Index

Show enters and exits. Hide enters and exits.

00:07:08chop3 enters the room.
00:08:25KirinDave leaves the room.
00:11:12KirinDav leaves the room.
00:11:45benburkert leaves the room.
00:24:31rubymaverick leaves the room.
00:25:22KirinDav enters the room.
00:30:15chop3 leaves the room.
00:32:45bremac enters the room.
00:33:05wycatsdbussink: wuttup
00:34:33KirinDave leaves the room.
00:37:12jrun enters the room.
00:37:47chop3 enters the room.
00:42:14chop3 leaves the room.
00:49:02KirinDav leaves the room.
00:58:58GMFlash leaves the room.
01:05:33miamoto_musashi enters the room.
01:12:37nicksieger enters the room.
01:27:57sholden enters the room.
01:32:21KirinDav enters the room.
01:35:22lopex leaves the room.
01:48:30chop3 enters the room.
01:49:07onkis enters the room.
01:49:44onkis leaves the room.
01:56:57ezmobius enters the room.
01:57:55mkescher leaves the room.
02:00:28MenTaLguY leaves the room.
02:00:55MenTaLguY enters the room.
02:07:00aasmith leaves the room.
02:10:35KirinDav leaves the room.
02:13:02ubiquitous enters the room.
02:20:00miamoto_musashi leaves the room.
02:20:03miamoto_musashi_ enters the room.
02:21:54miamoto_musashi_ leaves the room.
02:21:59miamoto_musashi enters the room.
02:30:54d2dchat leaves the room.
02:34:46brixenwow, I should go hiking more often
02:34:53brixendbussink went hog wild
02:37:08obvio enters the room.
02:42:13KirinDav enters the room.
02:43:07miamoto_musashi leaves the room.
02:44:32djwhittyeah, his oniguruma checkin seems to have broken the build on my system though
02:44:42djwhitthttp://pastie.caboo.se/163866
02:45:06brixenhmm, what platform?
02:45:13djwhittGentoo amd64
02:45:14brixenoh gentoo
02:45:17djwhittyeah
02:45:21brixenI'll try it
02:46:20_VVSiz_ enters the room.
02:47:01brixenchanging those external libs always seems to create some problems
02:49:05brixenwell, libtoolize does seem to fix it enough to build
02:49:18brixenI'll see what changed when it finished and test those changes on os x
02:52:24VVSiz_ leaves the room.
02:53:43TheVoice enters the room.
02:53:52_mutle enters the room.
02:54:18miamoto_musashi enters the room.
02:55:04djwhittcool
02:57:13miamoto_musashi leaves the room.
02:57:18miamoto_musashi_ enters the room.
02:59:52TheVoice leaves the room.
03:00:13TheVoice enters the room.
03:02:39GMFlash enters the room.
03:07:50dewd_ leaves the room.
03:09:27mutle leaves the room.
03:09:47d2dchat enters the room.
03:10:21benburkert enters the room.
03:16:29TheVoice leaves the room.
03:18:24hoopy_ enters the room.
03:20:46miamoto_musashi_ leaves the room.
03:24:42hoopy_ enters the room.
03:28:58webmat leaves the room.
03:33:42KirinDav leaves the room.
03:33:53nicksieger leaves the room.
03:35:28boyscout1 commit by Brian Ford
03:35:29boyscout * Fix for onig configure process on Gentoo 64.; 3f86dac
03:35:35brixendjwhitt: would you be able to test that?
03:35:47brixenit works for me on gentoo 64, os x intel, and ubuntu gutsy
03:39:11benburkert leaves the room.
03:43:50tlockney leaves the room.
03:44:51binary42 leaves the room.
03:48:18jinjing enters the room.
03:49:52rubuildius_ppcBrian Ford: 3f86dac22; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/163893
03:51:44xif enters the room.
03:58:13KirinDav_ enters the room.
03:58:49xif leaves the room.
04:02:06benburkert enters the room.
04:03:42tlockney enters the room.
04:13:16jbarnette enters the room.
04:15:45radarek leaves the room.
04:15:59radarek enters the room.
04:16:36jbarnette leaves the room.
04:21:53miamoto_musashi enters the room.
04:23:19crafterm enters the room.
04:29:44menator leaves the room.
04:32:44crafterm leaves the room.
04:37:52AndrewO leaves the room.
04:40:37chop3 leaves the room.
04:41:03benburkert leaves the room.
04:51:19nemerle leaves the room.
04:54:44nkpart leaves the room.
04:55:16djwhittbrixen: testing now
04:55:23brixenok, cool
04:57:33djwhittstill failing for me
04:57:34djwhitthttp://rafb.net/p/ncOJNf28.html
04:57:39djwhittdifferent error though
04:58:19djwhittbrixen: gotta sleep now, but I'll see if I can figure it out/help someone debug tomorrow
04:58:21brixenwell, I believe you can safely run that command
04:58:28brixenlibtoolize --copy --force
04:58:36brixenthen create a diff of the ltmain.sh that it creates
04:58:41rueTypinator
04:58:44brixenditch the other changes, and reapply that
04:59:14brixenI cannot imagine that libtool.m4 was the same on os x intel, gutsy, and the gentoo 64 system I generated that on
04:59:19brixenbut it worked on all of them
04:59:32brixenif you make that diff, you could pastie it and I could try it on some systems
04:59:56djwhittok, will do, just a sec
05:01:32jrun leaves the room.
05:01:32paulell enters the room.
05:02:13brixendjwhitt: if you can't pastie because of # of lines, email me at brixen at gmail
05:02:18brixengrabbing some dinner, bbiab..
05:02:22djwhittk
05:02:29djwhittbuilding right now
05:02:34djwhittassuming clean build I'll email it to you
05:02:47mediogre enters the room.
05:04:33wmoxam leaves the room.
