00:04 sheikra: I'm a little confused by the types. Why is the type of "1" :- One?
00:04 sheikra: and 2 :- Positive-Byte...
00:05 sheikra: and the type of any symbol the symbol itself ...
00:08 (quit) jonrafkind: Ping timeout: 240 seconds
00:20 (join) ckrailo
00:25 (join) mithos28
00:41 eli: sheikra: One is probably a recent addition due to the numeric tower work.
00:42 eli: As for symbols, you can "infer" either Symbol or the value -- it just happens that it's usually more useful to go for the latter.
00:42 eli: Unlike numbers.
00:42 eli: Except for 1, probably.
00:42 eli: Which is why they changed that, probably.
00:43 (nick) elliottcable -> paws
00:47 (nick) paws -> elliottcable
00:49 sheikra: eli: thanks. I just got back from lunch
00:50 sheikra: so is One a subtype of ...?
00:53 eli: sheikra: Of course.  It's still a Fixnum, Integer, Number, etc.
00:53 (nick) elliottcable -> everyone
00:53 (nick) everyone -> elliottcable
00:54 sheikra: eli: oh, the type One is "1" itself?
00:54 eli: elliottcable: Re the noise -- just to make sure, I really didn't mean to flame, but to find any kind of weak spots in the web page and documentation, so I really meant to ask those questions.
00:54 sheikra: It has only one element?
00:55 eli: sheikra: Yeah, that's my guess.
00:55 elliottcable: eli: hm?
00:55 elliottcable: eli: what noise?
00:55 eli: sheikra: Given that 1 is in it...
00:55 elliottcable: eli: the mailing list replies? Oh, I’ll get to them, I’m just not usually much of a mailing list person. I think that’s my … fifth mailing-list post *ever*.
00:55 sheikra: eli: hmm.. interesting
00:55 elliottcable: eli: were you one of the repliers?
00:55 eli: sheikra: You can also use `1' as a type (as well as any other literal), but for some types you get a different name, in case there are several options.
00:56 elliottcable: eli: if you want to get ahold of me faster, come hang in ##Paws. I’m usually there.
00:56 eli: sheikra: (BTW, I thought that you were @ brown...)
00:57 eli: elliottcable: Yeah, I had a long reply to your post, which might seem like a flame pile of rhetorical questions -- but no rhetorics intended.
00:57 eli: elliottcable: What's ##Paws?
00:58 elliottcable: eli: I think I might have already read that one. I read one or two replies before realizing there were like 10, and deciding “fuck this, I’ll come back later” :D
00:58 elliottcable: eli: my programming language. :3
00:59 elliottcable: but, honestly, we discuss language design in general (or have flame wars.) more often than anything in there
01:01 eli: elliottcable: What's special about it?
01:02 elliottcable: well, some random crap, like distributed object-space, and a fairly innovative interpreting preprocessor that pretty much leaves the syntax up to the user
01:03 elliottcable: eli: but the flagship feature is inherent asynchronicity via a sexy form of implicit and prevasive CPS
01:03 elliottcable: it was built to do the kinds of things Node.js does well, but better; and without all the damn typing of crazy nested explicit callbacks
01:05 elliottcable: oh, free concurrency.
01:05 (quit) yoklov: Quit: Leaving.
01:10 eli: elliottcable: My last impression of Node.js is a half-assed cover up for a lack of continuations and/or threads and such.  (It's a very vague impression.)
01:10 eli: What's with the syntax though?
01:10 eli: (And, any web pages?)
01:11 elliottcable: pretty much true, re: Node.js
01:11 elliottcable: but there’s benefits, primarily (IMO) the application of JavaScript to the problem
01:11 sheikra: eli: it seems that racket doesn't support overloading?
01:11 elliottcable: like I said, the syntax is non-existent. All that we define is a very minimal syntax for declaring scope and sub-expressions; everything else is defined in libspace with homoiconicity APIs available at preprocess-time
01:13 eli: elliottcable: Pointers?
01:13 eli: sheikra: What do you mean by overloading?
01:13 elliottcable: eli: hm?
01:14 sheikra: eli: I mean same-named functions dispatched by the types of their arguments. does typed racket support it?
01:15 eli: sheikra: You mean something like (U (A1 -> A2) (B1 -> B2))?
01:15 eli: elliottcable: ... for web pages that describe that.
01:15 elliottcable: ah, nope
01:16 elliottcable: still under dev.
01:16 elliottcable: Just a metric shitton of old code examples on gist.github, and the codebases in progress, and the knowledgebase of the people in the IRC channel (which is where everything meaningful is)
01:16 elliottcable: waiting to create a website and whatnot for when we have a finished product.
01:17 (join) no-name-
01:18 sheikra: eli: I mean (define f (lambda (x : Int) ...)) (define f (lambda (x : String)) ...) ... then (f 1) and (f "hi") dispatch to them respectively
01:18 (quit) mithos28: Quit: mithos28
01:19 sheikra: eli: I guess we are talking about the same thing..
01:19 sheikra: but I don't think they are unions
01:20 sheikra: (U (A1 -> A2) (B1 -> B2)) will not type check when applied on something of type A1
01:21 eli: sheikra: Ah.  You're talking about something that it will never do, since the very explicit design goal is to by a typed *racket* language, so it won't provide features that racket doesn't have.
01:21 eli: IOW, it doesn't support that, and will very likely continue to not do so...
01:23 eli: I thought that you were talking about making your own definition that dispatches on the type, and wanting that union type that doesn't quite work for reasons related to how a union cannot be a function (IIRC, a `case-lambda' type could be used in some cases).
01:24 (quit) no-name-: Ping timeout: 276 seconds
01:24 (join) no-name-
01:26 eli: elliottcable: Kind of hard to get excited over something like that...  (I tried to catch some stuffs, but didn't see anything from your web page or your github page.)
