Index

Show enters and exits. Hide enters and exits.

00:00:49luislavenafbuilesv: no problem, let me see.
00:03:38luislavenafbuilesv: 1 file, 3 examples, 3 expectations, 0 failures, 0 errors
00:03:46luislavenaruby 1.8.6 (2008-03-03 patchlevel 114) [i386-mswin32]
00:03:51luislavenathat will fail in mingw32, btw.
00:04:07luislavenaand mingw32 is the upcoming One-Click installer, just FYI
00:04:40fbuilesvluislavena: platform_is_not :mswin32, :mingw32 would satisfy it?
00:04:45dormando enters the room.
00:05:08luislavenafbuilesv: but will fail against mswin64 (there is a ruby build with VC9) :-P
00:05:13fbuilesvhah
00:05:24luislavenafbuilesv: mswin, mingw will be enough.
00:06:35luislavenaI'll see what could be wrong with the pipes on ruby later :-P
00:06:43luislavenais not good these aren't working :P
00:07:29kirindav_ enters the room.
00:09:33kirindav_ leaves the room.
00:10:57brainopia enters the room.
00:12:54boyscout1 commit by Federico Builes
00:12:54boyscout * Adds MinGW to the IO#popen spec guard; edd397c
00:13:08fbuilesvadamwiggins: I just pushed your patch, as rue said, thanks, pass your public key to evan if you want commit bit
00:13:25fbuilesvoops I missed something there
00:14:20luislavenafbuilesv: yep, you missed attribution ;-)
00:14:28fbuilesvluislavena: no, that was a minipatch I added
00:14:29brainopia enters the room.
00:14:36fbuilesvadam patch was applied under his name
00:14:43adamwigginsfbuilesv : Nice, you guys sure don't waste time
00:14:50fbuilesvbut somehow it didn't show up in here, although it is in http://git.rubini.us/?p=code;a=summary
00:15:36luislavenaadamwiggins: yeah, I love that... :)
00:17:03fbuilesvluislavena: sorry I missed your name in there, you'll find it here: http://rubinius.lighthouseapp.com/projects/5089-rubinius/tickets/506-specs-for-io-popen#ticket-506 -3
00:17:18luislavenafbuilesv: no need to include me.
00:18:01fbuilesvluislavena: still did, you'll bask in the glory of the almighty Lighthouse and you'll enjoy it!
00:18:13luislavenafbuilesv: lol
00:18:40lstoll enters the room.
00:19:01smparkes enters the room.
00:19:07twbray enters the room.
00:19:15luislavenaguys, it seems there is no Dir.tmpdir spec?
00:20:35luislavenaI want to be safe to require 'tmpdir' in mspec spec... instead of the hardcoded /tmp/path.
00:21:22rubuildius_amd64Federico Builes: edd397c82; 2090 files, 6570 examples, 22949 expectations, 0 failures, 0 errors; http://rafb.net/p/28Dfbt36.html
00:21:38evanlight_ enters the room.
00:22:01imajes enters the room.
00:22:45trythil enters the room.
00:22:51fbuilesvluislavena: it's not spec'd but it works under rbx (at least in Linux and OS X)
00:24:28luislavenafbuilesv: yep, the tmpdir.rb from MRI is there... btw, thre is Win32API code down there :-P
00:24:40rubuildius_ppcFederico Builes: edd397c82; 2090 files, 6573 examples, 22978 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/183954
00:26:52fbuilesvluislavena: will it always be windir/path? I thought the user was able to modify it
00:27:20luislavenafbuilesv: that's why the code uses the ENV...
00:27:29luislavenathe part of the Win32API is wrong anyway...
00:28:10fbuilesvI see
00:28:20luislavenafbuilesv: GetSystemWindowsDirectory works, and return properly setup, mine is D:/Temp :-D
00:29:26fbuilesvluislavena: woudln't that return $WINDIR or however it's called instead of the temp?
00:30:22twbray leaves the room.
00:30:23luislavenafbuilesv: only if this is true: @@systmpdir = temp if File.directory?(temp) and File.writable?(temp)
00:31:01luislavenafbuilesv: in Dir::tmpdir @@systmpdir is used as last resort...
00:32:09fbuilesvluislavena: nah, I meant outside the scope of tmpdir.rb, was my bad reading, thought you said tha tGetSystemWindowsDirectory returned the tmp dir
00:34:06luislavenafbuilesv: oh, sorry, I sometimes forgot to spoke out /write what I have in my brain :-P
00:34:24trythil leaves the room.
00:34:27trythil enters the room.
00:35:55evanlight_ leaves the room.
00:36:00evanlight enters the room.
00:37:15luislavenaanyone tried use reverse on the specs? just for the sake of spec individuality?
00:37:18evanlight leaves the room.
00:38:52crafterm enters the room.
00:44:18fbuilesvluislavena: Any suggestion on how to do that with mspec? I'd be happy to try it.
00:45:00luislavenafbuilesv: I'm done, need to update with your latests fixes :-)
00:45:08luislavenafbuilesv: about ot submit a patch to LH :-D
00:45:13demisone leaves the room.
00:45:26fbuilesv:-)
00:48:38adamwiggins leaves the room.
00:51:01luislavenafbuilesv: I need to use -k with format-patch ?
00:51:33luislavenafbuilesv: or I should submit numbered patches?
00:52:15fbuilesvluislavena: the -k is not necessary, you can just upload your patches unless there's a bazillion of them
00:52:29luislavenafbuilesv: only two...
00:52:38fbuilesvluislavena: just upload them to LH
00:53:15ndemonner enters the room.
00:53:35luislavenatoo bad LH only allows 1 attachment per ticket update/comment...
00:53:44luislavenahttp://rubinius.lighthouseapp.com/projects/5089-rubinius/tickets/507-patch-fixes-mspec-on-windows# ticket-507-2
00:54:52fbuilesvluislavena: you can upload 1 and then upload the other one
00:55:01luislavenafbuilesv: done that :-D
00:55:30luislavenafbuilesv: now I can trust the results of mspec :-)
00:55:34rphillips leaves the room.
00:55:57fbuilesvluislavena: awesome, let us know how many specs fail on a vanilla run :)
00:56:40luislavenafbuilesv: just -t ruby spec/ruby/1.8 :-)
00:57:32fbuilesvluislavena: I'm sure there are more hardcoded paths and stuff around
00:58:07luislavenawell, one step at time :-)
00:58:24luislavenaknwo that I know mspec works, I can check the other specs :-)
00:58:54luislavenafbuilesv: great:
00:59:24luislavenahttp://pastie.caboo.se/private/dyttvqhwkplspqkooycw
01:00:10fbuilesvhah
01:00:59fbuilesvFile.open('/tmp/empty.txt')
01:01:23fbuilesvas I told you, it'll be an interesting job trying to fix all the hardcoded strings in there but I'm sure it's necessary to start doing it
01:03:22luislavenafbuilesv: yep, that is a bad practice :-)
01:03:37luislavenafbuilesv: like assuming all your headers and libs are always inside /usr :-P
01:03:52fbuilesvluislavena: hah, all _my_ headers are there indeed :P
01:04:39crafterm enters the room.