05:07:31brixendjwhitt: k, thanks
05:07:49boyscout2 commits by Brian Ford
05:07:50boyscout * Allocate Hash in only one place and use Class.new machinery.; c1211a1
05:07:51boyscout * Removed unnecessary #setup from LookupTable.; fb0230a
05:17:18djwhittbrixen: sent patch
05:17:23djwhittgotta sleep now
05:17:31djwhittbut I'll be around tomorrow to debug/test more if necessary
05:20:14ezmobius_ enters the room.
05:27:43rubuildius_ppcBrian Ford: c1211a1d4; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/163928
05:29:01ezmobius leaves the room.
05:31:56d2dchat leaves the room.
05:34:39d2dchat_ enters the room.
05:35:40menator enters the room.
05:36:36menator leaves the room.
06:00:09benburkert enters the room.
06:05:01paulell leaves the room.
06:06:43srbaker leaves the room.
06:09:43kamal_fariz enters the room.
06:14:22sholden leaves the room.
06:15:23motion enters the room.
06:21:36context leaves the room.
06:23:03boyscout1 commit by David Whittington
06:23:04boyscout * Fix for onig config on Gentoo (second attempt); f7cb6d8
06:23:53brixenthat commit works for me on os x intel, gutsy, and an instance of gentoo 64
06:24:02brixenholler if it doesn't work for you
06:40:40rubuildius_ppcDavid Whittington: f7cb6d866; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/163942
06:47:23MenTaLguY leaves the room.
07:01:19jbarnette enters the room.
07:13:11quik_ enters the room.
07:14:33dbussinkmorning!
07:14:50dbussinkbrixen: i hope you mean hog wild in positive way ;)
07:15:05jbarnette leaves the room.
07:29:15lachie leaves the room.
07:29:46lachie enters the room.
07:30:58KirinDave leaves the room.
07:36:12boyscout1 commit by Dirkjan Bussink
07:36:13boyscout * Fix for failing unpack_sockaddr_in spec on Linux; e5512b2
07:37:28imajes enters the room.
07:56:12rubuildius_ppcDirkjan Bussink: e5512b2a7; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
08:20:09thehcdreamer enters the room.
08:22:23RyanTM leaves the room.
08:30:21benburkert leaves the room.
08:34:56foysavas leaves the room.
08:35:49foysavas enters the room.
08:36:34foysavas leaves the room.
08:40:05foysavas enters the room.
08:43:50boyscout1 commit by Dirkjan Bussink
08:43:51boyscout * Fix Socket.pack_sockaddr_in on BSD systems; 1a22f18
08:46:06zimbatm enters the room.
08:53:04ezmobius_ leaves the room.
08:56:19boyscout1 commit by Dirkjan Bussink
08:56:19boyscout * Fix TCPSocket#new spec, BSD systems make a distinction between IPv6/IPv4 localhost; 9e7fdf3
09:01:04octopod enters the room.
09:08:22aotearoa leaves the room.
09:11:43Fullmoon enters the room.
09:15:34rubuildius_ppcDirkjan Bussink: 9e7fdf3b0; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
09:15:35rubuildius_ppcDirkjan Bussink: 1a22f189d; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
09:33:06Arjen_ enters the room.
09:33:28miamoto_musashi leaves the room.
09:35:07boyscout1 commit by Dirkjan Bussink
09:35:08boyscout * Fix error handling in Socket; 477c40c
09:35:52octopod_ enters the room.
09:38:42perdix enters the room.
09:38:57crafterm enters the room.
09:43:13cavalle enters the room.
09:44:59rubuildius_ppcDirkjan Bussink: 477c40c8e; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
09:45:27crafterm leaves the room.
09:50:48imajes leaves the room.
09:51:41octopod leaves the room.
10:00:41brixen leaves the room.
10:11:18evandjwhitt: you around?
10:12:02boyscout1 commit by Dirkjan Bussink
10:12:03boyscout * Better fix for Socket error handling; be5b11c
10:13:23dbussinkevan: you wanted to ask about building on 64 bit? or something completely different?
10:13:44evanwanted to know why he checked in configure and friends rebuilt using a newer autoconf
10:15:37dbussinkwell, i updated oniguruma and that broke on 64 bit
10:15:44dbussinkso hence the new configure files
10:15:52evanwhy?
10:15:54evanwhy the new configure files?
10:15:59evani don't see why thats relativent
10:16:06evanand the commit message doesn't say why
10:16:36dbussinkreally don't know that, was asleep when it happened
10:16:44evanis it working now?
10:16:55dbussinkworks for me on every system i have
10:17:00aotearoa enters the room.
10:17:00evanok
10:17:03dbussinkos x, linux 64, freebsd, openbsd
10:17:12evantypically checking in a new configure is a mistake
10:17:15evanthats why I was curious.
10:18:01crafterm enters the room.
10:20:43craftermview from where the sydney sprint took place: http://www.flickr.com/photos/crafterm/2323895048/
10:20:47crafterm:)
10:21:21GMFlash leaves the room.
10:21:37dbussinklooks nice :)
10:23:04_mutle leaves the room.
10:24:24rubuildius_ppcDirkjan Bussink: be5b11cf3; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
10:24:29mutle enters the room.
10:26:17boyscout1 commit by Dirkjan Bussink
10:26:18boyscout * Be sure not to close the socket before the Errno.handle call; 25e3f23
10:27:21dbussinknow all ipaddr / socket fixes also work on other platforms :)
10:29:17dbussinkhas there ever been a sprint in europe?
10:39:35rubuildius_ppcDirkjan Bussink: 25e3f23e7; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
10:41:49binary42 enters the room.
10:45:00boyscout2 commits by Marcus Crafter
10:45:01boyscout * Extract dtrace probe implementations into a common file for better reuse and code ...; 1d19f68
10:45:02boyscout * Add example dtrace script for examining garbage collection; f646e40
10:46:06JimMc enters the room.
