00:02 (quit) bremner_: Quit: ZNC - http://znc.sourceforge.net
00:02 (join) ckrailo
00:02 (join) bremner_
00:03 (join) realitygrill_
00:07 (quit) realitygrill: Ping timeout: 260 seconds
00:08 (nick) realitygrill_ -> realitygrill
00:13 (quit) ray`: Remote host closed the connection
00:18 (join) Nanakhiel
00:21 Demosthenes: yoklov: hey, i make a sample parser after our talk
00:21 (quit) Lajla: Ping timeout: 260 seconds
00:24 (quit) Nanakhiel: Ping timeout: 258 seconds
00:37 (join) Lajla
00:43 (quit) yoklov: Quit: Leaving.
01:46 EM03: Hello
02:25 (quit) jonrafkind: Ping timeout: 246 seconds
02:25 (quit) ckrailo: Quit: Computer has gone to sleep.
02:35 (join) ckrailo
03:54 (quit) dnolen: Quit: dnolen
04:01 (quit) realitygrill: Ping timeout: 240 seconds
04:09 (join) leo2007
04:51 (quit) polypus: Remote host closed the connection
05:01 (quit) misterm: Read error: Operation timed out
05:16 (quit) EM03: Quit: EM03
05:29 (quit) leo2007: Ping timeout: 260 seconds
05:36 (join) leo2007
05:36 (join) mceier
06:02 (quit) rekahsoft: Ping timeout: 258 seconds
06:04 (join) Sisi
06:06 Sisi: hi, there is someone who used insert-image in class editor-canvas%? I have problem with gif
06:06 (join) MayDaniel
06:07 (quit) qha: Ping timeout: 240 seconds
06:27 (join) masm
06:29 (join) PLT_Notify
06:29 PLT_Notify: racket: master Casey Klein * dd58b45 (2 files in 2 dirs): Fixes domain checking of unioned reduction relations - http://bit.ly/kNXE57
06:29 (part) PLT_Notify
06:33 (quit) ckrailo: Quit: Computer has gone to sleep.
07:13 (join) tauntaun
07:21 (join) PLT_Notify
07:21 PLT_Notify: racket: master Casey Klein * cafcb47 (2 files in 1 dirs): Refactored reduction-relation-procs - http://bit.ly/iKQ66b
07:21 (part) PLT_Notify
07:39 (join) qha
07:40 (quit) Sisi: Ping timeout: 252 seconds
07:41 (quit) mceier: Quit: leaving
08:02 (quit) leo2007: Quit: rcirc on GNU Emacs 23.3.50.1
08:04 (join) lucian
08:26 (join) PLT_Notify
08:26 PLT_Notify: racket: master Matthew Flatt * 153d23e (1 files in 1 dirs): adjust sandbox memory-limit test timeout
08:26 PLT_Notify: racket: master Matthew Flatt * 0119365 (13 files in 5 dirs): internal reorganization of bytecode compiler and other bytecode tasks ...
08:26 PLT_Notify: racket: master Matthew Flatt * 09046b4 (31 files in 2 dirs): split up mzmark output
08:26 PLT_Notify: racket: master Matthew Flatt * 6950396 (6 files in 1 dirs): fixes for non-JIT build
08:26 PLT_Notify: racket: master commits cafcb47...6950396 - http://bit.ly/jTVREt
08:26 (part) PLT_Notify
08:43 (quit) lucian: Read error: Operation timed out
08:44 (join) lucian
08:58 (join) leo2007
09:02 (quit) MayDaniel: Read error: Connection reset by peer
09:05 (join) MayDaniel
09:12 (join) anRch
09:27 (quit) MayDaniel: Read error: Connection reset by peer
09:33 (join) mithos28
09:37 (quit) tauntaun: Quit: Ex-Chat
09:45 (join) rpr
09:46 rpr: Where does one report bugs?  e.g.  in-list is there but in-hash is missing in typed/racket.
