Index

Show enters and exits. Hide enters and exits.

00:21:16ragnard leaves the room.
00:25:05pietia enters the room.
00:26:54xmlhacker__ enters the room.
00:31:53pietia leaves the room.
00:32:25sfaxondoes anyone here know the status of the RbYAML library?
00:33:41DefilerThere are some reported bugs, and kwatch is apparently going to file some more he knows about
00:33:42sfaxonI'm pretty new to the project and looking for a place to get started
00:33:55DefilerSome of them seem to be assigned to me, but you are welcome to take a whack at them
00:34:16Defilerhttp://rubinius.lighthouseapp.com/projects/5089/tickets/322-bug-symbol-to_yaml-returns-incorrect-r esult
00:34:25Defiler..and http://rubinius.lighthouseapp.com/projects/5089/tickets/351-bug-module-tagurize-not-defined
00:35:08ataiar enters the room.
00:35:29ataiar leaves the room.
00:37:07sfaxonDefiler: thanks, is the current RbYAML pulled from somewhere else?
00:37:28DefilerI think it came from a tarball, not from a remote repo
00:37:39DefilerBut I can't remember, sadly. enebo or headius might know
00:37:53DefilerI don't think it was maintained, at the time
00:38:05headiusit hasn't been maintained for probably almost 2 years
00:38:39headiusit was a first step away from the old racc-based yaml.rb...once ola rewrote it into jvyaml we never bothered with it again
00:39:01headiusI think ola would be happy for someone to just take it over
00:50:02headius leaves the room.
00:53:12Fishy5 leaves the room.
00:54:21sfaxonDefiler: I have a patch for http://rubinius.lighthouseapp.com/projects/5089/tickets/351-bug-module-tagurize-not-defined
00:55:04DefilerCool. Please feel free to make an account at lighthouse and update the ticket with an attachment
00:55:16sfaxonbut I think the filed bug is pretty trivial in light of the other problems in RbYAML
00:55:28DefilerYeah. If you know of more specifics, please make tickets
00:55:45DefilerWe asked kwatch to do so with the ones he knows about, but he is on a different time schedule
00:56:06sfaxonyea, I'll dig around in a bit, I'm still trying to get the lay of the land
01:03:32brixenDefiler: did you install freebsd in parallels?
01:07:25imajes enters the room.
01:14:11crafterm enters the room.
01:21:44radarek leaves the room.
01:33:35MacRueAlmost done
01:35:13brixenwith os x on x61?
01:35:36MacRuebrixen: No, just the iMac setup. All kinds of quirks to get used to
01:35:53MacRuebrixen: Are you using Parallels Workstation or some other version?
01:37:33brixenI'm using parallels, but I would have gone with vmware fusion if I had known
01:37:43brixenlast time I bought vmware it was $300 :P
01:38:20brixengotta run, bbiab..
01:40:47ezmobiuspeople have osx running under Xen as a guest now
01:40:58ezmobiuspretty sweet
01:41:21evanreally?
01:41:22evanthats cool.
01:41:25evanmust be a hack though
01:41:36evansince i thought OS X had a bunch of guards in the kernel
01:41:52ezmobiuswell you have to get the osx86 torrent of osx to make it happen
01:44:34evanah yeah, that has the hacks
01:44:47srbakerwhat's vmware fusion?
01:45:29evanvmware for OS X
01:46:06macarthy enters the room.
01:47:08MacRueHm, is VMWare really better?
01:49:15MacRueevan: Oh, since you are around--what is the plan, should I keep end of March open? I need to start plotting my schedule
01:52:35evanoh, it's you rue
01:52:38evani wondered who this was
01:52:39evan:D
01:55:45MacRueHeh
01:55:56MacRueA really questionable McDonalds advertising venture?
02:01:46d2dchat enters the room.
02:03:28ezmobiusyou guys going to have another sprint here in SF anytime soon?
02:07:55macarthythere is only on mac
02:07:58macarthyone
02:17:16aotearoa enters the room.
02:24:52lopex leaves the room.
02:25:18ezmobius leaves the room.
02:28:07jinjing enters the room.
02:30:35gnufied enters the room.
02:30:47d2dchat leaves the room.
02:39:50GMFlash leaves the room.
02:44:05ezmobius enters the room.
02:45:02mae enters the room.