10:48:55dbussinkevan: when is your trip back to the us?
10:49:03craftermdbussink: i don't think so mate, something for to organise next :)
10:49:36dbussinki'd be willing to help with that :)
10:50:47dbussinkmakes it easy to do it just here in the netherlands :P
10:51:04crafterm:)
10:51:10evani'm headed back to the us in about 12 hours
10:53:58headiusevan: check this out
10:54:11boyscout1 commit by Evan Phoenix
10:54:11boyscout * Add subcommand's for rdoc and ri.; 49d7f1d
10:54:20headiussource: http://pastebin.com/m709f22f1
10:54:34rubuildius_ppcMarcus Crafter: 1d19f6842; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
10:54:41headiuscompiled output: http://pastebin.com/m561d03d4
10:55:57quik_ leaves the room.
10:56:03headiusthat may be hard to follow if you're not familiar with JVM bytecode
10:57:48evannot too hard, no
10:57:50evanpretty cool!
10:58:04warren_s leaves the room.
11:00:46dbussinkheadius: an experiment or did you guys already have this in?
11:00:56headiusI just wrote it
11:01:06headiusstarted about 10 hours ago
11:01:24headiusit's a type-inferring compiler for a ruby-like subset language to JVM bytecode
11:01:43headiuswith support for primitives as well
11:01:55dbussinkyeah, i see that :)
11:01:57dbussinknice :)
11:05:22rubuildius_ppcEvan Phoenix: 49d7f1d53; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors;
11:05:31evanheadius: the compiler is written in java?
11:05:41headiusno, ruby
11:05:52headiususes our Java integration layer pretty heavily though
11:05:56headiusand our parser
11:06:08evanso, i see that it bound String#+ to concat
11:06:11headiusyes
11:06:13evanwhat rules are you using to do that
11:06:14headiusmagic!
11:06:21evanmagic rules!
11:06:30headiusif I see a + method being called against String, I remap the name
11:06:40headiustypes are present throughout the system, so it's not that hard
11:06:43evanso, specific code that does that?
11:06:43evanie
11:06:54headiusit also goes out and looks at the target types to get their return values for additional inference
11:06:55evanif method_name == :+ and kind == :String; ...
11:07:07headiusyeah, roughly speaking
11:07:18evanso the rules are hardcoded?
11:07:18headiusit has to go through the inference logic to determine it's called on a string
11:07:38evanright, but the interence logic is a tree decorator
11:07:39headiuswell, for the moment...it would be trivial to provide a mapper
11:08:05headiusthe method name translation is just a "nice to have" I threw on
11:09:41evansure, just curious.
11:11:51headiusif it continues to pan out, there won't be any reason we can't rewrite parts of JRuby in Java
11:11:52headiuser
11:11:58headiusin "Ruby" in this form
11:14:08evanyou're using one of the java libraries that creates bytecode, yes?
11:14:34headiusyes
11:14:44headiusthough it's mostly to handle the bookkeeping
11:14:59headiusI call one method per bytecode I emit
11:15:13headiuswith some smoke and mirrors to map it all nicely in a Ruby wrapper
11:16:48cremesheadius: will these changes be part of 1.1 or are they an early look at 1.2 feature sets?
11:17:07headiusthis is for post 1.1 work
11:17:15cremesi see
11:17:18headiusthough the underlying files will probably sneak out in 1.1
11:18:01cremesi do quantitative programming for a trading company... i have near-term plans to use jruby for some trading algos that need to interface w/java
11:18:05headiusI'm just looking to a future where we can implement more of JRuby in "mostly Ruby" and have no perf hit at all
11:18:17headiuswell interfacing with Java is easy right now
11:18:32headiusthis is more if you want to implement some piece of code in ruby-like code that runs at Java speeds
11:18:45cremesi am just starting to look at it; i may drop by the jruby channel early next week for some help
11:18:55ctennis leaves the room.
11:19:33cremes"java speeds" sounds good!
11:20:32headiusyeah!
11:21:07geekounet leaves the room.
11:22:54headiusfor example
11:23:16lachieevan: http://pastebin.com/d283f497a here be my key!
11:23:27headiushttp://pastebin.com/m3ba79a4e
11:23:41headiusJRuby runs fib(35) in about 4-5s
11:24:21cremespretty slick
11:24:25evanstupid pastebin doesn't have a txt option.
11:24:30dewd enters the room.
11:24:39lachiepastie seems down
11:24:43lachietried there first
11:24:49evanno prob
11:25:12evanlachie: ok, added!
11:25:19evandon't screw up your git repo again!
11:25:25lachieevan: cool, thanks
11:25:32lachieheh, i'll try
11:25:35lachienot to
11:27:10dbussinklachie: problems at the sprint?
11:27:44lachiei think evan was the only one there who knew git
11:27:58lachiemy tree got a bit tangled somehow
11:29:26lachiedbussink: to fix it we reset it hard back to before the tangles :P
11:29:50dbussinkbrute force is always a good solution :)
11:30:47evanlachie: work on a branch
11:30:50evanyou wont get tangled.
11:30:58evanand use the rake tasks
11:31:01evanrake git:pull
11:31:04evanto update your branch
11:31:06evanand
11:31:08evanrake git:push
11:31:14evanto push the branch back to the world
11:31:56cjheath enters the room.
11:32:12lachieevan: ta
11:32:31rubuildius_ppc leaves the room.
11:33:30rubuildius_ppc enters the room.
11:33:33dbussinki can't live without the rake tasks
11:34:27GMFlash enters the room.
11:34:27rubuildius_ppc leaves the room.
11:34:38lachiei'm using git for almost everything now... not sure how to accelerate my learning of its finer points
11:34:42cremes leaves the room.
11:34:57cjheath leaves the room.
11:35:05evani'd read the using git tutorials a bit
11:35:18lachieor learning how to recover gracefully from the mess I got myself into... there must have been a better way
11:35:21evanlearn what the individual commands do, that helps a lot.