01:04:45luislavenafbuilesv: but that will fail from distro to distro... hardcoded paths are the root of evil :D
01:04:47crafterm enters the room.
01:04:58fbuilesvluislavena: no, that would be "vi" :)
01:05:37benny leaves the room.
01:06:29luislavenafbuilesv: hehehe
01:06:59fbuilesvluislavena: your patch looks fine to me, lets hope brixen doesn't find a problem including tmpdir in the specs and applies it
01:09:25qwert666 leaves the room.
01:09:36luislavenafbuilesv: ok, can wait :-)
01:12:50luislavenafbuilesv: it seems will need to work on the same branch for the time being :-)
01:13:51luislavenafbuilesv: I don't see a problem with including tmpdir... that is a MRI error have it split from core :-P
01:13:55imajes leaves the room.
01:14:34imajes enters the room.
01:15:21dblack enters the room.
01:15:24dblackhello
01:15:31fbuilesvdblack: hi
01:16:18fbuilesvluislavena: I don't see any problem with doing it either but since it's gonna be included in a lot of specs someone might think of a clever way to add it to mspec
01:16:50luislavenafbuilesv: yes, a helper will be good.
01:16:55fbuilesvluislavena: I we're gonna be using Dir.tmpdir for each spec that needs cross platform tests a helper could fix that
01:16:59fbuilesvyes, that :)
01:17:03luislavenafbuilesv: in any case, the hardcoded /tmp/ is not a real problem actually.
01:17:06imajes leaves the room.
01:17:34imajes enters the room.
01:17:43luislavenanot in this case really... I'm getting the rb_bug() thing no matter what filename I use :-P
01:18:47fbuilesvluislavena: is it only on the ungetc spec or are you getting it somewhere else?
01:20:07luislavenafbuilesv: nop, it cames from here:
01:20:39luislavenaspec/ruby/1.8/core/io/new_spec.rb
01:21:03luislavenalook at the @filename for the io_new shared specs.
01:22:18fbuilesvluislavena: what's the issue, the absolute path or the $$ ?
01:22:20luislavenawhen the fix I sent get applied, I'll work on getting all those annoying bits ironed.
01:22:29luislavenafbuilesv: the path, $$ is not an issue.
01:22:31obiejuan enters the room.
01:23:01fbuilesvluislavena: ok
01:23:41luislavenafbuilesv: thank you for your time man, I'm glad someone heard my screams :-D
01:23:42fbuilesvluislavena: I gotta run now but I'll make sure to talk to brixen to get something to make a tmpdir helper
01:23:46fbuilesvluislavena: np :)
01:26:50hornbeck enters the room.
01:35:02RyanTM leaves the room.
01:36:01RyanTM enters the room.
01:38:14seanstickle enters the room.
01:40:58seanstickleIs shotgun able to be used by other dynamic languages?
01:45:02agardinerseanstickle: it should be, but i don't know that anybody has tried very hard
01:45:12seanstickleah, ok.
01:45:19seanstickleI guess this is the time when I start to read the source code
01:45:24seanstickleAnd learn shotgun assembly code
01:45:27seanstickle:)
01:45:55agardineryeah, the easiest way is to just generate bytecode that shotgun can run
01:46:27agardinerwhat dynlang were you thinking of using with it?
01:46:59lopex leaves the room.
01:47:13seanstickleagardiner: scheme
01:47:26agardinerah, ok
01:47:26rubuildius_ppc leaves the room.
01:47:29rubuildius_ppc enters the room.
01:47:32seansticklebut mostly just playing with the idea
01:47:59agardinerok, well what you want to try and do is generate an .rbc file
01:48:17agardinerthen you can just point shotgun at that and have it run
01:48:36seanstickleFair enough.
01:48:38seanstickleThanks!
01:48:46agardinernp
01:49:33agardinercheck out http://rubini.us/doc/vm/ for details of the opcodes and .rbc file format
01:50:13agardinerbut bear in mind that these are still in a state of flux, and will change when the new vm is switched in
01:51:10seanstickleSure nuff.
01:51:23seanstickleAll part of my vm edumacation
01:51:36benny enters the room.
01:52:17lstoll_ enters the room.
01:57:20cremes enters the room.
02:01:22rubuildius_ppcFederico Builes: edd397c82; 2090 files, 6573 examples, 22978 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/183982
02:03:16brainopia leaves the room.
02:07:05Coderifous enters the room.
02:08:51lstoll leaves the room.
02:09:20lstoll enters the room.
02:11:17luislavena leaves the room.
02:14:51GMFlash leaves the room.
02:20:37antares enters the room.
02:21:05antares leaves the room.
02:27:34benburkert leaves the room.
02:30:16twbray enters the room.
02:30:29lstoll__ enters the room.
02:31:37imajes leaves the room.
02:32:45lstoll_ leaves the room.
02:37:30imajes enters the room.
02:38:10twbray leaves the room.
02:40:22lstoll__ leaves the room.
02:43:55antares enters the room.
02:45:25Fullmoon leaves the room.
02:46:07wdperson enters the room.
02:49:46lstoll leaves the room.
02:51:15wdperson leaves the room.
02:56:52headius enters the room.
02:57:57headiusyeehaw
02:58:06dblackhi headius
02:59:16headiushiya dblack
02:59:19Jzalae enters the room.
02:59:34headiusoff on any trips lately?
02:59:41dblackheadius: actually not for a while
02:59:46dblack(for a change)
03:00:02seanstickleheadius of jruby fame ?
03:00:07headiusnice...I just had a two-day trip to san jose, but otherwise nothing til JavaOne
03:00:23headiusseanstickle: of jruby, yes
03:00:32dblackseanstickle: do you know other headii? :-)
03:00:53dblack(i just had to get that latin plural in there somehow :-)
03:01:07dblackhas compiled ruby 1.8.7 preview 2
03:01:08seansticklecute
03:01:17dblacki'm still not *quite* sure what the goal of 1.8.7 is
03:01:17headiustwenty years ago when I was picking a BBS handle I guess I did ok
03:01:24headiushave not ever run into any
03:01:33headiusdblack: when you figure out, let me know
03:01:47headiusI understand wanting to backport some features, but at this rate 1.8.x will be a slower 1.9
03:02:17dblacki was kind of hoping that the 1.8 branch would continue to be developed without a lot of 1.9 features
03:02:31dblacki think ruby 1.8 is a great language and fully worth another version or two
03:02:37headiuswell, it certainly forces the issue for us getting 1.9 features implemented
03:02:42headiusthough so far they're the easy ones
03:03:14seanstickleWho's the lead on Cardinal?
