Index

Show enters and exits. Hide enters and exits.

00:00:17rueagardiner: Congrats on the move! You looking to be inside the M?
00:00:30agardineryeah, in paddington
00:00:36buddha_ leaves the room.
00:00:49agardinerbetween paddington stn and edgware rd
00:01:03agardineroh, and thanks! :-)
00:01:19rueDamned expensive, London. Otherwise a lot of fun
00:01:40agardineryeah, you got that right (the expensive bit)
00:01:52agardinerwell, both bits really! :-D
00:03:48zimbatm_ leaves the room.
00:05:30benburkert_ enters the room.
00:06:48rby_ leaves the room.
00:11:19Arjen_ leaves the room.
00:13:54mernen enters the room.
00:15:06trythil leaves the room.
00:16:21headius leaves the room.
00:19:28wycats enters the room.
00:19:55trythil enters the room.
00:20:54wycats_ enters the room.
00:21:12benburkert leaves the room.
00:21:28qwert666_ leaves the room.
00:22:25wycats leaves the room.
00:23:51joachimm leaves the room.
00:26:00boyscout8 commits by Ryan Davis
00:26:01boyscout * Added quickstart to readme; 54a407f
00:26:02boyscout * Added; 3c6e878
00:26:03boyscout * '?\ ' is horrid. use ?\s; 7e0cba7
00:26:04boyscout * trailing whitespace drives me nuts; ac21a76
00:26:05boyscout * simple refactorings; dddbf70
00:26:06boyscout ...
00:27:02brixenzenspider: when vlad prompts me for a password, is there an easy way to get it to tell me what it's trying to access?
00:28:57benburkert_ leaves the room.
00:29:04zenspiderit only prompts for sudo
00:29:36brixenhrm, wth is it not taking the pw then
00:32:14zenspiderno clue
00:32:25evanzenspider: it wont prompt for an ssh password?
00:32:40zenspiderI had some problems when ttys were still connected ... but that was blocking errors
00:33:07zenspiderevan: vlad assumes you're using ssh-agent or something similar
00:33:10drbrainvlad doesn't prompt for ssh key passwords, use keys or ssh-agent
00:33:14zenspiderbrixen: is it full on hanging?
00:35:09brixenzenspider: it prompts for pw but will not take it
00:35:52brixenzenspider: the first time it prompts, it accepts the pw, inits the scm/repo, then prompts again
00:36:28brixen2nd time - 5th time, it does not accept the pw, so I think it's trying to access something else
00:37:26rubuildius_amd64Ryan Davis: 54a407f96; 2097 files, 6777 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/MCvAhw71.html
00:38:11zenspiderbrixen: custom recipes? or straight up vlad?
00:39:06brixenzenspider: http://pastie.org/190226
00:39:37AndrewO enters the room.
00:39:43brixenzenspider: interestingly: http://pastie.org/190228
00:40:04enebo leaves the room.
00:40:18rubuildius_ppcRyan Davis: 54a407f96; 2097 files, 6779 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190229
00:40:37AndrewO leaves the room.
00:42:11drbrainbrixen: what's that password prompt on line 23?
00:42:30evanit's pretty clearly asking for an ssh password
00:42:45trythil leaves the room.
00:42:46brixendrbrain: on line 23 I gave it the same pw as on line 11-16
00:42:59drbrainbrixen: for ssh?
00:43:05brixenyeah
00:43:06drbrainyou need to set up your ssh key
00:43:26brixenso, why does it accept the pw on line 8?
00:43:31brixenI must have a key?
00:44:04drbrainit might be the sudo password prompt detector triggering
00:44:07drbrainmust
00:45:18zenspideryeah. this is almost all ssh problems... if you can't do ssh rubyspec ls w/o password, something is configured wrong
00:45:28zenspiderbtw, your domain value is JUST rubyspec... is that right?
00:45:39brixenrubyspec is in ssh/config
00:45:43KirinDav leaves the room.
00:45:45brixendidn't realize I had to have ssh key setup
00:46:19brixenthanks guys
00:46:44therealadam leaves the room.
00:46:50drbrainssh keys are Good for the Soul
00:47:46KirinDav enters the room.
00:47:52zenspiderbrixen: doco patches gladly accepted. ;)
00:48:05brixenif they're good for my *immortal* soul, I guess I'll comply
00:48:18brixenzenspider: it's just been so long since I've deployed with vlad
00:48:23brixendon't do much rails lately :)
00:48:32drbrainthink of all that time saved typing your password!
00:48:42ezmobius enters the room.
00:49:50ShayArnett enters the room.
00:51:20OccamRazor leaves the room.
00:51:48evanlife is too short to type your password more than once every 6 months
00:52:00brixenheh
00:52:09brixenahh, the sweet swoosh of deploy
00:52:25brixenI have keys almost everywhere, but it's not the first thing I set up
00:52:39brixensometimes I *like* to type my pw
00:52:46brixenmakes me feel important
00:52:54brixen(not really, that's total bs :)
00:53:02evanlike you're arming a nuclear missile
00:53:07brixenheh
00:53:10evan"Your passcode Mr. President."
00:53:24evan<boop> <boop> <boop> Goodbye Haiti!
00:53:40olabini enters the room.
00:53:41evannothing against Haiti, first country that came to mind.
00:53:48brixenI was going to say...
00:57:17drbrainits small enough that a nuke could obliterate it, too
00:57:38evanwouldn't be good for Florida though.
01:00:31eventualbuddha leaves the room.
01:00:50headius enters the room.
01:01:44yugui enters the room.
01:04:14wvdschel leaves the room.
01:06:44ShayArnett leaves the room.
01:11:45benburkert enters the room.
01:17:43benny leaves the room.
01:18:55ShayArnett enters the room.
01:40:28benstiglitz leaves the room.
01:46:17mkescher_ enters the room.
01:52:51hornbeck leaves the room.
01:53:11radarek leaves the room.
01:54:02trythil enters the room.
01:59:11KirinDav leaves the room.
02:02:41fbuilesv enters the room.
02:12:20kw enters the room.
02:18:07rue`bin/mspec ci -j` is very nice. I have to be mindful of the times though
02:20:22boyscout1 commit by Eero Saynatkari
02:20:23boyscout * Improved a spec description for String#index.; b700239
02:22:26mkescher_ leaves the room.
02:25:47lopex leaves the room.
02:32:24rubuildius_amd64Eero Saynatkari: b70023978; 2097 files, 6778 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/m7iarT60.html
02:34:21rubuildius_ppcEero Saynatkari: b70023978; 2097 files, 6780 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190287
02:42:19agile enters the room.
