgeo: I can only assume that dyoo is a Puella Magi Madoka Magica fan
00:58 (quit) mceier: Ping timeout: 255 seconds
01:02 (quit) ormaaj: Quit: WeeChat 0.4.0-dev
01:03 (join) Nisstyre-laptop
01:06 (join) ormaaj
01:23 (join) myx
01:23 (part) sw2wolf{away}: "ERC Version 5.3 (IRC client for Emacs)"
01:27 (quit) Fare: Ping timeout: 245 seconds
01:33 (join) Fare
01:51 (quit) Fare: Ping timeout: 245 seconds
01:59 Sgeo: Is there a way to add a type of number to Racket's number hierarchy?
01:59 Sgeo: Or numeric tower or whatever it's called?
02:00 mithos28: Sgeo: Sorta, but it involves deep internal manipulations
02:01 mithos28: there is work ongoing to add extended precision flonums on the mailing lists
02:01 mithos28: What would you want to add?
02:01 Sgeo: exact precision reals
02:02 Sgeo: *exact precision computable reals
02:02 mithos28: like CGAL?
02:02 Sgeo: Not familiar with CGAL
02:02 Sgeo: But I know there are Haskell libraries for it
02:02 Sgeo: Also, another type of number to add:  Dual numbers
02:03 mithos28: Sgeo: Well you could add your own structures, they just wouldn't work with the primitive numbers
02:06 mithos28: what would adding them to rackets number heirarchy get you that you want?
02:12 Sgeo: mithos28, automatic differentientation
02:12 Sgeo: (can't spell)
02:12 Sgeo: Unrelated:
02:12 Sgeo: #lang planet asumu/sweet racket
02:12 Sgeo: Does not seem to change the REPL
02:13 mithos28: Sgeo: Why do you need the core number system to support your extensions to get AD?
02:16 (join) neilv
02:16 (quit) neilv: Changing host
02:16 (join) neilv
02:16 (quit) neilv: Client Quit
02:16 Sgeo: Hmm. So I could easily use + like I'd expect as though it was a core number? So that I can, as long as I define what functions like sin mean on these numbers, can pass them to any Racket function that uses numbers and have them work properly?
02:17 mithos28: why cannot you just define your own + and sin, and use those?
02:19 mithos28: The issue with extension in the way that you are proposing is that it violates assumptions of other libraries that numbers are just the standard racket numbers
02:21 Sgeo: What is special about the standard Racket numbers that would be unfakable by my own kind of numbers?
02:22 mithos28: that the parts of a complex number are real numbers
02:23 mithos28: and that all numbers are complex
02:25 Sgeo: Why doesn't this work:
02:25 Sgeo: rudybot: (generator () (yield 1))
02:25 rudybot: Sgeo: your sandbox is ready
02:25 rudybot: Sgeo: error: #<undefined>:1:11: #%app: missing procedure expression; probably originally (), which is an illegal empty application in: (#%app)
02:26 Sgeo: I'm looking at http://docs.racket-lang.org/reference/Generators.html
02:26 mithos28: you are missing the require
02:26 mithos28: (require racket/generator)
02:26 Sgeo: Ah
02:27 Sgeo: Hmm.
02:28 Sgeo: #%app is sort of like a macro, right?n It seems to check certain requirements, but what does it expand into? It can't expand into itself, can it?
02:28 Sgeo: It's run before evaluation clearly
02:28 Sgeo: rudybot, (if #f () "false")
02:28 rudybot: Sgeo: error: #<undefined>:1:7: #%app: missing procedure expression; probably originally (), which is an illegal empty application in: (#%app)
02:28 mithos28: #%app is a macro
02:29 mithos28: #%app is one of the core forms actually
02:29 mithos28: like if, define-values, lambda
02:29 Sgeo: So, it's just run and then, if it doesn't error, it ... stays there?
02:29 Sgeo: (The default #%app I mean)
02:30 mithos28: yes, it corresponds to a function call in fully expanded syntaxt
02:32 Sgeo: rudybot, (define whee (let loop () loop)) ((((((((((whee))))))))))
02:32 rudybot: Sgeo: http://jtnimoy.net/workviewer.php?q=178   "In Tron, the hacker was not supposed to be snooping around on a network; he was supposed to kill a process. So we went with posix kill and also had him pipe ps into grep. I also ended up using emacs eshell to make the terminal more l33t. The team was delighted to see my emacs performance -- splitting the editor into nested panes and running different modes. I was tickled that
02:32 Sgeo: wat.
02:32 mithos28: rudybot: be nice to Sgeo
02:32 rudybot: mithos28: it has a repl in drscheme.