03:03:40headiuscardinal is dead
03:03:44headiusbut it was kevin tew
03:03:45seanstickleorly?
03:04:19dblacki guess it was dan sugalski (sp?) originally
03:04:21headiusyeah, nothing's been done for well over a year, and parrot's probably been completely redesigned in the meantime :)
03:04:27seanstickleFair enough
03:04:32dblackor am i confusing cardinal with parrot?
03:04:38seanstickleSo no other Ruby on Parrot implementations currently?
03:05:03AndrewO enters the room.
03:05:21headiusnot that I'm aware of
03:06:41seanstickleSweet
03:07:04AndrewO leaves the room.
03:07:07brainopia enters the room.
03:09:32seanstickleI guess I have my next project set up then
03:10:49AndrewO enters the room.
03:26:08headius leaves the room.
03:30:11dewd enters the room.
03:31:14imajes leaves the room.
03:41:10ruivaldo enters the room.
03:42:10imajes enters the room.
03:42:52kw enters the room.
03:48:53seanstickle leaves the room.
03:50:10ruivaldo leaves the room.
03:51:26twbray enters the room.
03:51:39twbray leaves the room.
04:01:55benburkert enters the room.
04:09:30twbray enters the room.
04:10:06RyanTM leaves the room.
04:11:24imajes_ enters the room.
04:12:10dblack leaves the room.
04:12:27imajes leaves the room.
04:15:51twbray leaves the room.
04:18:31twbray enters the room.
04:22:08shanghaililly enters the room.
04:28:03twbray leaves the room.
04:31:38d2dchat enters the room.
04:41:38rphillips enters the room.
04:41:49rphillips leaves the room.
04:42:05d2dchat leaves the room.
04:42:07rphillips enters the room.
04:43:02obvio enters the room.
04:45:16imajes_ leaves the room.
04:45:44wmoxam_ leaves the room.
04:46:02shanghaililly leaves the room.
04:52:01wmoxam enters the room.
05:01:12dewd leaves the room.
05:02:56Coderifous leaves the room.
05:11:00lstoll enters the room.
05:21:14rudebwoy enters the room.
05:27:09agardiner leaves the room.
05:27:31AndrewO leaves the room.
05:27:52anteaya_ leaves the room.
05:32:55d2dchat enters the room.
05:34:21trythil_ enters the room.
05:34:21trythil leaves the room.
05:37:29be9 enters the room.
05:40:23jero5 leaves the room.
05:42:49agardiner enters the room.
05:51:08wmoxam leaves the room.
05:56:33srbaker enters the room.
06:05:22trythil_ leaves the room.
06:05:34trythil enters the room.
06:09:26fbuilesv leaves the room.
06:19:07rphillip1 enters the room.
06:20:24kirindav_ enters the room.
06:21:35rphillips leaves the room.
06:27:01jicksta leaves the room.
06:27:21jicksta enters the room.
06:33:40jtoy enters the room.
06:34:38srbaker leaves the room.
06:38:52MenTaLguY leaves the room.
06:45:01dlee enters the room.
06:46:17srbaker enters the room.
06:55:11srbaker leaves the room.
06:58:20dlee leaves the room.
07:04:11d2dchat leaves the room.
07:04:59headius enters the room.
07:05:15benburkert leaves the room.
07:05:52dctanner enters the room.
07:06:08agile leaves the room.
07:14:57agile enters the room.
07:15:19kirindav_ leaves the room.
07:20:03lstoll_ enters the room.
07:22:47dctanner leaves the room.
07:29:33lstoll leaves the room.
07:35:39crafterm leaves the room.
07:47:37benny leaves the room.
07:48:18headius leaves the room.
07:57:55zenspiderboyscout: you ok?
07:58:06zenspideroh. my push failed... hrm
07:58:53boyscout1 commit by Ryan Davis
07:58:54boyscout * Added cluster_test, a distributed CI system; 991c6e6
08:04:29trythil leaves the room.
08:04:39trythil enters the room.
08:06:20rubuildius_amd64Ryan Davis: 991c6e613; 2090 files, 6570 examples, 22949 expectations, 0 failures, 0 errors; http://rafb.net/p/RaLXqH14.html
08:08:52headius enters the room.
08:12:04rubuildius_ppcRyan Davis: 991c6e613; 2090 files, 6573 examples, 22978 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/184069
08:15:53thehcdreamer enters the room.
08:19:38trythil_ enters the room.
08:19:38trythil leaves the room.
08:23:19ruezenspider: boyscout has been a bit hazy today, too
08:28:20dctanner enters the room.
08:29:40yaroslav enters the room.
08:31:06cypher23 enters the room.
08:34:19agardiner leaves the room.
08:34:35octopod enters the room.
08:36:46cypher23 leaves the room.
08:37:09cypher23 enters the room.
08:40:31mitchellvriley enters the room.
08:41:20dctanner leaves the room.
08:43:06trythil_ leaves the room.
08:48:02headius leaves the room.
08:56:52brainopia leaves the room.
08:58:22Maledictus enters the room.
09:00:42mutle enters the room.
09:03:56lstoll_ leaves the room.
09:06:32mitchellvriley leaves the room.
09:07:56qwert666 enters the room.
09:34:06kw leaves the room.
09:38:37VVSiz enters the room.
09:46:15dctanner enters the room.
09:50:59Arjen_ enters the room.
10:00:14obvio leaves the room.
10:05:28yaroslav leaves the room.
10:06:44yaroslav enters the room.
10:15:00demisone enters the room.
10:23:43tockitj enters the room.
10:38:03JimMc enters the room.
10:42:58zimbatm enters the room.
10:46:45imajes enters the room.
10:55:44imajes__ enters the room.
11:04:10imajes leaves the room.
11:05:35Fullmoon enters the room.
11:07:56ttmrichter_ leaves the room.
11:08:32Fullmoon leaves the room.
11:09:02ttmrichter_ enters the room.
11:09:08yaroslav leaves the room.
11:09:43yaroslav enters the room.
11:15:43dctanner leaves the room.
11:25:45dblack enters the room.
11:32:34lstoll enters the room.
11:42:33ttmrichter enters the room.
11:44:10webmat enters the room.
11:50:51ttmrichter_ leaves the room.
12:05:01jtoy leaves the room.
12:12:10radarek enters the room.