02:42:41yugui leaves the room.
02:45:44VVSiz_ enters the room.
02:46:03dschn leaves the room.
02:52:59yugui enters the room.
02:56:53dlee leaves the room.
03:04:07VVSiz leaves the room.
03:06:50MenTaLguY enters the room.
03:10:00Fullmoon enters the room.
03:11:40kw leaves the room.
03:15:03AndrewO enters the room.
03:17:07ezmobius leaves the room.
03:24:25yugui leaves the room.
03:24:25benburkert leaves the room.
03:24:33yugui enters the room.
03:30:56Fullmoon leaves the room.
03:41:00Fullmoon enters the room.
03:55:56antares enters the room.
03:55:56bitbang_ leaves the room.
04:09:12fbuilesv leaves the room.
04:10:31AndrewO leaves the room.
04:10:32trythil leaves the room.
04:10:45trythil enters the room.
04:16:51anteaya leaves the room.
04:19:08boyscout1 commit by Adam Gardiner
04:19:09boyscout * Add support for conditional breakpoints; 7b54945
04:19:31hornbeck enters the room.
04:27:25rubuildius_amd64Adam Gardiner: 7b54945cb; 2097 files, 6778 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/g5Oxow49.html
04:29:20imajes enters the room.
04:33:46rubuildius_ppcAdam Gardiner: 7b54945cb; 2097 files, 6780 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190316
04:53:26headius leaves the room.
04:53:52be9 enters the room.
04:58:27KirinDave enters the room.
05:02:40mernen leaves the room.
05:10:10lstoll enters the room.
05:14:12ezmobius enters the room.
05:17:45imajes leaves the room.
05:22:32nicksieger leaves the room.
05:22:58hornbeck leaves the room.
05:26:47rueHaha, awesome
05:26:55DefilerWhat is?
05:26:57rue"The End Justifies the Memes"
05:27:26Defilerboom tisk
05:33:54headius enters the room.
05:35:28kw enters the room.
05:40:05blakewatters enters the room.
05:40:24GMFlash leaves the room.
05:44:13knowtheory enters the room.
05:44:29nicksieger leaves the room.
05:52:45ryantmulligan leaves the room.
05:55:41yaroslav enters the room.
05:59:55headius leaves the room.
06:00:29headius enters the room.
06:07:44DefilerThere are so many gems I've never heard of now
06:07:45Defilerwild
06:10:47ezmobius leaves the room.
06:11:27trythil leaves the room.
06:11:44trythil enters the room.
06:16:12crafterm enters the room.
06:17:01hornbeck enters the room.
06:19:50dctanner enters the room.
06:20:41shanghaililly enters the room.
06:23:50shanghaililly leaves the room.
06:24:58yugui leaves the room.
06:30:15lstoll_ enters the room.
06:30:55lstoll leaves the room.
06:37:02agardiner enters the room.
06:37:23lstoll_ leaves the room.
06:37:29dctanner leaves the room.
06:37:59lstoll enters the room.
06:38:01agardiner leaves the room.
06:39:48KirinDave leaves the room.
06:40:52rueMan, they fix the weirdest things in MatzRuby
06:40:57rue"Step may be bignum"
06:41:40dysingerhas anyone started on the net/ specs ?
06:41:45dysingerI don't see any
06:43:00KirinDave enters the room.
06:47:02evandysinger: thanks again for the kind words about my presentation
06:47:10dysingernp :)
06:47:20evani really do hope that it starts to seep in for people
06:47:21dysingerIt was inspiring
06:47:26dysingerand I feel the same way
06:47:34evanthat their life and their projects life is better when they let go
06:47:47dysingerI have been getting older and leading teams for a long time. I am trying to get it down to a science.
06:47:58dysingerand "enable" people
06:48:34dysingerWhen I was younger I used to fall into that Ego trap of "I am my code" "don't talk bad about my code"
06:48:37rueOh cool
06:48:45rueI may have won $25 million
06:48:47dysingerI still struggle with it - everyone does
06:48:51dysingerWTF
06:48:58rueI just need to send them my bank account number..
06:49:03dysingerhaha
06:49:50dysingerSo this spec is going to be gnarly and I imagine that's why nobody has tackled it yet - but we have a need on my project for net/imap.rb specs.
06:50:30rueYeah, net/ will be pretty nasty. I really think we should consider just rewriting the *library* but it is not exactly high priority
06:50:40dysingeranyone started that yet ? I didn't see it - but I am noob to actually trying to contribute. I have been git-following the project since Aug last year or so.
06:51:17rueIf we had any, it would be in spec/ruby/1.8/library/net.. and looks like no
06:53:24tarcierirue: heh, sounds like fun :)
06:53:42ruedysinger: Having a single section to start from will help, though. Have you worked with BDD/RSpec and such otherwise?
06:55:59wycats leaves the room.
06:56:08KirinDave leaves the room.
06:58:27benny enters the room.
07:08:56dysingersorry I was out
07:09:10dysingerYes I am a BDD fiend and have been doing rspec for a year.
07:09:49KirinDave enters the room.
07:09:53dysingerrue so I have never attacked something so complex as networking so I may FLAIL!
07:10:17dysingerbut I was thinking of just mocking everything on the outside of net/imap.rb
07:10:31dysingerconnecting to a server and seeing what complains
07:10:39dysingerstubing and should_receiving as I go
07:11:12dysingerIt would be a big task though I imagine.
07:12:14ruedysinger: Yeah, definitely the best way for as much as you can
07:15:30dysingerMan the rubinius specs take three times as long to run as one of my fat rails projects. Heh.... :)
07:15:35dysingerI thought I had it bad :)
07:16:37dysingerI need a new 8-core mac pro w/ 32GB RAM. I have been salivating over those. hey! Excuse to buy one :)
07:17:21dysingerIf only Ruby had native threads and autotest / rspec could use them.
07:18:29dbussinkwe have the -j option now for mspec
07:18:52dbussinkruns stuff in parallel
07:19:34dysingersweet
07:19:53dysingerlike -j2 or -j4 etc ?
07:20:23wycats enters the room.
07:23:58rueI think just -j.. never tried anything else :)
07:24:13rueMorning, dbussink ;)
07:24:33rueYou all packed up yet?
07:24:41dbussinkpacked up?
07:25:00rueYeah, only a month to go!