02:33 mithos28: Sgeo: you want a : not a ,
02:33 (join) univyrse
02:33 Sgeo: rudybot: (define whee (let loop () loop)) ((((((((((whee))))))))))
02:33 rudybot: Sgeo: unstable/automata found by google :-2 whee
02:33 Sgeo: ...
02:33 mithos28: and you need your parens to match
02:33 mithos28: you can use eval and it will give better error messages
02:33 Sgeo: rudybot, eval (define whee (let loop () loop)) ((((((((((whee))))))))))
02:33 rudybot: Sgeo: ; Value: #<procedure:loop>
02:34 Sgeo: Is it because I'm using two forms?
02:34 Sgeo: rudybot: 1 2
02:34 rudybot: Sgeo: eh?  Try "rudybot: help".
02:34 Sgeo: rudybot, eval 1 2
02:34 rudybot: Sgeo: ; Value: 2
02:34 mithos28: yes
02:34 Sgeo: rudybot, (begin (define whee (let loop () loop)) ((((((((((whee)))))))))))
02:34 rudybot: Sgeo: ; Value: #<procedure:loop>
02:35 univyrse: the official racket guide is surprisingly well written, and teaches a lot of things in small spaces
02:37 Sgeo: The official Racket guide also doesn't mention a lot of cool things, like events and generators
02:37 Sgeo: At least, that I've seen, haven't read the whole thing
02:37 univyrse: I'm on 8.1
02:38 univyrse: just chugging my way until I finish it
02:38 univyrse: and i come here when I get really stuck
02:40 Sgeo: :( video I'm watching just called let a function
02:40 Sgeo facepalms
02:53 mithos28: Sgeo: It is. Just at phase 1.
02:58 univyrse: i never liked video tutorials
02:58 univyrse: i found myself copying down the code and not learning
02:59 univyrse: in addition, most video teachers dont teach, they just commentate alongside footage of themselves slowly building a program
03:02 Sgeo: I wish Racket was more immutability-focused like Haskell or Clojure
03:02 Sgeo: I like immutability
03:03 mithos28: Sgeo: What is missing?
03:03 Sgeo: Everything immutable by default
03:03 mithos28: What is not?
03:04 Sgeo: and somestruct-thing-set as a functional "setter", for example
03:04 Sgeo: Rather than somestruct-thing-set!
03:04 univyrse: haskell makes my dog cry
03:04 univyrse: and me too
03:04 mithos28: By default the set! functions are not defined
03:04 mithos28: and there is struct-copy if you want functional setters
03:05 univyrse: once you define (struct thing (x)) then (thing-x-set! newvalue) is autmatically defined
03:05 mithos28: functional setters have issues with inheritence though
03:05 mithos28: univyrse: No
03:05 mithos28: rudybot: init racket
03:05 rudybot: mithos28: your sandbox is ready
03:05 mithos28: rudybot: (struct foo (x))
03:05 rudybot: mithos28: Done.
03:06 mithos28: rudybot: set-foo-x!
03:06 rudybot: mithos28: you mean you want set-foo-x! to dispatch to your own functions?
03:06 mithos28: rudybot: foo-x
03:06 rudybot: mithos28: ; Value: #<procedure:foo-x>
03:07 Sgeo or someone should make an applicative language, where values can define a lift function to be applied to any functions that try to use that value as an argument
03:07 univyrse: yeah someone should do something about a thing
03:08 Sgeo: I think that would be a sort of generalization of lazy and frtime
03:09 Sgeo: With that model though, order of arguments could make a significant semantic difference :(
03:12 ozzloy: my girlfriend looked at the wescheme games and says they are too slow
03:13 ozzloy: sorry, that's out of context
03:13 mithos28: ozzloy: I remember the context
03:13 ozzloy: oh cool
03:13 mithos28: Has she ported hers over, or just looking at the ones she found on the web?
03:14 ozzloy: for the benefit of everyone else: she wrote a game, but wants to make a one-click installer for every platform
03:14 ozzloy: she looked at http://www.wescheme.org/run?publicId=wares-bogus-watch-swarm-dimly
03:14 ozzloy: and a couple others
03:14 ozzloy: also she has mouse interaction in her game
03:14 (quit) jonrafkind: Ping timeout: 245 seconds
03:14 ozzloy: and none of the examples we found have mouse interaction
03:15 mithos28: That ran pretty fast for me
03:15 ozzloy: i'm wondering if the lag is a result of the racket compiling to javascript, then being run
03:15 Sgeo: Wait, is it real Racket, or some sort of ...