12:24:56antares leaves the room.
12:32:42antares enters the room.
12:33:42lstoll leaves the room.
12:46:00zimbatm leaves the room.
12:47:05anteaya enters the room.
12:57:03wdperson enters the room.
12:57:45tockitjQ: Are there avaliable web based test runners (for Test::Units or RSpec) ?
12:58:51Fullmoon enters the room.
13:07:14ruetockitj: Probably. Check #ruby-lang and #ror
13:07:16rueNite
13:08:54tockitjrue, maybe i have not expressed myself well - i'd like to run Test::Units from web browser (i'm not refering to testing web apps) :-)
13:11:48skaar enters the room.
13:11:56rueI understood, and I am pretty sure there is a remote method available. Just dunno where and looks like everyone else is out still over here
13:12:06rueActually going now :P
13:12:47rueThose two might have more folks awake
13:13:09tockitjthanks anyway
13:13:11tockitjnite nite
13:13:27Maledictustockitj: look at cruisecontrol.rb
13:18:05RyanTM enters the room.
13:27:26mutle leaves the room.
13:27:47mutle enters the room.
13:28:28obiejuan leaves the room.
13:33:17lstoll enters the room.
13:43:27dblack leaves the room.
13:46:35imajes__ leaves the room.
13:47:28lstoll leaves the room.
13:47:59imajes enters the room.
13:50:45antaresRSpec does have a stories server. It is different from CC.rb in a sense you can edit those stories, and it only runs when you want it, not reacting to commit events.
13:52:17Illocution leaves the room.
13:53:17headius enters the room.
13:55:13Illocution enters the room.
13:58:49peeja leaves the room.
13:59:32Coderifous enters the room.
14:02:49Fullmoon leaves the room.
14:02:54yaroslav leaves the room.
14:05:12yaroslav enters the room.
14:08:51AndrewO enters the room.
14:10:29cremes__ leaves the room.
14:14:20headius_ enters the room.
14:14:37cremes_ enters the room.
14:15:21headius leaves the room.
14:35:36peeja enters the room.
14:37:37wmoxam enters the room.
14:37:37antares leaves the room.
14:39:48GMFlash enters the room.
14:46:34agile leaves the room.
14:47:38dewd enters the room.
14:51:21Fullmoon enters the room.
14:56:50moofbong enters the room.
15:00:30jtoy enters the room.
15:04:39headius leaves the room.
15:04:52headius enters the room.
15:10:16yipstar enters the room.
15:13:40enebo enters the room.
15:14:23d2dchat enters the room.
15:16:02Jzalae leaves the room.
15:17:41chris2 enters the room.
15:21:03yipstar leaves the room.
15:29:45fbuilesv enters the room.
15:31:38yipstar enters the room.
15:32:00jtoy leaves the room.
15:32:04tockitj leaves the room.
15:41:11yugui enters the room.
15:42:23trythil enters the room.
15:42:35twbray enters the room.
15:44:17twbray leaves the room.
15:47:52peeja leaves the room.
15:48:48qwert666_ enters the room.
16:00:49agile enters the room.
16:01:37therealadam enters the room.
16:01:56anteaya_ enters the room.
16:02:42peeja enters the room.
16:03:20twbray enters the room.
16:05:37ctennis leaves the room.
16:06:33qwert666 leaves the room.
16:09:26anteaya leaves the room.
16:10:56benstiglitz enters the room.
16:11:31peeja leaves the room.
16:14:41benny enters the room.
16:15:01macournoyer enters the room.
16:21:26rype enters the room.
16:21:27brainopia enters the room.
16:23:52wmoxam leaves the room.
16:24:51twbray leaves the room.
16:25:40peeja enters the room.
16:28:54wmoxam enters the room.
16:31:12benburkert enters the room.
16:33:29lopex enters the room.
16:34:19jlindley enters the room.
16:41:53wmoxam leaves the room.
16:43:47Fullmoon leaves the room.
16:46:14w1rele55 leaves the room.
16:51:18wmoxam enters the room.
16:52:46w1rele55 enters the room.
16:54:46fbuilesv leaves the room.
16:57:52benny leaves the room.
16:59:46benburkert leaves the room.
17:02:01benburkert enters the room.
17:05:12anonuser enters the room.
17:08:13FoobarGuy enters the room.
17:08:17FoobarGuyhello all
17:08:38Defiler<insert appropriate locale-dependent greeting here>
17:08:59d2dchat_ enters the room.
17:09:22mutle leaves the room.
17:11:06benstiglitz_("GREETING")
17:11:17imajes leaves the room.
17:11:50FoobarGuyI was reading up on how to contribute, i was wondering if theres any room for contributions in the compiler end of the project
17:12:50evanFoobarGuy: sure
17:13:11wycats leaves the room.
17:14:05wycats enters the room.
17:14:46FoobarGuygoodie, I got the impression all there was left was libraries :(
17:14:54DefilerDefinitely not
17:15:03dctanner enters the room.
17:15:18wycats leaves the room.
17:15:23wycats_ enters the room.
17:15:39wycats_ leaves the room.
17:16:18wycats enters the room.
17:16:37Defilerevan: Can you lay eyes on this when you have a moment? http://rafb.net/p/EAiIrO74.html
17:17:22twbray enters the room.
17:17:24Defilerevan: I feel like the way RBX ends up at EOF early (line 55 in paste) is at the heart of the problem. Does that suggest anything to you?
17:18:11d2dchat leaves the room.
17:18:33obvio enters the room.
17:18:37FoobarGuyso I guess the way to jump in is just to pull the repository, compile it and fuck around with it eh? hehe
17:18:45evanDefiler: hm.
17:19:03DefilerFoobarGuy: Sure. Definitely read the compiler code and try to get a handle on it
17:19:13DefilerFoobarGuy: It isn't as simple as it could be, but it is easier to get into than it once was
17:19:19evanto get an EOF like that, i'd think that the socket would have had to be closed on the rbx side.
17:19:48DefilerAccording to dtruss, it doesn't seem to be calling close on it at that point
17:20:18DefilerHere's a much more verbose run of the same thing, including all the calls socat makes
17:20:22Defilerhttp://rafb.net/p/7jwvBr95.html
17:20:32Defilervs the same in MRI (startup crap left off, sorry) http://rafb.net/p/c2MHmD52.html
17:21:20Defilerline 61 in the RBX paste matches line 1 in the MRI paste, specifically. Everything before that is socat spam
17:22:11evanis line 110 the client writing to the server?
17:22:19evanor the server writing to the client?
17:22:22evanie, who is fd 3?
17:23:22Defiler3 is the server socket, 4 is the client
17:23:58Defiler(lines 39 and 40 give the details in the c2MHmD52 paste)
17:24:11evanso a read from 3 is the server talking to the client?