07:25:57dysingerlol the whole internets is filled with "twitter dosn't ruby & rails. Ruby sucks. Rails sucks. I told you so" today
07:25:57dbussinki'm not a woman :P
07:26:28dysingertwitter (doesn't like)
07:28:53crafterm leaves the room.
07:29:12rueWhat is Twitter again?
07:30:20dbussinkhehe
07:34:02rueI am experimenting in literate programming.
07:34:08tarcierirue: Slow motion IRC
07:35:11ruechar buffer[SocketType::HOW_MANY_BYTES_I_SHOULD_SEND_ONCE];
07:35:24tarcieriheh
07:35:54tarcierian order of magnitude above the MTU is good :)
07:36:44rueApache uses 8kiB everywhere for whatever reason
07:38:14tarcierisounds good
07:38:15tarcieriheh
07:38:40evanstd::string this_one_time_I_told_this_story;
07:41:30Arjen_ enters the room.
07:42:02rueSomething like that :P
07:44:10chris2 enters the room.
07:47:56dbussinkone time at band camp
07:49:27headius_ enters the room.
07:50:35headius leaves the room.
07:55:59GMFlash enters the room.
07:57:45KirinDave leaves the room.
08:03:34mkrauskopf enters the room.
08:06:13chris2 leaves the room.
08:06:13ixx leaves the room.
08:08:56dysinger leaves the room.
08:18:00chris2 enters the room.
08:18:00ixx enters the room.
08:19:12rueMorning
08:27:50Maledictus enters the room.
08:27:56joachimm enters the room.
08:29:19octopod enters the room.
08:29:51danlucraft enters the room.
08:40:52dbussinkrue: VVSiz only says morning in #jruby ;)
08:40:55yaroslav leaves the room.
08:41:04trythil leaves the room.
08:41:05VVSizheheh :)
08:42:21dctanner enters the room.
08:42:41dctanner leaves the room.
08:43:06brainopia enters the room.
08:46:25dbussinkVVSiz: nickname hinting is always fun :)
08:46:56VVSiz:)
08:48:03VVSizheh, new specs comitted, new failures on JRuby. the story of my life (lately) ;)
08:48:42qwert666 enters the room.
08:51:16lstoll leaves the room.
09:01:09dbussinkwell, keeps you busy so to speak
09:01:12dbussinkwhat's failing?
09:02:03mutle enters the room.
09:14:38VVSizdbussink: some recent File stuff. most probably, easy to correct on JRuby side
09:15:00VVSizheh, nice commit message: "Added" ! :)
09:15:16VVSizand nice path in senders.rb: File.expand_path("~/Work/p4/zss/src/ParseTree/dev/lib")
09:15:45dbussinkhehe
09:16:47VVSizthat's 3c6e8789ab39fa2d1735c4dd5476aae50463cbc8 by zenspider(?)
09:20:30kschiess enters the room.
09:20:45dbussinkzenspider yeah
09:21:12TheVoice leaves the room.
09:23:55zimbatm enters the room.
09:26:15kw leaves the room.
09:42:07thehcdreamer enters the room.
09:50:52benny leaves the room.
09:51:16benny enters the room.
09:59:23_mutle enters the room.
10:00:32headius_ enters the room.
10:09:26chris2you guys went c++?
10:10:01headius leaves the room.
10:10:19rueYep
10:11:18mutle leaves the room.
10:13:00rueApache2 bucket brigades are the stupidest, most ass-backward implementation in the field of http servers
10:15:51rueWhy I invariably start these fabulous conversion projects at 3am I dunno.
10:17:01euleBetter than 4 am (http://www.ted.com/index.php/talks/view/id/148)
10:19:04rueI dunno, I think the stars may have aligned better at 4
10:34:16SpakmanI'm trying to install rubinius on 64bit Fedora. When I rake install I'm told that glibtool cannot be found. What is glibtool? Seems to be an Apple tool.
10:39:57wycats_afk enters the room.
10:40:04wycats leaves the room.
10:40:46dbussinkSpakman: it's part from glib, pretty stardard stuff
10:40:49dbussinknot apple specific
10:41:16dbussinki have rubinius running in a fedora system, so it should work
10:41:24dbussinkmake sure you have all the dependencies though
10:42:38dbussinkon fedora you need libtool btw
10:42:42dbussinknot glibtool
10:43:28eule leaves the room.
10:50:52aasmith leaves the room.
10:51:05aasmith enters the room.
10:59:55rueWell, specifically, it IS GNU Libtool
11:04:38zenspider leaves the room.
11:13:46flori leaves the room.
11:13:52flori enters the room.
11:19:14yugui enters the room.
11:20:29GMFlash leaves the room.
11:40:30anteaya enters the room.
11:43:38headius_ leaves the room.
12:05:06wmoxam enters the room.
12:13:49antares enters the room.
12:16:33saywatmang enters the room.
12:17:02saywatmangwats rubinius' policy on adding new features to the language that wont break 1.8 compat ?
12:17:49MaledictusI don't know, just curious what that would be?
12:18:44saywatmanglike say.. rescue/ensure available in all blocks
12:19:00saywatmangw/o begin/end needed
12:19:17Maledictusah
12:19:36saywatmangor... being able to rescue specific exceptions inline
12:19:49saywatmangblah rescue(BlahError) p :blah_error_caught!
12:20:02saywatmangstuff like that...
12:20:07MaledictusI must say I'm not a big fan of such incompatibilities, but if Matz would implement that it could be cool
12:20:22saywatmangwaht do u mena incompats?
12:20:44saywatmangoh you just mean adding stuff that 1.8 doenst have?
12:20:54Maledictusif rubinius had this feature and someone uses it, one would not be able to run it on mri or jruby
12:20:56Maledictusyes
12:21:00saywatmangright right
12:21:02saywatmangi agree...
12:21:09saywatmangso i must convince matz then.. ok :P
12:21:19Maledictushehe, would be the best
12:21:30saywatmang leaves the room.
12:26:14chris2rubinius commits around the clock! http://rubyurl.com/1K9O
12:26:41Maledictusheh, cool
12:27:06Maledictusan now by country ;9
12:27:10chris2heh
12:27:18chris2that's the script, http://pastie.caboo.se/190425
12:28:20Maledictusnice
12:29:10chris2google charts is pretty nasty :P
12:29:28Maledictusnever worked with it. But looks nice imo
12:29:48chris2yeah, but you need to get the format exactly right
12:31:43MaledictusOh, that's positive I think. I don't like the "be gentle with input and strict with output" theme. I would be always strict
12:32:16chris2but then show a fricking error message and not just a white png :P
12:32:29Maledictusyeah, sure :)
12:32:34Maledictuslol.