01:28 eli: (BTW, IMO markdown has far worse problems than it's treatment of *...* etc, or quotes...)
01:29 eli: (That's re http://blog.elliottcable.name/posts/markright.xhtml)
01:30 (quit) martinhex: Read error: Operation timed out
01:30 (join) martinhex
01:35 (quit) no-name-: Ping timeout: 276 seconds
01:35 (join) no-name-
01:36 (quit) sheikra: Read error: Connection reset by peer
01:37 elliottcable: eli: I’m interested in hearing them
01:37 elliottcable: eli: let’s not clog up here; come hang in ##Paws
01:38 (join) sheikra_
01:39 (join) pattern
02:13 (join) neilv
02:26 (join) realitygrill_
02:29 (quit) realitygrill: Ping timeout: 246 seconds
02:29 (nick) realitygrill_ -> realitygrill
02:31 (quit) realitygrill: Quit: realitygrill
02:32 Sgeo: Argh
02:32 (quit) sheikra_: Ping timeout: 246 seconds
02:33 (join) hkBst
02:58 (quit) neilv: Ping timeout: 240 seconds
02:58 eli: elliottcable: Sounds like it would be off topic there too...
02:59 elliottcable: eli: yeah, but we always are, so meh
02:59 (join) neilv
02:59 (quit) neilv: Changing host
02:59 (join) neilv
03:00 bremner_: make a new channel? ;)
03:00 elliottcable: pff
03:00 elliottcable: it’s too much fun
03:00 elliottcable: I tend to collect intelligent people there, so the off-topic conversations are always … interesting.
03:01 neilv: better than collecting intelligent people in one's basement
03:01 elliottcable: are you sure?
03:01 elliottcable: perhaps you could generate brainpower by osmosis.
03:01 elliottcable: or by, you know, eating brains.
03:01 elliottcable: either way.
03:02 neilv: that research space is pretty well exhausted
03:04 eli: elliottcable: It's not too bad here too, and I prefer to reduce the number of bits that I need to sift through...
03:05 elliottcable: hah
03:05 bremner_ goes to #emacs for offtopic chatter ;)
03:05 eli: The problem of markdown is very visible in wikipedia and such -- as soon as it's used for something moderately important and/or more than some personal blog or random comments, things break down pretty badly.
03:06 eli: In wikipedia, you can see it in the huge amount of ad-hoc templates and a salad of syntaxes that mean all kinds of things.
03:06 eli: SO also breaks too often on all kinds of corner cases that are "not popular enough".
03:07 eli: (I always have problems with our documentation URLS which have parens in them...)
03:19 (quit) no-name-: 
03:22 (quit) Demosthenes: Ping timeout: 276 seconds
03:24 (join) no-name-
03:24 (join) Demosthenes
03:25 (quit) cpach: Ping timeout: 240 seconds
03:33 (quit) ckrailo: Quit: Computer has gone to sleep.
04:14 Demosthenes: anyone here using rackunit, know how to make it provide positive acknowledgement of tests?
04:23 (quit) Sgeo: Read error: Connection reset by peer
04:24 (join) Sgeo
04:32 (quit) neilv: Ping timeout: 260 seconds
04:32 (join) sheikra
04:52 (quit) Demosthenes: Ping timeout: 264 seconds
05:00 (join) neilv
05:15 (quit) sheikra: Read error: Connection reset by peer
05:18 (join) sheikra
05:19 (quit) neilv: Ping timeout: 260 seconds
05:20 (join) masm
05:31 (quit) sheikra: Quit: Leaving
05:34 (join) sunnyps
05:52 (part) sunnyps: "Leaving"
06:41 (quit) rekahsoft: Ping timeout: 276 seconds
06:42 (join) rekahsoft
07:00 (quit) martinhex: Read error: Connection reset by peer
07:01 (join) martinhex
07:04 (quit) no-name-: 
07:38 (join) MayDaniel
07:51 (join) yoklov
07:51 (join) PLT_Notify
07:51 PLT_Notify: racket: master Robby Findler * 71770ee (1 files in 1 dirs): one more collection-path => collection-file-path in scribble - http://bit.ly/mvHEnt
07:51 (part) PLT_Notify
07:52 (join) yoklov1
07:52 (quit) yoklov: Read error: Connection reset by peer
08:00 (join) lucian
08:10 (join) sstrickl
08:15 (quit) MayDaniel: Read error: Connection reset by peer
08:18 (quit) yoklov1: Quit: Leaving.
08:24 (join) PLT_Notify
08:24 PLT_Notify: racket: master Matthew Flatt * ffbaaa3 (1 files in 1 dirs): kFreeBSD fix from David Bremner - http://bit.ly/jxCeyv
08:24 (part) PLT_Notify
08:46 (join) realitygrill
08:47 (join) yoklov
09:02 (quit) yoklov: Quit: Leaving.
09:05 (join) tfb
09:24 (quit) lucian: Remote host closed the connection
09:25 (quit) realitygrill: Quit: realitygrill
09:51 (join) outworlder
09:51 (part) outworlder
09:54 (join) MayDaniel
09:55 (join) ckrailo
09:59 (join) drewr
10:05 (join) sisi
10:06 sisi: hi :'p
10:08 sisi: omg :'( bande de doupet..
10:09 sisi: scheme < all ..
10:09 (quit) sisi: Client Quit
10:09 _p4bl0: okaaaay
10:10 (join) dnolen
10:15 (quit) dnolen: Client Quit
10:33 (quit) rekahsoft: Read error: Operation timed out
10:34 (join) rekahsoft
10:37 (join) Qerub
10:39 (quit) ckrailo: Quit: Computer has gone to sleep.