17:24:15seanstickle enters the room.
17:24:35evanif so, then why did mysql close the socket?
17:24:43DefilerYes
17:24:44evanbecause it's fd 3 that has an EOF
17:24:52evanoh
17:24:57evanbecause 5 seconds past?
17:25:02DefilerYeah
17:25:04evanwhy didn't rubinius write any data?
17:25:09Defilerrbx is blocking somehow/somewhere
17:25:13Defilerand mysql times out the login request
17:25:20evanthat should be straight forward to figure out
17:25:34demisone leaves the room.
17:25:39headiusheheh
17:25:39DefilerLead me, warleader
17:25:42evanhave you instrumented mysql.rb to figure that out?
17:25:51Defilerextensively, but I haven't found it
17:25:51evanyou just just put 'p :stage' in mysql.rb
17:26:03evanseems like it's in the handshake
17:26:13evanwhich mysql.rb are you using?
17:26:40FoobarGuythats ok Defiler, I didn't expect it to be easy, when a guy comes in and says he wants to work on a compiler, you have to assume he doesn't enjoy making his life easy :)
17:28:12Defilerevan: rails trunk
17:28:19Defilerevan: http://rafb.net/p/uV0tIc75.html
17:29:05evanok
17:29:06evanso
17:29:09evanyou perform a read
17:29:14evanand the read takes 5 seconds to return?
17:29:20evanie, if you put
17:29:22evanp :read_start
17:29:26evan<code for read>
17:29:28evanp :read_end
17:29:35dbussinkthe problem probably is that rbx doesn't send the login request
17:29:44dbussinkso mysql disconnects after a 5 sec timeout
17:29:51evanthe 2 :read symbols will be output 5 seconds apart?
17:30:00Defilerhttp://rafb.net/p/zLUYZP61.html
17:30:02dbussinkor is the login request send to the socket?
17:30:23DefilerThe login request is sent, but it is sent 5 seconds too late
17:30:32evanwhat is buf.join?
17:30:40benburkert leaves the room.
17:30:42evanis that String#join?
17:30:51Defilerbuf is an Array, and gets joined into a String
17:31:02DefilerI added a local for it so I could inspect it, but the code is otherwise unmodified
17:31:03evanright ok.
17:31:20dbussinkDefiler: but you haven't found what it's waiting for?
17:31:21evanthis seems clear to me.
17:31:29Defilerdbussink: Nope
17:31:33yaroslav leaves the room.
17:31:38evanour read event doesn't think it's gotten +len+ bytes
17:31:46evanso it's still waiting for more.
17:32:04dbussinkand the disconnect probably triggers it, so the message gets send than
17:32:08evanyep
17:32:09dbussinkbut then it's too late
17:32:36wycats_ enters the room.
17:32:44DefilerYeah. It wants to read 57 bytes
17:33:26DefilerSo, this is where I run out of steam.. I know it wants to read from a socket that totally has the required data available
17:33:32Defilerbut it blocks unexpectedly
17:34:10yaroslav enters the room.
17:34:25evanyou need to set a gdb breakpoint for _cpu_wake_channel_and_read
17:34:30evanand walk through it
17:34:33evanas it's receiving data
17:34:37evanand see what happens.
17:34:43Defilermysql times out when I do that, though, so I can't successfully connect under gdb
17:34:52evando that at the beginning
17:34:55evannot at the end.
17:34:59evandon't attach
17:35:01headiusbump mysql's timeout up
17:35:03evanstart under gdb.
17:35:11Defilerheadius: oh, can I do that?
17:35:17headiusI would assume you can
17:35:42headiusyou can tweak things like max packet size, so I assume there's a timeout setting somewhere
17:35:54evanDefiler: better yet
17:35:55headiusmight make it easier to hunt down
17:36:00evancall IO#sysread instead of #read
17:36:03evanand see what happens.
17:36:23evanthat will isolate it even more
17:36:27evanif sysread returns right away
17:36:34evanit's a bug in the buffering inside IO#read
17:37:13Defilerchanging it to sysread causes it to raise EOFError, at least, which is better
17:37:28evanafter 5 seconds?
17:37:48DefilerYes
17:38:36evanso, should i just pull down rails trunk to try this myself?
17:38:42evanand what test script are you using?
17:39:11Defilerwith mysql.rb copied to lib/ from the activerecord 'vendor' directory, just doing this is enough: ./shotgun/rubinius -r"mysql" -e "Mysql.new"
17:39:32DefilerMysql.new('127.0.0.1') if you want to force a loopback TCP connection instead of a domain socket
17:39:39dbussinkDefiler: maybe simulate it using netcat?
17:39:51evanyes, did you ever simulate this?
17:39:53DefilerThe problem is that there is a connection sequence returned by the 'hello' response
17:39:59DefilerSo you can't connect via a replay
17:40:01evanie, NOT using the mysql client
17:40:23DefilerI was not able to produce a valid netcat, though I did give it the old college try
17:40:43thehcdreamer leaves the room.
17:40:58dbussinkbut isn't this something that occurs even with a simple test where you try so send out multiple messages to a socket
17:40:59dbussink?
17:41:06DefilerNo, it is not
17:41:17benburkert enters the room.
17:41:19DefilerI can connect to a UNIXServer process and send data back and forth forever
17:42:14headiusare you sure it's not using a unix socket in the 127.0.0.1 case? I've had some trouble getting it to *not* use the unix socket
17:42:25DefilerI verified it in wireshark
17:42:27headiusok
17:42:35headiusmust be the 0.0.0.0 vs 127.0.0.1 thing
17:42:41DefilerIt does use a unix socket if you pass it 'localhost' though
17:42:48headiusbug
17:43:01headiusvery annoying sometimes
17:43:02DefilerNope. Deliberate 'support rails-style database.yml' crap
17:43:12dbussinkwell, it's a mysql standard
17:43:21headiusthen mysql is wrong
17:43:22dbussinki've seen it in every mysql client app
17:43:45headiusit sounds like a hack to improve performance without requiring people to know about unix socket
17:43:53dbussinkyeah, but doing it differently in rbx results in everyone blaming rbx
17:44:03dbussinkheadius: probably is yeah
17:44:04headiusstart at the top!
17:44:11headiusfix mysql!
17:44:12benny enters the room.
17:44:13headius:)
17:44:16headiusbreak everyone's apps
17:44:19evanWTF
17:44:22DefilerWell, we don't get to modify this code, so the point is moot. :)
17:44:27evanyou have to give personal info to download mysql now!?
17:44:35DefilerThere's a tiny tiny link to avoid that
17:44:45evanha!
17:44:46evanfound it.