02:45:57gnufiedezmobius, hi
02:46:09ezmobiushey
02:46:38gnufiedabout push a new release of bdrb!
02:46:58ezmobiuscool
02:47:03gnufieddid you see new docs that we are putting: http://backgroundrb.gnufied.org
02:48:14gnufiedlet me know, if i missed anything.
02:51:32ezmobiussite looks nice
02:53:00mutle enters the room.
02:55:06gnufiedi think this will be the 1.0 as it should have been! i just hurried for 1.0. Worked on test cases, removed packet from bdrb tree.
03:08:27_mutle leaves the room.
03:13:37context leaves the room.
03:17:15gnufiedi really wish to attend one of those coding sprints on rubinius.
03:19:03cavalle_ enters the room.
03:19:46papadimo enters the room.
03:20:06papadimohelloy
03:22:35papadimo leaves the room.
03:26:01cavalle__ enters the room.
03:26:01cavalle_ leaves the room.
03:35:06cavalle leaves the room.
03:37:20srbaker leaves the room.
03:51:03Defilerbrixen: I haven't yet.. forgot. doh! (re: FreeBSD)
03:59:42wmoxam leaves the room.
04:03:31srbaker enters the room.
04:20:26srbaker leaves the room.
04:27:23aotearoa_ enters the room.
04:33:28therealadam leaves the room.
04:35:02aotearoa leaves the room.
04:42:05therealadam enters the room.
04:44:08crafterm leaves the room.
04:46:40mae leaves the room.
04:49:56headius enters the room.
05:06:14therealadam leaves the room.
05:16:59dewd enters the room.
05:17:09mernen leaves the room.
05:21:36TheVoice enters the room.
05:30:14be9 enters the room.
05:30:34roo enters the room.
05:31:44rueHooray
05:31:45jptix leaves the room.
05:32:01mae enters the room.
05:32:09MacRueColloquy sucks, *wave
05:33:32TheVoicerue: whats going down?
05:34:32rueA heist
05:34:58TheVoicehow much?
05:36:37rueThousands, maybe even tens of dollars
05:37:37headiuscolloquy's nice
05:37:55rueIt plays sounds!
05:38:26headiusyou can turn the sounds off of course
05:39:11rueThat is what THEY want you to believe
05:39:17rueBut I can still hear them..
05:39:56TheVoicewhats my share?
05:42:28headiusa pittance
05:42:44headiusto be paid in 20 equal installments of 1/20th of a pittance
05:43:41rueYay, 5 channels open simultaneously
05:47:25TheVoiceheadius: I will have to remove your ankles with a specially designed ankle saw should my share not be above 1/15th of a pittance.
05:47:56rueBe right back
05:47:59rue leaves the room.
05:52:37crafterm enters the room.
05:54:20rue enters the room.
05:54:27rue leaves the room.
05:59:49wmoxam enters the room.
06:02:26rue enters the room.
06:13:29RyanTM leaves the room.
06:14:22RyanTM enters the room.
06:14:58aotearoa enters the room.
06:16:22macournoyer enters the room.
06:16:34nitay_ enters the room.
06:17:53tizianobis enters the room.
06:18:10macournoyer leaves the room.
06:22:13gnufied_ enters the room.
06:24:24gnufied leaves the room.
06:27:38nitay leaves the room.
06:28:21rue leaves the room.
06:28:58rue enters the room.
06:28:59rue leaves the room.
06:29:41rue enters the room.
06:31:15RyanTM leaves the room.
06:34:31mae leaves the room.
06:35:03mae enters the room.
06:35:12be9hi there. who can tell me what's libmquark for?
06:36:41crafterm leaves the room.
06:43:31aotearoa__ enters the room.
06:44:17aotearoa_ leaves the room.
06:45:46tizianobis leaves the room.
06:55:00ezmobius leaves the room.
07:04:27ruebe9: libmquark is a 'quark' implementation.. basically it is supposed to be a unique identifier for a given string
07:04:30rueInteger specifically
07:06:31ruehttp://files.kittensoft.org/public/irc.png
07:06:40rue^^^ That is a good IRC client, headius :)
07:06:50be9rue: ok, already figured out.
07:07:51GMFlash enters the room.
07:07:58headiusgreat
07:09:49ezmobius enters the room.
07:11:57aotearoa leaves the room.
07:15:15cavalle__ leaves the room.