10:43 (quit) MayDaniel: Read error: Connection reset by peer
10:57 (quit) hkBst: Read error: Connection reset by peer
11:04 (join) ckrailo
11:15 (join) realitygrill
11:18 (join) anRch
11:19 (join) mithos28
11:44 (quit) realitygrill: Read error: Connection reset by peer
11:44 (join) dnolen
11:45 (join) realitygrill
11:45 (join) PLT_Notify
11:45 PLT_Notify: racket: master Vincent St-Amour * e64ba22 (1 files in 1 dirs): Have the TR test harness skip emacs unsaved file backups.
11:45 PLT_Notify: racket: master Vincent St-Amour * 4cd0ad4 (2 files in 1 dirs): Coerce reals in float expressions when valid.
11:45 PLT_Notify: racket: master commits ffbaaa3...4cd0ad4 - http://bit.ly/lZkhOc
11:45 (part) PLT_Notify
11:55 (join) jonrafkind
11:58 (quit) anRch: Quit: anRch
12:04 (join) anRch
12:10 ohwow: http://nice.sourceforge.net/visitor.html hmm
12:14 (quit) realitygrill: Read error: Connection reset by peer
12:15 (join) realitygrill
12:17 Qerub: ohwow: Are you enjoying the idea of multimethods? :)
12:20 ohwow: Yes it's nice
12:21 ohwow: Racket doesn't have dynamic dispatch @ core, does it?
12:21 ohwow: I've managed to find http://planet.racket-lang.org/package-source/murphy/multimethod.plt/2/1/planet-docs/main/index.html tho
12:21 rudybot: http://tinyurl.com/3zsn26m
12:22 Qerub: ohwow: Hehe, yes, nice indeed.
12:22 Qerub: ohwow: Actually, it has dynamic dispatch, but only single (not multi).
12:22 ohwow: please elaborate
12:22 Qerub: http://docs.racket-lang.org/reference/mzlib_class.html
12:22 ohwow: what's the difference between single dispatch and multiple distpactch?
12:23 Qerub: Start here: http://en.wikipedia.org/wiki/Multiple_dispatch
12:23 ohwow: I've noticed that method defined by (define-method) accepts.. symbols?
12:23 ohwow: I would assume it accepts structs
12:24 jonrafkind: single dispatch uses the dynamic type of the "first" argument, the foo in foo.blah(...)
12:25 ohwow: Qerub: I know what's multiple dispatch, but what's single dynamic dispatch?
12:25 jonrafkind: multidispatch uses the dynamic type of all arguments, x, y, z, in foo.blah(x, y, z)
12:25 ohwow: ah
12:26 Qerub: ohwow: I'm not familiar with that multimethod-library. Reading the page you linked now.
12:29 (join) lucian
12:29 (quit) tfb: Ping timeout: 264 seconds
12:31 Qerub: ohwow: It seems like the library author wanted something more general than multimethods for structs/classes and added "ad-hoc hierachies of values".
12:31 Qerub: ohwow: Are you familiar with Common Lisp Object System? It has multimethods aswell.
12:32 ohwow: no, not really
12:33 Qerub: I hope I don't get kicked for bringing up CL here ;-)
12:36 Qerub: In the multimethods.plt example, I find this pretty confusing: (define-multimethod (beats? a b) :: (vector-immutable a b))
12:37 Qerub: What does vector-immutable mean in that context?
12:37 Qerub: (rhetoric question)
12:38 (join) superjudge
12:41 (quit) anRch: Quit: anRch
12:45 (nick) samth_away -> samth
12:45 samth: Qerub, you can check out swindle for CLOS-style programming in racket
12:46 Qerub: Oh, right.
12:46 Qerub: ohwow: Listen to samth :)
13:04 (join) elliott
13:05 elliott: Are the nightly builds stable enough to use day-to-day or are the release builds preferable?
13:06 jonrafkind: nightly is probably ok, they are equivalent to what the developers usually run (i.e, build from HEAD each day or whatever)
13:06 jonrafkind: i mean im usually 3-4 days lagged behind HEAD but i havent had any issues
13:07 elliott: OK, thanks.
13:15 (quit) mithos28: Quit: mithos28
13:17 (join) mithos28
13:18 (quit) realitygrill: Read error: Connection reset by peer
13:19 (join) realitygrill
13:27 (join) MayDaniel
13:28 (join) realitygrill_
13:29 (quit) realitygrill: Ping timeout: 246 seconds
13:29 (nick) realitygrill_ -> realitygrill
13:37 (quit) realitygrill: Read error: Connection reset by peer
13:37 (join) realitygrill
13:43 (quit) realitygrill: Read error: Connection reset by peer
13:43 (join) realitygrill
13:51 (quit) MayDaniel: Read error: Connection reset by peer
13:55 (quit) realitygrill: Read error: Connection reset by peer
13:56 (join) realitygrill
14:04 (join) anRch
14:21 (quit) pattern: Read error: Connection reset by peer
14:24 (join) PLT_Notify
14:24 PLT_Notify: racket: master Vincent St-Amour * bdf1073 (1 files in 1 dirs): Fix optimizer expected log format. ... - http://bit.ly/m0fI9Q
14:24 (part) PLT_Notify
14:25 (join) Fare
14:32 (join) dnolen_
14:33 (quit) elliott: Quit: Leaving
14:34 (join) elliott
14:35 (quit) dnolen: Ping timeout: 251 seconds
14:41 (quit) mithos28: Quit: mithos28
14:42 (join) heeki
15:06 (quit) anRch: Quit: anRch
15:35 (join) PLT_Notify
15:35 PLT_Notify: racket: master Sam Tobin-Hochstadt * 18e83f2 (1 files in 1 dirs): Fix result contract of `stx->list'.
15:35 PLT_Notify: racket: master Sam Tobin-Hochstadt * 744d192 (1 files in 1 dirs): Remove duplicate definitions; reprovide from `racket/private/stx' instead.
15:35 PLT_Notify: racket: master Sam Tobin-Hochstadt * c2291e0 (2 files in 1 dirs): Add types for some of `syntax/stx' ...