17:44:47DefilerI think it is white on white
17:44:48evanbastards
17:44:50evaneyes headius
17:44:51evan:)
17:45:07Defilerheadius: Yeah.. you are in a better position to fix MySQL than we are :)
17:45:35DefilerWhat with being nestled in the crook of Big Brother's arm heh.
17:46:12wycats leaves the room.
17:46:21headiusit's warm and cozy here
17:47:11DefilerWe have always been allied with MySQLonia.
17:47:20peeja leaves the room.
17:47:23evanhaha
17:47:40evanapplies for a job at the Ministry of Truth
17:48:07DefilerIf evan finds and fixes this problem immediately after downloading this file, I am going to spend the rest of the day getting drunk in the sun, I think
17:48:43evanbe sure to wear sunscreen.
17:48:48Defilerword.
17:48:48headiusI bring you love
17:49:02Defilerhttp://www.wunderground.com/cgi-bin/findweather/getForecast?query=32312
17:49:33evanDefiler: looks like here
17:49:35trythilwow, that's some nice weather
17:49:43Defiler\m/
17:50:14brixenDefiler: horrid, it was 12 degrees before windchill at 7000 ft yesterday, we hiked to 8000
17:50:37Defilerbrixen: Awesome. Where did you hike?
17:50:40brixenwe'll be lucky to see 45 today
17:50:52brixenDefiler: 1/2 up mt hood
17:50:57brixentraining hike
17:51:09brixenreally only 2000 ft, we started at 6000 :)
17:51:11DefilerExcellent. You'll be the perfect weapon soon.
17:51:16brixenheh
17:51:29evanbrixen: how was the snow?
17:51:36evanor was this just a hike?
17:51:39brixenevan: awesome, 2 hrs of snowboarding first
17:51:49brixen7" new soft dry powder
17:51:52evannice.
17:51:59seanstickle leaves the room.
17:52:29wycats enters the room.
17:52:44anteaya leaves the room.
17:55:03evanok, duplicated the connection problem here.
17:55:52evandeath to tabs+spaces
17:56:16jlindley leaves the room.
17:56:21peeja enters the room.
17:56:34imajes enters the room.
17:57:19wycats leaves the room.
17:57:35wycats_ leaves the room.
17:58:03wycats enters the room.
17:58:47dbussinkbrixen: i have a day to spare in portland in june, maybe a short hike would be nice :P
17:59:34yugui leaves the room.
17:59:38peeja leaves the room.
18:00:30yaroslav leaves the room.
18:01:11yaroslav enters the room.
18:01:17yaroslav leaves the room.
18:02:30yaroslav enters the room.
18:03:31brixendbussink: sure, lots of great hikes in the columbia gorge
18:03:48Defilerevan: Oh yeah, mysql.rb has the worst formatting
18:03:59Defilerevan: ..and not even vim/emacs hint comment thingies
18:04:11rueMorning
18:04:18brixenmorning rue
18:04:27dbussinkbrixen: been a while, when i was in the us we went to Yosemite, grand canyon, zion and bryce, was really nic
18:04:29marnen enters the room.
18:04:32dbussinkbrixen: nice
18:04:34rueIt is almost technically morning too
18:04:37dlee enters the room.
18:04:43DefilerI'm going to sneak out for some lunch. Back swiftly I hope
18:04:43dbussinkbrixen: but i'll bring my shoes then just in case :p
18:07:17hornbeck leaves the room.
18:07:43hornbeck enters the room.
18:07:45rueFoobarGuy: Were there any particular docs that we should update that you recall?
18:08:50FoobarGuythink you got the wrong guy :)
18:10:16djwhittrue: perhaps you're thinking of FoobarWidget?
18:10:36brixendbussink: this link has some popular hikes: http://www.nwhiker.com/gorge1.html
18:11:13brixendbussink: multnomah falls is kind of touristy, but a beautiful waterfall
18:11:26dbussinkbrixen: well, it's only a wild idea :)
18:11:28rueFoobarGuy: You mentioned you had gotten the impression that the compiler was complete, sorry. I am catching up
18:11:36dbussinkbrixen: we need to get there in the first place probably :p
18:12:27imajes_ enters the room.
18:12:59brixendbussink: sure, transportation wouldn't be a problem unless we had 5+ people
18:13:32rueI thought you were supposed to hike, not take a car like some wuss :P
18:13:35dbussinkbrixen: well, i'd love to just do a simple track for a few hours, nothing really straining, but just something to get out :)
18:13:58brixendbussink: yep, sounds good. we'll figure it out
18:14:10VVSiz_ enters the room.
18:14:47rue*wave
18:15:12TheVoice enters the room.
18:16:51dgtizedso Kernel#fork seems to work, yet we have no specs for it, and it messes with spec output if you try do a basic test?
18:18:03benburkert leaves the room.
18:19:19octopod leaves the room.
18:20:39brixendgtized: yeah, not sure how to handle fork yet
18:20:56suguru enters the room.
18:21:08suguruhi
18:21:18brixendgtized: may need to start a new process and have it fork
18:21:31brixendgtized: rather than forking the spec-running process itself
18:21:35dgtizedbrixen: I was trying to ensure that IO.pipe worked across a fork boundary
18:21:35brixensuguru: hello
18:21:46dgtizedbrixen: and it works, but doubles the spec output
18:21:58brixendgtized: right, hence my suggestionn ;)
18:22:39brixendgtized: why don't you work on a general approach to #fork related stuff that starts a new process, uses fork in that process, and gathers the result for the expectation
18:22:59evanah ha!
18:23:16brixenDefiler: probably should start mixing your drinks :)
18:23:59evanDefiler: you never checked the size of the buffer in IO#read, did you?
18:24:33rueThink he took a head start
18:28:28imajes leaves the room.
18:30:35suguru leaves the room.
18:30:58VVSiz leaves the room.
18:32:36evanhouston: we have mysql.
18:32:56rueWee :)
18:33:08evanlet me do a quick query to double check.
18:34:43mernen enters the room.
18:35:04brixenhmm, it could be argued that vim is inhumane: http://nitpicker.pbwiki.com/The+Humane+Interface
18:35:24dgtizedbrixen: I'll consider the forking thing, but mostly it just seems like a pain in the ass
18:35:32brixendgtized: heh, indeed
18:35:52dgtizedevan: are you opposed to people commiting to the cpp branch as yet?
18:35:57brixen(oh the jokes that could be writ)
18:35:59evandgtized: no, go ahead.
18:37:28dgtizedwhere is this "nicer.rb"?
18:37:34evanah crop.
18:37:35evancrap.
18:37:38evanjust delete using it.
18:37:41evanit was an experiment
18:37:41dgtizedk
18:37:44evani didn't mean to commit it.