11:35:29evannot really
11:35:32evanyou really tangled it.
11:36:10lachieor learning the git-discipline not to get there in the first place (rake tasks notwithstanding)
11:36:33lachieyeah, i've had a look at a few
11:36:37lachiegot the peepcode :)
11:38:10olabini leaves the room.
11:41:34rubuildius_ppc enters the room.
11:48:32boyscout1 commit by Marcus Crafter
11:48:33boyscout * Refactoring, added _source_location for obtaining the current source location and ...; dc95ff1
11:53:39quik_ enters the room.
11:54:32Fullmoon leaves the room.
12:00:01binary42 leaves the room.
12:03:37rubuildius_ppcMarcus Crafter: dc95ff1cd; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/163970
12:03:38rubuildius_ppcEvan Phoenix: 49d7f1d53; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/163968
12:09:49chop3 enters the room.
12:11:21tizianobis enters the room.
12:25:36krisps enters the room.
12:30:04chop3 leaves the room.
12:59:32d2dchat_ leaves the room.
13:00:05gnufied leaves the room.
13:00:53d2dchat enters the room.
13:05:38olabini enters the room.
13:05:50tizianobis leaves the room.
13:08:49ubiquitous leaves the room.
13:09:11d2dchat leaves the room.
13:10:43wmoxam enters the room.
13:11:24cavalle leaves the room.
13:24:26boyscout1 commit by Vladimir Sizikov
13:24:27boyscout * Added JRuby speciifc guard to singleton rubyspecs (JRUBY-2239).; 218cc7f
13:28:57dodecaphonic enters the room.
13:33:01wmoxam leaves the room.
13:33:22dewd leaves the room.
13:33:46rubuildius_ppcVladimir Sizikov: 218cc7fbd; 1741 files, 5924 examples, 20092 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/163993
13:34:30skaar enters the room.
13:35:26pauldix enters the room.
13:37:04webmat enters the room.
13:45:01cavalle enters the room.
13:49:35quik_ leaves the room.
13:49:55krisps leaves the room.
14:00:09jinjing_ enters the room.
14:04:14crafterm leaves the room.
14:04:43jinjing__ enters the room.
14:06:01AndrewO enters the room.
14:09:30AndrewO leaves the room.
14:12:16jinjing leaves the room.
14:19:16jinjing_ leaves the room.
14:19:23AndrewO enters the room.
14:22:58wmoxam enters the room.
14:25:16cremes enters the room.
14:26:13jinjing__ leaves the room.
14:28:55pauldix_ enters the room.
14:36:07djwhittevan: I'm here now
14:37:50djwhittevan: I actually didn't check that in. I sent the patch to brixen to test and he checked it in ;)
14:38:31djwhittdbussink: you around?
14:42:13enebo enters the room.
14:45:19pauldix leaves the room.
14:47:00jinjing enters the room.
14:49:00benburkert enters the room.
15:01:02d2dchat enters the room.
15:12:34chris2 enters the room.
15:14:13dewd enters the room.
15:14:55srbaker enters the room.
15:15:30nicksieger enters the room.
15:24:01benburkert leaves the room.
15:31:15webmat leaves the room.
15:36:18panpot enters the room.
15:41:09pauldix_ leaves the room.
15:46:15tizianobis enters the room.
15:48:35riffraff enters the room.
15:48:39riffraff leaves the room.
15:56:29yipstar enters the room.
15:57:49dbussinkdjwhitt: i am now
15:58:59djwhittyeah, I'm trying to figure out what exactly happened with that onig change... after looking at it further I don't think I needed to rebuild the config stuff
15:59:57djwhittit did break after the new version of onig but when looking at the diffs I can't figure out why
16:00:26djwhittI wonder if something is being gitignored that shouldn't be
16:01:24lopex enters the room.
16:01:31dbussinki really don't know, i didn't have any problems with the old version on 32 bit systems
16:01:52dbussinkbut it works now on a whole variety of systems, so for me it's ok now
16:02:33djwhittyeah, seems to be building fine everywhere right now. I just want to figure out why it broke in the first place
16:11:22dbussinkVVSiz: i see you updated some specs, you might need to look at one of my commits from yesterday, ad8c630662
16:11:30dbussinkVVSiz: dunno what jruby does there
16:11:41dbussinkVVSiz: the bug has been confirmed for mri
16:11:54webmat enters the room.
16:17:22srbaker leaves the room.
16:17:25srbaker_ enters the room.
16:32:14olabini leaves the room.
16:35:07VVSizdbussink: thanks for the update, will do. In fact, I was already doing that (some new JRuby failures)
16:39:21wifelette_ enters the room.
16:39:22wifelette leaves the room.
16:39:24aotearoa leaves the room.
16:52:22binary42 enters the room.
16:54:35RyanTM enters the room.
16:55:11brixen enters the room.
16:57:55chris2 leaves the room.
16:58:32VVSizbrixen: good morning :) Is there a way to print out a list of excluded entries, *including* the comments?
16:59:22brixenhey VVSiz
16:59:26brixenhmm, not yet :)
16:59:31VVSiz:)
16:59:35brixenyou can use -fs --dry-run to print the spec strings
16:59:41brixenbut nothing prints the tags
16:59:49VVSizyeah, that's what we use right now
17:00:03brixenwouldn't be hard to add one though
17:00:54panpot leaves the room.
17:01:19VVSizcurrently, we add the JRUBY-XXXX issue tracker references to every new exclude, and we have 'show-excludes' command, so it might be pretty useful to see not only actual spec strings but associated bug numbers
17:01:46brixensure, makes sense
17:02:02brixenwhat to tack a whack at implementing it, since you've got the use case? :)
17:02:09brixenor take*
17:02:39VVSizI'm not touching this stuff today, have a high temperature and don't trust myself much :)
17:02:47brixenheh, ok
17:02:54brixenneed to run some quick errands
17:02:58VVSiz:)
17:03:00brixenI'll take a look a bit later
17:03:07brixenI think I understand what you need
17:03:12VVSizthanks!!