15:35 PLT_Notify: racket: master commits bdf1073...c2291e0 - https://github.com/plt/racket/compare/bdf1073...c2291e0
15:35 (part) PLT_Notify
15:36 (join) Anchor
15:37 Anchor: Hi, any advices how to install Racket in Ubuntu 11.04?
15:37 Anchor: It looks that I have to build it from the sources
15:38 Anchor: Are there any live persons?
15:38 ohwow: Anchor: http://packages.ubuntu.com/oneiric/racket
15:38 ohwow: I am not sure what is Ubuntu 11.04
15:39 ohwow: if it's an older version, I assume you have some kind of backporting system?
15:39 (join) mithos28
15:39 Anchor: I'm not sure
15:39 samth: Anchor, go here: http://racket-lang.org/download/
15:39 Anchor: apt-get doesn't know about it
15:39 samth: select Ubuntu
15:40 samth: download, and follow the instructions
15:41 samth: alternatively, go here: https://launchpad.net/~plt/+archive/racket
15:41 Anchor: what about x86-64 - there is i386 only
15:42 Anchor: in downloads list
15:42 samth: i don't think there are prebuilt binaries for x86-64 right now
15:42 Anchor: I see x86-64 only for Fedora-14
15:42 samth: there is: http://pre.racket-lang.org/installers/
15:42 samth: which has debian installers for 64 bit
15:42 samth: but no ubuntu yet
15:43 Anchor: are there any troubles if I'll try to build it by myself?
15:45 jonrafkind: Anchor, did you realize you can install the ppa for ubuntu as samth suggested?
15:47 Anchor: I'm not sure that Debian's package fits fresh ubuntu
15:49 jonrafkind: what?
15:49 jonrafkind: a ppa is for ubuntu
15:52 (quit) Anchor: Quit: Ухожу я от вас (xchat 2.4.5 или старше)
16:03 (join) elliott_
16:05 (join) gmarceau
16:07 (quit) elliott: Ping timeout: 240 seconds
16:10 (quit) superjudge: Quit: superjudge
16:16 (quit) Qerub: Quit: Qerub
16:24 (quit) elliott_: Ping timeout: 250 seconds
16:28 (join) Anchor_
16:31 tonyg: Are .zo files architecture-independent? If I compile on one kind of machine, and transplant the .zo files to another architecture, will things work out OK?
16:31 bremner_: probably.
16:32 bremner_: in principle, racket could compile differently on different architectures
16:32 (quit) Anchor_: Ping timeout: 240 seconds
16:32 bremner_: as far as I know, nothing in collects does that currently.
16:32 jonrafkind: there is (get-platform) or something which returns 'unix or 'win32
16:32 tonyg: cool. I'm about to try it: I'm building Racket for OpenWRT
16:33 bremner_: tonyg: is that a debian like thing?
16:33 tonyg: yep
16:33 tonyg: well no
16:33 tonyg: sorry
16:34 tonyg: it's linux
16:34 (join) Anchor
16:34 bremner_: hah
16:34 tonyg: but it uses "*.ipkg"
16:34 tonyg: a la many other linuxy embedded systems
16:35 bremner_: right. well, I currently build debian packages this way (PLT_SETUP_OPTIONS="--no-zo --no-docs") most architectures.
16:36 tonyg: Interesting! Thanks bremner_, that looks like a useful tip
16:36 jonrafkind: does that mean the collects get built without zo's bremner_ ?
16:36 bremner_: jonrafkind: no, it just means I only build .zo's once
16:37 jonrafkind: im lost
16:37 jonrafkind: in what way do the zos get built multiple times?
16:37 bremner_: because racket gets compiled for 10 architectures
16:38 tonyg: bremner_: so you somehow sleight-of-hand the previously-compiled .zos in when building for most of the 10?
16:38 jonrafkind: oh I see
16:38 bremner_: well, I split into two packages, racket-common has the .zo's and racket is built everywhere.
16:39 tonyg: aha! ok that looks like a sensible choice. On a router w limited memory I may have to subdivide more finely than that eventually even
16:39 bremner_: there is some work underway to make racket more modular, I heard.
16:39 tonyg: bremner_: where should I look for your deb scripts? I'd love to see the details of how you approach this
16:39 tonyg checks the regular tree to see if there's a debian/ directory there :)
16:40 tonyg: (nope, there isn't)
16:40 bremner_: git://git.debian.org/git/collab-maint/racket
16:40 tonyg: thanks, I'll check that out
16:46 tonyg: whoa, ok, it has complete sources in it
16:50 tonyg: ah, perhaps because of dfsg?
16:50 bremner_: ah, more that disk space is cheaper than my time. sorry about the big download.
16:50 tonyg: :) np
16:52 tonyg: bremner_: when you're cross-compiling, does the normal ./configure dance Just Work? I don't see any special code in the control file
16:53 tonyg: I mean rules
16:53 bremner_: I don't cross compile, I (well, debian) have a lot of computers
16:53 tonyg: ah :) ok
16:53 tonyg: hmm
16:53 tonyg predicts pain
16:54 tonyg: pleased to see mips support in there
16:55 (quit) mithos28: Quit: mithos28
16:55 bremner_: I think you will hit the problem of diskspace without some hacking and slashing.
16:55 offby1: Don't suppose anyone has taught the racket profiler to emit output suitable for kcachegrind?