07:15:46cavalle enters the room.
07:19:17aotearoa enters the room.
07:23:31boyscout1 commit by oleg dashevskii
07:23:31boyscout * Fixes for libev & libmquark to compile on Mingw/Windows.; 0f0cce1
07:33:47jinjing leaves the room.
07:35:50rubuildius_ppcoleg dashevskii: 0f0cce1ad; 1723 files, 5814 examples, 19809 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/160034
07:41:41ezmobius leaves the room.
07:47:27aotearoa__ leaves the room.
07:53:19agile leaves the room.
08:09:24wmoxam leaves the room.
08:14:21cavalle leaves the room.
08:15:02cavalle enters the room.
08:18:32boyscout1 commit by Evan Phoenix
08:18:33boyscout * Remove lightning and custom stubs; c836948
08:19:54headiushey rue, now you're on mac, you can try jruby
08:23:22aotearoa_ enters the room.
08:25:54jinjing enters the room.
08:27:48aotearoa_ leaves the room.
08:28:59rueheadius: Oo, that is true!
08:29:23headiusI recommend soylatte jdk6, but the built-in jdk5 is just fine too
08:29:25rubuildius_ppcEvan Phoenix: c8369486f; 1723 files, 5814 examples, 19809 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/160038
08:29:53headiuslightning is dead, long live lightning!
08:30:47rueShould I get the Venti or the Grande?
08:32:36headiusgrande!
08:32:49headiusoh wait, venti(tm) is their bigger size
08:36:01halorgiumheadius: yes, but they take a chunk of your liver for it
08:36:16headiusI refuse to say venti(tm) when I go in there
08:36:47rueI just say "the small cup"
08:37:07rueOnly once has the "barista" argued with me over it
08:37:22rueheadius: Incidentally, have you looked at Buildr?
08:37:37headiusyes, I think they recently updated it so it works all correctly with jruby
08:38:05headiushaven't migrated yet...monkeying with the build isn't exactly a priority right now
08:41:06rueMonkeying with stuff is always a priority!
08:46:12rueI really wish FF was as fast as Camino :/
08:49:01crafterm enters the room.
08:50:20aotearoa leaves the room.
08:56:33headiusFF3?
08:57:01langenberg enters the room.
09:04:54sfaxon leaves the room.
09:05:20langenberg leaves the room.
09:10:28rue2
09:10:36rueI suspect 3 will be, if anything, slower
09:15:21headius_ enters the room.
09:15:21headius leaves the room.
09:19:55dodecaphonic leaves the room.
09:29:01headiusit's been snappier for me
09:29:08headiusfewer runaway javascripts too
09:32:34aotearoa enters the room.
09:40:03thehcdreamer enters the room.
09:40:37radarek enters the room.
09:45:33headiushmmm
09:46:30headius& isn't calling to_proc
09:46:33headiusis that intentional?
09:47:55headiusok, looks like it's been filed already
09:49:28headiushttp://rubinius.lighthouseapp.com/projects/5089/tickets/363-yield-block-problem
09:49:38headiusI guess that also hits the order-of-eval problem too
09:49:46headiusbugger
09:49:49headiusI wanted to try something
09:51:31aotearoa_ enters the room.
09:55:22context enters the room.
09:57:08maharg leaves the room.
10:01:20aotearoa_ leaves the room.
10:03:18GMFlash leaves the room.
10:18:36mae_ enters the room.
10:18:39mae leaves the room.
10:19:06mae_ leaves the room.
10:19:21aotearoa leaves the room.
10:19:35mae enters the room.
10:21:23mae leaves the room.
10:21:40mae enters the room.
10:22:37maharg enters the room.
10:36:12chris2 enters the room.
10:37:02boyscout1 commit by Charles Nutter
10:37:03boyscout * Add a spec for #363, & not coercing using to_proc.; 60f9544
10:43:56crafterm leaves the room.
10:49:31rubuildius_ppcCharles Nutter: 60f9544ad; 1723 files, 5815 examples, 19810 expectations, 0 failures, 1 error; http://pastie.caboo.se/paste/160046
10:52:00crossblaim enters the room.
11:04:37ragnard enters the room.
11:24:18maduyb__ enters the room.
11:37:06Silex leaves the room.
11:41:41headius leaves the room.
11:46:55ragnard enters the room.