09:52 (quit) mithos28: Quit: mithos28
09:59 rpr: And I don't think hash-ref sigs correct.
10:01 (join) PLT_Notify
10:01 PLT_Notify: racket: master Matthew Flatt * 9188e36 (2 files in 2 dirs): try to fix thread-local variables for Max OS X Lion - http://bit.ly/mqidxh
10:01 (part) PLT_Notify
10:10 (part) Lajla: "I am gay for you"
10:19 (join) mithos28
10:22 mithos28: rpr: There is a menu item 'Submit Bug Report...' under the help menu in DrRacket.
10:24 (join) yoklov
10:41 (quit) mithos28: Quit: mithos28
10:44 (join) lucian_
10:44 (join) mithos28
10:44 (quit) lucian: Disconnected by services
10:45 (nick) lucian_ -> lucian
10:45 (nick) lucian -> lucian_
10:45 (nick) lucian_ -> lucian
10:50 rpr: Ahh... There be the problem.  Not using DrRacket.  I'll fire it up.
10:57 (quit) anRch: Quit: anRch
11:02 (join) tauntaun
11:06 rpr: If there is a type signature for 'in-hash' which works with 'for/fold' it is well obscured.
11:17 (join) Lectus
11:23 (join) realitygrill
11:23 mithos28: rpr: I think it just isn't supported yet
11:28 (join) realitygrill_
11:29 (quit) realitygrill: Ping timeout: 260 seconds
11:29 (nick) realitygrill_ -> realitygrill
11:40 (join) dnolen
11:42 (quit) Lectus: Quit: Ex-Chat
11:49 (join) jonrafkind
12:02 (quit) realitygrill: Quit: realitygrill
12:14 (quit) lucian: Remote host closed the connection
12:20 (quit) leo2007: Ping timeout: 260 seconds
12:21 (quit) yoklov: Quit: Leaving.
12:23 (join) realitygrill
12:25 (join) MayDaniel
12:40 (join) yoklov
12:45 (join) leo2007
13:09 (quit) realitygrill: Quit: realitygrill
13:09 (quit) tauntaun: Quit: Ex-Chat
13:16 rpr: I'm getting slaughtered here.  Seems optional args don't work either.
13:18 jonrafkind: rudybot: (let () (define (x [y 5]) (+ y 2)) (list (x) (x 10))
13:18 rudybot: jonrafkind: Y OCTAVE NO WORK
13:19 jonrafkind: rudybot: (let () (define (x [y 5]) (+ y 2)) (list (x) (x 10)))
13:19 rudybot: jonrafkind: ; Value: (7 12)
13:19 jonrafkind: rudybot: (let () (define (x #:y [y 5] #:z z) (+ y z)) (list (x #:z 9) (x #:z 10 #:y 1)))
13:19 rudybot: jonrafkind: ; Value: (14 11)
13:31 (nick) samth_away -> samth
13:34 (quit) Demosthenes: Ping timeout: 240 seconds
13:35 (join) Demosthenes
13:45 rpr: Just typed R6RS bytevectors.
13:47 (join) realitygrill
14:01 (join) tauntaun
14:02 (quit) realitygrill: Quit: realitygrill
14:09 rpr: HaHa take that type checker. https://gist.github.com/961546  Took the entire morning...
14:15 (quit) MayDaniel: Read error: Connection reset by peer
14:18 samth: rudybot, init typed/racket
14:18 rudybot: samth: error: typed/racket: standard-module-name-resolver: collection not found: "typed" in any of: (#<path:/usr/local/racket-textual/collects>) in: typed/racket
14:18 (quit) yoklov: Quit: Leaving.
14:18 samth slaps rudybot
14:19 samth: rpr, some suggestions:
14:19 samth: 1. put your code on planet
14:20 samth: 2. instead of the macros for fx=?, just do (define fx=? fx=)
14:21 samth: also, you can do (rename-in racket/fixnum [fx= fx=?])