03:15 Sgeo: hmm
03:15 ozzloy: it runs pretty fast for me too, but sometimes when pressing left a few times in a row, it hiccups for about 1/3 of a second
03:15 mithos28: http://hashcollision.org/whalesong/examples/raphael-demo/raphael-demo.html
03:16 ozzloy: that looks mesmerizing
03:16 mithos28: http://hashcollision.org/whalesong/
03:16 mithos28: has more examples
03:17 ozzloy: for the benefit of everyone else: she wrote this game in pygame, but as far as we can tell, there's no easy way to cross compile a windows installer for a pygame project from ubuntu
03:18 ozzloy: mithos28, so you're saying any lag is probably the fault of the specific game implementation algorithms?
03:18 mithos28: I'm guessing that a lot of the lag is that
03:18 (join) mceier
03:19 mithos28: not that it will be easy to have low lag, but that most of those games are programmed by 12-14 year olds (from the bootstrap project).
03:19 mithos28: Thus I doubt they were optimized much past getting them working
03:22 Sgeo: Hmm, found a post that purports to explain Scheme style macros for CL people
03:25 ozzloy: mithos28, cool
03:26 (join) feste
03:26 ozzloy: mithos28, i'll tell her that
03:26 ozzloy: oh, hi there
03:26 feste: hi
03:26 feste: tell me what?
03:26 ozzloy: that the games were probably not optimized past getting them to work
03:26 ozzloy: passed?
03:26 feste: ah
03:26 feste: past
03:32 (quit) feste: Quit: Page closed
03:34 (quit) mithos28: Quit: mithos28
03:36 Sgeo: I seriously cannot get over let's loop thingy being a first-class procedure
03:36 Sgeo: Clojure may have poisoned my brain slightly
03:39 (quit) Kaylin: Quit: Leaving.
03:40 ozzloy: Sgeo, "let's loop thingy ..." what's let's loop thingy
03:41 Sgeo: As in, the foo in (let foo () ... )
03:47 ozzloy: Sgeo, oh, named let
03:51 Sgeo: I seriously need to find a way to test let-syntax
03:54 Sgeo breathes
03:55 Sgeo: HAH
03:55 Sgeo: AHAHAHHAHA
03:55 ozzloy: ?
03:55 Sgeo: This is awesome
03:56 Sgeo: http://pastie.org/5601436
03:56 Sgeo: The result was "HONK"
03:57 Sgeo: (The name 1-arg-app was when I wanted to use the other thing that is an example of replacing #%app, but that uses define-syntax
03:58 Sgeo: erm, define-syntax-rule
03:58 ozzloy: i sure don't know begin-for-syntax and let-syntax
03:58 ozzloy: idk macros.  this is macros, yes?
03:59 Sgeo: This is macros and replacing #%app
03:59 ozzloy: what's #%app? function application?
04:00 Sgeo: It's a macro that is implicitly around every function application
04:00 ozzloy: oic
04:00 univyrse: dont you people sleep
04:00 Sgeo: (+ 1 2) is really (#%app + 1 2)
04:00 ozzloy is reading reference page with #%app now
04:01 univyrse: i am drunk and i am going to sleep now
04:01 ozzloy: does (#%app + 1 2) turn into (#%app #%app + 1 2) ?
04:01 (part) univyrse: "Leaving"
04:01 ozzloy: gnight u<tab>
04:01 Sgeo: ozzloy, #%app isn't a function, it's a macro
04:01 Sgeo: And a special form
04:03 ozzloy: what's a special form?  manual search procudes typed racket docs
04:03 Sgeo: Um, terminology in other Lisps
04:04 Sgeo: By "special form", I meant form recognized by whatever is processing the language as a primitive form
04:04 Sgeo: As in, not a macro (and not a function application)
04:04 ozzloy: oic
04:04 ozzloy: i didn't know there was anything other than applications and macros
04:05 (join) bitonic
04:05 ozzloy: macros that get expanded to applications
04:07 Sgeo: Think about if
04:07 ozzloy: oh wait
04:07 ozzloy: yeah
04:07 ozzloy: haha, that's what i was just going to say
04:07 ozzloy: i remember now how if's args aren't all evald
04:07 ozzloy: critically not
04:07 ozzloy: because recursion
04:08 Sgeo: ozzloy, it's not just that though. Macros don't have to expand their args. But there needs to be some fundamental way to make a choice
04:08 Sgeo: Hmm, actually, I might be wrong about that
04:08 Sgeo: But in any case, as it happens, if is in fact a special form in Racket, rather than just a macor
04:08 Sgeo: *macro
04:09 ozzloy: good times
04:09 ozzloy: i need to finish sicp
04:09 ozzloy: clearly
04:10 ozzloy: that will help me with stuff like this, i think
04:11 Sgeo: Well, #%app is Racket-specific
04:11 Sgeo: And I'm not sure if SICP discusses macros
04:12 Sgeo: Should at least be discussion of function application vs special forms though, I would... guess. I haven't actually read SICP
04:12 ozzloy: i think it goes over normal versus applicative order
04:12 ozzloy: which i vaguely remember as being relevant to This-Sort-Of-Thing
04:36 Sgeo: rudybot, help
04:36 rudybot: Sgeo: bug <more> ..., help [<what>], version, quote, source, url, seen <nick>, uptime, t8 <from> <to> <text> ..., init [<lang>], eval <expr> ..., give <to> <expr> ..., apropos <str> ..., desc <id>, doc <id>
04:36 Sgeo: rudybot, help give
04:36 rudybot: Sgeo: give <to> <expr> ...: evaluate and give someone the result
04:36 Sgeo: rudybot, give Sgeo (+ 1 1)
04:36 rudybot: Sgeo: error: Talk to yourself much too?