11:48:47random8r enters the room.
11:50:25maduyb__ leaves the room.
11:53:07jptix enters the room.
12:02:02VVSiz enters the room.
12:02:22FrihD enters the room.
12:02:29FrihDhello
12:02:36random8rheya
12:05:43FrihDi've a brand new computer :) now it'll be fast enough to do some dev' :D
12:06:55random8rhehe
12:30:57boyscout1 commit by Vladimir Sizikov
12:30:58boyscout * Adjusted Kernel#catch test a bit, to make it more generic.; b1caeea
12:34:32gnufied_ leaves the room.
12:35:11TheVoice leaves the room.
12:44:36rubuildius_ppcVladimir Sizikov: b1caeeac6; 1723 files, 5815 examples, 19810 expectations, 0 failures, 1 error; http://pastie.caboo.se/paste/160077
12:49:42nmb enters the room.
12:52:47random8r leaves the room.
12:53:40jacen_ enters the room.
12:55:30nmb leaves the room.
12:57:58crossblaim leaves the room.
12:58:58Fullmoon enters the room.
12:59:07mediogre enters the room.
13:00:34Fullmoon leaves the room.
13:01:00Fullmoon enters the room.
13:19:52chris2 leaves the room.
13:21:04Silex enters the room.
13:26:04Fullmoon leaves the room.
13:48:26Fullmoon enters the room.
13:56:31Fullmoon leaves the room.
13:57:30Fullmoon enters the room.
13:59:01jptix_ enters the room.
13:59:39jptix leaves the room.
14:11:05geekounet leaves the room.
14:11:08geekounet enters the room.
14:14:40zimbatm enters the room.
14:23:31lopex enters the room.
14:29:15dodecaphonic enters the room.
14:31:29jinjing enters the room.
14:36:20sfaxon enters the room.
14:37:21Fullmoon leaves the room.
14:45:29sfaxon leaves the room.
14:54:32chris2 enters the room.
14:55:18srbaker enters the room.
15:01:42srbaker leaves the room.
15:03:37Fullmoon enters the room.
15:12:25RyanTM enters the room.
15:19:51macarthy enters the room.
15:22:51benny leaves the room.
15:32:47ataiar enters the room.
15:34:15ataiar leaves the room.
15:34:45ataiar enters the room.
15:36:09Fullmoon leaves the room.
15:38:31Fullmoon enters the room.
15:40:56Fullmoon leaves the room.
15:41:36srbaker enters the room.
16:06:13jptix enters the room.
16:06:58dodecaphonic leaves the room.
16:07:55jptix_ leaves the room.
16:19:09enebo enters the room.
16:34:05GMFlash enters the room.
16:40:22jinjing leaves the room.
16:42:19jptix leaves the room.
16:45:02rubuildius_ppc leaves the room.
16:45:06Arjen_ enters the room.
16:47:14jptix enters the room.
16:53:20enebo leaves the room.
16:55:34headius enters the room.
17:01:56ataiar leaves the room.
17:02:09dewd leaves the room.
17:11:25GMFlash leaves the room.
17:20:43dewd enters the room.
17:21:40jp_tix leaves the room.
17:21:44jp_tix enters the room.
17:25:24Jzalae enters the room.
17:26:03jptix leaves the room.
17:26:55jptix enters the room.
17:40:03headius leaves the room.
17:49:09jptix leaves the room.
18:02:28mernen enters the room.
18:11:54crossblaim enters the room.
18:14:43benburkert leaves the room.
18:21:00mae leaves the room.
18:21:31mae enters the room.
18:22:17mae leaves the room.
18:22:48mae enters the room.
18:23:28DefilerCrazy. My working copy doesn't build, even though 'git status' is clean
18:23:31Defiler..but a fresh clone works fine
18:24:05scooprlocal commits?
18:24:22scoopr(I've yet to crack git, I've got no idea what I'm talking about)
18:25:25DefilerOK, something bad has changed recently in the Makefile world, I think
18:25:35Defilera fresh clone followed by 'rake' leaves some modified files showing in git status
18:25:48DefilerI'll look at it later, I guess
18:41:16evanDefiler: what files?
18:42:47brixenevan: looking at it too, rake distclean didn't work. I'll have a pastie with VERBOSE=1 in a sec
18:43:10brixenI did a local clone and it built fine, git status shows no files
18:46:15gnufied enters the room.