12:33:06yugui leaves the room.
12:33:35MaledictusIs there any documentation for --pretty in git?
12:33:42chris2man git-log
12:33:48Maledictusman git-log doesn't really help
12:34:06Maledictusonly says there is medium, full, etc
12:34:32chris2uh
12:34:37chris2either your git is really old
12:34:37Maledictusah, got it
12:34:40chris2http://linux.die.net/man/1/git-log
12:34:41Maledictusyou're right
12:34:50Maledictusit has it's own section
12:34:53chris2or your search skills suck :-)
12:34:57Maledictus:P
12:36:40anteaya leaves the room.
12:40:26yugui enters the room.
12:43:29ctennis leaves the room.
12:43:40brainopia leaves the room.
12:48:22webmat enters the room.
12:56:31naeu enters the room.
12:57:00naeu leaves the room.
12:57:14naeu enters the room.
12:57:35ctennis enters the room.
12:59:10wdperson enters the room.
13:13:31manveruDefiler: any updates for me? ^^;
13:16:26ryantmulligan enters the room.
13:32:01dctanner enters the room.
13:42:38robertpostill enters the room.
13:45:08radarek enters the room.
13:46:46zimbatm leaves the room.
14:05:40boyscout1 commit by Vladimir Sizikov
14:05:41robertpostill leaves the room.
14:05:41boyscout * More tests for BigDecimal#floor.; 3cf6c1e
14:08:20webmatI have an awesome patch to submit
14:08:40AndrewO enters the room.
14:09:30webmatIt allows you to turn off the awesome_backtrace colorizing off with 'No' and 'nO' in addition to the currently supported 'no' and 'NO' ;-)
14:09:36webmathttp://pastie.caboo.se/190461
14:10:30webmatI'm not sure whether Evan thought of code patches THAT small for first time contributors. Hehe
14:10:34danlucraftwell rubinius is ready for the enterprise then
14:10:57webmatyep
14:11:09webmatmuch more stable now, too ;-)
14:11:09danlucraftit could be smaller. You could correct spelling in a comment or something.
14:11:25webmator remove a space at the end of a comment line
14:11:53webmatI can't think of anything more insignificant than that
14:11:55danlucrafthmm. you win.
14:11:58webmatlol
14:15:41joachimm leaves the room.
14:16:40benny leaves the room.
14:17:23rubuildius_amd64Vladimir Sizikov: 3cf6c1e03; 2097 files, 6778 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/3fUK9w31.html
14:19:09rubuildius_ppcVladimir Sizikov: 3cf6c1e03; 2097 files, 6780 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190462
14:19:32probablycorey enters the room.
14:24:11moofbong enters the room.
14:27:09VVSizdamn you, BigDecimal!
14:43:30GMFlash enters the room.
14:43:40zimbatm enters the room.
14:45:59srbaker enters the room.
14:46:03brainopia enters the room.
14:46:32KirinDave enters the room.
14:50:48wmoxam leaves the room.
14:59:34chammiya enters the room.
14:59:38chammiya leaves the room.
15:00:00wmoxam enters the room.
15:06:31joachimm enters the room.
15:12:29agile leaves the room.
15:13:39dambalah leaves the room.
15:21:09_sk enters the room.
15:33:24imajes enters the room.
15:34:35dctanner leaves the room.
15:34:50dctanner enters the room.
15:37:14imajes leaves the room.
15:37:39imajes enters the room.
15:37:48headius enters the room.
15:39:19ryantmulligan leaves the room.
15:40:28ryantmulligan enters the room.
15:41:23OccamRazor enters the room.
15:44:56therealadam enters the room.
15:45:26boyscout1 commit by Vladimir Sizikov
15:45:27boyscout * A couple of test cases for BigDecimal's #quo, #div, #/.; b9806e0
15:48:52headius_ enters the room.
15:51:33ttmrichter leaves the room.
15:55:20trythil enters the room.
15:55:34headius leaves the room.
15:57:00joachimm leaves the room.
15:57:24rubuildius_amd64Vladimir Sizikov: b9806e0ef; 2097 files, 6778 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/Rm9DKs76.html
15:59:04rubuildius_ppcVladimir Sizikov: b9806e0ef; 2097 files, 6780 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190509
16:05:14joachimm enters the room.
16:05:38dc_ leaves the room.
16:10:03benstiglitz enters the room.
16:14:51antares_ enters the room.
16:19:21pauldix enters the room.
16:19:35enebo enters the room.
16:19:47_sk leaves the room.
16:31:12srbaker leaves the room.
16:32:42srbaker enters the room.
16:34:33srbaker leaves the room.
16:37:16joachimm leaves the room.
16:38:35antares_ leaves the room.
16:45:27yaroslav enters the room.
16:51:08srbaker enters the room.
16:52:16srbaker leaves the room.
16:53:14blakewatters leaves the room.
16:58:18agile enters the room.
16:59:18benny enters the room.
17:02:23enebo leaves the room.
17:05:05marnen enters the room.
17:05:32nicksieger enters the room.
17:12:56pauldix leaves the room.
17:26:16naeu leaves the room.
17:28:00fbuilesv enters the room.
17:29:39radarek leaves the room.
17:32:12yipstar enters the room.
17:32:36enebo enters the room.
17:33:55headius_ leaves the room.
17:34:08headius enters the room.
17:44:29kw enters the room.
17:44:36binary42 leaves the room.
17:49:14boyscout1 commit by Vladimir Sizikov
17:49:15boyscout * New BigDecimal#divmod specs.; 854a011
17:49:25TheProkrammerhttp://twitter.com/ev/statuses/801530348
17:49:31thehcdreamer leaves the room.
17:49:32TheProkrammerSince it was brought up yesterday here.
17:49:53evanyep.
17:50:03evanArrington is still sticking to his story
17:50:06evanbecause he's a douche.
17:50:09VVSizwell, it's on Slashdot now, so it official!
17:50:37TheProkrammerThe story is that they abandoned RoR for their /backend/ MONTHS ago... yay. :)
17:50:57evanyeah
17:50:59evanlong ago.
17:51:02TheProkrammerIt's so sad that I barely ever ever read slashdot any more.
17:51:08evanthe backend hasn't been rails maybe ever.
17:52:03VVSizone thing for sure though, Twitter's performance is not that great
17:52:18evangrowing pains.
17:52:22evanplus, it's twitter.