04:37 Sgeo: rudybot, give rudybot (+ 1 1)
04:37 rudybot: Sgeo: error: I'm full, thanks.
04:37 Sgeo: rudybot, give Sgeo_ (+ 1 1)
04:37 rudybot: Sgeo_: Sgeo has given you a value, say "rudybot: eval (GRAB)" to get it (case sensitive)
04:37 (nick) Sgeo -> Sgeo_
04:37 Sgeo_: rudybot, (* 2 (grab))
04:37 rudybot: Sgeo_: your sandbox is ready
04:37 rudybot: Sgeo_: error: reference to an identifier before its definition: grab in module: 'program
04:37 Sgeo_: rudybot, (* 2 (GRAB))
04:37 rudybot: Sgeo_: ; Value: 4
04:37 Sgeo_: rudybot, (GRAB)
04:37 rudybot: Sgeo_: ; Value: 2
04:37 Sgeo_: Cool
04:52 (quit) ormaaj: Ping timeout: 264 seconds
04:54 (join) kofno
05:01 (quit) Nisstyre-laptop: Quit: Leaving
05:02 (join) Nisstyre-laptop
05:09 (join) ormaaj
05:19 asumu: SICP discusses macros but not in a particularly useful sense.
05:21 asumu: PLAI has a brief but good discussion of macros.
05:23 (join) hash_table
05:38 (quit) Nisstyre-laptop: Quit: Leaving
05:39 Sgeo: Ok, the whole module+ test thing is going to mean that I'm actually likely to write tests in this language, unlike any other language I worked with, I think
05:39 (quit) ormaaj: Ping timeout: 255 seconds
05:40 (join) add^_
05:54 (join) ormaaj
06:14 (join) noam_
06:23 (join) noam__
06:26 (quit) noam_: Ping timeout: 265 seconds
06:33 (quit) hash_table: Ping timeout: 244 seconds
06:40 (join) jeapostrophe
06:52 (quit) serhart: Quit: Leaving.
06:59 (join) MayDaniel
07:00 (quit) cdidd: Remote host closed the connection
07:01 (join) masm
07:19 (join) cdidd
07:20 (join) serhart
07:21 (quit) serhart: Read error: Connection reset by peer
07:21 (join) serhart
07:26 (join) serhart1
07:26 (quit) serhart: Ping timeout: 255 seconds
07:55 (quit) bartbes__: Ping timeout: 264 seconds
07:55 (join) bartbes_
07:59 (quit) serhart1: Quit: Leaving.
07:59 (join) soegaard
08:05 (quit) cdidd: Remote host closed the connection
08:34 (quit) bitonic: Ping timeout: 255 seconds
08:47 (quit) sethalves: Quit: ERC Version 5.3 (IRC client for Emacs)
08:48 (join) sethalves
08:50 (quit) ormaaj: Read error: Operation timed out
08:54 (quit) soegaard: Quit: soegaard
08:54 (join) bitonic
09:05 (join) ormaaj
09:21 (join) nejucomo
09:24 (join) mizu_no_oto
09:29 (quit) bitonic: Remote host closed the connection
09:34 (nick) sethalves -> sethAway
09:37 (quit) Nisstyre: Quit: Leaving
09:41 (join) Shviller
09:54 (join) Anonymous6708067
09:54 Anonymous6708067: Hello everyone!
09:54 (nick) Anonymous6708067 -> Anonymous708980
09:55 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
10:01 (join) anonymou1
10:01 (nick) anonymou1 -> Anonymous7898789
10:11 (join) mizu_no_oto
10:29 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
10:34 (part) nejucomo
10:45 bremner: eli: I'm just figuring out pl.plt for the coming term.  I guess I can probably reverse engineer collects/pl from the pl.plt you distribute?