18:37:52evanwhat you adding?
18:38:13smparke1 enters the room.
18:38:24rueBunnies!
18:38:26dgtizedmore Fixnum stuff + specs, it looked like that might be easy
18:39:11dgtizedit's been a while since I did any c++ so thought it worth to take a low hanging fruit or two
18:39:29cypher23 leaves the room.
18:40:09evantest first.
18:40:19evanand only if it's needed
18:40:34evanie, don't add a bunch to the VM if no one calls it.
18:40:56dgtizedright but mult and div seem useful
18:41:02evanthey're not.
18:41:06evanactually
18:41:17evanwhat do the primitives call?
18:41:28dgtizedlet me check, but I thought we used primitives for that
18:41:32evanwe do
18:41:41evanyeah, actually, go for it.
18:43:49dgtizedis there good reason that all of the Fixnum primitives have fixnum_ prefixing them accept for add and sub?
18:43:59evanyes.
18:44:06evanbecause they only operate on fixnums
18:44:09evanadd and sub are old
18:44:15evani haven't yet reworked the primitives
18:44:23evanbut i'm going to standardize them
18:44:39dgtizedright, so would it make sense to fix the add and sub primitives in the current shotgun to have that prefix?
18:47:47evanno
18:47:49evando not do that.
18:48:04evanwe are not retrofitting shotgun with changes.
18:48:10evanthats just causes churn.
18:48:18evanneedless
18:48:29evanVM is cleaning up and fixing all
18:48:35evancpp rather.
18:49:13obvio171 enters the room.
18:49:56dgtizedbut we aren't going to change the primitive hookup routine in kernel/bootstrap right? so we will still need to change those names when we hook them back up to the new vm?
18:50:26obvio leaves the room.
18:51:44dgtizedalso I'm getting a bunch of compiler warnings about SIGEMT not being declared in scope
18:52:05dgtizedsorry not warnings but full blown errors from builtin_task.cpp:251
18:54:15rueShould be signal.h
18:59:53evanif it's not, replace SIGEMT with another rare signal.
19:00:28evandgtized: yes, we'll have to change the names, but the point is that the cpp work doesn't effect current work
19:00:43evanheadius: check ruby-core
19:00:48evani created a wiki for the meeting tonight.
19:01:23peeja enters the room.
19:01:25headiusok
19:04:08headiuswe should consider having someone to guide/lead the meeting
19:04:36evanagreed.
19:05:18evanI added a leader: section to the meeting page
19:06:34rueTake notes! No, wait.. talk about stuff!
19:08:22xmlhacker leaves the room.
19:09:26imajes_ leaves the room.
19:11:08xmlhacker enters the room.
19:12:03dbussinktoo bad it's not a europe friendly time :(
19:16:19obvio171 enters the room.
19:16:25dbussinkguess all the info will be available somewhere eventually?
19:16:37dbussinkbut i'm off for now, back later
19:16:39evanyea
19:19:03marnenevan: I understand I should talk to you about getting a commit bit now that I have some patches accepted...
19:19:20evanok, i just need an ssh public key
19:19:32marnensure, one second
19:20:51jero5 enters the room.
19:21:53marnenhttp://pastie.caboo.se/184348
19:22:44marnenis that the one you need?
19:24:01evanyep
19:24:14marnenqool
19:24:25evanok, added!
19:24:26evanwelcome!
19:24:34marnenthanks!
19:24:56marnenso I should just be able to commit to origin now?
19:25:05evanyou need to do
19:25:09evanrake git:committer
19:25:10evanfirst
19:25:12evanbut then, yes.
19:25:19evanplease read the howto use git page
19:25:30marnenI've been...still getting comfortable with Git.
19:25:32evanabout using topic branches
19:25:36evanand using rebase.
19:25:45marnenyeah, I've been following that procedure.
19:26:22marnenOK, just ran the rake script. Thanks!
19:27:43obvio leaves the room.
19:27:53peeja leaves the room.
19:30:54marnenand now Git is asking me for a password. Where do I set that? With git-config?
19:31:29evanum
19:31:30evanno
19:31:35evanno password
19:31:41evanthat means that ssh isn't using your key
19:31:59marnenyeah, I would have thought so.
19:32:05marnenLet me check my ssh options.
19:33:00marnenwaitaminute...does git.rubini.us need an RSA key rather than DSA?
19:33:16evanno
19:33:20evaneither is fine.
19:33:28marnenHmm
19:33:57marnen'cause what I gave you was DSA.
19:34:03marnenok...checking a few more things
19:39:24peeja enters the room.
19:41:58evanhm
19:42:12evanhow the heck does MRI set eof for
19:42:25evanf = File.open("test"); f.read(File.size("test"))
19:42:35evanit's inside stdio
19:42:35be9 leaves the room.
19:44:01rueIt is in the nastiness of io.c
19:44:30evanit's that code that has Sockets broken.
19:45:16peeja leaves the room.
19:45:37rueevan: In our IO?
19:45:41evanyep
19:45:52evanour code to have the same eof behavior as mri
19:46:34rueYou would get the EOF from the OS, no? Whether sockets or files
19:46:45evansay there are 4 bytes in the file
19:46:46rubuildius_amd64there are 4 bytes in the file
19:46:46rubuildius_ppcthere are 4 bytes in the file
19:46:47boyscoutthere are 4 bytes in the file
19:46:50evanhahah
19:46:50rueHaha
19:47:00evanawesome.
19:47:01evananyway
19:47:03brixenok, so we have 4 bytes in the file
19:47:05peeja enters the room.
19:47:08evanif you do a read of 4
19:47:13evaneof isn't set
19:47:17evanbecause you haven't hit it
19:47:24evanyou have to basically read past the end to see eof
19:47:36fbuilesv enters the room.
19:49:10peeja leaves the room.
19:49:47jlindley enters the room.
19:49:52rueHm, really?
19:49:55evanyep
19:50:11evanread(fd, buf, 4) => 4
19:50:16evanread(fd, buf, 4) => 0
19:50:31evanthere is no way to test for eof other than performing another read
19:50:36evanat least that i can find.
19:50:51benburkert enters the room.
19:51:23fbuilesv leaves the room.
19:51:30rueAlways peeking seems wasteful
19:52:02evanyou have to toggle non-blocking flags
19:52:10evanand save any data that you accidentally get.
19:52:13evanit's a pain.
19:52:24wycats leaves the room.
19:52:41eneboevan: Do you need to add each individual to People page on Ruby deisgn wiki? I see no edit link
19:52:57evanhm, you should have one...
19:53:04evanare you logged in?
19:53:29eneboyeah...but I just created an account
19:53:33evanhm, maybe i need emails to allow editting
19:53:35rueWait, why is that not in the size?