17:03:22brixensure, thanks for all the help
17:05:54Jzalae leaves the room.
17:06:13vborja enters the room.
17:10:57DefilerHas anyone other than crafterm worked on the Rubinius DTrace integration?
17:11:57DefilerNever mind. According to 'git log -i -Sdtrace' the answer is no
17:16:38djwhitt enters the room.
17:19:42lopex leaves the room.
17:23:35dbussinkbrixen: i haven't gone haywire today :P
17:30:52lopex enters the room.
17:32:48KirinDave enters the room.
17:33:25Defilerheadius: Typo in your latest blog: 1: invokespecial #15; //Method java/lang/Object."&ltinit>":()V
17:39:01rubymaverick enters the room.
17:42:33tmornini enters the room.
17:43:46tmorninihello all
17:43:55tmornini leaves the room.
17:44:07tmornini enters the room.
17:44:43thehcdreamer leaves the room.
17:46:43DefilerMorning
17:50:26benburkert enters the room.
17:52:29jbwiv_ enters the room.
17:58:59cavalle leaves the room.
18:01:52headiusDefiler: thank you
18:02:17headiusblogger doesn't auto-escape angles, very irritating
18:04:53octopod leaves the room.
18:05:33jbarnette enters the room.
18:07:00zimbatm leaves the room.
18:07:32xif enters the room.
18:18:05cjheath enters the room.
18:21:45yipstar leaves the room.
18:22:07obvio leaves the room.
18:23:59headiusfixed!
18:24:18MenTaLguY enters the room.
18:25:20crossblaim enters the room.
18:25:30crossblaimhi
18:27:02djwhittheadius: do you think there is any possibility Duby could be made generic enough to serve as a low-level language for Rubinius as well
18:27:38headiusthe language itself, definitely
18:27:49headiusI've talked about it with evan before, but he didn't seem interested at the time
18:28:00djwhittseems like it's very similar to what garnet is supposed to be
18:28:30headiusyeah, I think it was, but garnet's never gone anywhere, and squeak (for example) has its language "slang" that's much, much lower-level
18:28:45headiusI'm not sure if garnet is intended to be more at the slang level or the duby/rpython level
18:28:51djwhittinteresting, I've heard of slang before but never looked at it myself
18:29:30djwhittwould you need something much lower level than duby/rpython?
18:29:39djwhittseems like the PyPy guys are doing pretty well with rpython
18:29:42headiusslang basically is a super limited subset of smalltalk that gets compiled to integer operations against the in-memory image
18:29:52tarcieriDuby?
18:29:58tarcierisounds like drug paraphenalia
18:30:01headiusso it can be made really fast, but it's way too low-level to be useful on the JVM
18:30:03djwhitttarcieri: http://headius.blogspot.com/
18:30:09tarcieriyeah, reading
18:30:31djwhittah, I see
18:31:26djwhittif I understand correctly though the pypy guys get away with working at a pretty high level with rpython
18:31:42headiusI think the hardest part of implementing Duby for Rubinius would be the type inference
18:31:54headiussince there's little or no reflective capabilities in normal C
18:31:57tarcierirpython needs a better name
18:32:04djwhittrython ?
18:32:08tarcierimonty!
18:32:09headiustarcieri: I almost called Duby PRuby
18:32:30headiusfor them I'd look to genus and species names
18:32:40headiusworld's smallest species of python or something
18:33:04tarcierireally likes monty :)
18:33:44probablycorey enters the room.
18:35:47headiusAntaresia perthensis
18:35:56headiussmallest member of the python family
18:36:02headius"Perthensis"
18:36:18headiusmaybe mutate it into Parenthesis
18:36:24djwhittheh
18:36:33djwhittno I think that's lisp ;)
18:36:52tizianobis leaves the room.
18:37:30tarcierithey could always just call it pygmy
18:39:04headiusyeah, that's not bad either
18:52:20headiusat any rate, I will modify Duby to have some basic aliases for common types like strings and integers, and a name-importing mechanism to shorten other type declarations
18:52:49headiusso outside of specific types entering the name imports, the base set of types ought to be translatable to C and used by rubinius
18:53:27perplexes enters the room.
18:54:16TheVoice enters the room.
18:56:32wycats leaves the room.
18:59:26dkubb leaves the room.
19:00:21radarek leaves the room.
19:00:36TheVoice leaves the room.
19:03:02TheVoice enters the room.
19:13:32thehcdreamer enters the room.
19:15:28yipstar enters the room.
19:16:50d2dchat leaves the room.
19:17:20d2dchat enters the room.
19:19:11d2dchat leaves the room.
19:21:21d2dchat enters the room.
19:22:12TheVoice leaves the room.
19:22:39TheVoice enters the room.
19:25:53srbaker_ leaves the room.
19:27:37cavalle enters the room.
19:33:43srbaker enters the room.
19:38:27wycats enters the room.
19:38:36ezmobius_ enters the room.
19:38:44d2dchat leaves the room.
19:39:14d2dchat enters the room.
19:39:45pietia enters the room.
19:40:28Fullmoon enters the room.
19:45:11perdiy enters the room.
19:46:29dkubb enters the room.
19:46:59Fullmoon leaves the room.
19:47:18TheVoice leaves the room.
19:48:29TheVoice enters the room.
19:48:40TheVoice leaves the room.
19:49:41TheVoice enters the room.
19:50:04TheVoice leaves the room.
19:50:56TheVoice enters the room.
19:51:19TheVoice leaves the room.
19:52:12TheVoice enters the room.
19:52:35TheVoice leaves the room.
19:53:28TheVoice enters the room.