10:47 (quit) spiderweb: Quit: ERC Version 5.3 (IRC client for Emacs)
10:49 (quit) kofno: Remote host closed the connection
10:49 (join) kofno
10:49 (join) cdidd
10:52 (quit) kofno: Remote host closed the connection
11:03 greghendershott: ICYI you can tweet Racket #code2012. http://www.ioncannon.net/projects/code2012/
11:04 greghendershott: https://twitter.com/search?q=%23code2012
11:17 (join) mithos28
11:18 (join) Fare
11:19 (join) anRch
11:23 (quit) cdidd: Remote host closed the connection
11:25 (join) cdidd
11:29 (quit) cdidd: Read error: Connection reset by peer
11:29 (quit) jeapostrophe: Ping timeout: 255 seconds
11:29 (join) RacketCommitBot
11:29 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/HCCXgQ
11:29 RacketCommitBot: racket/master 8f300b9 Robby Findler: add rotary dial to the image guide
11:29 (part) RacketCommitBot
11:57 (join) cdidd
12:09 (quit) anRch: Quit: anRch
12:09 (quit) cdidd: Remote host closed the connection
12:10 (join) Kaylin
12:15 (join) cdidd
12:21 (quit) cdidd: Ping timeout: 245 seconds
12:23 (join) Nisstyre-laptop
12:25 (join) cdidd
12:29 (quit) cdidd: Remote host closed the connection
12:47 mithos28: anyone on knowledgable about racket's optimizations? Optimization coach is not providing me with enough info.
12:48 (join) neilv
12:50 mithos28: https://gist.github.com/4421561 is my program, dequeue! is a lot slower than fast-dequeue! and the only difference is a set-node-value!
12:53 (join) RacketCommitBot
12:53 RacketCommitBot: [racket] plt pushed 2 new commits to master: http://git.io/1rM7JQ
12:53 RacketCommitBot: racket/master 2a83c21 Robby Findler: add some color to the image guide's fractals...
12:53 RacketCommitBot: racket/master a948f1b Robby Findler: fix commit c818eb3ca3325fb7eac7f2a8105ea0121013214c...
12:53 (part) RacketCommitBot
13:02 (join) francisl
13:03 mithos28: Fixed it, by keeping track of different things so that the set! was not necessary for space safety
13:12 (join) cdidd
13:13 (quit) francisl: Remote host closed the connection
13:13 (join) francisl
13:25 (quit) cdidd: Read error: Connection reset by peer
13:27 (join) cdidd
13:38 (join) mizu_no_oto
13:39 (join) jonrafkind
13:40 (quit) Kaylin: Read error: Connection reset by peer
13:41 (quit) cdidd: Ping timeout: 276 seconds
13:41 (join) cdidd
13:50 mithos28: wow contracts are expensive
13:52 Nisstyre-laptop: mithos28: well runtime type checking takes time
13:52 Nisstyre-laptop: mithos28: if you want type checking for everything use typed racket, or Haskell
13:53 mithos28: Nisstyre-laptop: it is a simple tag check that I am doing
13:53 (quit) cdidd: Ping timeout: 245 seconds
13:54 mithos28: (queue? any/c -> any) is much more expensive than (any/c any/c -> any) and queue? is a struct predicate
13:55 Nisstyre-laptop: mithos28: that seems obvious
13:56 mithos28: Really? I don't see why a single check would be as expensive as the entire operation I am doing
13:57 mithos28: well actually it is because the contract system is broken with respect to arity checking
13:59 Nisstyre-laptop: mithos28: that checking that it is *anything* is less time consuming than checking for a specific type seems obvious
13:59 mithos28: Nisstyre-laptop: its nat that it takes time, its the amount of time that it takes
13:59 Nisstyre-laptop: mithos28: what kind of struct are you making?
14:00 (quit) snorble_: Ping timeout: 255 seconds
14:00 mithos28: (struct queue (head tail) #:property prop:sequence (…))
14:01 Nisstyre-laptop: mithos28: I don't know why that would take a long time to check in a contract
14:01 mithos28: Nisstyre-laptop: Because chaperones are slow
14:02 mithos28: Which is my annoyance
14:02 mithos28: hand coding the check at the begining of the function, is about the speed I expected (lost in the noise)
14:02 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
14:04 (join) mizu_no_oto
14:04 (join) RacketCommitBot
14:04 RacketCommitBot: [racket] plt pushed 3 new commits to master: http://git.io/Mtc6Iw
14:04 RacketCommitBot: racket/master ba973a3 Matthew Flatt: add `configure' test for mmap() and mprotect()...