17:52:46evanas tbray pointed out, people seem to think of twitters uptime like it were a missile defense system
17:53:20yaroslav leaves the room.
17:55:42blakewatters enters the room.
17:57:19evanso, i think i've come up with a nice way to eliminate ivar_as_index
17:57:21rubuildius_amd64Vladimir Sizikov: 854a01132; 2097 files, 6778 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/J44leE19.html
17:57:24evanand keep everything nice and simple
17:58:28brixenohh, do tell :)
17:59:26evanit's based on what dgtized and I were discussing the other day
17:59:50evanbasically, don't even expose that, say, name is a slot in CompiledMethod
18:00:09evanjust let the VM have a per class registery that maps ivars to slots.
18:00:22evanlet methods be generated as normal
18:00:41evanie, compile a method for CompiledMethod that uses @name into using, say, get_ivar @name
18:01:12evanbut when the method is attached, quickly go through and mutate all ivar accesses to ivars which are slots into instructions which access the slots directly
18:02:15evana simple version can easily handle our existing usage of slots
18:02:40evanand it can easily be extended in the future to dynamicly assign ivars to slots
18:03:55rubuildius_ppcVladimir Sizikov: 854a01132; 2097 files, 6780 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190584
18:04:17mutle leaves the room.
18:04:22brixenhmm
18:04:29evanbasicly, dipping our toe into the runtime optimization pool
18:04:35brixenright
18:05:19brixenhow do we specify which are slots in this "quickly go through and mutate..." ^^^
18:05:34evanat first, hard code it.
18:05:37evanwhen the VM boots itself
18:05:46evanmake a LookupTable that sits in each class
18:05:49brixenoh, so something like hints is still necessary?
18:05:50evanthat provides a @name => 1
18:05:53brixenok
18:05:57evanit's runtime hints though
18:06:00evanrather than compile time.
18:06:04brixengotcha
18:06:11evanwhich makes it much much more powerful
18:06:16evanyet, still pretty simple
18:06:23brixenyeah
18:06:45evanthis is fairly easy because the get_ivar and push_my_field instructions both take one operand
18:06:56evanso they can be directly swapped in and out in the instruction stream
18:07:01brixennice
18:07:29evanplus, since i've starting to push more things deeper, into the VMMmethod object in the VM
18:07:42evanthe CompiledMethods first class InstructionSequence object remains unchanged
18:08:04evanwe're just manipulating the 'compiled' version thats stored in VMMethod
18:08:17brixenoh, that's slick
18:08:37evanthe other tricky thing i've been considering is this:
18:09:04evanfor increased type safety, bulitins declare their slots of a certain type
18:09:15evanfor example, CompiledMethod has name declared as
18:09:17evanSYMBOL name
18:09:26evanthus it only expects to find a SYMBOL there
18:09:28evanno ducktyping
18:09:37evanbecause you can't ducktype in the VM
18:09:52binary42 enters the room.
18:10:00evanI've got 2 options to remain type safe:
18:10:31evan1) change stuff like name back to OBJECT, and access it only through an accessor function, which performs a cast
18:10:37evanOBJECT name_obj;
18:10:49evanSYMBOL name() { as<Symbol>(name_obj); }
18:11:03evanas<> raises an exception if the value isn't what it exepcts
18:11:11evanwhich would be translated back into ruby
18:11:18evanso thats option 1
18:11:21evanoption 2:
18:11:34evanwhen allowing slots to be set directly from ruby
18:11:39evanie, the set_my_field instruction
18:11:49evanit doesn't set the slot directly in memory
18:12:11evanit dispatches it through a little C++ code, which would perform a type check on name
18:12:19evaner. a type check on the in coming value
18:12:24evanbefore assigning it to the name slot
18:12:33evanagain, raising an exception if things aren't as they seem
18:12:48brixenthat's primarily for VM stability, right?
18:13:00evanyeah
18:13:07brixenhow expensive is that? estimate?
18:13:16KirinDave leaves the room.
18:13:25evanone thing that contributed to shotgun instability is that slots would be assumed to be of a certain form
18:13:32evanbut never checked that they actually were
18:13:38brixenright
18:13:50evanand we'd get crashes because it would try and treat a symbol like an array
18:14:03brixenit seems like option 2 could be faster
18:14:05naeu enters the room.
18:14:08evanoption 2? well.
18:14:22marnenVVSiz: stop with the new specs -- that means I have to actually implement them! :D
18:14:29evanit's considerably more expensive then just setting it
18:14:33VVSizmarnen: hehehe :)
18:14:34evanbut on the other hand, it's only on assignment
18:14:37brixenmarnen: heh
18:14:38evanwhich is more rare than access
18:14:48evanoption 1 slows down all access to the slots
18:14:50brixenevan: just thinking 2 is faster than cast + access
18:14:53brixenyeah
18:14:56evanbecause they're constantly casting
18:14:59binary42_ enters the room.
18:14:59brixenyep
18:15:23binary42 leaves the room.
18:15:36evanoption 2 would be using the object's type to lookup an object with virtual functions, dispatching to a virtual function, which would contain a switch statement
18:15:38VVSizmarnen, brixen: one more BigDecimal sickness for you: BigDecimal('1').divmod(BigDecimal('3E-9'))[1] #==> '0.3E9'
18:15:43zimbatm leaves the room.
18:15:47brixenthe more general approach in 1 (i.e. cast and access) doesn't seem necessary
18:15:53VVSizbut should be 0.3333333333E9 :) so much for precision :)
18:15:57brixenbut maybe I'm missing something
18:16:15evanbrixen: we don't need them both
18:16:23evanbut we have to have one or the other
18:16:29brixenevan: right, I think 2 over 1
18:16:31brixenoption 2
18:16:33evanok.
18:16:54evan2 has the benefit of being able to done largely via generate code
18:17:08evani wrote some code last night to parse out the types of all the slots in the builtin classes
18:17:25evanso i'd just autogenerate the function that contains the switch to perform type checks
18:17:28marnenVVSiz: no, actually, I think neither result is correct, unless you meant [0] instead of [1]
18:17:45VVSizdamn, I meant [0] indeed :)
18:18:29evanhm, i need to actually sign that maid service up
18:18:39brixenevan: so, what does this slot type stuff look like in the class?
18:18:40VVSizthe diff between actual result and calculated one: 33333333 :)
18:19:00evanbrixen: which?
18:19:04marnenyeah
18:19:05marnenoy
18:19:26brixenevan: the types of the slots in the built in classes, is it the same ivar_as_index syntax?
18:19:35brixeni.e. a hash in the class body?