17:06 (join) tauntaun
17:20 (join) yoklov
17:33 (join) lucian_
17:35 (quit) lucian: Ping timeout: 240 seconds
17:49 (join) PLT_Notify
17:49 PLT_Notify: racket: master Robby Findler * 4d4a94a (3 files in 2 dirs): removed EoPL from the language dialog (and adjust the docs a little ... - http://bit.ly/kO24Ph
17:49 (part) PLT_Notify
18:04 (part) yoklov
18:07 (join) mithos28
18:10 (quit) tauntaun: Quit: Ex-Chat
18:15 (nick) lucian_ -> lucian
18:22 (quit) dnolen_: Ping timeout: 252 seconds
18:26 samth: offby1, we'd love you forever if you did :)
18:28 (quit) sstrickl: Quit: sstrickl
18:29 offby1: Now that I've played with the profiler, I either don't understand it, or it doesn't work well enough for me :-|
18:30 offby1: its call tree seems to be less deep than I'd expected: I have a reasonably complex function which shows up as a leaf node, even though it calls a couple of other functions.
18:31 offby1: I suspect this has to do with how  continuation-mark-set->context works, but don't really know
18:32 offby1: so: this particular function is where all my time is going, but I don't understand -why-
18:34 samth: can you up the sampling rate?
18:46 (join) Rodolfo
18:46 Rodolfo: hello
18:47 Rodolfo: could anyone help me figure out what's wrong with a bundled game?
18:47 Rodolfo: I am trying to run "Paint by Numbers" on Racket 5.1.1
18:48 Rodolfo: and I get a dialog box with the following:
18:48 Rodolfo: min-height in area<%>: expected argument of type <exact integer in [0, 10000]>; given 26587/128   === context === /opt/racket/collects/racket/private/more-scheme.rkt:265:2: call-with-exception-handler /opt/racket/collects/games/paint-by-numbers/gui.rkt:66:4 do-make-object /opt/racket/collects/games/paint-by-numbers/paint-by-numbers.rkt:211:4 do-make-object /opt/racket/collects/games/paint-by-numbers/paint-by-numbers.rkt:502:4: playe
18:52 offby1: sounds like a simple programming error.
18:52 offby1: Welcome to Free Software! :-)
18:53 jonrafkind: Rodolfo, what os are you on
18:53 jonrafkind: oh linux I guess
18:53 jonrafkind: well it works for me..
18:54 Rodolfo: yeah, ubuntu 10.10
18:54 Rodolfo: I would like to fix and send a patch
18:54 offby1: jonrafkind: what do I need to do in drracket to run that?
18:54 Rodolfo: but the context info didn't help me much
18:55 jonrafkind: the games? i dunno, i just ran it from the command line. bin/plt-games
18:56 Rodolfo: I opened some of the *.rkt from racket/collects/games/paint... but I don't know how to run them from DrRacket.
18:56 offby1: ah.
18:56 offby1: ooh ooh I see what Rodolfo is seeing
18:57 jonrafkind: looks like you have to instantiate the game@ unit and run (player)
18:59 Rodolfo: which file to open/run on drracket?
18:59 jonrafkind: what
18:59 Rodolfo: main.rkt is not "runnable " out of the box: Module Language: only a module expression is allowed, either     #lang <language-name>  or     (module <name> <language> ...)
19:00 offby1: I just submitted a bug report; when I get the URL I'll holler
19:00 offby1: http://bugs.racket-lang.org/query/?cmd=view&pr=11924
19:01 Rodolfo: well, is there a link to this bug tracker in the main rack-lang.org pages? I didn't find it before...
19:01 offby1: if you're asking "how would I have submitted a bug like that", the answer is: run drracket, click on the "Help" menu, and choose "Submit bug report"
19:01 jonrafkind: hm thats true, the bugs page isn't linked from anywhere
19:03 Rodolfo: aham! Thanks, now I know (that was indeed my question)
19:04 Rodolfo: So should I try to submit my first bug? "The bug page isn't linked from anywhere"
19:06 jonrafkind: thats probably done on purpose
19:06 jonrafkind: when you use drracket to submit bugs all sorts of information is sent along
19:06 jonrafkind: if you use the web page that information is lost and it makes debugging harder for the developers
19:07 offby1 nods gravely
19:08 jonrafkind: probably assuming people will find the bug report mechanism in drracket is a mistake, since not everyone uses drracket and the whole procedure is highly unique in the software world, but such is life
19:08 Rodolfo: ok, so I will postbug my bug report for later
19:09 Rodolfo: *postpone
19:09 (join) bdamos
19:09 Rodolfo: BTW, I have an announcement to make :)
19:10 Rodolfo: I will introduce Racket to some of my friends from Dojo Rio (http://dojorio.org) this Wednesday
19:10 jonrafkind: is this some sort of famous thing?
19:11 Rodolfo: this is a Coding Dojo in Rio de Janeiro, maybe the most influential in Brazil
19:12 jonrafkind: ive never heard of a coding dojo, sounds like a lug or something
19:12 (nick) samth -> samth_away
19:15 Rodolfo: A Coding Dojo is a place to deliberately practice programming skills
19:17 Rodolfo: we have a short "about" http://dojorio.org/about/
19:18 Rodolfo: and brian d foy, Perl hacker, visited us last week [http://blogs.perl.org/users/brian_d_foy/2011/05/fun-at-the-dojo-rio.html]
19:19 jonrafkind: a 4 minute timer.. rofl
19:19 Rodolfo: basically we meet every week to solve programming puzzles using TDD, pair programming, ...
19:20 (join) Rolaulten
19:20 Rodolfo: the timer is to turn pair programming into "group programming"
19:20 Rodolfo: if you have never tried something like a dojo, you should give it a shot.
19:20 Rolaulten: Hello everyone - I'm still new to programing in racket/scheme and have a quick question - how do I instantiate and use a new library?
19:21 jonrafkind: (require "some-library.rkt")
19:21 Rodolfo: it is impressing how your typing/coding skills change when you get ten people looking at your face and your code :P
19:21 jonrafkind: or if its a library that comes with racket you dont need the quotes, (require some-library)
19:21 Rolaulten: and it the code must be a .rkt? (its been one of those days for me were my brain is dead...)