14:04 RacketCommitBot: racket/master dc9643a Matthew Flatt: fix typo in error message
14:04 RacketCommitBot: racket/master cdfc491 Matthew Flatt: racket/base: add exn:fail:filesystem:errno and exn:fail:network:errno...
14:04 (part) RacketCommitBot
14:06 Nisstyre-laptop: mithos28: okay, it will still be faster with typed racket though
14:06 Nisstyre-laptop: (or Haskell)
14:07 mithos28: Nisstyre-laptop: If I wanted to be using haskell I wouldn't be using racket
14:07 Nisstyre-laptop: mithos28: I know
14:07 Nisstyre-laptop: mithos28: Typed Racket is a useful choice though
14:07 Nisstyre-laptop: I'm not sure if you can mix normal Racket with it or not
14:08 mithos28: Nisstyre-laptop: I help maintain TR, I'm not using it because I know exactly why its contract generation will be worse than my manual contract generation
14:08 Nisstyre-laptop: mithos28: okay, fair enough
14:16 Sgeo: Hrm. Would a Tcl-like surface syntax for Racket make sense?
14:17 jonrafkind: are you looking for a non-sexpression syntax?
14:19 Sgeo: Yes. I'm aware that sweet exists, although couldn't get it to work at the DrRacket REPL
14:19 Sgeo: Only seemed to work in the module body
14:19 mithos28: Sgeo: have you looked at honu?
14:20 Sgeo: No
14:21 Sgeo: Actual Tcl as a Racket language would also be interesting, but it would likely be ... problematic
14:26 jonrafkind: honu is a javascript-like syntax that supports macros, i work on it
14:30 Sgeo: Cool
14:35 Sgeo: I wonder if it's a criticism of Racket that certain languages would not work well at all as a Racket language
14:35 Sgeo needs to think about it more
14:35 mithos28: What issues do you see arising?
14:36 Sgeo: Just the impedence mismatch between using macros to do metaprogramming and using eval to do metaprogramming
14:36 Sgeo: Tcl is largely centered around eval
14:37 Sgeo: Hmm, well, I guess all Tcl commands would be usable from the Racket side, even if not all of them made sense to use in that manner
14:37 mithos28: So, just use eval. What is the problem with that
14:37 Sgeo: Racket functions would be ... somewhat usable from the Tcl side
14:38 Sgeo: But Tcl commands intended to accept code would only be able to accept Tcl code, I ... think. Rather than being usable in a macro-y way
14:38 Sgeo: And couldn't really use Racket macros comfortably
14:39 mithos28: So you are saying that embededing TCL in racket would be easy, but using racket from Tcl would be hard.
14:39 mithos28: I think that speaks about Tcl more than racket
14:40 (join) tfb
14:40 Sgeo: Actually, wait. More problems. Everything in Tcl is fundamentally a string (from the code's point of view). So passing a string to a Racket function that, say, expects a number would be "fun"
14:43 Sgeo: I should port Factor's EBNF stuff to Racket
14:46 Sgeo: I just remembered that Racket does have parsing stuff already
14:59 (join) dyoo
15:03 dyoo: Sgeo: just checking: I'm developing an EBNF-style tool.  Not released yet, but see: http://hashcollision.org/ragg
15:04 dyoo: Is this something that is useful to you?
15:04 dyoo: (currently working on test cases, documentation, error handling...)
15:06 neilv: sgeo: i think an interesting exercise would be to implement tcl in an optimized way in racket: so that lots of evaluation happens with native racket code, and you only go to tcl string expansion evaluation when necessary
15:08 neilv: Sgeo: although, tcl is dead.  it was popular on unix for a while, due to tk. then it had brief popularity for web server side, as a saner alternative to perl, and then (iirc) sun hired ousterhout and tcl vanished
15:08 asumu: Sgeo: only reason why the sweet package doesn't support the REPL is that I don't know how.
15:08 asumu: Anyone know how to hook into #%top-interaction for non-sexp languages?
15:09 dyoo: I think this is something that jpolitz and co were also fighting with regards to pyret
15:09 mithos28: asumu: You don't want #%top-interaction
15:09 mithos28: you want to hook into the reader
15:09 dyoo: They wanted to get the python grammar to work with the repl, and fought with drracket's detection of end-of-expression
15:09 dyoo: I don't know if they resolved that problem already.
15:09 asumu: mithos28: ah, ok. BTW, did you ping me the other day?
15:10 asumu: dyoo: hmm, ok, so sounds like further work is needed to make this an easy thing to implement.