19:53:51TheVoice leaves the room.
19:54:43TheVoice enters the room.
19:55:06TheVoice leaves the room.
19:55:58TheVoice enters the room.
19:56:21TheVoice leaves the room.
19:57:13TheVoice enters the room.
19:57:36TheVoice leaves the room.
19:58:27TheVoice enters the room.
19:58:50TheVoice leaves the room.
19:59:27perdix leaves the room.
19:59:42TheVoice enters the room.
20:00:04TheVoice leaves the room.
20:00:57TheVoice enters the room.
20:01:20TheVoice leaves the room.
20:17:39mediogre leaves the room.
20:22:18srbaker_ enters the room.
20:23:20srbaker leaves the room.
20:23:23srbaker_ leaves the room.
20:24:00srbaker enters the room.
20:26:29srbaker_ enters the room.
20:26:40srbaker leaves the room.
20:30:52brixen_ enters the room.
20:31:50pietia leaves the room.
20:32:04benburkert leaves the room.
20:32:27brixen_ leaves the room.
20:34:21seydar enters the room.
20:34:45headius_ enters the room.
20:39:26seydari'm sorry for not doing anything recently
20:39:34seydarplease forgive me! i have finals :-(
20:39:48seydarin like, an hour
20:39:52seydar leaves the room.
20:40:34TheVoice enters the room.
20:40:39headius leaves the room.
20:41:07headius enters the room.
20:41:56imajes enters the room.
20:42:51brother_rspec enters the room.
20:43:28srbaker_ leaves the room.
20:47:35mae leaves the room.
20:51:31lachie leaves the room.
20:51:59cremes_ enters the room.
20:52:36thehcdreamer leaves the room.
20:54:13quik_ enters the room.
20:54:42jbwiv_ leaves the room.
20:56:04_eric leaves the room.
20:56:06wmoxam leaves the room.
20:57:13headius_ leaves the room.
21:06:02GMFlash leaves the room.
21:14:41skaar leaves the room.
21:18:31TheVoice leaves the room.
21:18:42TheVoice enters the room.
21:27:46paulell enters the room.
21:31:01benburkert enters the room.
21:31:36headiushey, I dunno if you guys have read this one: http://research.sun.com/techrep/2000/abstract-88.html
21:31:45headiusabout the concurrent mark/sweep GC in JVM
21:32:02headiusit's a bit slower than other GCs but it does the most to minimize pauses
21:32:26headiusI think azul has a pauseless concurrent collector too, but I haven't looked for any papers on it
21:32:45ezmobiusheadius: that duby stuff looks pretty interesting
21:33:23headiusahh yeah, evening project, but it seemed to go well
21:33:46headiusideally it will make it possible to implement much more of JRuby in Ruby, and still have it perform as it does today
21:35:43crafterm enters the room.
21:38:35rmm5t enters the room.
21:44:01edward_g leaves the room.
21:49:34Defilercrafterm: Yo. Got a sec to talk about DTrace?
21:50:04brother_rspec leaves the room.
21:50:14lachie enters the room.
21:50:42Defilercrafterm: Where are the names of the probes that Rubinius supports defined? I don't see, for example, 'function-entry' anywhere.. and DTrace says it can't find a probe with that name when I try to use the example scripts under tools/dtrace
21:56:50agardiner enters the room.
21:56:58benburkert leaves the room.
21:58:20paulell leaves the room.
21:59:49benburkert enters the room.
22:01:30probablycorey leaves the room.
22:01:31dodecaphonic leaves the room.
22:10:04webmat leaves the room.
22:10:40cavalle_ enters the room.
22:13:30TheVoice leaves the room.
22:13:30dkubb leaves the room.
22:13:30ezmobius leaves the room.
22:13:30wycats leaves the room.
22:13:30cavalle leaves the room.
22:13:30tmornini leaves the room.
22:13:30lopex leaves the room.
22:13:30rubuildius_ppc leaves the room.
22:13:30geekounet leaves the room.
22:13:30jero5 leaves the room.
22:13:30nitay leaves the room.
22:13:30cout leaves the room.
22:13:30kAworu leaves the room.
22:13:30fearoffish leaves the room.
22:13:30ko1_away leaves the room.
22:13:30lenbust leaves the room.
22:13:30rudebwoy leaves the room.
22:13:57TheVoice enters the room.
22:13:57dkubb enters the room.
22:13:57ezmobius enters the room.
22:13:57wycats enters the room.
22:13:57tmornini enters the room.
22:13:57lopex enters the room.
22:13:57rubuildius_ppc enters the room.
22:13:57geekounet enters the room.
22:13:57jero5 enters the room.
22:13:57kAworu enters the room.
22:13:57rudebwoy enters the room.
22:13:57lenbust enters the room.
22:13:57nitay enters the room.
22:13:57cout enters the room.
22:13:57fearoffish enters the room.
22:13:57ko1_away enters the room.
22:25:03benburkert leaves the room.
22:25:27benburkert enters the room.
22:34:03perdiy leaves the room.
22:36:17crossblaim leaves the room.
22:36:49crossblaim enters the room.
22:37:02mkescher enters the room.
22:43:54Arjen_ leaves the room.
22:46:56boyscout1 commit by Wilson Bilkovich
22:46:57boyscout * Correctly set the enclosing class for evaled code.; 746d89d
22:48:19geekounet leaves the room.
22:50:19benburkert leaves the room.
22:50:29rmm5t leaves the room.
22:55:38josb leaves the room.
22:56:51dgtizedanyone else getting a segfault?
22:57:39DefilerNope. Paste it?
22:58:03dgtizedhmm, ran it again and got a different error
22:58:12dgtizedi'll see if I can reproduce
22:58:53dgtizedwhat happened to -C to clean out the .rbc files?
22:59:17DefilerIt is still there, but the help output is totally retarded now
22:59:37brixen-C is no longer there
22:59:40d2dchat leaves the room.