18:19:40evanno
18:19:43evanit's not at all in ruby code
18:19:47brixenahh, doh
18:19:49evanit's hidden entirely in the VM
18:19:50brixengotcha
18:19:51brixenheh
18:19:56evanyou can access it from ruby
18:19:59brixenall these classes floating around now :)
18:20:00evanyou'd do something like
18:20:12evanCompiledMethod.mapped_ivars
18:20:22brixenI'm not used to thinking of classes in the VM still :)
18:20:28evan# => <LookupTable @name => 1, @file => 2 ... >
18:25:44octopod leaves the room.
18:31:45thehcdreamer enters the room.
18:32:58headius_ enters the room.
18:32:58headius leaves the room.
18:35:10Fullmoon leaves the room.
18:36:57Arjen_ leaves the room.
18:37:07binary42 enters the room.
18:38:19binary42_ leaves the room.
18:40:55wdperson enters the room.
18:42:01boyscout1 commit by Vladimir Sizikov
18:42:02boyscout * More BigDecimal#divmod rubyspecs.; b677164
18:42:58TheVoice enters the room.
18:45:17zimbatm enters the room.
18:47:31thehcdreamer leaves the room.
18:49:44binary42 leaves the room.
18:50:19headius_ leaves the room.
18:50:45binary42_ enters the room.
18:51:48TheProkrammerheadius needs to use screen...
18:52:05evanor a proxy
18:52:23rubuildius_amd64Vladimir Sizikov: b6771644d; 2097 files, 6778 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/T8SM9n12.html
18:53:24djwhittfyi ctrlproxy works great for me
18:53:42marnen leaves the room.
18:53:53evandjwhitt: me too
18:54:03evanhasn't crashed yet like dircproxy
18:54:13evanplus it's nice just telling irssi to use it as a proxy
18:54:28djwhittyeah, seems to gobble up a lot of memory, but that might just be logging settings or something
18:54:34djwhittlikewise, hasn't crashed at all for me
18:55:35binary42 enters the room.
18:55:57evanhm
18:56:06evani'm thinking about simplifying primitives a bit too.
18:56:10rubuildius_ppcVladimir Sizikov: b6771644d; 2097 files, 6780 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190608
18:56:19evanhow headius and I were talking about
18:56:30binary42 leaves the room.
18:56:38evanjust make it so that
18:56:43binary42 enters the room.
18:56:49evan /* Ruby.primitive :fixnum_add */
18:57:09evanINTEGER Fixnum::add(STATE, INTEGER other) { ...
18:58:31binary42_ leaves the room.
18:58:54evanauto generates glue code to call automatically call that C++ function
19:00:15djwhittwell, since no one else is speaking up, I'm not a huge fan of that kind of thing being in comments, but I'm not sure there's a better way in C++
19:00:30joachimm enters the room.
19:00:31evanyeah, i'm all ears for a better way
19:00:31djwhittJRuby uses annotations right?
19:00:33evanyeah
19:00:48evanI like that locatily though
19:00:58evanrather than stashing it off in another file like we do now
19:01:34evanthere aren't any form of C++ anonations, right?
19:01:41djwhittno that I know of
19:01:48djwhittC++ has pretty much 0 introspection
19:02:07evanyeah
19:02:46macournoyer enters the room.
19:03:09evanspeaking of which
19:03:28evananyone know wtf is with the style of putting /* blah */ comments IN the argument list?
19:03:33evanIronRuby seems to do that in C# code
19:03:48benny leaves the room.
19:05:56marnen_ enters the room.
19:06:17joachimmis there a guide for building the c++ vm and the tests and coverage reports for it?
19:06:43evancd vm/test;
19:06:51evanrake test;
19:06:54evanmake coverage;
19:06:57evan:)
19:07:12evanmake coverage creates a coverage/ directory, just open the index.html to see the report
19:07:36djwhittis there an equivelant of heckle for C++ I wonder
19:07:44djwhittit would be sweet to run something like that on the VM
19:07:52evanhm...
19:08:14joachimmevan: is that supposed to pass without errors?
19:08:20evanyes
19:08:28evanwell, on os x
19:08:34evanthere are some errors on linux currently
19:08:41marnen_ leaves the room.
19:08:42joachimmevan: I have problems on ppc osx leopard.
19:08:47djwhitthttp://www.mutationtest.net/twiki/bin/view/Resources
19:09:10djwhittthere are some C++ mutation tools listed, but they're commercial
19:09:42evanwell, thats not a deal breaker
19:09:57evansince I have a company credit card
19:10:05djwhitthehe, nice
19:10:09djwhittdeal breaker for me though ;)
19:10:15djwhittunless they're really cheap
19:13:06evani wonder how much insure++ is
19:13:10evani can't seem to find a price
19:14:08evanyikes.
19:14:13evanfound the price of an older one
19:14:14evan$2k
19:16:02evani think this falls under the category of if you have to ask what the price is, you con't afford it.
19:18:17joachimmevan: is the c++ vm build with debug info by default?
19:18:25evanyep
19:18:29evanwhat error you getting?
19:25:08joachimmevan: http://pastie.textmate.org/private/l2kweo1m4brfloywmybltw not exactly descriptive.
19:25:35evanoh
19:25:38evanthat means it segfaulted
19:25:42evanfire up
19:25:44evangdb ./runner
19:27:00binary42 leaves the room.
19:27:20joachimmevan: http://pastie.textmate.org/private/r70yymj3lqsew3gxketvw
19:27:46evanhit c
19:27:50evanthats expected
19:27:53evanthats during the events tests
19:27:59evannot a segfault
19:29:48joachimmevan: http://pastie.textmate.org/private/pxas8i0mgtzcdwbfoloq
19:30:04evanbt
19:30:13evanthats a real segfault
19:30:20evanhit bt
19:31:04thehcdreamer enters the room.
19:31:09joachimmhttp://pastie.textmate.org/private/jekumy5iave58bllk2jcq
19:31:21benburkert enters the room.
19:33:14evanhm, interesting.
19:33:29evanoh oh
19:33:32evancif is NULL there
19:33:44evansomething isn't getting checked likely
19:33:59evanyou're on ppc 10.4 or 10.5?
19:34:02joachimmevan: dinner with girlfriend, will be back in 40 min.
19:34:04jarib enters the room.
19:34:05joachimmevan: leopard
19:34:09evanjoachimm: no prob chief
19:34:24jarib leaves the room.