15:10 mithos28: asumu: Yes, you might want to look at port-read-handler
15:10 mithos28: are you the current maintainer of pfds?
15:10 asumu: Yes, though I've only done some minimal maintenance.
15:11 mithos28: ok, because it currently has issues with typechecking
15:12 (quit) francisl: Ping timeout: 255 seconds
15:12 asumu: Hmm, drats. Is that on 5.3.1 or HEAD?
15:12 mithos28: HEAD, about a week old now I'd guess
15:12 asumu: Okay, that's odd and I'll look into it.
15:14 mithos28: Is there a way to require a submodule from the command line
15:15 mithos28: raco -s is giving me issues
15:15 mithos28: or is using -e the best option
15:16 (join) apeman_
15:16 apeman_: in GRacket REPL, how do you get previous line?
15:17 asumu: mithos28: raco test can do it.
15:17 (quit) tfb: Quit: sleeping
15:17 mithos28: asumu: Sorry thats what I meant was having issues: raco test -s
15:17 mithos28: it kills my gui
15:18 mithos28: apeman_: Control-P I believe
15:18 asumu: Oh, that's annoying. Yeah, I don't know then.
15:18 (join) anRch
15:19 asumu: Hmm, dammit. The pfds thing probably has something to do with type variable scope.
15:20 asumu: Actually, that's just because my git repo has weird commits in it. *switches to clean branch*
15:27 mithos28: rackets inliner is pretty bad
15:29 Sgeo: asumu, the racklog language does it somehow
15:29 Sgeo: Don't know if there may be other considerations
15:31 asumu: Looks like racklog does it with a DrRacket info hook.
15:31 mithos28: procedure-rename makes my program go slower
15:49 (join) DarkTiger
15:49 DarkTiger: http://paste.lisp.org/display/134359 — why does this give an ‘unbound identifier’ error?  The docs at http://pre.racket-lang.org/docs/html/reference/match.html#(form._((lib._racket/match..rkt)._define/match)) seem to indicate that define/match should be available from the racket/match library.
15:51 asumu: DarkTiger: Are you using 5.3.1?
15:51 (join) mye
15:51 DarkTiger: Yes
15:53 asumu: DarkTiger: the pre-release docs are for the pre-release builds. 5.3.1 doesn't have `define/match`, but the next release will.
15:54 (quit) mye: Client Quit
15:54 (join) tfb
15:54 DarkTiger: Oops.  Thanks.
15:55 DarkTiger: I blame Google!
15:57 greghendershott: DarkTiger: I've done that myself. However if you want to try a nightly build of the prerelease, they're at http://pre.racket-lang.org/installers/
15:58 greghendershott: In which case you _do_ want http://pre.racket-lang.org/docs/* instead of http://docs.racket-lang.org/*  :)
16:01 asumu: Ah, pfds is broken because it relied on broken scoping rules. Yay...
16:03 (quit) neilv: Quit: Leaving
16:04 (quit) mizu_no_oto: Quit: ["Textual IRC Client: www.textualapp.com"]
16:04 DarkTiger: greghendershott: Thanks — might do that
16:05 ozzloy: asumu, thanks
16:05 ozzloy: Sgeo, module+ test ftw!
16:06 ozzloy: Sgeo, being able to write tests just above the tested procedure is surprisingly encouraging for me
16:07 greghendershott: mithos28: Re the slow contract, I'm guessing (any/c -> any) is optimized to a no-op, and  only with ("not any/c" -> any) do you incur the real overhead?
16:07 (quit) sirdancealot: Quit: KEEP SPARKS. FLAME AWAY.
16:08 mithos28: greghendershott: Yes, I realized that. It is because contracts are broken with respect to arity checking
16:08 (join) sirdancealot
16:09 greghendershott: mithos28: FWIW I've experimented with having the contract declaration syntax expand to simply _be_ something like handwritten checks: https://github.com/greghendershott/termsheet
16:09 mithos28: If I call a function with a contract (any/c -> any), with 2 arguments I 'should' get a contract violation
16:10 greghendershott: Oh? Huh.
16:10 mithos28: The should there, is in the theory of contracts
16:10 (join) mizu_no_oto
16:10 mithos28: but in the implementation you get a primitive arity error
16:11 greghendershott: Hmm I'm getting a contract violation for that
16:12 greghendershott: e.g. (define/contract (f x) (any/c . -> . any) x)
16:12 greghendershott: (f 0) => 0
16:12 greghendershott: (f 0 0) => "f: arity mismatch; …."
16:12 greghendershott: Oh duh I see what you're saying.