19:22 Rodolfo: Rolaulten, it could be something else
19:22 jonrafkind: you can name it anything you want, (require "foo.bar")
19:22 jonrafkind: if you end it with .rkt then the system will do some convenience things for you
19:23 jonrafkind: such as assuming .rkt is there
19:24 Rolaulten: Ok - right now I have (#% require (lib /path/foo.scm))
19:24 Rolaulten: so just change it to (require foo.scm) and it will work
19:24 jonrafkind: where did you learn about such syntax?
19:25 Rolaulten: poking around the documentation/google
19:25 Rodolfo: (require "/path/foo.scm")
19:25 Rolaulten: kk
19:25 Rodolfo: don't forget the quotes
19:25 jonrafkind: you should probably never use #%, and the (lib ...) is from an older version of racket (the 4.x and before series)
19:26 tonyg: damn. the boehm gc runs C programs to decide how to build. makes cross compilation... tricky.
19:26 Rolaulten: its for a class - we need to use r5rs
19:26 jonrafkind: oh
19:27 jonrafkind: also in case you didn't realize, you should put '#lang r5rs' at the top of your files
19:27 Rolaulten: tis there
19:28 Rodolfo: back to the bug in paint-by-numbers, how to get my hands wet into exterminating that bug?
19:28 Rolaulten: but no - for r5rs the syntax is right or do I have something wrong - cause the error message I get when I add the quotes it a bad path
19:29 jonrafkind: Rodolfo, printf debugging
19:29 jonrafkind: just toss them in randomnly
19:30 jonrafkind: Rolaulten, paste the require line
19:31 Rolaulten: (#%require (lib "file:///home/dorian/.racket/5.0.1/srfi-1.rkt"))
19:31 Rodolfo: is there any page describing how to send a patch?
19:32 Rodolfo: (addressing a given bug)
19:32 (quit) Checkie: Ping timeout: 260 seconds
19:32 Rolaulten: I'm geting the bug "lib: bad module path in: (lib "file:///home/dorian/.racket/5.0.1/srfi-1.rkt")
19:33 tonyg: augh. Does Racket's embedded Boehm GC use Makefile.am at all or is it vestigial?
19:33 jonrafkind: instead of (lib "file:///home/dorian/...") just do "/home/dorian/..."
19:34 Rolaulten: still getting the same error
19:34 jonrafkind: (#%require (lib "srfi/1"))
19:34 jonrafkind: this works better
19:34 jonrafkind: Rodolfo, send a mail to dev@racket-lang.org or just announce your patch here
19:35 Rolaulten: wonderful! thankyou, so now I can just call any of the procedures within the like most any other language?
19:35 jonrafkind: http://docs.racket-lang.org/reference/require.html?q=r5rs#(form._((lib._racket/private/base..rkt)._lib))
19:35 jonrafkind: yea
19:35 Rodolfo: ok... I just subscribed to it. I noticed I was only in the google groups mirror for the users list
19:36 jonrafkind: technically #%require isn't part of r5rs, so its not clear what your teacher wants when they say to use r5rs
19:37 Rolaulten: the prof is somewhat all over...at one point he frogot that it was not '94 and modern computers can do the Fibonacci sequence to n=20 without crashing
19:38 Rodolfo: ok, thanks jonrafkind, thanks eveyone - i'm leaving for now
19:46 (join) sstrickl
19:53 (part) Rodolfo
19:57 (quit) Rolaulten: Quit: Leaving
19:59 (join) sethalves
19:59 (join) dnolen
20:09 (quit) Fare: Quit: Leaving
20:10 (quit) ckrailo: Ping timeout: 260 seconds
20:44 (join) PLT_Notify
20:44 PLT_Notify: racket: master Matthew Flatt * 11fa781 (1 files in 1 dirs): fix future-scheduling problem related to custodians - http://bit.ly/mGtNEF
20:44 (part) PLT_Notify
20:45 (join) realitygrill_
20:45 (quit) mithos28: Quit: mithos28
20:46 (part) sstrickl
20:48 (quit) realitygrill: Ping timeout: 246 seconds
20:48 (nick) realitygrill_ -> realitygrill
21:14 (join) bohanlon
21:17 (join) no-name-
21:20 (quit) jonrafkind: Ping timeout: 260 seconds
21:23 (join) mithos28
21:33 (quit) lucian: Remote host closed the connection
21:33 eli: offby1: Missing functions might be a result from inlining.
21:34 eli: offby1: And re that tool -- making a different output from the profiler should be pretty easy, you could start with just thet plain text output and hack it.
21:35 eli: But given that they refer to a bunch of conversion scripts, and given thet the output is basically similar to other profilers, it might be easier to hack such a conversion thing instead.
21:36 (join) ckrailo
21:39 offby1: I'm more interested in getting more-informative profile data than in adapting the output for kcachegrind.
21:40 offby1: I'd be surprised if it was all about inlining.  The function in question is "string->utterance", visible at https://github.com/offby1/rudybot/blob/master/log-parser.rkt#L17
21:44 eli: And which function did you expect to see in the call tree but didn't?
21:49 offby1: any of them
21:50 offby1: regexp, read, make-utterance
21:50 offby1: in other words: I know that that function is slow, but I don't know why
21:50 offby1: is it make-utterance?  Is it "read"?  Is it something else?
21:50 offby1 shrugs
21:51 (join) PLT_Notify
21:51 PLT_Notify: racket: master Asumu Takikawa * 47db06a (1 files in 1 dirs): round -> ceiling in paint-by-numbers - http://bit.ly/lVFBZG
21:51 (part) PLT_Notify
21:53 (join) Demosthenes
22:01 eli: offby1: There is no `regexp' call there -- it's a pattern which `match' expands to some code; the code will use `regexp-match' but you won't see that since it's a primitive function.
22:01 eli: Same goes for `read'.
22:02 eli: My guess is that `make-utterance' is a constructor, therefore it takes a very small amount of time (just the allocation and setting of the results), so that will be hard to catch, and it is likely to be inlined.