19:38:27boyscout1 commit by Vladimir Sizikov
19:38:28boyscout * Added news specs for BigDecimal#mult.; d401159
19:40:04pauldix enters the room.
19:40:15twbray enters the room.
19:43:33evananyone around?
19:43:40evanwant to bounce an idea off someone
19:45:18drbrainok
19:45:40kw leaves the room.
19:45:45evani'm thinking about primitives
19:46:04evanand starting to realize that having syntax for them in bootstrap is maybe not the greatest idea
19:46:13evanbecause it's not as though you can mix and match them.
19:46:22drbrainwith what?
19:46:30thehcdreamer leaves the room.
19:46:36evanyou can translant a primitive into another class
19:46:41evanit can really only go one place.
19:46:45evaner.
19:46:46evanyou can't
19:47:18evanthe name of the method it's called as is the only real mutable part
19:47:19drbrainok
19:47:49evanso, i'm thinking, to simplify things, just moving registration of the primitives into the VM
19:47:52rubuildius_amd64Vladimir Sizikov: d4011595a; 2097 files, 6778 examples, 24248 expectations, 0 failures, 0 errors; http://rafb.net/p/NJGrmO38.html
19:48:00evanso when the VM boots up, the primitives are already attached at the proper places.
19:48:05evanno bootstrap that just attaches them
19:48:34drbrainhow would the VM know where to attach them?
19:48:55evanit would calculate them via an annotation in the C++ source
19:48:59riffraff enters the room.
19:49:04evansince things are organized by clasess in C++ that match the Ruby class names
19:49:11evanclass Fixnum {
19:49:16drbrainthen I think that's a terrible idea
19:49:17evan // Ruby.primitive :add
19:49:22evanok, why?
19:49:40drbrainright now the primitives ask the VM for a feature it provides
19:49:50drbrainso the VM can be generic in its function
19:50:03evannot sure how ya maen
19:50:05drbrainnow the VM has to know about the upper layer
19:50:08evancould you give me an example
19:50:33drbrainnow bootstrap asks the VM to do work for it via primitives
19:50:39evancertainly, the primitive is a generic unit of work
19:50:45evanbut they're still class oriented
19:50:50evanbecause they all use the receiver
19:51:28drbrainI think the VM knowing about classes is a convenience for us programmers
19:51:33drbrainbut is really unimportant
19:51:34evanso, you like having an ruby exposed bootstrap?
19:51:36riffraff leaves the room.
19:51:48evancertainly thats true
19:51:56evanbut the primitives still care
19:52:04drbrainbut the VM needing to know about how Ruby is implemented feels like too much coupling
19:52:13evan99% of the primitives have a guard on the receiver
19:52:17evanto make sure it's of a certain type
19:52:28drbrainin the Future, primitives will be implemented in ruby and translated, right?
19:52:34evansure
19:52:52evanbut this change doesn't make than any more difficult
19:53:00evanbecause the translated ones still would have to be linked into the VM
19:53:02drbrainI think its ok for the primitives to have a guard on the receiver
19:53:24evanwhat about attaching the primitives in Ruby do you like?
19:53:33evanthat you can go read bootstrap and see what are primitive methods?
19:53:33rubuildius_ppcVladimir Sizikov: d4011595a; 2097 files, 6780 examples, 24274 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/190641
19:53:51drbrainsince primitives are supposed to be written in ruby, I don't think the VM (which isn't ruby) is the place to hook them up
19:54:16evanwhat they're written in actually doesn't matter
19:54:21drbrainzenspider: wrote a tool yesterday that uses the current system to figure out what methods call down to primitives
19:54:27evanwhat matters (in my mind) is that they're VM execution units
19:55:04evanright, so in that case, it's the ability to look into bootstrap and see what ruby methods are hooked up as primitives
19:55:08evanit increases the visibility
19:55:14drbrainyes
19:55:25evanok
19:55:29evanyou're right.
19:55:30evanit stays.
19:55:54drbrainit also feels like its the wrong side of the VM/ruby coupling boundary to put that information
19:56:02evantrue
19:56:13evani buy that argument
19:56:28evani'm still going to make the primitives more object oriented on the VM side
19:56:40evanbut they'll remain hooked up via the Ruby.primitive construct in bootstrap
19:56:49drbrainthat's cool
19:57:09drbrainI think it is tempting because you've OO-ified the VM so much
19:57:15evanyeah
19:57:26evanthats why i needed to bounce the idea off someone
19:57:30evanto check my thinking.
19:57:30ShayArnett enters the room.
19:58:07drbrainhappy to help
19:59:47dysinger enters the room.
20:01:16fbuilesv leaves the room.
20:12:41riffraff enters the room.
20:20:15lopex enters the room.
20:21:22chris2 leaves the room.
20:22:17kw enters the room.
20:24:32evansweet
20:24:38evanthis glue code generation is going to work well I think.
20:26:16benny enters the room.
20:26:22benburkert_ enters the room.
20:31:51riffraff leaves the room.
20:32:35ndemonner leaves the room.
20:33:56joachimm leaves the room.
20:34:36benburkert leaves the room.
20:37:38rueMuu
20:37:40enebo_ enters the room.
20:44:07rueevan: See if Dinkumware has anything
20:44:22nicksieger leaves the room.
20:44:31enebo leaves the room.
20:45:17naeu leaves the room.
20:49:00rueGut feeling, annotations are a slippery slope. Java has to use annotations because the language sucks, to horribly generalise
20:49:30rueThere are certainly some practical applications, but you have to be careful
20:49:57rueBecause then you have THREE languages you are working with. Ruby, C++ and the annotations.
20:49:59anteaya enters the room.
20:52:42ndemonner enters the room.
20:52:53naeu enters the room.
20:53:11enebo enters the room.
20:59:10ShayArnett leaves the room.
20:59:10macournoyer leaves the room.
20:59:10agile leaves the room.
20:59:10Maledictus leaves the room.
20:59:10rubuildius_amd64 leaves the room.
20:59:10rue leaves the room.
20:59:49macournoyer enters the room.
20:59:49agile enters the room.
20:59:49Maledictus enters the room.
20:59:49rue enters the room.
20:59:49rubuildius_amd64 enters the room.
21:03:52naeu leaves the room.
21:05:09qwert666_ enters the room.
21:06:31enebo_ leaves the room.
21:07:21EugZol enters the room.
21:16:07be9 leaves the room.
21:16:20joachimm enters the room.
21:17:21ctennis leaves the room.
21:19:23riffraff enters the room.
21:21:19evanrue: true
21:21:29evani'm using a single formatted comment for now.