14:21 samth: in the require
14:21 rpr: Thanks.
14:22 samth: 3. use internal define instead of (let ([x (lambda ...)]) ...)
14:23 samth: 4. use `for' and its variants instead of `do'
14:23 rpr: I've having a bit of trouble getting for/xxx comprehensions past the type checker.
14:23 samth: 5. use `let*' instead of nested `let'
14:24 rpr: Specifically the 'in-hash'
14:24 samth: rpr, some of them might not work in all cases -- it turns out to be hard
14:24 samth: does it work if you just leave out the `in-hash'?
14:24 rpr: For non regular sequences yes.
14:25 rpr: Sorry for non- in-hash sequences.
14:25 rpr: Ouch.  error-display-handler is not typed in typed/racket.
14:26 samth: rpr, you can use `require/typed' even with identifiers from `racket/base'
14:26 samth: but i'll fix that
14:26 rpr: The big one that killed me this morning is typing optional args.
14:27 samth: rpr, yes, that's an unfortunate current limitation
14:27 rpr: samth: BTW FWIW I thinked typed/racket is the best thing I've seen in years.
14:27 samth: rpr, thanks!
14:28 rpr: Very very cool.
14:28 samth: rpr, this seems to work for hashes: (for ([(k v) #hash((1 . 2))]) (display k))
14:30 rpr: I tried to poke into it I think it has to do with in-hash and values.
14:31 rpr: (All (a b) (Sequenceof (values (-> a) (-> b)))))
14:31 rpr: sorta kinda
14:32 rpr: Other then in-hash the other for/xxx stuff has been ok.
14:34 samth: in-hash should have the type (All (a b) (Sequenceof a b))
14:35 rpr: tried that and similar :(
14:36 rpr: https://gist.github.com/961579
14:36 rpr: Line 108 the for/fold seems to depend on (values k v) down deep.
14:37 rpr: https://gist.github.com/961575
14:38 rpr: Other stuff is going pretty well.  (Embarrassed to say I've forked the source files.)
14:46 samth: rpr, you can't make `in-hash' typecheck
14:46 samth: so don't feel bad that it didn't work
14:46 samth: it requires deep black magic
14:47 rpr: yea sorta figured.
14:48 rpr: error-display-handler would be nice to get on the list.
14:48 samth: error-display-handler you should be able to do w/ require/typed
14:48 samth: but i'll fix it
14:49 rpr: I'm going to do that now.  But I have it _everywhere_.
14:50 mithos28: samth: error-display-handler requires a type for a function which can return any number of values. How is that represented in TR?
14:50 samth: hmm, that'll be hard to do with `require/typed'
14:51 mithos28: can it currently be done on the internal side?
14:53 rpr: I could be just "doing it wrong"  But I've got quite a few (with-handlers ...) and I often dump out the error with ((error-display-handler) "Oh crap" ex)
14:53 samth: mithos28, i think so
14:53 samth: the problem is generating contracts for it
14:54 samth: rpr, that seems like a reasonable thing to do
14:55 rpr: FWIW 80% of the code behind http://goo.gl/tPPJz is Racket.  I intend to type the whole lot of it.
14:56 DT``: samth, why is a for loop better than a do loop using unsafe-c[ad]r?
14:58 rpr: Style thing.  Functional comprehensions vs. looping constructs.
14:58 samth: DT``, what rpr said, plus it's usually easier to read, and it does the unsafe optimization for your
14:58 samth: and it's more extensible
14:58 (join) realitygrill
14:58 samth: rpr, awesome!
14:59 DT``: I see, thank you!
15:10 samth: rpr, once my build finishes, i'll push those fixes
15:21 (quit) mithos28: Quit: mithos28
15:23 rpr: samth: ok.  If I get this latest module to type check I'm hitting the pool.  But I'll pull the latest from git tonight.