22:03 eli: But in all of these, the only possible and likely offender is the regexp matching, so that's where I'd focus my attention if the goal is just to speed it up.
22:05 eli: offby1: Here's an example:
22:05 eli: (let ([s (string-append (make-string 100000000 #\a) "b")]) (time (regexp-match? #rx"a*?b" s)))
22:05 eli: vs
22:05 eli: (let ([s (string-append (make-string 100000000 #\a) "b")]) (time (regexp-match? #rx"a*b" s)))
22:06 eli: The second is about twice faster on my machine.
22:07 eli: The dangerous thing with things like ".*?" is that they'll generate all possible matches until it reaches that "b", whereas ".*" goes for it directly.
22:13 eli: offby1: I was about to say that this: (let ([s (string-append (make-string 100000000 #\a) "b")]) (time (regexp-match? #rx"(a*?)b" s)))
22:13 eli: would be even worse, but I didn't realize how much worse it would be, and took a while before I could even kill the process...
22:28 offby1: eli: thanks!
23:10 Demosthenes: i'm still curious how to enable positive affirmation of passed tests in rackunit ;]
23:13 offby1: well, when I use text-ui (or whatever it's called) it prints the number of successes and failures to current-output-port
23:15 Demosthenes: wonder if that requires run-tests
23:15 Demosthenes: i'm just doing check-X
23:17 offby1: ah, I suspect it does.
23:17 offby1: you'll probably have to make a test suite
23:17 Demosthenes: hrm
23:20 eli: offby1: http://bugs.racket-lang.org/query/?cmd=view&pr=11925
23:21 offby1: oho
23:22 offby1: hmm.  How do I get my email address onto that bug, such that I get emailed if it gets edited?
23:23 eli: offby1: you can't do it yourself...
23:24 eli: Are you still using first.last@gmail.com?
23:24 Demosthenes: offby1: ok, so i wrapped it in define-test-suite, and it gives a summary
23:24 offby1: eli: yep, tx
23:24 Demosthenes: i'm still looking for a "test:OK" per check
23:24 offby1: actually, it's alpha.omega@gmail.com
23:25 offby1: Demosthenes: sucker!
23:25 offby1: Demosthenes: I'm kidding.  I sympathize.
23:25 eli: offby1: You're on it now.
23:25 Demosthenes: ;]
23:25 offby1: eli: tx^2
23:25 offby1: Demosthenes: Be nice if you could hack rackunit to follow the "test anything protocol".
23:25 offby1: Probably not that hard, actually
23:26 offby1: maybe you can write a custom test-runner -- that'd be the clean way to do it
23:26 Demosthenes: i was looking at the "around" funcs...
23:26 Demosthenes: just surprised it wasn't a builtin
23:27 offby1: eli: I should mention that the racket-lang web docs are so fast that I'm no longer tempted to look at my local copy.  This is in contrast to the Python docs, which are always infuriatingly slow
23:27 eli: Demosthenes: You should really ask on the list since the people who are doing rackunit are not here frequently.
23:27 Demosthenes: ah
23:27 Demosthenes: i may do that
23:27 eli: offby1: Yeah, I've argued to go on-line only...
23:28 Demosthenes: actually, i really appreciate my local copy of the docs....
23:28 offby1: Demosthenes: another idea is just to hack it a little so that the 'verbose argument to run-tests does what you want
23:28 Demosthenes: makes coding on an airplane easier
23:28 offby1: eli: you can say "even offby1, who whines about everything, says they're fast enough!"
23:28 eli: Demosthenes: Well, my full proposal was to make them optional, so you could install them locally if you want to, and default to using them on-line.
23:29 eli: Doing things *only* on-line is impractical since there are schools and labs where that won't work.
23:31 Demosthenes: sure
23:31 Demosthenes: i also like the pdfs
23:31 Demosthenes: i've loaded them in my new ebook reader
23:31 offby1: eww
23:31 offby1: me holds his nose and waves his other hand in front of his face
23:31 offby1: he said "pdf"
23:31 eli agrees with offby1
23:31 eli: But some people like it...
23:31 eli: It became much more tolerable since I started using chrome.
23:31 eli: Turns out acrobat was a large part of why I hated pdfs,
23:31 eli: s/,/./
23:31 Demosthenes: != acrobat
23:32 Demosthenes: hrm, i like the ebook reader. i don't like the lack of drm-free ebooks
23:32 eli: Demosthenes: I've tried replacing it, but then there's still obnoxiousness left in firing up some other process.
23:32 Demosthenes: i bought a pair of R books for $50, no drm. i want to buy a recent sci-fi title for $9? drm.
23:33 eli: And yeah, the adobe focus on DRM in PDFs is obviously making it as appealing as an ipad.
23:33 Demosthenes: eli: yeah, i'm using xpdf :P
23:33 offby1: some day our grandchildren will mock us for putting up with all this DRM crap
23:33 Demosthenes: otoh PDF is an iso standard now...
23:33 Demosthenes: the drm is an extension they own
23:33 Demosthenes: offby1: i didn't buy a nook or kindle. i refused.
23:33 Demosthenes: i wont buy drm crippled goods
23:34 offby1: I will :-(
23:34 offby1: if they're shiny enough
23:34 Demosthenes: i bought a Pocketbook 902, bigger than the kindle dx ;]
23:34 Demosthenes: awesome screen, and just added a 8 GB microsd
23:35 eli: I certainly hope that our grandchildren will do that, rather than admire us as the last generation of free stuffs roaming the earth.
23:35 Demosthenes: i keep a copy of "The right to read" by Stallman on my ebookreader
23:35 Demosthenes: :P
23:36 eli: Crap as literature, but effective...
23:36 Demosthenes: yep
23:36 offby1 subtly mentions that he re-upped with the FSF a few weeks ago
23:36 Demosthenes: my position in life is simple, no machine is allowed to control human behavior, or tell me no.