21:22:06qwert666 leaves the room.
21:25:23headius enters the room.
21:27:07agile leaves the room.
21:30:07yaroslav enters the room.
21:30:33yaroslav leaves the room.
21:30:46yaroslav enters the room.
21:31:20whitethunder922 enters the room.
21:32:00whitethunder922i'm having trouble installing rubinius:
21:32:02whitethunder922(in /home/mwhite/rubinius-daily)
21:32:02whitethunder922fatal: Not a git repository
21:32:02whitethunder922make vm
21:32:02whitethunder922cd shotgun; make rubinius
21:32:02whitethunder922make[1]: Entering directory `/home/mwhite/rubinius-daily/shotgun'
21:32:03whitethunder922cd lib; make library;
21:32:05whitethunder922make[2]: Entering directory `/home/mwhite/rubinius-daily/shotgun/lib'
21:32:07whitethunder922bison -o grammar.c grammar.y
21:32:09whitethunder922make[2]: bison: Command not found
21:32:11whitethunder922CC grammar.o
21:32:13whitethunder922cc: grammar.c: No such file or directory
21:32:15whitethunder922cc: no input files
21:32:17whitethunder922make[2]: *** [grammar.o] Error 1
21:32:19whitethunder922make[2]: Leaving directory `/home/mwhite/rubinius-daily/shotgun/lib'
21:32:21evanyou need to have bison istalled
21:32:21whitethunder922make[1]: *** [lib/librubinius-0.8.0.so] Error 2
21:32:23whitethunder922make[1]: Leaving directory `/home/mwhite/rubinius-daily/shotgun'
21:32:24evanstop.
21:32:25whitethunder922make: *** [vm] Error 2
21:32:25rueBison
21:32:27evanplease
21:32:27whitethunder922rake aborted!
21:32:31rue^D
21:32:34rue:D
21:32:45evanwhitethunder922: do not paste into channel
21:33:05whitethunder922sorry, never been here before
21:33:18evanwhitethunder922: you need bison installed.
21:33:29Defilermanveru: No fix yet for that wacky included hook. right now I am only touching bugs that block merb or rails
21:33:36zimbatm leaves the room.
21:34:10ctennis enters the room.
21:36:32whitethunder922ok, i installed bison but now there are all sorts of errors with readline.c after doing "rake build" again
21:36:41atmos leaves the room.
21:37:13evanwhitethunder922: you need the readline headers installed
21:39:53ruewhitethunder922: http://rubinius.lighthouseapp.com/projects/5089/installation also
21:39:55binary42 enters the room.
21:40:09rueIt may need to be updated so let me know if anything is missing from there
21:40:40dbussinkwhitethunder922: the bison missing is in the faq there
21:40:55rueI think the INSTALL doc may be closeish
21:43:25whitethunder922i think it's good now, thanks
21:43:28manveruDefiler: :(
21:43:47dmpk2k leaves the room.
21:44:35rueWtf, there were tons of posts that just appeared in spams
21:45:26Defilermanveru: I'll still fix it if nobody else does first.. RailsConf is coming up soon, though
21:46:04manveruoh well... i just want to be able to run ramaze :P
21:46:42Defilerwait, real code does that stupid thing?
21:46:43DefilerWhy?
21:46:50Defilerand where?
21:47:24manveruincluded does extend the object it was included in
21:47:30DefilerWhy?
21:47:33manverui thought that was a normal pattern...
21:47:41evanno
21:47:48evanthe pattern is to include a seperate module in extended
21:47:52manveruso you can get normal and singleton methods from a module
21:48:05evanthe one that rails has put forth, rather.
21:48:21dlee enters the room.
21:48:42dalee enters the room.
21:48:43manverui never expected it to become a problem since it runs just normal in 1.8/1.9 ...
21:48:46DefilerCrap, it is common
21:48:52Defilerhttp://www.google.com/search?hl=en&safe=off&client=firefox-a&rls=org.mozilla%3Aen-US%3 Aofficial&hs=DuM&q=%22extend+self%22+%22def+self.included%22&btnG=Search
21:48:57dalee leaves the room.
21:49:09Defilershit and rails does it doesn't it
21:49:20manveruno idea...
21:49:22Defilertime to read the MRI code for it I guess
21:49:30evancalling extend in included
21:49:30evanyeah
21:49:34evanthats the rails pattern
21:49:43Defilercalling extend self apppears to be a common rails pattern
21:49:48evani've never seen someone extend the thing that was included though too
21:50:04manveruevan: did i say that?
21:50:07yaroslav leaves the room.
21:50:16manveruif so, sorry, i'm tired :)
21:50:17Defilerhttp://blade.nagaokaut.ac.jp/cgi-bin/vframe.rb/ruby/ruby-talk/188468?188228-198790
21:50:20Defilere.g. ^^
21:50:31dmpk2k enters the room.
21:50:37manverumy specs shows it really nicely
21:50:55DefilerYeah, I know.. but I thought it was a spec for a synthetic bug report
21:51:01Defilernot real code in the wild
21:51:16manveruall my specs are from trying to run ramaze on rubinius
21:51:41manverui just have to synthesize the parts that offend out of it
21:53:09DefilerOK. I apologize for not understanding earlier
21:53:42DefilerI had you filed in the "crazy neighbor with cats" part of my mental bug tracker
21:53:59manveru^^;
21:54:25riffraff leaves the room.
21:54:35rueDamnit, now I will be left there by myself again
21:55:10evanthat being said
21:55:18evanyou should be able to extend something that was included.
21:55:19evanand not crash.
21:55:21evan:)
21:55:24DefilerFor sure, yeah
21:55:30DefilerI was just thinking 'June'. Heh
21:55:40evanit's easy
21:55:42evanit's a quick check.
21:55:45evanit's like 2 minutes.
21:55:46DefilerSadly no
21:55:50evanwhy not?
21:55:51dbussinkafaik it's indeed pretty commonly used in rails
21:55:51Defilerryan and I worked on it for hours
21:56:02evanwhy is it complicated?
21:56:04moofbong leaves the room.
21:56:07manverumetaclasses are nasty :|
21:56:12DefilerWell, the code that should work is easy
21:56:21dbussinkDefiler: i feel a "the master will fix it in 5 minutes" moment coming up ;)
21:56:27Defilerbut it seems like the ancestors list only contains the included module AFTER included fires
21:56:36evanthats a bug.
21:56:41evanit shouldn't.
21:56:41Defilernot before.. but even if you check it early, we couldn't figure out why