15:24 (join) mithos28
15:37 (join) PLT_Notify
15:37 PLT_Notify: racket: master Sam Tobin-Hochstadt * 2f3e9fc (2 files in 2 dirs): Add types for `in-hash' etc.
15:37 PLT_Notify: racket: master Sam Tobin-Hochstadt * 27aa3f7 (1 files in 1 dirs): Add type for `error-display-handler'
15:37 PLT_Notify: racket: master commits 9188e36...27aa3f7 - http://bit.ly/inKgbq
15:37 (part) PLT_Notify
15:42 rpr: samth: Thx.  I'm gone swimming.
15:42 (part) rpr: "ERC Version 5.3 (IRC client for Emacs)"
16:00 (quit) leo2007: Ping timeout: 248 seconds
16:16 tauntaun: Is there a roughe estimate of how many times faster the JVM is than the Racket VM?
16:16 tauntaun: s/roughe/rough
16:17 jonrafkind: theres the language shootout
16:19 samth: tauntaun, it strongly depends on what you're doing
16:20 jonrafkind: some guy posted on the list that method dispatch in the class system in racket is like 2 orders of magnitude slower than the jvm
16:20 tauntaun: samth: yes, I figured.  But I thought maybe there was a meaningful average out there.
16:20 samth: jonrafkind, that's tony garnock-jones, a grad student here
16:20 samth: tauntaun, i don't think there's a meaningful way to average these things
16:22 samth: tauntaun, you can look at this: http://shootout.alioth.debian.org/u32/benchmark.php?test=all&lang=java&lang2=racket
16:22 rudybot: http://tinyurl.com/44rg8jc
16:22 tauntaun: samth: ok, thanks.
16:23 samth: but as you can see from that chart, it strongly depends on the program
16:23 jonrafkind: are most of those benchmarks numerical?
16:25 samth: jonrafkind, they vary -- strings, numbers, binary trees, a few other things
16:25 (quit) Demosthenes: Quit: Lost terminal
16:25 jonrafkind: it would be nice if there was a class system based test
16:33 samth: jonrafkind, that wouldn't make much sense in Haskell or C
16:34 jonrafkind: then there would be no benchmark for those languages
16:34 tauntaun: FWIW, Haskell's type classes have "methods".
16:36 (quit) ravi_: Ping timeout: 248 seconds
16:37 bremner_: huh. I never knew that terminology
16:50 samth: jonrafkind, the point is that the benchmarks are supposed to be about algorithms, not coding styles
16:51 (join) teliti
16:52 teliti: Hello. I just installed racket on ArchLinux. I was astonished just to see a shell. Is racket for Linux without gui?
16:53 bremner_: teliti: try "drracket", it works fine in linux
16:53 teliti: bremner: thank you :-)
16:54 (quit) teliti: Client Quit
16:58 (join) ravi_
17:03 (quit) ravi_: Ping timeout: 252 seconds
17:05 (join) misterm
17:07 (quit) mithos28: Quit: mithos28
17:07 bremner_: no problem
17:29 (join) mithos28
17:50 (join) rotty
17:54 rotty: hi folks. is there a way (that is not manual pruning) to get a minimal install of racket? I'd just need Core + R6RS + SRFIs.
17:55 jonrafkind: I guess racket textual is the closest you canc ome
17:55 rotty: (any superset of that would be OK as well, but all the GUI stuff is totally superfluous for this installation)
17:56 jonrafkind: but at this point it doesnt seem to mean much since all the gui stuff is in racket anyway
17:56 jonrafkind: http://download.racket-lang.org/racket-textual-v5.1.1.html
17:56 rotty: oh, I want to build myself
17:57 rotty: ah, right there's a source package as well; can I somehow get to that from the git checkout?
17:57 jonrafkind: is that a constraint or just a fact
17:58 jonrafkind: well you can get essentially the source package if you omit the tests
17:58 rotty: I need to build for Debian Lenny, i386
17:59 jonrafkind: so why dont you want to download the 5.1.1 unix source package?