16:13 (quit) mizu_no_oto: Client Quit
16:15 asumu: mithos28: pfds should be fixed now
16:15 mithos28: asumu: Thanks
16:15 (join) mye
16:16 mithos28: anyone know of a good concurrent queue benchmark?
16:18 (quit) dyoo: Quit: dyoo
16:20 (quit) tfb: Quit: sleeping
16:25 mithos28: Well my queues perform a lot worse than thread unsafe queues, and then I added an operation that took 1ms per queue operation and they are near indistinquishable
16:25 mithos28: so if you are doing real work and not a micro benchmark they are great
16:27 (quit) jonrafkind: Ping timeout: 245 seconds
16:34 (quit) anRch: Quit: anRch
16:34 (quit) mye: Quit: mye
16:35 (join) mye
16:41 (quit) mithos28: Quit: mithos28
16:42 (join) mithos28
16:44 (quit) mithos28: Client Quit
16:50 (quit) Nisstyre-laptop: Read error: Connection reset by peer
17:10 (join) jonrafkind
17:15 (join) dyoo
17:19 (quit) dyoo: Read error: Connection reset by peer
17:21 (join) RacketCommitBot
17:21 RacketCommitBot: [racket] plt pushed 3 new commits to master: http://git.io/wX7Miw
17:21 RacketCommitBot: racket/master e06f31c Neil Toronto: More precise return type for `magnitude' (it always returns nonnegative)
17:21 RacketCommitBot: racket/master f5fa935 Neil Toronto: Moar `math/matrix' review/refactoring...
17:21 RacketCommitBot: racket/master c8e3b45 Neil Toronto: Added QR decomposition tests (mostly a pretty thorough randomized test);...
17:21 (part) RacketCommitBot
17:30 (join) tfb
17:38 (join) neilv
17:46 (quit) apeman_: *.net *.split
17:46 (quit) mceier: *.net *.split
17:46 (quit) SeanTAllen: *.net *.split
17:46 (quit) SHODAN: *.net *.split
17:46 (quit) wkelly: *.net *.split
17:46 (join) SHODAN
17:46 (join) wkelly
17:46 (join) mceier
17:46 (quit) tfb: Quit: sleeping
17:46 (join) apeman_
17:47 (join) SeanTAllen
17:47 (quit) myx: Ping timeout: 255 seconds
17:51 (join) Nisstyre-laptop
17:53 (quit) DarkTiger: Quit: Lost terminal
17:55 apeman_: in scheme, a web page claims that (if uses normal order evaluation. Is this technically accurate, since (if short circuits?
17:56 jonrafkind: if has its own evaluation rules that are neither normal nor lazy
17:59 (join) jeapostrophe
17:59 (quit) jeapostrophe: Changing host
17:59 (join) jeapostrophe
17:59 neilv: in schemes, short-circuit evaluation is done with special syntax, not with argument order of evaluation rules
18:13 (join) mithos28
18:21 (quit) neilv: Quit: Leaving
18:33 (join) apeman__
18:35 (quit) apeman_: Ping timeout: 265 seconds
18:35 (nick) apeman__ -> apeman_
18:45 (quit) mithos28: Quit: mithos28
18:46 (join) mithos28
19:00 (join) netrino
19:01 (quit) mye: Quit: mye
19:04 (quit) mithos28: Quit: mithos28
19:19 (join) mye
19:26 (quit) Nisstyre-laptop: Quit: Leaving
19:28 (quit) add^_: Quit: The Garbage Collector got me...
19:30 (join) RacketCommitBot
19:30 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/iXmmug
19:30 RacketCommitBot: racket/master 8e8cc30 Matthew Flatt: fix dist spec
19:30 (part) RacketCommitBot
19:37 (join) mithos28
20:01 (quit) mithos28: Quit: mithos28
20:09 (join) mye_
20:11 (quit) mye: Ping timeout: 240 seconds
20:11 (nick) mye_ -> mye
20:19 (quit) jonrafkind: Ping timeout: 255 seconds
20:42 (join) mithos28
20:50 (quit) MayDaniel: Read error: Connection reset by peer
21:37 asumu: It's neat when job postings mention Racket experience as a plus: http://www.linkedin.com/jobs?viewJob=&jobId=4483494&srchIndex=4
21:42 (quit) masm: Remote host closed the connection
21:52 mithos28: yay segfault
22:33 chandler: asumu: It doesn't seem like they're actually using it, though.
22:36 (quit) jeapostrophe: Ping timeout: 244 seconds
23:10 (quit) mithos28: Quit: mithos28
23:21 (quit) Shviller: Ping timeout: 260 seconds
23:22 (join) Shviller
23:41 (join) mithos28
23:52 (quit) Fare: Ping timeout: 244 seconds