22:59:43brixenunless someone added it back
22:59:48dgtizedwhy isn't it?
23:00:04brixenbecause it's not effective, since it only looked for _spec.rbc
23:00:13brixenwe need a force compile flag to the compiler
23:00:28dgtizedor just set the compile time flag to now
23:00:33brixenwhy do you need -C?
23:00:36Defilerbrixen: Would you make a ticket for that and assign it to me?
23:00:42brixenDefiler: sure
23:00:44dgtizedbecause the first time I ran it bin/mspec in a while it segfaulted
23:00:45rubuildius_ppcWilson Bilkovich: 746d89d6d; 1741 files, 5916 examples, 20084 expectations, 0 failures, 0 errors;
23:00:54dgtizedand the second time it did something else
23:01:02dgtizedthere should be a way to clean slate the specs
23:01:16brixenhave you run it since the .rba's were changed from zip to ar?
23:01:26dgtizedI last recompiled friday
23:01:27brixenif not, rm -rf spec && git checkout spec
23:01:42brixenI don't know what last friday was in terms of timeline ;)
23:02:37brixenanyone know what TZ lighthouse is in?
23:02:43brixenthinks we're tuesday
23:02:57DefilerProbably UTC
23:03:30dbussinkit is tuesday ;)
23:03:34crossblaim leaves the room.
23:03:34quikyes, it is
23:03:44mutle leaves the room.
23:03:48DefilerYeah, it is Tuesday in more places than it is Monday, at the moment
23:04:07brixenDefiler: so, I see you resolved #354, are you going to push the changes in the instance_eval branch, or shall I?
23:05:02mutle enters the room.
23:05:07dbussinkbrixen: i was wondering about replacing the RUBY_CONFIG hash with lookuptable, but i guess that means even more c functionality needs to be exposed?
23:05:31dbussinkor is that not going to be replaced?
23:06:41Defilerbrixen: I haven't looked at that branch. Might want to hold off on it, though.. just got re-acquainted with another case that doesn't work
23:06:45DefilerThough that one at least is fixed
23:07:14brixendbussink: that's not going to be converted, we do too much with the keys in Ruby code
23:07:38brixendbussink: there's no real advantage, it's not performance critical and we don't really need key.to_s all over
23:07:48brixenDefiler: ahh ok
23:08:12dbussinkbrixen: ah ok, well, any other candidates for being replaced?
23:08:38brixenthe last one I'm looking at is the symbol table, then I'll clean up the Hash C/Ruby a bit
23:09:34brixendbussink: want to take a stab at integrating the FreeBSD gdtoa dir?
23:10:13dbussinkbrixen: it's horrible :P
23:10:18brixendbussink: looks like 1.9 uses both strtod and dtoa, but they imported the version before the gdtoa dir
23:10:33dbussinkbrixen: well, it's a pretty big mess
23:10:37brixenyeah, I know, but unless you have a better way to get x-platform MRI compatibility :P
23:10:51dbussinkbrixen: i was thinking we might grab an older version from freebsd/netbsd
23:11:06dbussinkbrixen: that is sufficient
23:11:32brixenwell, I looked at freebsd 7.0 code, that ways it is in its own dir
23:11:38brixeneither way it's a bit of a mess
23:11:56brixenthere must be a reason freebsd 7.0 went to all that trouble
23:11:58brixen;)
23:12:27dbussinkhehe, well, i don't know where we want to draw the line wrt for example platform support
23:12:43brixenright, kind of following mri lead here
23:13:17brixeniirc, there was something in that changelog about using the imported strtod to avoid locale hell
23:13:36dbussinkyeah, that's what i've seen as the major argument for this
23:14:00dbussinki think the fact that it was more platform independent, was just a side advantage
23:14:03dgtizedok doing rm -rf spec && git checkout fixed it
23:14:40brixendgtized_: yeah, that was because of the change to ar, -C wouldn't have helped you any ;)
23:15:23cored enters the room.
23:15:52dbussinkbrixen: but i'm not gonna take this on atm, first some sleep
23:16:07dkubb leaves the room.
23:16:08brixendbussink: heh, ok, night!
23:16:10dbussinkbrixen: tomorrow last day of preparation for my thesis talk
23:16:40brixenohh, that should be priority :)
23:17:04dbussinkyeah, but i sometimes need to get my mind of it for some time :P
23:17:06jrun enters the room.
23:17:32brixendbussink: heh, we're always here for some diversion :D
23:19:04dbussinkhehe, bit strange always those changes after some years, officially not a student anymore
23:20:01craftermDefiler: ping?
23:21:27craftermthe probes are defined in the rubinius.d file, which a header is generated from - however dtrace support is only compiled in if you build rubinius with a ./configure --enable-dtrace at the moment
23:21:50brixencrafterm: I did that, but still no luck
23:21:58brixencrafterm: http://rafb.net/p/odoIhf24.html
23:22:14craftermah, ok
23:22:15brixenI did: ./configure --enable-dtrace; make
23:22:19brixenthen ran that
23:22:23craftermdo shotgun/rubinius.local.bin
23:22:32brixenk, sec...
23:23:03craftermthe shotgun/rubinius runs a script (which dtrace will get as the PID) which starts rubinius
23:23:23craftermwhereas rubinius.local.bin is the vm binary directly
23:24:14brixenhmm, must need some env: Unable to find a bootstrap to load!
23:24:39brixenso, I just need to make sure dtrace gets the right PID?
23:24:57craftermyes.. the -c bit is the command to run, that PID gets passed to the dtrace script as the $target variable
23:25:36brixenok
23:26:59Defilercrafterm: So, are you saying that DTRACE=1 rake isn't the right way to enable it?