17:59 rotty: I have already done a build of 5.1.1 via git, but the resulting install is much larger than I'd like.
18:00 rotty: jonrafkind: since, if I have to fix/tweak something, I'd like to be able to switch to the latest release, hack on it, and do a minimal install for the target machine.
18:00 rotty: s/latest release/latest git master/
18:00 jonrafkind: im not sure of an easy way to do that, for now you could find the largest directories in collects/ and erase the biggest ones
18:01 jonrafkind: i can probably tell you which directories you dont need
18:01 jonrafkind: in the future there will be some distribution system that will make something like what you want (core racket, gui racket, etc)
18:01 rotty: how is this "textual" tarball created? I guess that's not a manual process?
18:02 jonrafkind: well its probably scripted somewehre, let me check the racket scripts
18:02 rotty: I'd also be OK with being able to create a stripped-down tarball myself from git.
18:02 rotty: thanks.
18:03 jonrafkind: hm I dont see the scripts in plt repository, maybe its just on eli's machine
18:03 rotty: yeah, I'm looking forward to that (more modular racket distribution).
18:04 rotty: well, then, for now, I'll strip down the installation manually.
18:04 jonrafkind: do you have a target for how small you want the installation?
18:04 jonrafkind: in terms of megabytes
18:05 rotty: not really, but significantly below 100MB would be nice.
18:05 rotty: (it's shared, hosted server, and I don't want to waste disk space if I can easily avoid it)
18:05 jonrafkind: ok
18:06 jonrafkind: for even more space savings, in theory you can erase the .rkt files after compiling to .zos
18:06 jonrafkind: I wonder if the racket vm could support gzipping .zo files too..
18:07 (join) lucian
18:07 bremner_: they are not already? I thought that was what zlib was used for
18:07 bremner_: just a vague idea I had
18:08 jonrafkind: oh I dunno, running 'file' on the .zo doesn't say 'gzipped data'
18:09 rotty: I guess I can drop the "scribblings" collection?
18:09 samth: rotty, that's all the documentation
18:09 rotty: (plus "tests", "redex", "games")
18:10 rotty: yeah, I don't need documentation there.
18:10 bremner_: rotty: how much does doing "raco setup --no-docs" help?
18:10 samth: rotty, the distribution scripts are in collects/meta/dist-specs.rkt
18:10 samth: that'll tell you what's in textual
18:10 (nick) samth -> samth_away
18:11 rotty: samth_away: thanks
18:11 bremner_: for me, collects is 50M after "raco setup --no-docs"
18:15 rotty: yeah, I should have done that, apparently.
18:18 rotty re-installs
18:19 bremner_: oops, let me double check that number
18:22 rotty: (I'd be surprised if .zo files were compressed -- that'd significantly affect startup time vs. just mmap-ing them)
18:22 jonrafkind: its true, but if you were ultra concerned about disk space usage it might be worthwhile
18:23 rotty: well, I'm certainly not ultra-concerned :-)
18:27 bremner_: Hmm I lied/was confused. It's already up to 73M and not finished yet.
18:28 (quit) em: Read error: Connection reset by peer
18:29 (join) emma
18:32 rotty: well, after eliminating the biggest obvious targets, I'm down to 107MiB; that's OK for starters
18:36 rotty: (still quite huge, if I might say so)
18:37 rotty: bremner_: was that 50MiB figure a full tarball you built from, or a "textual" install?
18:39 rotty: bremner_: nm, I didn't see your last message
18:49 bremner_: rotty: right, the correct number for no-docs is 132M
18:50 bremner_: that's built from (roughly) the unix-source tarball
18:51 (join) chturne
18:52 chturne: Are boxes specific to Racket? I thought they might have come from Scheme systems, but I can't find any documentation on why they're even used.
18:57 DT``: chturne, I'm pretty sure that's not Racket-specific, but I might be wrong.