18:46:32brixenwell, that's not much help: http://pastie.org/160173
18:46:42brixenjust dies on the link step for librubinius
18:46:43evanhrm.
18:47:17evanwell, i removed lightning last night...
18:47:39evanhow did 'rake distclean' not work?
18:48:05brixennot sure
18:48:11evani mean
18:48:16brixenrake --trace doesn't give any better info
18:48:17evanwhy do you say it didn't work?
18:48:32brixenbecause I thought it would be necessary after you removed lightning
18:48:36brixenso I tried it
18:48:51evanand it did what to tell you it didn't work?
18:49:09brixenwell, it fails in the same place, clean, distclean don't make a difference
18:49:24evanfails trying to link?
18:49:38evango into shotgun and type 'make'
18:49:39brixenyep, like the 2nd part of the pasite
18:50:16brixenupdated the pastie
18:50:28evanwtf.
18:50:37evanwhy is it working for me....
18:50:49evanrun 'make clean' in shotgun
18:50:50evanand try
18:50:53brixenk
18:51:38brixensame failure
18:51:52evanwhat the hell.
18:51:53brixenoddly, it doesn't emit anything for the link step
18:52:05brixenit fails before that with no message of any kind
18:52:17brixenexcept: make: *** [lib/librubinius-0.8.0.dylib] Error 2
18:52:18evanstill working fine here...
18:52:43evanhrm hrm hrm.
18:52:58evanok, go into shotgun/lib
18:52:59evanand do
18:53:02evanmake library
18:53:19brixendid that, no errors
18:53:36brixenI just rm'd shotgun and git reset --hard
18:53:52gnufiedare we talking about latest build failure that i am seeing?
18:53:59evanperhaps.
18:54:43brixengnufied: can you git clone local_rbx_dir other_dir
18:54:48brixengnufied: and then diff the shotgun dirs?
18:55:12gnufiedone sec
18:55:24evanstrange, you're right, it's not even getting to the link stage
18:55:53brixenyeah, I think there's a left over file somewhere that .gitignore is hiding
18:56:08brixenbecause rm shotgun, git reset --hard worked, i.e. it's building now
18:56:46brixenhmm, I wonder
18:56:49evanok
18:56:53evani've got the error here
18:56:55evanon linux
18:56:55evanone sec.
18:56:58brixengnufied: do you still have a lightning dir in shotgun?
18:57:06evanoh ack!
18:57:11evanwhy is that still there..
18:57:16gnufiedbrixen, no
18:57:22brixenbecause git does not track directories
18:57:30evanmine is full
18:57:38brixenand .gitignore'd files won't be affected either
18:57:52evanack
18:57:55brixenfull of build products or actual source?
18:57:59ezmobius enters the room.
18:58:01evansource
18:58:06evanas though i hadn't deleted it
18:58:08brixenhmm
18:58:14evanand git status doesn't show it
18:58:17ezmobius leaves the room.
18:58:27brixenit's not here after the rm + reset
18:59:16evanvery odd that git didn't remove it..
18:59:37evancourse, removing it doesn't solve the problem
18:59:49brixenyeah, must be other files
19:00:06brixenI'll jump back to the commit before and try this again
19:00:23evanok, make in shotgun/lib is returning with code 2
19:00:55evan2 means there were errors..
19:01:12evanoh crap.
19:01:14evani know whats wrong.
19:01:29evani've got it
19:01:43gnufied?
19:01:44evandistclean isn't removing .d files from shotgun/lib/subtend
19:01:57evanand there are some .d files in there that have lightning files as deps
19:02:43evanone sec, i'll commit the Makefile fix
19:03:01evanthanks a lot make, for telling me that the reason you bombed out was because of a missing dep
19:03:09evankicks make in the ass
19:06:19boyscout1 commit by Evan Phoenix
19:06:20boyscout * Remove all .d files; a1801c1
19:06:34evanok, that fixes it. you'll still need a 'rake clean' though
19:07:00mutle leaves the room.
19:07:40mutle enters the room.
19:10:02brixenevan: did you see the new fonc/vpri paper?
19:10:41brixendang, no time to read now, off to snowboarding
19:10:51brixenback much later :)
19:12:26jacen_ leaves the room.
19:12:35evanlater
19:17:33gnufiedjust wanted to know, so what are the major things on which work is left?