23:27:24craftermright - ./configure --enable-dtrace && rake should set all that up now
23:28:21brixenDefiler: odd, the --dtrace command to shotgun is complaining: dtrace: failed to open -p: No such file or directory
23:28:30brixenyet dtrace --help says -p is for a PID
23:29:37brixenoh wait, you can't do it like that, can you? nothing check whether you are in child or parent
23:29:38rmm5t enters the room.
23:30:02brixenneeds to review fork in Ruby
23:31:24crafterm"Unknown primitive archive_files" - you guys seen that before?
23:31:52brixenok, fork with a block, nm
23:32:01brixencrafterm: what are you running?
23:32:20brixencrafterm: also, you might try: rm -rf spec && git checkout spec
23:32:25brixenif you are trying to run specs
23:32:48craftermbrixen: http://rafb.net/p/ujgjxA45.html
23:32:52Arjen_ enters the room.
23:35:25brixencrafterm: I think your stables got wonky, iirc drbrain removed that a few days ago
23:35:31brixenle'me see if I can find the commit
23:35:33craftermhmm
23:35:34craftermok
23:35:40crossblaim enters the room.
23:35:42craftermi'll clean everything out
23:36:20brixencrafterm: yeah, a63f45782
23:36:25brixenrake clean build
23:36:37craftermcool mate
23:38:05RyanTM leaves the room.
23:39:35dgtizedman we have a lot of primitives now
23:39:50RyanTM enters the room.
23:40:08crossblaim leaves the room.
23:40:35headiusyou noticed that eh?
23:41:00dgtizedit's too bad FFI is slower then primitives, it would be really nice for things like lookuptable if they were just basically there own C extension
23:41:45headiusI haven't looked at the primitives list in a while
23:41:49djwhittI don't think that would be possible with Lookuptable anyway
23:41:50GMFlash enters the room.
23:41:56djwhittbecause it's needed inside the VM
23:42:10rubymaverick leaves the room.
23:42:37dgtizedwell it could dynamic link
23:43:00dgtizedbut expose it's interface to kernel code using FFI?
23:43:10headiuswow, that is a long list
23:43:12dgtizedI don't know the load order timing on these things
23:43:23cremes_ leaves the room.
23:43:33dgtizedyea, and now we have lots of related primitives that aren't next to eachother
23:43:51brixenyou can reorder those if necessary
23:43:57brixenit calculates the index
23:44:01dgtizeddo you have to rebuild stable?
23:44:22brixenI don't believe so
23:44:28brixeneasy enough to try
23:44:35Defilerbrixen: You have to set the SCRIPT env variable to a path to a dtrace script right now
23:44:41Defilerbrixen: we should change it to --dtrace=foo.d
23:44:49brixenDefiler: yeah, just realized that
23:44:57brixenI don't know how to set SCRIPT with sudo
23:45:15Defilersudo env SCRIPT=foo ./shotgun/rubinius blah
23:45:17brixen--dtrace=foo.d would be much better
23:45:19brixenahh env
23:45:20Defiler(terrible)
23:45:26brixenthat helpful little command
23:45:37dgtizedbrixen, could we just add a macro to primitives.rb that generates the index when primitives.rb get's loaded in so we don't need to explicitely define the primitives list in a second place?
23:46:09brixendgtized_: you could try that
23:46:45dgtizedk, I'll see what I can do
23:46:52brixencool
23:47:02rmm5t leaves the room.
23:47:32dgtizedbrixen: wait so I'm guessing the primitive index is put into compiled files or something right?
23:47:42dgtizedbrixen: instead of the primitive name?
23:47:54brixendgtized_: take a look at the .gen files
23:50:30binary42 leaves the room.
23:50:55dgtizedbrixen: system_primitives.gen?
23:51:22dgtizedbrixen: I understand how it's using the index for the opindex offset, what I"m not certain about is if that gets propegated to compiled .rbc files
23:51:32dgtizedor if the rbc file just uses the name
23:51:34agardinerdgtized_: primitive indexes are not in .rbc files
23:51:43agardinerthese store the primitive symol
23:51:57agardinerbut this is converted to the primitive index when the rbc is unmarshaled
23:52:23dgtizedok so the only time the index is used is at runtime with respect to the compiled order of the offsets in the VM?
23:52:40agardineri believe that's correct
23:53:08agardineryou can reorder the primitive_names.rb and just recompile shotgun and things should work
23:53:38agardineri think evan has a todo to just eliminate primitive_names.rb
23:54:02dgtizedagardiner: ok, then i will just do that now if I can
23:54:08agardineryeah, sounds good
23:54:35KirinDave leaves the room.
23:54:36Arjen_ leaves the room.
23:57:06agardinerdgtized_: ahh, just checked when the primitive index is assigned
23:57:15brixenagardiner: in the debugger, with d, it doesn't print the label for 'goto'.
23:57:33cored leaves the room.
23:57:37agardinerit isn't at unmarshal time, its when the primitive is encountered in a compiled method in _cpu_try_primitive
23:57:40craftermDefiler: 'Unable to find a bootstrap to load!' - was that what you were getting mate?
23:58:05brixencrafterm: I got that trying to use rubinius.local.bin
23:58:20craftermbrixen: ok - i get it as well now
23:58:49agardinerbrixen: the label isn't saved in the compiled method
23:59:08agardinerthey are resolved to IP addressed when the bytecode is generated
23:59:10brixencrafterm: seems like getting the --dtrace cmd line opt to shotgun would be handy
23:59:13boyscout1 commit by Wilson Bilkovich
23:59:14boyscout * Fix class_open scoping inside EvalExpressions; a9f5d7a
23:59:23Defilercrafterm: No, that is not what I was getting
23:59:25brixenagardiner: ok, had a feeling
23:59:35craftermbrixen: definitely
23:59:46Defilercrafterm: I just got dtrace telling me it couldn't find a probe called rubinius<the_pid>:::function-entry
23:59:48agardineri could assign my own labels though, and tag lines with those labels if you think it would help