18:57 DT``: they're useful if you want to pass something by reference.
18:58 chturne: DT``: I can't find the word "box" in the R5RS index
18:58 chturne: Do you know if they're not from Scheme?
18:59 DT``: I'm not sure if they're Racket-specific, sorry.
18:59 (join) rpr
19:00 chturne: I'll have to look at some examples in the Racket distribution to understand how they're used to pass references, that's currently confusing me a bit :-)
19:01 chturne: Thanks for your comments DT`` :-)
19:01 DT``: np.
19:02 DT``: chturne, btw, boxes are referenced here too: http://community.schemewiki.org/?scheme-faq-language
19:02 chturne: Hm, seems to be a non-standard extension then.
19:02 chturne: Thanks again!
19:08 (quit) qha: Ping timeout: 240 seconds
19:09 (join) PLT_Notify
19:09 PLT_Notify: racket: master Matthew Flatt * 02c2c17 (1 files in 1 dirs): Prevent #\uFFFF and #\uFFFE from being passed through to Pango ...
19:09 PLT_Notify: racket: master Matthew Flatt * db75ddd (1 files in 1 dirs): fix decompiler's listing of captured top- and module-level variables
19:09 PLT_Notify: racket: master Matthew Flatt * 979ea86 (2 files in 1 dirs): experiment (unsuccessfully) with inlining heuristic ...
19:09 PLT_Notify: racket: master commits 27aa3f7...979ea86 - http://bit.ly/jBD1in
19:09 (part) PLT_Notify
19:11 eli: rotty: What do you need this for?
19:27 (nick) emma -> em
19:30 (join) Checkie
19:46 rpr: Anyone know syntax for typed/racquet for a keyword proc.  Is it even currently possible?   e.g. (define connect (lambda/kw (#:key (user #f) (pass #f))))
19:47 (quit) mithos28: Quit: mithos28
19:54 eli: rpr: I don't think that it's possible yet.
19:54 eli: And it's "racket", BTW.
19:57 rpr: They changed the name _again_!!   Geeeez.
19:58 rpr: thx
19:59 bremner_: it never was racquet
20:00 bremner_: maybe you were making a joke, in which case, sorry
20:00 rpr: t'was a small one for sure (joke).
20:00 bremner_: fair enough.
20:05 jonrafkind: this is prison camp. no one can have any fun
20:16 rpr: Ok its giving me error msgs like it knows what its talking about so it might be in there already just not documented. In spgsql there is a class connection%  Anyone know how to require/typed it and then its method?  i.e. (send conn ...)
20:22 rpr: I suspect it is not possible.  This is a show stopper.
20:35 rpr: No, just tired.  Just need to wrap 1/2 a dozen or so method calls is all.
20:36 (quit) rpr: Remote host closed the connection
20:47 (quit) lucian: Remote host closed the connection
21:52 (quit) masm: Quit: Leaving.
22:00 (join) Lectus
22:55 elliottcable: Hey all.
22:55 elliottcable: Trying to build racket from source; using advice from http://lists.racket-lang.org/users/archive/2011-May/045388.html
22:57 elliottcable: hm. just answered my own question before I could ask it.
22:57 elliottcable: ignore me! :D
22:57 (quit) tauntaun: Quit: Ex-Chat
23:01 (join) ckrailo
23:20 (join) PLT_Notify
23:20 PLT_Notify: racket: master Robby Findler * e88f86e (0 files in 0 dirs): a start on the scribble/lncs language - http://bit.ly/knaxEv
23:20 (part) PLT_Notify
23:20 (join) mithos28
23:30 (join) PLT_Notify
23:30 PLT_Notify: racket: master Robby Findler * f60f234 (0 files in 0 dirs): added a drracket test - http://bit.ly/iGKkP6
23:30 (part) PLT_Notify
23:52 (quit) mithos28: Quit: mithos28