19:17:50gnufied1.0 will have green threads and no utf ?
19:18:14gnufieds/utf/unicode
19:22:59olabini leaves the room.
19:25:24evanyes
19:25:27evanit will.
19:25:36evanbecause we're targetting 1.8 for 1.0
19:26:21benburkert enters the room.
19:27:23wmoxam enters the room.
19:32:17dodecaphonic enters the room.
19:37:49GMFlash enters the room.
19:46:11mediogre leaves the room.
19:53:10TheVoice enters the room.
19:57:03vruz enters the room.
19:59:57srbaker leaves the room.
20:04:33be9 leaves the room.
20:05:01jptix enters the room.
20:06:58srbaker enters the room.
20:07:59jptix_ enters the room.
20:15:14jptix__ enters the room.
20:16:17jptix leaves the room.
20:16:56KirinDav enters the room.
20:17:21jptix enters the room.
20:17:23jptix_ leaves the room.
20:17:28KirinDav leaves the room.
20:17:34srbaker leaves the room.
20:17:56KirinDave enters the room.
20:20:30benburkert leaves the room.
20:23:09aotearoa enters the room.
20:24:52thehcdreamer leaves the room.
20:24:58jptix_ enters the room.
20:27:27benburkert enters the room.
20:28:00jptix___ enters the room.
20:28:58wmoxam leaves the room.
20:31:20Defilerevan: so...
20:31:45Defilerevan: http://rafb.net/p/QbhUkh22.html
20:31:58Defilerevan: That is what I get after a fresh clone from the master, and then 'rake'
20:32:07Defiler(from earlier today, before the .d commit)
20:33:59wmoxam enters the room.
20:34:35jptix__ leaves the room.
20:35:26w1rele55 enters the room.
20:40:19flori_ leaves the room.
20:40:26flori enters the room.
20:43:33DefilerI cloned that directory to another, and can't reproduce it.. some it seems like a first-time-only thing? Bizarre
20:43:56benburkert leaves the room.
20:44:07benburkert enters the room.
20:44:21jptix_ leaves the room.
20:46:32jptix leaves the room.
20:46:35hassox enters the room.
20:58:32KirinDave leaves the room.
21:02:41jptix___ leaves the room.
21:25:37wmoxam leaves the room.
21:29:21ijcd_ enters the room.
21:33:04langenberg enters the room.
21:33:57aotearoa leaves the room.
21:39:06wmoxam enters the room.
21:39:38aotearoa enters the room.
21:40:52boyscout1 commit by Wilson Bilkovich
21:40:53boyscout * Tag headius's new to_proc spec as failing; 0af27d1
21:43:26Fullmoon enters the room.
21:43:39ijcd enters the room.
21:45:46agardiner enters the room.
21:54:36ijcd_ leaves the room.
21:59:24Fullmoon_ enters the room.
22:05:23drbraingit is so slow :(
22:05:53Defilerreally?
22:05:59drbrain6 minutes to checkout rubinius!
22:06:09drbrain(on my FreeBSD 7 paralles image)
22:06:16headius enters the room.
22:06:17DefilerMine maxes out my internet connection doing a clone
22:06:23DefilerSo sounds like something is up
22:06:33olabini enters the room.
22:06:47drbrainmost of that was counting objects, I think
22:07:19headiusyeehaw
22:07:20dewd leaves the room.
22:07:32agardinergiddy up
22:07:39Defilerheadius: FYI, when you add a failing spec, please tag it as failing
22:07:45headiuswhy?
22:07:48headiusI added it for jruby
22:07:50DefilerI did it for you, but CI was broken after your push
22:08:03Defileradded what for jruby?
22:08:08headiusI thought the only requirement was that MRI pass
22:08:35DefilerThat's how you know it is a good spec, but it needs to be tagged as failing in order for it to be excluded from CI
22:08:42headiusI added that spec because it wasn't covered, and I was twiddling that stuff in jruby
22:08:58headiusperhaps ci shouldn't be synced to the leading edge of specs
22:09:01headiusbrixen mentioned changing that
22:09:06headiusI don't know what you guys have set up for ci
22:09:20DefilerAt the moment, CI just runs the specs in HEAD that are not tagged as failing
22:09:41headiusso every time someone adds a spec, they have to also run with rubinius and add failing tags if necessary?
22:09:55DefilerWell, they do if they are a rubinius committer
22:09:59headiuscould I request that every time you guys add a spec, you please run with jruby and let me know about it? :)
22:10:03DefilerIf they just submit a patch, no
22:10:12headiusI'm a spec committer
22:10:31DefilerI hear what you are saying.. but at the moment the process is not perfect
22:10:33FrihD leaves the room.
22:10:37headiusso fix the process
22:10:50headiusit's simple to have the ci bots pull a specific revision and update them periodically
22:10:58headiusthat's what vvsiz has been doing for us
22:11:03DefilerWell, it would be if I had commit access to the CI configuration
22:11:20drbrainautomatic is a bad idea
22:11:25drbrainit'll just hide problems
22:11:38wifelette_ enters the room.
22:11:40wifelette leaves the room.
22:11:42headiusautomatic what?
22:12:01DefilerAnyway, I just wanted to clarify how it works right this second.
22:12:06DefilerI am sure everyone agrees that it should change.
22:12:10headiusI'm saying you don't automatically update the tags
22:12:17drbrainok
22:12:20headiusyou freeze CI to a specific revision of specs so it can stay green
22:12:29headiusand you manually update revision + tags periodically
22:12:52headiusotherwise you're trying to keep CI (which should stay green) synced with specs (which should ideally keep finding breakages)
22:15:35headiusand I think having everyone add tags when they find breaking specs also masks breakages
22:16:01headiusor otherwise every new spec that comes in needs to be accompanied by failing tags and bugs...and that's a lot of hassle
22:16:10rhubarb leaves the room.
22:16:35Fullmoon_ leaves the room.
22:16:36DefilerThe original idea was that the people submitting the specs are best-equipped to know whether they are currently passing on rbx
22:16:44drbrainhrm, ruby seems faster on my Parallels FreeBSD 7 image than it does native
22:16:49DefilerAs in, whether they are there to round out the 'ruby spec', or to expose a rubinius bug
22:18:40headiusDefiler: ci is to let you know when someting breaks, not when someone finds new breaking behavior
22:20:19DefilerRight
22:20:22Defilerhence tags
22:21:22headiusCI shouldn't be synced to the leading edge
22:21:23ijcd_ enters the room.
22:21:24headiusperiod
22:21:32ubiquitous enters the room.
22:21:52headiusit's to tell you when new things break, not when new specs don't pass...and you should update tags when you want to update to a more recent set of spec
22:21:54DefilerI don't think it is that black and white
22:22:09DefilerWe could sync it to HEAD if we switched from 'fails' to 'passes' as the tag it uses
22:22:23DefilerThat way new specs wouldn't be included until they were tagged as 'safe for ci'
22:22:32langenberg leaves the room.
22:22:42headiuswhich would basically be the same thing
22:23:19headiusyou're specifying a set of specs you expect to pass in either case...either those listed as passing on HEAD or those not listed as failing on a specific revision
22:23:43headiusCI should be run against a set of specs you expect to always pass, not a set of specs that are changing day by day
22:24:11headiusput simply, I shouldn't be able to break CI by adding specs
22:24:59rueIt really is not much of a difference
22:26:06rueEither all committed specs are automatically excluded and CI must be updated manually or then all committed specs are automatically added to CI and must be manually excluded
22:26:43headiusno
22:26:46headiusboth are wrong
22:27:02headiuswait
22:27:12headiusyou're confusing me
22:27:30headiusin the former, are you talking about CI always being on head but not including new specs until they've been vetted?
22:28:04headiusI think CI should be a manual update always, because you want it to always be running the same set of tests
22:28:36evanso
22:28:48evanthe problem here is that the term 'ci' is overloaded
22:29:10evanDefiler is talking about 'bin/mspec ci' which we use constantly, to test that things aren't busted before a checkin
22:29:36evanheadius: you're talking about some developer-agnostic task
22:29:44headiuswell, I'm talking about the same goal
22:29:52evanthat occurs in the background, checking on the state of the current work
22:29:55headiusci in whatever form should be run against a set of tests you expect to always pass
22:30:00evanthey're similar, but not the same.
22:30:04headiusso you know you aren't regressing
22:30:11headiusme adding a new spec rubinius doesn't pass isn't rubinius regressing
22:31:24evanthis is a problem because the current 'ci' system is exclusionary, rather than inclusionary
22:32:48ijcd leaves the room.
22:33:02headiuswhich is why having it tied to a revision would work just fine
22:33:10evannot really
22:33:17headiussure it would
22:33:20headiusthat's what we're doing on jruby
22:33:33evanwe'd never be testing the current code, we'd be always testing the past
22:33:34headiusa specific revision, periodically updated and retagged
22:33:41headiusno, you're always testing the current code
22:33:45evanunless you mean a specific version of the specs
22:33:47headiusbut you're not testing against new specs
22:33:48evannot the entire checkout
22:33:49headiusthat's what I mean
22:33:53headiusspecific version of specs
22:34:12headiusyeah, there'd be no point to use a specific version of the whole repo for ci
22:34:21headiushead of rubinius, specific revision of specs
22:34:27headiusruby specs, anyway
22:35:06evanwhen does VVSiz move the revision forward?
22:35:09evanand does only he do it?
22:35:38evanand what uses that mechanism? just the background ci, or do you use it to run specs while you're doing dev
22:35:43headiusI've done it before, but he's mostly been doing it now
22:35:51headiuswe use it to spec during dev too
22:36:17headius"ant spec" runs the specs tied to a specific revision, for watching regressions
22:36:17evanhe moves it forward once a week?
22:36:19headius"ant spec-all" runs head of specs
22:36:20evanonce a day?
22:36:22evanonce a month?
22:36:25headiussomething like once a week
22:36:27headiusso far
22:36:50rueWell, you must remember that CI has an exact role
22:36:57headiushe's also been actively fixing spec failures as they come up, so that's kinda been his role
22:37:06rueEveryone trying to fix things should be running just plain mspec
22:37:12headiusabsolutely
22:37:22rueCI is there only to make sure nothing currently working breaks
22:38:45evanheadius: if you run all specs/tests, do you have a large number of failing ones?
22:38:56evanheadius: or are you typically green across the board
22:39:00headiushovers around 100
22:39:04headiusfor the full run
22:39:10headiusF+E
22:39:30headiusdepends on jruby flags too....if I turn on fork the process specs look a lot better, but the JVM starts freaking out
22:39:32evando you just know the ones that normally F/E and ignore them then?
22:39:46headiusthere's only expected to be F/E for spec-all
22:39:59headiuscI runs against a specific revision of specs
22:40:02headiusCI
22:40:43evanyour dev cycle is what then? edit; compile; ant run-ci-specs; repeat
22:40:43evan?
22:40:55headiusyep, basically
22:41:07headiuswe also have our normal test run, which takes longer
22:41:15evanwhere 'ant run-ci-specs' uses VVSiz's checkpoint
22:41:21headiusyep
22:41:38headiusso it stays green
22:41:56headius(though right now there's 13E because some things differ on mac and I haven't fixed them yet)
22:42:50headiusVVSiz is on linux
22:48:28flori_ enters the room.
22:49:34flori leaves the room.
22:49:43Fullmoon_ enters the room.
22:51:07ijcd_ leaves the room.
22:51:33Fullmoon_ leaves the room.
22:52:00ijcd enters the room.
22:53:10Fullmoon_ enters the room.
23:00:41drbrain leaves the room.
23:03:26drbrain enters the room.
23:05:39nicksieger leaves the room.
23:08:30dodecaphonic leaves the room.
23:18:12macarthy leaves the room.
23:19:29macarthy enters the room.
23:19:34rares enters the room.
23:19:53benburkert leaves the room.
23:20:17benburkert enters the room.
23:22:56benburkert leaves the room.
23:23:16benburkert enters the room.
23:25:34drbrain leaves the room.
23:28:18drbrain enters the room.
23:31:41pd enters the room.
23:32:18drbrain leaves the room.
23:33:36crafterm enters the room.
23:35:04drbrain enters the room.
23:35:38lstoll enters the room.
23:40:23rares leaves the room.
23:47:25benburkert leaves the room.
23:47:52Fullmoon_ leaves the room.
23:48:41lstoll_ enters the room.
23:48:54benburkert enters the room.
23:49:08Arjen_ leaves the room.
23:57:42Fullmoon leaves the room.
23:58:03wifelette leaves the room.
23:59:05lstoll leaves the room.
23:59:56djwhittgot a spec question...