23:36 offby1: I used to think I had an admirably low member number, but it turns out their membership has always been tiny
23:37 eli: offby1: ?
23:37 offby1: eli: !
23:37 eli: "re-upped with the FSF?"
23:37 offby1: renewed my membership, which I'd let lapse years ago
23:37 eli: Demosthenes: Get an ipad -- as someone who never used any isomething, it was absolutely horrifying.
23:37 (quit) masm: Ping timeout: 240 seconds
23:38 eli: offby1: Ah.
23:38 (join) ZachBeta
23:39 Demosthenes: eli: try that again?
23:39 eli: Demosthenes: Huh?
23:39 Demosthenes: i still use a linux laptop, and put up with the hardware issues, because apple likes drm... otherwise i'd have one of those aluminmum macbooks in a heartbeat
23:40 Demosthenes: oh, the pocketbook runs linux, and is easily rootable ;]
23:41 eli: Demosthenes: personally, I'm not using *any* apple stuff (well, since 1988, at lease) -- then we somehow got to own an ipad, and exactly because I've never used anything it was horrifying just how bad apple is.
23:41 Demosthenes: they have some talented people, and some of the new stuff rocks...
23:41 Demosthenes: but the lack of control is disturbing
23:42 (join) masm
23:42 Demosthenes: though if tmobile is bought by at&t, verison is going to have a new iphone subscriber, i refuse to gibve money to at&t
23:42 offby1: what with cloud computing, it's all becoming quite vague
23:42 eli: Demosthenes: *Nothing* justifies such a fascist nightmare.
23:43 Demosthenes: yeah, forget the cloud. no security or control....
23:43 ZachBeta: but convenience
23:43 offby1: ah, but the convenience and
23:43 Demosthenes: eli: well, i'd argue macos is better than windows ;]
23:43 offby1: curse ZachBeta and his fast fingers
23:43 ZachBeta: huzzah!
23:43 offby1: ZachBeta: get your own brain, willya
23:43 offby1: I'm using mine
23:43 eli: ZachBeta: Especially not any illusion of convenience.
23:43 ZachBeta: eli, illusion?
23:44 eli: Demosthenes: I'd strongly argue differently.
23:44 eli: ZachBeta: Yes.
23:44 ZachBeta: eli, please explain
23:45 eli: ZachBeta: Like any fascist government, things are convenient as long as you're on the same side, but flip around fast once you're not.
23:46 Demosthenes: erm, i'd rank MS higher on the fascist scale
23:46 eli: Demosthenes: Apple is so much worse, it's really surprising.
23:46 offby1: remember, "fascist" is just a metaphor when applied to companies.  Companies don't have armies with guns.
23:47 eli: Much due to the inflated ego of the man on top.
23:48 Demosthenes: eh
23:48 ZachBeta: maybe I'm just one of the sheeple... but I like using gmail and gdocs and dropbox and all cloudy stuff
23:48 Demosthenes: jobs is competent. gates left ms in the hands of a dancing chimp
23:49 ZachBeta: Demosthenes, DEVELOPERS DEVELOPERS DEVELOPERS DEVELOPERS
23:49 Demosthenes: omg
23:49 offby1: ZachBeta: me too
23:50 offby1: I don't think Benjamin Franklin ever weighed in on the wisdom of sacrificing a little bit of liberty for rather a whole lot of convenience
23:50 ZachBeta: it's an optimization really...
23:50 ZachBeta: and are you ever really 100% free?
23:51 eli: Chimp or not, I consider that kind of barbaric attempt at getting hackers excited much better than the two-faced speeches of jobs speaking about how peolpe hate DRM while providing one of the most restricted platforms in the history of computers.
23:51 ZachBeta: eli, for the time being he's a benevolent dictator
23:51 eli: Or the speech about how people hate advertisements and then reaching a conclusion that it's somehow good for mankind to have all ipad advertising ... go through apple.
23:51 eli: ZachBeta: I couldn't agree less.
23:52 ZachBeta: eli, I'm not exactly an apple scholar, but my perspective is a company with VERY consistent products
23:52 eli: Or the speeches about how great it is to have a central place to distriobute software from, when it's obvious that the 60% or whatever that is left in apple's pockets is the main reason behind that.
23:53 eli: Or even banning someone from the apple bookstore because he wrote an unflattering biography of jobs.
23:53 eli: Or deleting links to consumerreports right after they publish an unfavorable review of the iphone 4.
23:54 ZachBeta: I'm not going to ask for sources, but I definitely rescind "benevolent"
23:55 eli: Or making people publicly apologize for making jokes about apple.  (Like what's her name who made a skit about iphone, and the next morning started with an apology about doing that.)
23:55 (quit) masm: Ping timeout: 240 seconds
23:56 eli: ZachBeta: The funny thing is that I was never some anti-apple zealot, and I'm really not someone who believes conspiracy theories...
23:56 ZachBeta: some of your statements don't all pass my bullshit filter, but enough of them do
23:57 eli: Tell me which one, and I'll tell you where I read about it.
23:57 ZachBeta: they deleted all links to CR and banned people from the bookstore?
23:58 ZachBeta: I've heard the apple store is real strict
23:58 ZachBeta: and I get the impression that they run things pretty tightly, but I can't imagine you can possibly bring a company into fortune 100 without being evil
23:59 eli: Ah -- Ellen DeGeneres.
23:59 ZachBeta: even google
23:59 ZachBeta: eli, you heard that reference on ellen?
23:59 ZachBeta: ??????
23:59 eli: ZachBeta: I agree, but the piles of crap is overwhelming.
23:59 eli: ZachBeta: No, she's the one that apologized.
23:59 eli: http://news.cnet.com/8301-13579_3-20004186-37.html
23:59 offby1: shame, I kinda like her