19:53:45evanenebo: what email did you use?
19:53:49evanrue: why is what?
19:53:50enebotom.enebo@gmail.com
19:53:52wycats enters the room.
19:53:55atmos enters the room.
19:54:05evanok, reload
19:54:15enebothank you :)
19:54:22evanthats probably better
19:54:27evanso that the wiki doesn't get hijacked.
19:54:46rueecho "1234567890" > test.txt; wc -c test.txt
19:54:55rue11 bytes
19:55:22peeja enters the room.
19:55:32ruestat will definitely have st_size 11
19:55:42boyscout1 commit by Evan Phoenix
19:55:43boyscout * Fix IO#read from blocking; 6b8b0d8
19:55:43evanso?
19:55:51evanDefiler is going to hate me.
19:55:55evanbut mysql works now.
19:56:31evanrue: pretend now that the fd points to a pipe or socket.
19:56:55rueYou cannot know how "large" a socket is
19:57:47rueSo by definition you are reading until reaching EOF or connection breaks or whatever
19:58:15yaroslav leaves the room.
19:58:23marnenOy. There's something screwy with the SSH keys right now. SSH recognizes my keys when I use them for other hosts, but not git.rubini.us . According to ssh -v, my client is offering the key, but the server is not accepting it.
19:58:32fbuilesv enters the room.
19:58:44evanrue: so perhaps IO needs to have logic to know what the fd points to
19:58:50evanto support the proper semantics.
19:59:30rueIs DSA enabled?
19:59:30evanOH SO BOGUS.
19:59:48evanmarnen: paste me your id_rsa.pub
19:59:49evanwe'll try that.
20:00:14marnenI don't have one at the moment, but give me a minute and I'll make one.
20:00:26rueevan: That is why there is no IO{#,.}size :)
20:00:35evanrb_io_eof() waits for readability, then uses getc() to check for eof, then uses ungetc() if it's not eof.
20:00:47evanrue: true dat.
20:01:54evanironiclly, the hack we have in eof? now is quite close to the logic in MRI
20:02:29marnenok, try this: http://pastie.caboo.se/184365
20:02:53danlucraft enters the room.
20:02:59rueevan: I am not sure if that should be necessary; I may be missing the use case
20:03:06danlucrafthello
20:03:29evanrue: it is to match MRI's semantics
20:03:43evanwhich is that eof? blocks if the IO has no data
20:04:09evanmarnen: oh ack.
20:04:15marnen?
20:04:15evanyou're putting hard line returns in there
20:04:17evanor pastie is.
20:04:28evanso the key is fucked up in the authorized_keys file
20:04:38chris2 leaves the room.
20:04:40marnenack indeed. Sorry 'bout that
20:04:43radarek leaves the room.
20:04:47evanrepaste your dss
20:04:51marnenprobably happened when I copied from Terminal.
20:04:54evanyeah
20:04:56evandon't do that.
20:04:56marnenok, trying again.
20:06:21rubuildius_amd64Evan Phoenix: 6b8b0d8a1; 2090 files, 6570 examples, 22949 expectations, 0 failures, 0 errors; http://rafb.net/p/27HuNW29.html
20:06:45marnenDSA without the silly returns: http://pastie.caboo.se/184372
20:07:28evanok, looks good
20:07:30evangive it a shot.
20:07:34evanyou'll need to rebase though
20:07:37evansince i did a commit
20:07:54evanmy tummy says it's lunch time.
20:07:56evanbbiab.
20:08:15rubuildius_ppcEvan Phoenix: 6b8b0d8a1; 2090 files, 6573 examples, 22978 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/184375
20:08:21marnenthanks, that worked
20:08:30marnensorry about being so st00pid
20:08:45evanno prob.
20:12:00jlindley leaves the room.
20:12:22rueMatzRuby strikes again
20:12:52rueAlthough it is easier for the programmer, I suppose
20:14:31twbray leaves the room.
20:15:10EugZol leaves the room.
20:15:36hornbeck leaves the room.
20:16:27marnen leaves the room.
20:17:37danlucraft leaves the room.
20:43:16twbray enters the room.
20:44:24NoKarma enters the room.
20:44:36NoKarmaHeya all
20:44:42NoKarmafbuilesv: Heya :)
20:44:50NoKarmafbuilesv: We both got accepted, yay
20:44:51fbuilesvNoKarma: Hey :)
20:44:58fbuilesvNoKarma: That's what Im seeing :)
20:45:18fbuilesvat least they gave the answer, I've been outside of classes with no lunch for like two hours now
20:45:45NoKarmafbuilesv: hehe
20:46:19evanaccepted to what?
20:46:20fbuilesvNoKarma: I won't be 100% sure until I see it on the dashboard too :)
20:46:26fbuilesvevan: google summer of code
20:46:35fbuilesvevan: we both passed proposals to improve the specs for Rubinius :)
20:46:45danlucraft enters the room.
20:46:47evanhurrah!
20:46:52fbuilesv:-)
20:47:30NoKarmaevan: I was thinking the same :)
20:50:46NoKarmaactually, there are three Spec related proposals, that got accepted
20:51:23fbuilesvNoKarma: where are you looking at them?
20:51:36NoKarmahttp://code.google.com/soc/2008/ruby/about.html
20:51:51zimbatm enters the room.
20:52:44fbuilesvawesome
20:52:53fbuilesvok, I gotta go get some junk food right now, talk to you tonight
20:53:06fbuilesvNoKarma: grats :)
20:53:14fbuilesv leaves the room.
20:55:07brainopia leaves the room.
20:56:39yaroslav enters the room.
20:57:48imajes enters the room.
20:58:43Defilerevan: haha funny story
20:59:05Defilerevan: When I went to look at the diff of your fix, I didn't notice the scroll indicator at first, and I thought it was just "done = false"
20:59:06twbray leaves the room.
20:59:17evanhah
20:59:27evanit was only slightly more complicated than that
20:59:49evani just had to add an extra line to pull directly from the buffer if it contained at least the ammount of data requested
20:59:58evanwhen you were doing @sock.read(4)
21:00:02evanthe buffer gets filled with like 70 bytes
21:00:24evanbecause the server streams it all directly
21:00:25jlindley enters the room.
21:00:47evanthen, stupidly, IO#read tries to fill the buffer with MORE data
21:00:50evaneven though it's got enough.
21:01:20DefilerI guess I stared at the socket.rb code for so long, I forgot to look for insanity in io.rb
21:01:23Defiler=(
21:01:47DefilerAnyway, awesome. Now we're just waiting on eval order for the final Rails push
21:01:53evanyou started at the very bottom
21:01:55evaninstead of the very top
21: