Index

Show enters and exits. Hide enters and exits.

00:01:33Arjen_ leaves the room.
00:04:52zenspiderhuh... I can repro running on the cmdline but only if I run the whole suite
00:05:03zenspiderrepro 'multiruby -v -S mspec 1.8/library/readline > x 2>&1'
00:05:19brixendrbrain: http://pastie.org/223005
00:05:22zenspideroops. no, that is NOT a repro
00:05:26brixendrbrain: tell me what I'm missing
00:05:34zenspiderchomp off library/readline and it'll rerpo
00:05:46drbrainzenspider is working on it, be patient
00:06:20zenspider1.8/library won't repro either...
00:06:24zenspiderinteresting
00:07:27wmoxam leaves the room.
00:09:53zenspidersmallest repro I have is: ~/.multiruby/install/mri_1_8/bin/ruby -v -S mspec 1.8
00:10:01michalw leaves the room.
00:10:29zenspidernope... damnit.
00:12:45zenspiderdamnit damnit damnit
00:13:06zenspiderI've backed out all my changes and I can't hand repro this at all
00:14:37ryanlowe_ enters the room.
00:18:56zenspiderbut it repros perfectly using ci.rb... damn.
00:19:11zenspidereven the ruby -e 'system "..."' hack I did before has no effect
00:19:34zenspiderI'll wait 5-10 minutes to make sure it isn't just horking the IO
00:20:17zenspidertho... I'm at 0 CPU. :/
00:20:54zenspideryup. and I kill it and the log fills up
00:21:46rue leaves the room.
00:22:30rue enters the room.
00:22:40taf2 leaves the room.
00:22:57EugZol leaves the room.
00:25:02ezmobius leaves the room.
00:25:13zenspiderok... best I can do at this stage... (damnit)
00:25:49zenspiderremoving requires to readline gets me full test results... anything else does not. I don't have a better/smaller repro
00:25:58evanzenspider: do you have a repro I can look at?
00:27:25evanwhats the command you're running?
00:27:37evani haven't been following, was on the phone.
00:27:38imajes leaves the room.
00:27:59zenspiderevan: repro involves code not pushed yet
00:28:03evank.
00:30:25zenspiderevan: I guess I push my ci code... I'm stuck.
00:30:29zenspiderwell...
00:30:39zenspiderI'm stuck, short of removing readline specs
00:30:50zenspiderremoving readline seems to work fine
00:30:58zenspiderbut it is a sucky solution
00:31:24ryanlowe leaves the room.
00:31:32zenspiderone I'm willing to go with at this point in time tho
00:34:06zenspiderevan: make a call
00:34:16evansorry, one sec.
00:34:45evanso, readline on MRI hangs?
00:35:03evanperhaps just quarantine the specs for now
00:35:04drbrainyes
00:35:10evanthen we'll sort it out
00:35:14evanand unquarantine them.
00:37:18evandoes that work?
00:37:50cored enters the room.
00:38:41zenspiderevan: quarantine how?
00:38:57evani think there is Kernel#quarantine in mspec
00:39:00evantakes a block
00:39:02evanjust doesn't run it
00:39:04evanotherwise just
00:39:08evanif false
00:39:10evanout the whole thing
00:39:17zenspiderk
00:39:36evanwith a comment please
00:39:56zenspiderkk
00:40:01RyanTM leaves the room.
00:40:16lopex leaves the room.
00:45:26zenspiderevan: http://rafb.net/p/d4F8Ax26.html that's my diffs so far. huge because of indention, but whatever
00:45:35zenspiderthe kill specs prolly need better treatment
00:45:36evandon't indent it.
00:45:37evanjust put
00:45:41evanif false
00:45:42evanat the top
00:45:44evanend at the bottom
00:46:02zenspiderI hate you. :P
00:46:38evanI know you do.
00:46:46evanalso, remove the comments is Proces
00:46:49evanand quaratine them.
00:46:56MenTaLguYis there a reason not to delete stuff rather than comment it out?
00:47:07evanwell, for these, we need to fix them.
00:47:10brixenwait, why are we quarantining readline specs that run just fine
00:47:11anteaya leaves the room.
00:47:27evanbrixen: they don't seem to run fine on MRI
00:47:31brixenthey do
00:47:34brixenjust not in ci.rb
00:47:50evanwho's problem is it then? ci.rb or the specs?
00:47:53brixenci.rb
00:48:05brixenno one else is having a problem with them
00:48:17evani'm trying to unblock zenspider
00:48:20brixenwe cannot just quarantine rubyspecs that we have a problem running
00:48:23zenspiderum. no. not true
00:48:35zenspiderrubuildiuses are having problems as well
00:48:52zenspiderand yes, we can, and should. they're not stable
00:49:20brixenI beg to differ
00:49:49shame leaves the room.
00:50:09evanrather than fighting about it
00:50:25evani'd rather than zenspider quaratine them AND provide a proper repro
00:50:38brixenwell, I strongly object
00:50:40evanso that either ci.rb can be fixed or the specs can be fixed.
00:50:46brixenthere are multiple users of the rubysepcs
00:50:59brixenwe cannot go quarantining them when we can't run them
00:51:01drbrainit's both ci.rb and rubuildius having this problem
00:51:15brixendrbrain: show me where the bot is having a problem with readline
00:51:33drbrainit's in your scrollback
00:51:45drbrainmy client doesn't have search, so I can't pull it up
00:51:50drbrainsearch for 'rake'
00:51:54drbrainor rubuildius
00:52:11brixenI'll check on the ey slice and on rubuildius_amd64
00:52:32evank
00:52:44evanzenspider: hold off for now.
00:52:54zenspider<djwhitt> thank goodness. I don't know what is causing it, but I've been
00:53:05zenspider having a terrible time getting rubuildius_amd64 to complete a run
00:53:09zenspider
00:53:13brixenzenspider: that doesn't mean anything
00:53:19brixenI want to see a repro
00:53:28brixenevan: I can't access ey slice
00:53:34evanhrm.
00:54:07zenspiderI'm so done with this... I'm going home
00:54:25drbrain[2008/06/26 15:37:38] <djwhitt> zenspider: yeah, regarding the two processes thing, that's my experience too. which is why I don't run the specs using rake in rubuildius
00:54:44drbraincorundum: botsnack
00:54:44corundumdrbrain: schweet!
00:55:36drbrainwe're getting kicked out of vivace, bbiab
00:58:05brixendjwhitt: ping
00:59:27brixendjwhitt: nvm, I got in
01:05:18yasuhito enters the room.
01:08:33qwert666_ leaves the room.
01:11:01jero5 leaves the room.
01:11:08headius leaves the room.
01:12:16heycarsten enters the room.
01:16:31brixendjwhitt: fwiw, I just completed a bin/mspec ci -B full.mspec run on the amd64 machine
01:16:42brixendjwhitt: but I saw stuff like this go by several times: debug2: channel 0: window 999367 sent adjust 49209
01:16:49brixendjwhitt: any idea what that is?
01:18:20drbrainbrixen: that's from ssh
01:18:44evanyeah, thats ssh debug stuff
01:18:49drbrainyou probably still have -vv set
01:18:53brixendjwhitt: I'm still getting the perm issues, is your cleanup script available
01:19:04brixendrbrain: ahh, indeed I do, thanks
01:19:06evanbrixen: i think that the hang was rubuildius_amd64 actually running it
01:19:17drbrain... via rake
01:19:27evannot just via bin/mspec at a prompt
01:19:43brixenI'll run it as a subprocess with system
01:21:42rue leaves the room.
01:23:01rue enters the room.
01:25:12evanbrixen: ey slice is back up
01:25:26brixenruby -e "system('bin/mspec ci -B full.mspec -V > spec.out')"
01:25:30brixenthat completed fine
01:25:33brixenevan: thanks
01:25:45brixenhow else would you like me to try running them?
01:25:56drbraintry running specs via rake
01:26:02drbrainvia system
01:26:03brixenI know there is an issue with rake
01:26:11brixenbut what do I assume from that?
01:26:19drbrainwhat issue?
01:26:30brixenI'll run it and see
01:26:34brixenI've seen it hang
01:26:48brixendoesn't rake just #system the command?
01:27:51twbray leaves the room.
01:28:20drbrainultimately, yes, but not with shell
01:28:27drbrainerr, optionally not with shell
01:28:32drbrainsystem(*cmd)
01:28:41brixenwhat does 'sh "bin/mspec ci -B full.mspec"' do?
01:29:35drbrainlooks like system "bin/mspec ci ..."
01:29:44evanif it does that
01:29:46evanit uses exec
01:29:47evannot system.
01:30:04drbrainno, Rake#sh uses system
01:30:09drbrainI just looked at ti
01:30:12drbrainit
01:30:18evanyeah but
01:30:25evanMRI uses exec(2) for #system
01:30:32evanif there are no shell characters.
01:31:03evanif it does a system(*ary), whats in the ary?
01:31:22drbrainif ary is a string, it uses sh
01:31:54drbrainit's basically def sh(*cmd) system(*cmd) end
01:32:03drbrainwith a bunch of printy stuff wrapped around it
01:32:15brixenit's hanging on spec/frozen/1.8/library/net/http/get_spec.rb
01:32:31brixenall the net specs likely need to be checked for races
01:32:37brixenthese are indeterminate hangs
01:32:48brixenI tagged a bunch already to sync the specs
01:34:38brixenI'll run again with those excluded
01:38:03cored leaves the room.
01:38:57benburkert leaves the room.
01:41:18ryanloweI'm not really concerned about speed -- but should Rubinius be able to run a real Rails app? I thought it might be a good way to find problems .... if so, has anyone documented how to use Rubinius with Rails?
01:41:28wmoxam enters the room.
01:41:59crafterm enters the room.
01:42:16drbrainryanlowe_: it should, and it should work just the same as you would with ruby, but use 'rbx' instead of 'ruby'
01:42:24brixenrunning the specs via rake while excluding Net::* specs works fine
01:42:35crafterm enters the room.
01:42:57ryanloweok i'll try to get a skeleton working with Rails 2.1.0
01:43:39drbrainryanlowe_: of course, we may have regressed :)
01:43:55ryanloweoh ya? would now not be a good time in the cycle to try this?
01:44:18drbrainwell, regression can happen at any time
01:44:26ryanloweI'm not really familiar with Rubinius' version numbers, iterations, etc
01:44:32drbrainI don't think now is worse than any other day
01:44:34ryanloweI was just going to try it with master :)
01:44:45ryanloweor would that be a bad idea? :)
01:44:54drbrainactual version of rubinius doesn't matter
01:44:58drbrainyes, try master
01:45:02drbrainor head
01:45:04ryanloweok
01:45:14anteaya enters the room.
01:45:56yasuhito leaves the room.
01:46:27yasuhito enters the room.
01:46:58yasuhito leaves the room.
01:49:29ryanlowedrbrain: is anyone else doing this already?
01:49:40ryanloweor is it kind of ad hoc?
01:49:56drbrainad-hoc
01:50:04drbrainpeople pick what they like to do
01:50:26ryanloweno I mean -- is anyone specifically testing Rails support with a real project?
01:50:26rubbish enters the room.
01:50:34ryanloweand sharing experiences, etc
01:50:41naeu leaves the room.
01:50:41drbrainnot that I've heard of
01:50:48ryanloweok
01:50:50drbrainwe did some rails work before RailsConf
01:51:22cezarsa leaves the room.
01:52:43ryanloweI saw that, that's why I'm here ;)
01:52:47shame enters the room.
01:54:26jtoy enters the room.
01:55:01ryanlowegreat progress btw, I was pleasantly surprised
01:55:04sudoer enters the room.
01:56:44drbrainthanks!
01:59:49boyscout3 commits by Chad Fowler
01:59:50boyscout * Merge branch 'master' of git@git.rubini.us:code; 9a01c92
01:59:51boyscout * resolve merge conflicts. enabled Array#__rescue_match in rescue splat compilation; 8ea5cb1
01:59:52boyscout * Implemented __rescue_match__ for Array and call that from the compiler in splat rescu ...; 43628b1
02:06:55rubuildius_ppcChad Fowler: 9a01c926d; 2533 files, 8434 examples, 28074 expectations, 0 failures, 0 errors
02:07:26brixenhm, ppc bot is fine
02:07:34brixenI just killed a process on the amd64 bot
02:07:41brixenit's currently running chad's commit
02:08:47djwhittbrixen: oh yeah, I've been having a terrible time with that bot hanging recently
02:09:08brixendjwhitt: it's hanging in the net/** specs
02:09:20brixenI've tagged some of them, I'm going to tag the rest for now
02:09:38djwhittk
02:09:38brixendjwhitt: you should run bin/mspec ci -B full.mspec -V > something
02:09:45brixenso that you can cat something to see where it's at
02:10:27djwhittyou mean the bot should do that?
02:10:39brixenyeah
02:10:45lstoll enters the room.
02:10:46djwhittI can add that in
02:10:49brixenmight be helpful
02:11:04brixenit's obvious the thing is hung and not just spinning cus the cpu was at 2.1%
02:11:10djwhittyeah
02:11:11brixennormally it's +90%
02:12:17brixenos x seems not to have the issues
02:12:17jtoy leaves the room.
02:12:23brixenI'm going to try on ubuntu
02:12:29blakewatters enters the room.
02:20:09heycarsten leaves the room.
02:20:25brixenugh parallels crashed on me
02:21:42rue leaves the room.
02:22:31rue enters the room.
02:22:50brixendjwhitt: interesting to watch the cpu util on this spec run, it has slowly degrated from 90% to 22% over several minutes
02:22:59brixener degraded
02:23:57VVSiz_ enters the room.
02:25:50qrushI'm not sure if this is the best place, but I've got a Rails question. #rubyonrails doesn't have all of the ruby/rails experts ;)
02:26:30drbrainthis is certainly the wrong place
02:26:36qrushAlright. :) figured i'd ask.
02:27:10ryanlowehas a working rbx ... woot
02:27:59trythil enters the room.
02:29:39ryanlowedrbrain: I'm digging into the Rails guts to figure out where to replace 'ruby' with 'rbx'...
02:29:59drbrainwell, start with
02:30:04drbrainrbx gem install rake
02:30:09drbrainthen rbx rails blah
02:30:47drbrainerr, rbx gem install rails --no-rdoc --no-ri
02:30:49ryanlowehmm ... I already have rake but with ruby 1.8.6
02:30:53drbrainsince RDoc is still quite slow
02:30:58djwhittryanlowe: I tried Rails out a couple weeks ago asside from weirdness with ActiveRecord things seemed to generally work
02:31:02ryanlowecan they co-exist on the same machine?
02:31:11drbrainyes
02:31:21drbrainrubinius installs it do a different path
02:31:25drbrainto
02:31:28ryanloweahh rock on
02:31:53drbrainrbx rake
02:31:58drbraininstead of rake
02:37:39ryanloweok done
02:37:46ryanlowetrying out: rbx script/server -p 9999
02:38:44ryanloweworks
02:39:21c0sin leaves the room.
02:39:41drbraincool
02:40:32ryanloweuses WEBrick by default
02:40:52twbray enters the room.
02:41:49VVSiz leaves the room.
02:46:06ryanloweI tried: rbx gem install mongrel ... but when I run: ruby script/server mongrel -p 9999 it says mongrel is not on my path
02:46:49ryanlowethe mongrel gem 1.1.5 seemed to install OK with rbx
02:47:09djwhittryanlowe: if you installed the mongrel gem successfully script/server should use it by default
02:47:15ryanlowehmm
02:47:23djwhittryanlowe: though I haven't tried it in rbx
02:47:46ryanlowewhere does rubinius store the gems I install with rbx?
02:48:09ryanloweI didn't get an error message from rbx gem install mongrel, so I figured I was all good :)
02:49:14ryanloweI ask where it's stored because perhaps it's not on a PATH rbx can see
02:49:41ezmobius enters the room.
02:50:05drbrainryanlowe: rbx gem env
02:50:27drbrainprefix anything you'd ordinarily run via the command line with rbx
02:50:38drbrainso, `rake` for ruby's rake, `rbx rake` for rubinius rake
02:51:01drbrain&c
02:51:27ryanlowek
02:51:35ryanlowemongrel is on: rbx gem list
02:52:10drbrainso, maybe `rbx mongrel_rails` ?
02:52:49ryanlowek i'll try that ... maybe it just doesn't like script/server
02:57:26ryanlowemongrel_rails is using 100% CPU ... not sure if it's locked up or just starting up :)
02:58:49wmoxam leaves the room.
03:00:07drbrainsometimes you can open up Activity Monitor and sample rbx to see if it's hung
03:00:21ryanlowek
03:00:25drbrainor, if it's the VM that hung vs some ruby code that's spinning furiously
03:00:38drbrainif you've got cpu_run a bunch, it's still trying to do stuff :)
03:00:50ryanloweit gave me a backtrace when I interrupted, would you like me to pastie it?
03:01:00drbrainsure
03:01:17ezmobiusyou have to be patient to boot a rails app
03:01:22ezmobiusit can takea minute or two ;)
03:01:34drbrainhis WEBrick app came up pretty quick
03:01:39ryanloweWEBrick took longer than MRI, but not that long
03:01:48ezmobiusahh
03:01:57ryanlowemongrel_rails was using 100% CPU for 5 minutes :)
03:02:26ryanlowehttp://www.pastie.org/223078
03:02:56twbray leaves the room.
03:03:52drbrainodd...
03:04:09ryanlowebtw when I start WEBrick with script/server I get something almost immediately as output:
03:04:12ryanloweclass Float held ownership of the tag:yaml.org,2002:float tag
03:04:40drbrainwe have yaml issues that are yet to be resolved
03:04:44drbrainsomeday!
03:04:50ryanloweWEBrick took 50 seconds of CPU time to start
03:06:27ryanlowewhat is a *.rbc file?
03:06:42drbrainI need to add a ^T handler to rubinius :/
03:06:43ryanlowethey are littering my project in vendor/rails now :)
03:06:43elight leaves the room.
03:06:49drbrainit is the compiled bytecode
03:07:04ryanloweare they always alongside the source file?
03:07:16drbraincompiling is still a rather time-consuming process, so rubinius caches its work
03:07:20elight enters the room.
03:07:20drbrainyes
03:07:39ryanloweinteresting
03:07:52ryanloweplop that into .gitignore
03:09:21lchin enters the room.
03:09:22drbrainI'm going to run mongrel_rails for a while and see if it finishes doing anything
03:09:27drbrainit seems to be trying
03:09:29blakewatters leaves the room.
03:09:31ryanloweok
03:09:42ryanloweI'm on a lowly Mac Mini 2Ghz :)
03:10:48ryanlowehttp://github.com/ryanlowe/unfoo/tree/master/RUNNING
03:10:50drbrainI'm not much faster a 2.6GHz
03:10:54twbray enters the room.
03:11:13ryanloweif anyone else tries to get a Rails project running, you can point them there
03:11:27heycarsten enters the room.
03:11:32heycarsten leaves the room.
03:11:58ryanloweI'm going to see if I can get this running publicly on Slicehost ... the data won't be sensitive
03:12:18ryanloweperformance should be interesting if I need to use WEBrick tho :)
03:13:26ryanlowethanks for your help, drbrain
03:13:52drbrainmine's still churning away, running 250.37u 2.98s
03:14:33ryanlowe:)
03:14:52ryanlowethat backtrace seemed to be saying it was still figuring out dependencies 5 minutes into mine?
03:15:01ryanlowemaybe a cyclic problem?
03:15:09twbray leaves the room.
03:16:28elight leaves the room.
03:16:35kli enters the room.
03:16:41drbrainyeah, it seems to be stuck in RubyGems
03:16:58drbrainI just killed mine after running 411.44u 5.08s
03:17:04drbrainstill in RubyGems
03:17:15drbrainI need to fix my RBX first
03:17:18drbrainsec
03:17:26kli leaves the room.
03:18:09ryanloweI'll bbl, short break
03:20:14kli enters the room.
03:20:48elight enters the room.
03:23:08headius enters the room.
03:23:47elight enters the room.
03:28:12drbrainwhy can't I trap INFO?
03:28:18drbrainit does nothing, and makes me sad
03:29:25drbrain... and works in MRI
03:31:00ezmobius leaves the room.
03:31:30twbray enters the room.
03:33:10drbrainok, so signals don't get handled unless I sleep
03:33:26drbrainand caller from trap is not useful
03:34:58twbray leaves the room.
03:35:07drbrain(in rbx)
03:37:02crafterm leaves the room.
03:37:27evandrbrain: how were you waiting if not sleep?
03:37:53wmoxam enters the room.
03:37:56drbrainhttp://rafb.net/p/nwGJYC87.html
03:38:10drbrainif I take the sleep out in rbx, ^T never gets handled
03:39:09drbrainwhat I really want to do his hit ^T and see what the method I was just running in is
03:39:21drbrainthat script (sleep removed) works that way in MRI
03:40:33drbrainhttp://rafb.net/p/OiKZNp31.html
03:40:37drbrainWTF IS THAT!!!
03:40:44drbrainbegin ... retry
03:40:53cremes_ enters the room.
03:41:09drbrainno checking, just assume it fails?
03:43:04kevwil enters the room.
03:45:26evandrbrain: you want to busy wait?
03:45:34evanwe have pre-empting
03:45:41evanperhaps it's not checking events when it pre-empts.
03:45:55drbrainhere's my Grand Plan
03:46:22drbrainnormally, you hit ^T and the shell spits out your pid and load and system/user times
03:46:29evanyep.
03:46:33evanwant that to work in rubinius?
03:46:36evanbacktrace wise
03:46:44drbrainI'd like for Rubinius to have a ^T handler that also spits out what method is currently running (at least)
03:46:50qrush leaves the room.
03:46:53headius leaves the room.
03:46:54evansure, easy enough.
03:47:02evanlet me check the preempt code
03:47:02drbrainfull backtrace is too much
03:47:10evani'll bet it's just not polling event
03:47:11evans
03:47:18drbrainthe loop part there was just to see where it was to know it was working
03:47:28evanah.
03:47:32evanwell, don't forget
03:47:34evanin rubinius
03:47:43drbrainit's hard to test if you immediately exit
03:47:43evansignals aren't delivered right away
03:47:47evansame with MRI
03:47:57evanthe fact that the signal occured is marked
03:48:00drbrainwell, it's ok if you just get close
03:48:07evanand at some point in the future, it's checked.
03:48:27evanclose might be 10 milliseconds away
03:48:29drbrainespecially for rubinius, to know "I was over in X, so I've gotten this far"
03:48:35evanok.
03:49:06qrush enters the room.
03:50:24drbraineven 100ms would be fine
03:51:28drbrainfew people could tell that their ^T got handled too late
03:52:45boyscout6 commits by Eric Hodel
03:52:46boyscout * Fix fastthread gem so that fastthread.rb can be required. Fixes mongrel.; 5b00438
03:52:47boyscout * Fix struct generation from .ffi files; f0c5d2d
03:52:48boyscout * Move ConstGenerator data sources inside their respective blocks; 931147a
03:52:49boyscout * Move TypesGenerator to lib/ffi; 2ccc23b
03:52:50boyscout * Move ConstGenerator out of rakelib, reduce generation code in platform.rake.; b709d88
03:52:51boyscout ...
03:52:52drbrainI got a china iPhone spam
03:53:08drbraingo rubuildius go!
03:53:20drbrainparticularly amd64
03:53:58drbrainI rearranged a lot of rakelib/platform.rake, and not all the include files for the const generator were required on OS X
03:54:04drbrainwhich makes me puzzled
03:54:05djwhitthmm... that's a tall order of late. looks like their's a process hung right now. I'll see if I can fix it
03:54:08elight leaves the room.
03:54:11djwhitt*there's
03:54:17kevwil leaves the room.
03:54:26rubuildius_amd64 leaves the room.
03:55:30rubuildius_amd64 enters the room.
03:56:01imperator enters the room.
03:56:41wvdschel enters the room.
03:57:31rubuildius_ppcEric Hodel: 5b0043852; bin/ci failed! http://rafb.net/p/zAF4PA53.html
03:58:04rubuildius_amd64Eric Hodel: 5b0043852; build failed! http://rafb.net/p/rlaBCS40.html
03:58:12gnufied leaves the room.
04:01:12drbrainhaha, I broke the world!
04:01:40drbrain... Assertion failed: (("ev_io_start called with negative fd", fd >= 0)), function ev_io_start, file ev.c, line 1714.
04:01:46drbrainbut I don't think that's me
04:02:14djwhittdrbrain: just verified that amd64 error by hand on my laptop too
04:02:35drbraindjwhitt: do this:
04:02:48drbrainrm runtime/platform.conf; rake runtime/platform.conf
04:03:32djwhitttrying it ...
04:03:41drbrainoh, and FWIW, I've watched this 6 times today: http://www.youtube.com/watch?v=X8Nc8RCLy1s
04:04:47djwhittdrbrain: same error
04:05:10drbraindid the platform.conf generate correctly?
04:05:19drbraincan you paste it to me if there were no errors?
04:05:25djwhittsure
04:05:40eventualbuddha enters the room.
04:06:00eventualbuddha leaves the room.
04:06:30drbrainhow do you make a video loop in youtube?
04:06:48djwhittdon't do it man. think of the children
04:08:13djwhittdrbrain: http://rafb.net/p/hIJFEc90.html
04:08:52twbray enters the room.
04:09:20drbrainah-ha! I'm missing a bunch of your flags
04:10:49rubbish leaves the room.
04:11:26drbraindjwhitt: does this fix? http://rafb.net/p/UR3BSa62.html
04:11:36drbrainI am looking for which header would have your TCP_NODELAY
04:11:39drbraintoo
04:11:46drbrainI think you should have one of those
04:12:01kevwil enters the room.
04:12:35drbraindjwhitt: try http://rafb.net/p/gy9Q5w42.html
04:12:43drbrainit may have TCP_NODELAY and so-on filled in too
04:12:48gnufied enters the room.
04:12:51djwhittdrbrain: k
04:12:54drbraingrep TCP_NODELAY runtime/platform.conf
04:15:42masswaves hello
04:15:52drbrainhi mass!
04:15:56drbrainI broke it!
04:15:59djwhittyep has TCP_NODELAY set to 1 now
04:16:16drbrainbetter
04:16:22djwhitttrying build now
04:16:23masswhatcha breaking/fixing? :)
04:16:37drbraingrep -v ' = [a-z0-9-]' runtime/platform.conf
04:16:48drbraincan you send me the output of that, to see if there's something I missed?
04:17:04drbrainmass: I reworked how FFI looks up C structs and constants and types a bit
04:17:29drbrainso that FFI extensions don't need to depend on rubinius' build architecture
04:18:13drbrainbut I took out too many headers during my reorganization
04:18:47djwhittdrbrain: http://rafb.net/p/xrQzf845.html
04:19:04djwhittthough, the build succeeds now
04:19:15drbrainsome will be blank
04:19:26drbrainnot everybody has SIGDANGER, for example
04:20:31drbraincan you grep -r SO_REUSEPORT /usr/include
04:20:39drbrainyou should definitely have that one
04:21:41djwhittheh, this is funny
04:22:01djwhittjust a sec ..
04:22:29drbrainFWIW, I have 49 missing
04:22:33djwhitthttp://pastie.org/223096
04:24:33drbrainwow, that's really bizarre
04:24:57drbrainyou are on linux?
04:24:59djwhittyep
04:25:04djwhittGentoo
04:25:11cremes_ leaves the room.
04:25:31drbrain"As under Linux SO_REUSEPORT is not implemented and SO_REUSEADDR does allow multiple processes to utilize the same host:port pair without having to set ..."
04:25:36drbrainso says google
04:25:41drbrainso, I guess you're good then
04:25:46drbrainI will commit this
04:26:22djwhittyeah, ci (not full) ran fine too
04:26:29drbrainok
04:26:52drbrainI just wanted to be sure for something that isn't necessarily covered by ci
04:27:45boyscout1 commit by Eric Hodel
04:27:46boyscout * Restore includes for linux. My Bad. Pair: David Whittington.; 184fbf3
04:28:14drbrainryanlowe: update, I fixed mongrel
04:28:27ryanlowedrbrain: ohhh neat
04:28:57ryanlowewhat was the issue?
04:29:15drbrainmongrel has a purposeful infinite loop in it
04:31:16ryanloweand it was stuck in that loop?
04:31:26drbrainyup
04:31:41ryanloweok i will pull and recompile
04:35:26drbrainI filed a bug
04:35:55ryanlowe...for?
04:36:09smparkes leaves the room.
04:36:48drbrainmongrel
04:37:39ryanlowelink?
04:39:12drbrainhttp://mongrel.rubyforge.org/ticket/37
04:39:54ryanlowethx
04:39:59ryanlowemongrel works for me now with rbx
04:40:07drbrainyay!
04:40:14ryanlowetakes about 60 seconds to load up tho
04:41:13ryanloweworks: rbx mongrel_rails start -p 9999
04:41:29drbrainyay!
04:41:30ryanlowedoesn't work: rbx script/server mongrel -p 9999
04:41:54ryanlowethe latter gives: PROBLEM: Mongrel is not available on your system (or not in your path)
04:42:11drbrainah
04:42:13ryanloweperhaps I'm using the params right
04:42:19ryanloweI'm not* using...
04:42:24drbrainno, it is correct
04:43:17drbrainyou would need to add the executable directory (in rbx gem env) for it to show up
04:43:34drbrainadd it to PATH
04:44:59ryanloweEXECUTABLE DIRECTORY: /usr/local/lib/rubinius/0.9/bin
04:45:25drbrainyes
04:46:30ffeynman enters the room.
04:46:44ryanlowenope, that didn't do it
04:49:05ryanlowehmmm just a sec ... my PATH isn't sticking...
04:49:07rubbish enters the room.
04:49:46kevwil leaves the room.
04:53:58ryanloweif the executables have the same name as MRI executables, do I want the Rubinius dir at the end of my path?
04:54:09ryanlowewhat does it do to my MRI support?
04:55:07drbrainend, yes
04:55:20elight enters the room.
04:55:25drbrainMRI will get picked first, which may not be what you want
04:55:41drbrainwhich is one of the reasons why we went with rbx rake and so-on
04:55:58ryanloweI had Rubinius first and script/server still couldn't find mongrel
04:56:08ryanloweit's not a big deal I'll just use: rbx mongrel_rails
04:56:15drbrainthen there may be a different problem
04:56:16ryanlowethanks for fixing it :)
04:56:27drbrainnp
04:56:40ryanloweis this a bug you'd like me to report?
04:56:48ryanloweor do you guys care about Rails bugs yet?
04:56:59gnufied leaves the room.
04:57:23drbrainyes
04:57:43ryanloweroger that, I will do my best
04:57:48drbrainit'll help us most if you can distill it into a minimal testcase, though
04:57:56ryanlowehmmm
04:57:58elight leaves the room.
04:58:22ryanloweI don't know much about Rubinius or script/server up to this point :) maybe I should sit on it?
04:58:43ryanloweor by testcase do you mean to reproduce manually?
04:59:11drbrainsomething that can be run with a minimum amount of setup to reproduce what shouldn't happen
04:59:28ryanlowegotcha
04:59:34drbrainso we can ruby blah.rb and rbx blah.rb and see different output until we fix rubinius
05:00:02drbrainif you want to figure out the path thing, I'd start by going to wherever that message gets generated and poking around
05:00:12drbrainthen working backwards until you find something Extra Suspicious
05:00:42ryanlowegood tip, thanks
05:01:53ryanloweif I make changes to my Rails code in development, will Rubinius know the source has changed and recompile the changed file(s) for Rails? or will I have to restart the server?
05:01:53kli leaves the room.
05:02:18drbrainit should behave exactly like rails
05:02:23drbrainthat you're used to
05:02:36drbrainas far as .rbc files go, pretend they don't exist
05:02:47ryanloweuses his limited imagination
05:02:52evanwe could make the .rbc . files
05:02:56evanblah.rb => .blah.rbc
05:03:01evanto increase the illusion.
05:03:10drbrainheh
05:03:23ryanlowethen they would be insidious like .svn directories :S
05:03:41drbrainillusory but evil?
05:03:52ryanlowerbx script/generate controller took a while but appears to have worked
05:05:56drbrainevan: if you have a spare while, can you review rakelib/platform.rake?
05:06:09evansurely my fine seattle fellow.
05:09:23evanlight enters the room.
05:10:17ryanloweI got a backtrace from: rbx rake, running my 2 test suite
05:10:34evandrbrain: I read your diff
05:10:41evanlooks like moved some stuff from rakelib to lib/ffi
05:10:46evanbut you changed and moved in the same commit
05:10:47drbrainyes
05:10:50evanso i can't exactly tell what you did
05:11:07anteaya leaves the room.
05:11:07loincloth_ enters the room.
05:11:11drbrainI namespaced it under FFI, and did some minor cosmetic changes
05:11:22drbrainlike, take a block, have a name, to simplify platform.rake
05:12:08evank
05:12:10evanseems fine
05:12:19evanthe only change to platform.rake looks like move the constants
05:12:26evanand using stuff from now lib/ffi
05:12:27evanyes?
05:12:31drbrainyes
05:12:32loincloth leaves the room.
05:12:40evank
05:12:44evanseems on the up and up.
05:12:48tarcieriso I don't suppose if any of you know offhand if the parser bug that prevented me from installing Rake on rbx got fixed
05:12:49tarcieriheh
05:12:53tarcierirebuilds, retries
05:12:53drbrainok
05:12:59RyanTM enters the room.
05:13:01ryanlowerbx rake basic test run backtrace, regex blowup: http://www.pastie.org/223113
05:13:10evandrbrain: I didn't review every line though.
05:13:14ryanlowetarcieri: I installed rake with rbx today
05:13:16evanwas there something specific you wanted me to double check?
05:13:20drbrainit should be easy to reuse the const stuff in zlib.rb.ffi and so-on
05:13:23evanlight leaves the room.
05:13:27evanryanlowe: that warms my heart.
05:13:29drbrainno, I just wanted your impression
05:13:30tarcierisweet
05:13:44evanto know that someone can just happily install rake under rbx without having to hack anything.
05:13:51evani can probably die happy knowing that.
05:13:52tarcieriheh
05:13:56evan(probably)
05:14:02tarcieristill wants to put together a mongrel-rbx gem
05:14:05evan(we'll see, i'm only 28, there is still a ways to go)
05:14:11tarcieriwith the Rubinius.asm parser
05:14:21evansurely!
05:14:36tarcieriso if Rake works, that's possible
05:14:36tarcieriheh
05:14:37ryanloweI did as drbrain suggested today: sudo rbx gem install rake --no-rdoc --no-ri
05:14:50evanis going to keep doing that until someone says "ok, and don't call me shirley"
05:14:51drbrainryanlowe: pull out the regular expression from SelectorAssertions
05:15:01tarcieriyeah, earlier I was hitting some horrible bug in the (JRuby?) parser
05:15:16tarcieribut that was a few weeks ago
05:15:18drbrain ActionController::Assertions::SelectorAssertions.__module_init__ (SelectorAssertions) at /Users/ryanlowe/workspace/rails/unfoo/vendor/rails/actionpack/lib/action_controller/assertions/selec tor_assertions.rb:599
05:15:26drbrainthere's probably something funny going on in there
05:15:51tarcieriaight, rbx rebuilt, let's try Rake
05:15:51ryanlowemight be my use of assert_routing
05:15:55drbrainok, I'm going to watch a movie while it's still DAYLIGHT
05:16:01ryanloweI forgot I was supposed to take baby steps :)
05:16:24drbrainryanlowe: sounds like a baby step
05:16:27drbrain:)
05:16:32drbrainadd one test, right?
05:16:36drbrainanyhow, movie!
05:16:49ryanlowedrbrain: which movie?
05:17:11tarcieriSuccessfully installed rake-0.8.1
05:17:11tarcieri1 gem installed
05:17:13tarcieriw00t
05:17:30ryanlowetarcieri: I also installed rails with rbx today the same way
05:17:30evanbadabing!
05:18:29drbrainryanlowe: it's Unforgiven vs. L.A. Confidential
05:18:33drbrainI chose Unforgiven
05:18:35ryanloweooooh
05:18:42ryanlowetough choice, but a good decision
05:18:47ryanloweyou'll be gone a while :)
05:18:52drbrainboth of these movies are fantastic, so hard to choose
05:19:15ryanlowe131 minutes ... hmm, thought it was longer
05:21:07headius enters the room.
05:21:25tarcieriI'd like to experiment with Mongrel on multiple VMs
05:21:25rue leaves the room.
05:21:38tarcierithis will allow me to make a gem that doesn't need a C extension
05:22:01tarcieriand uhh, release it
05:22:02rue enters the room.
05:22:06ryanlowedrbrain: the regexp is Regexp.new("#{RJS_STATEMENTS[:any]}\\(\"([^\"]*)\", #{RJS_PATTERN_HTML}\\)", Regexp::MULTILINE)
05:22:07tarcieriEvan Weaver said he'd maintain it
05:22:18evansweet.
05:22:33evancries at dregx
05:23:41tarcierievan: at least, until you can pull of crazy LLVM hacks to make C extensions with globals work across VMs
05:23:42tarcieriheh
05:23:51evan:D
05:23:59evanI think i'll be able to make it work
05:24:07tarcierithat'd be so fucking sick
05:24:07tarcieriheh
05:24:07evanwhich will be o/~ aaaweeesooomeeee o/~
05:24:08evanlight enters the room.
05:24:17tarcieriyes aweeeeeeeeesome
05:24:58tarcieriof course, thing thing that'd REALLY be sick
05:25:02tarcieriis Rubinius::VM.fork :)
05:25:21tarcieriI honestly don't know if I can realistically pull anything off without it
05:25:25tarcierire: multi-VM mongrel
05:26:34evanlight leaves the room.
05:27:25AndrewO leaves the room.
05:27:46lchin leaves the room.
05:30:10ryanloweeven with a placeholder test that just asserts true, I get that backtrace
05:30:38ryanlowehttp://www.pastie.org/223113
05:31:38ryanlowegotta love Google indexing GitHub...
05:33:11crafterm enters the room.
05:33:34ryanloweis this a problem with regexp.rb's group_part_class? (line 389) or a Rails bug?
05:35:24lchin enters the room.
05:40:13ryanloweI just commented out the lines referencing that broken regex and the backtrace goes away ... test suite doesn't raise anything but doesn't run the one test
05:40:25ryanlowethis was part of the output this time: require 'test/unit/testcase' has been deprecated
05:40:33massryanlowe: its good google is indexing it; github doesn't really have search functionality
05:40:47ryanlowemass: agreed :) it's very handy
05:41:08ryanloweindexing that much code would be a heck of a job for GitHub
05:42:05mahargI have a feeling that multi-vm is going to be one of those things that everyone talks about and no one ends up actually using
05:44:28ffeynman leaves the room.
05:44:58loincloth_ leaves the room.
05:45:50ryanloweif I have a backtrace for a lighthouse bug, how should I attach it?
05:45:58evanplease
05:46:13ryanlowejust as a text file I guess eh
05:46:20evanor in the ticket is fine
05:46:23loincloth enters the room.
05:46:24evanlight enters the room.
05:46:26random8r enters the room.
05:46:29ryanloweok
05:47:21evanlight leaves the room.
05:47:28ryanloweI'm going through Rails to find bugs, places that need support to get Rubinius running Rails
05:47:46random8rhey I was wondering how do I compile a ruby script?
05:47:48ryanlowestarted a project at: http://github.com/ryanlowe/unfoo for it
05:48:03random8rif i have a ruby script called "blah.rb" how would I compile it using rubinius? (I've downloaded and built rbx)
05:48:21random8rI was looking for a shotgun command, but I couldn't find one
05:48:32random8ris it the rubinius executable in /shotgun
05:48:38fbuilesvrandom8r: yes
05:48:46fbuilesvrandom8r: just shotgun/rubinius your_script.rb
05:49:04RyanTM leaves the room.
05:49:40ryanlowehow do I indicate the Rubinius version number on a bug report?
05:49:41headiusmaharg: multiple vms are used in JRuby all the time
05:49:57ryanloweis it even useful or would people just go by the date it was reported?
05:50:22ryanloweversion number could be a Git commit number maybe
05:50:26random8rthanks
05:51:23fbuilesvryanlowe: by date should be fine but you can include the last commit hash if you want
05:51:32ryanlowek
05:51:46rubbish leaves the room.
05:54:37evanryanlowe: the output of `rbx -v` is useful.
05:54:56ryanlowerubinius 0.9.0 (ruby 1.8.6 compatible) (9a01c926d) (06/26/2008) [i686-apple-darwin9.2.2]
05:55:10ryanlowe9a01c926d is that last commit hash?
05:55:12loincloth leaves the room.
05:56:43fbuilesvryanlowe: not the last commit but it was made a few hours ago
05:57:02ryanlowewell the last commit I have that I'm running here :)
05:57:05fbuilesv:)
05:57:10evanyeah, thats a 'bug'
05:57:28evanif the file containing that hash isn't recompiled, it doesn't update.
05:57:31random8rjust did a time test between compiled and MRI, (1.8.6) and the complied shotgun code comes out 6 times slower!
05:57:32random8rwhy?
05:57:51evanrandom8r: what was the code?
05:57:58random8roh just a loop through an array
05:58:01random8rhang on i'll pastie
05:58:18evandoesn't probably mean much other than our Array is slower
05:58:23random8ryeah fair enough.
05:58:31random8rhttp://pastie.org/223135
05:58:39random8rrange rather
05:59:00random8rooh my code is deaded lol
05:59:33random8rhow can I make array faster? coz I use array quite a bit ;-)
05:59:41ryanlowehere's my bug, let me know if you need more details: http://rubinius.lighthouseapp.com/projects/5089-rubinius/tickets/616-rails-rake-test-run-couldn-t- determine-group-part-type-to-instantiate
05:59:43evanwell, thats not array
05:59:45evanthats Range.
05:59:48random8rtrue
06:00:12headius leaves the room.
06:00:16TheVoic1 enters the room.
06:01:26ezmobius enters the room.
06:02:05twbray leaves the room.
06:02:08wmoxam leaves the room.
06:02:14evanah ha!
06:02:18evanthats a bug for chad.
06:02:24evanthats his new Regexp#to_s
06:02:30evanwhich, btw, is a HUGE bitch.
06:02:40ryanloweyou're welcome :)
06:02:41ryanlowelol
06:03:04ryanloweseems to be used in a minor method in Rails ... but it prevents that class from compiling
06:04:04evancould you attach the code around selector_assertions.rb line 99 as well?
06:04:10evanto the ticket
06:04:14ryanlowe599?
06:04:18evanjust so the next person doesn't have to hunt.
06:04:21evaner. yeah
06:04:22evan599
06:04:40ryanlowek :)
06:05:11ryanloweif I put code in a comment, can I format it?
06:05:16evanyeah
06:05:16evanput
06:05:17ryanloweor is a file attachment better?
06:05:18evan@@@
06:05:20evancode blah
06:05:22evan@@@@
06:05:25evaner, 3 @'s
06:05:27ryanlowek
06:08:19ryanloweannnnnd, boom goes the dynamite
06:09:12loincloth enters the room.
06:09:21evanwould you humor me and track down what RJS_STATEMENT[:any] and RJS_PATTERN_HTML are too?
06:09:25evan:D
06:09:44ryanloweonly if you spell it humour like the rest of us do ;)
06:09:51evansure!
06:09:58evancolour it red too!
06:10:16ryanloweheh
06:10:23ryanlowebravo
06:10:51dary leaves the room.
06:12:11massif its spelled colour, do you also spell tumor as 'tumour'?
06:12:14dary enters the room.
06:12:21ryanloweyessir
06:12:26evanand thus the american zing arrives!
06:12:30ryanloweit's a whole family of disfunction
06:13:12massI think americans spell english words correctly. We had the benefit of being as far away as possible from the evil cloud of influence of Webster
06:13:23massor as I like to spell it, Webstre
06:13:51evanor how we spell it now, Websta
06:14:12ryanloweI was kidding, I'm not a fan of prescriptive grammars ;)
06:14:13masswhatcha talkin bout, willis?
06:14:27ryanloweevan: bug 616 updated with the other constants
06:15:01evanwow, it's just a black hole of dregx, isn't it.
06:15:22evanthe idea is to try and let the person that fixes the bug construct a standalone case to be included as a spec
06:15:32evanso they need as much info about the specific case as possible
06:15:33ryanlowegotcha
06:15:54ryanlowethat's a good goal ... if I knew more regex I'd make the spec myself :)
06:16:19evanoh actually.
06:16:29evanyou know what would be super totally awesome and stuff
06:16:35ryanloweyeah I just noticed there are more constants ... blerg
06:16:42evanrun that under MRI
06:17:04evanput put in a 'p RJS_PATTERN_EVERYTHING' somewhere
06:17:12evanso we can see what MRI has the regex as
06:19:45ryanlowegot it
06:19:47ryanloweit's a doozy
06:19:55mitchellvriley enters the room.
06:22:09ryanloweevan: how's that?
06:23:14evanwow.
06:23:25evankittens are crying right now because of it.
06:23:33ryanlowebows
06:23:46ryanlowehas that affect on kittens worldwide
06:23:50massdid you just find out RJS is actually just one large regex?
06:24:10evanpretty much.
06:24:24evanand that kittens hate RJS.
06:24:32TheVoic1baby kittens
06:24:37evanheh
06:24:40evanyes, baby kittens.
06:24:48antares leaves the room.
06:24:52evanTheVoic1: hey, hows things?
06:24:54ezmobiuspugs too
06:24:55massevery time you generate javascript with rails, a kitten is sent to a forced labor camp
06:24:58TheVoic1melting
06:25:03TheVoic1how about you?
06:25:10evanit's a bit warmish, yes.
06:25:14ezmobiusrjs is such a monumentally bad idea
06:25:14evansupposed to be hot again this weekend.
06:25:40massezmobius: there have been far worse javascript-related ideas
06:25:56TheVoic1I'm leaving this place. So much sin here.
06:26:07massTheVoic1: in #rubinius?
06:26:09evanhehe
06:26:11TheVoic1lol
06:26:15TheVoic1los angeles
06:26:27evanTheVoic1: going to alaska where it's cool?
06:26:40TheVoic1I'm heading to indianapolis, to grab whatever corn is left.
06:26:54massstate motto: Have you Hugged a Polar Bear Today?
06:27:24massI don't plan to visit indianapolis until they have the first all-biodiesel indianapolis 500
06:27:43evanI thought it was "Does Mr. Polar Bear need a Hug? I Think So!"
06:28:43dary leaves the room.
06:29:15kdc enters the room.
06:29:48massof course, defiler can vouch that the Florida Motto is "Florida: because 40,000,000 cockroaches can't all be wrong"
06:30:16ryanlowedown to 40M eh, nice improvement
06:30:54massI'm only counting the non-flying variety
06:31:05evanin Jacksonville.
06:31:24TheVoic1evan, jump out of any planes recently?
06:32:06massI want to jump out of a plane soon
06:32:38evanTheVoic1: recently? no.
06:32:40evanjust the once.
06:32:58evanthat was earlier this year
06:33:13evaneveryone should try it once.
06:33:19TheVoic1I thought you were planning to do it again
06:33:20evani'm ok not doing it for a while.
06:33:25evannah.
06:33:47evanmy stomach didn't take too well to it.
06:33:51TheVoic1ah
06:33:54evani've got a weak one.
06:34:10evanmade it down ok, but if we'd been in the air another 20 seconds
06:34:14evani'd probably have lost my lunch.
06:34:37TheVoic1those planes they take you up in are usually old or small or both
06:34:53evanthe plane was ok, I had the adrenaline pumping
06:34:56evanand the fall was fine
06:35:05evanit was gliding at the end.
06:37:22TheVoic1how do you think that would compare to hand gliding?
06:37:25random8r leaves the room.
06:38:53evanhaven't been hang gliding
06:39:06evanthough, with my stomach, if i'm in control, everything is fine
06:44:33ryanloweI'm getting this on "rbx rake test": require 'test/unit/testcase' has been deprecated
06:44:54evanyeah, thats miniunit complaining about the rails tests
06:45:15ryanlowewhich one is broken? :)
06:45:34ezmobiusguess ;)
06:45:56ryanloweRubinius is aiming at Ruby 1.9?
06:47:44ryanlowesorry for the dumb questions, I'm a big noob ... just started playing with Rubinius today
06:47:45fbuilesvryanlowe: 1.8.6 p111 (or 114?)
06:48:15ryanloweahh i says right here: rubinius 0.9.0 (ruby 1.8.6 compatible)
06:48:29ryanloweMRI 1.8.6 doesn't complain
06:48:52ryanloweright after that deprecated line I also get: ARGH! someone defined Test::Unit::TestCase rather than requiring
06:48:52ryanloweARGH! someone defined Test::Unit::Assertions rather than requiring
06:49:13ryanloweI suppose that's miniunit too?
06:49:15fbuilesvryanlowe: I _think_ it might have something to do with using miniunit but you'd have to talk to zenspider about that.
06:49:35ryanloweok
06:49:58ezmobiuswow i didn't know you can use Proc.new.call in place of yield to call the block attached to a method
06:50:26ryanloweI have one functional test with "assert true" in it :) and get those messages .... the test isn't run though (works in MRI 1.8.6)
06:50:44fbuilesvyeah, I saw the bug report but I'm blind in that area :P
06:50:57ryanloweoh there's already a bug
06:51:33evanezmobius: yeah, Proc.new returns the current block as a Proc
06:51:38evani'm pretty sure thats from like ruby 1.3
06:51:43evanbefore the &block syntax
06:51:54ezmobiuscrazytown
06:51:58evanbigtime.
06:52:24TheVoic1ezmobius: did you end up getting that loft in the end?
06:52:40ezmobiusnot a loft no, but a townhouse
06:52:57ezmobiusnow tom's going to get booted from the 4th floor
06:53:25TheVoic1did they start building out the office space on the 3rd?
06:53:36ezmobiusyeah the 3rd flloor is full now
06:53:40ezmobius3rd floor is engineering
06:53:44TheVoic1awsome
06:53:45antares enters the room.
06:53:54evanezmobius: i'll be up on monday, btw.
06:53:59ezmobiusoh sweet!
06:54:04ezmobiusfor how long?
06:54:18evantil tuesday afternoon is all
06:54:26ezmobiusthats cool
06:54:27evanwell, late morning.
06:55:35TheVoic1I think I'll probably drop by for a visit around thanksgiving, see what you guys are up to.
06:55:47ezmobiussounds good
06:57:19ryanloweit looks like there were some 'rails hacks' in lib/test/unit/assertions.rb and then miniunit was updated afterwards, which could have changed things
07:18:17Yurik leaves the room.
07:18:17rubuildius_amd64 leaves the room.
07:20:56rue leaves the room.
07:21:33rue enters the room.
07:25:01jack leaves the room.
07:36:00NoKarma enters the room.
07:37:59qwert666 enters the room.
07:40:32lchin_ enters the room.
07:40:32lchin leaves the room.
07:41:53MenTaLguY leaves the room.
07:43:33drbrainryanlowe: can you expand all the bits in that pattern so we can reconstruct it?
07:43:42drbrainNameError: Missing or uninitialized constant: RJS_PATTERN_HTML
07:43:52ryanlowem'ok
07:44:18drbrainoh, I see you did
07:45:31ryanloweoh I thought you meant print it with MRI like I did with RJS_PATTERN_EVERYTHING
07:46:30wireless_ enters the room.
07:46:38drbrainRuntimeError: Couldn't determine Group part type to instantiate
07:46:42drbrainit is good
07:46:52w1rele55 leaves the room.
07:47:05drbrainhttp://www.youtube.com/watch?v=X8Nc8RCLy1s
07:47:17drbraindamn, I've watched this ten times in the last twelve hours
07:48:20evanhehe
07:48:24evani was humming it earlier
07:48:54drbrainI'm going to watch it another three times right now, at least
07:51:24drbrainsweet! $7 gas!
07:51:35drbrainI'd be able to justify an electric scooter then
07:52:08TheVoic1what ever happened to that google electric car?
07:52:22drbrainI already have a car
07:52:28drbrainwhy would I want another one?
07:52:59evandrbrain: where $7?
07:53:17TheVoic1to show your competency at car ownership.
07:53:33drbrainstupid cbs news article: http://www.cbsnews.com/stories/2008/06/26/eveningnews/main4214022.shtml?source=RSSattr=HOME_421402 2
07:53:37drbrainaww, 2010
07:54:52evanfinally, americans are going to find out the hard way what economists mean when they say "market forces"
07:55:05ryanloweyeah seriously
07:57:06drbrainTheVoic1: I am very competent at car ownership: http://flickr.com/photos/drbrain/151487121/
07:57:40TheVoic1what does that run on?
07:58:30ryanloweleaded gas :)
07:58:33drbrainsouls of children
07:58:51TheVoic1baby children?
07:58:52drbrainnah, but it has an "UNLEADED ONLY" sticker still
07:59:01drbrainryanlowe: /(?-ix:)/.to_s is the minimal repro
07:59:01ryanlowe:)
07:59:16jammiwe have $10 gas
07:59:23drbrainTheVoic1: they have to be under the age of 7
07:59:27ryanlowedrbrain: nice one
07:59:38lchin_ leaves the room.
07:59:40evandrbrain: put that in the ticket
07:59:47evandrbrain: chad wrote that code at japan
08:00:22lchin enters the room.
08:00:33jammi(1.6e/l)
08:01:00TheVoic1should I be paranoid and upgrade MRI for that security thing? Couldn't they wait until after I was done with the rails 2.1 upgrade.
08:01:29evanTheVoic1: if you must, you should apply the patches manually
08:01:33evanotherwise rails is busted
08:01:35evanfrom what I understand
08:02:09drbrainTheVoic1: FreeBSD just backported the patches
08:02:15drbrainhrm, I haven't restarted him
08:02:23TheVoic1yeah if only I was still on FreeBSD, boy I miss it
08:02:30ryanloweafaik, it's MRI that's busted :)
08:02:45drbrainugh, this code is ugly
08:03:15drbrainhttp://pasternak.superalloy.nl/pastes/604
08:03:28drbrainhrm, that paste site sucks too
08:03:31TheVoic1I didn't get the feeling that the security threat was bad enough to warrant caring that much.
08:03:57drbrain@source[i, 4].index ':'
08:04:00