Index

Show enters and exits. Hide enters and exits.

15:13:59Enlenl
15:14:14sometestenl
16:06:56kronos_vanoenl
16:07:01kronos_vano:(
16:07:06Enl:))
16:09:03brixenkronos_vano, Enl: que pasa amigos? :)
16:09:29Enleverything is good ;)
16:09:34brixenheh
16:11:34evankronos_vano: hello
16:11:43kronos_vanohi
16:12:24kronos_vanobrixen, oO
16:12:57kronos_vanoevan, Where is rubinius 1.0? )
16:13:22evanI got sick and went on vaca
16:13:23evansoon!
19:12:42boyscoutAdd spec showing IO#reopen creates files - 7773d8a - Evan Phoenix
19:12:43boyscoutAllow IO#reopen to create files - 6fceb30 - Evan Phoenix
19:17:13boyscoutUnwrap args for Method#to_proc. Fixes #277. - 5c8c23c - Evan Phoenix
19:20:54boyscoutCI: rubinius: 6fceb30 successful: 3449 files, 13485 examples, 41050 expectations, 0 failures, 0 errors
19:33:21boyscoutAdd spec for instance_eval scoping with string - ca93088 - Evan Phoenix
19:33:21boyscoutUse the receiver metaclass as the scoped class. Fixes #276. - 28fef11 - Evan Phoenix
19:41:41boyscoutCI: rubinius: 28fef11 successful: 3449 files, 13487 examples, 41052 expectations, 0 failures, 0 errors
19:53:23boyscoutAdd ability to disable execinfo even if the system has it - ff32708 - Evan Phoenix
19:54:55scooprEvan is on fire =)
19:55:07brixenhurry, get the hoses!
19:55:15brixenwe don't want evan burning up
19:55:35evan:D
19:55:42evanall issues are closed.
19:55:42scooprNah, maybe it's just the sunburns
19:55:51brixenheh
19:56:04evangoing to grab some lunch now!
19:56:25brixengood plan
19:56:33cremesquick, someone add some issues so he isn't bored upon return!
19:56:34brixenis eating a sandwich
19:56:40evancremes: haha
19:56:47evana number of them a targeted for post 1.0
19:56:58evanbecause they're just too big of changes to do right now
19:57:07cremesmakes sense
19:58:11cremeslet's see, ironruby 1.0 came out a few weeks back, macruby 0.6 a few days ago, jruby 1.5 is imminent and now rbx 1.0
19:58:17cremesgood time to be a rubyist
19:58:38evanyep!
19:58:49evanwe've already got a great list of stuff post 1.0
19:59:06evanie, there are some ugly performance problems in a few areas that we just can't fix atm.
19:59:35cremesheh, i've probably run into a few...
19:59:41cremesi'll start posting perf related issues again
20:01:27evancremes: well, wait until post 1.0
20:01:41evanonly critical stuff should go into issues right now.
20:01:44boyscoutCI: rubinius: ff32708 successful: 3449 files, 13487 examples, 41052 expectations, 0 failures, 0 errors
20:03:22cremesthere's no way to tag them as post 1.0? if not, i'll hold off
20:03:34evanno, there's not atm.
20:03:38cremesk
20:03:41evanactually though
20:03:48evanbring them up on the mailing list
20:03:55evanthats a better place to discuss performance stuff anyway
20:03:59cremesk
20:05:06brixencremes: the #pack/#unpack issue I have a strategy for
20:05:29brixenlet us know about other perf issues you hit
20:05:33cremesbrixen: great, that was going to be my first one :)
20:06:06brixenI'm writing a Ragel based version that all Ruby impl can use if they want
20:06:12evanyaeh, pack is something that brixen has a good strategy for
20:06:21evanbut it's definitely post 1.0 work
20:06:21brixensince Ragel emits Java, C/C++, C#, Ruby
20:06:24evanprobably directly post
20:13:07cremesbrixen: have you looked at the ruby emitted by ragel? it is Ugly (capital U)
20:14:23scooprto be fair, the C ain't that much better
20:14:42Defilerhave you looked at the C emitted by bison/yacc?
20:16:54cremesDefiler: how do you think most people win the obfuscated C contest? bison/yacc is their secret weapon! :)
20:17:16DefilerIt takes a twisted human mind to get that last 10% though
20:18:56mjijacksonhi all. is rubinius 1.0.0rc4 known to have issues using rubygems?
20:19:26mjijacksoni'm trying out rbx right now using rvm, but can't get me rubygems working properly
20:20:01mjijackson$LOAD_PATH is empty (at least in irb)
20:20:16mjijacksonwell, not totally empty. there's "lib" and "."
20:20:59brixenmjijackson: rc4 is pretty old, we're about to release rc5
20:21:05brixenmjijackson: so, you should be using master
20:21:20brixenrvm install rbx-head or build (install) from git
20:21:43Defilerthat being said, rubygems should work in rc4 :)
20:21:45mjijacksonbrixen: holy crap ... master? what if my computer blows up? ;)
20:21:51brixencremes: we wouldn't be using the Ruby output
20:22:04brixenmjijackson: we keep master sane, it should always work for you
20:22:20mjijacksonbrixen: cool. i'll give it a shot
20:22:23brixenif it blows up your computer, borrow a nearby co-worker's and let us know :)
20:23:30brixencremes: the state machine formalism in Ragel gives us a much better medium to express the algorithms than any of the programming languages
20:23:42brixencremes: we'd just use a C++ primitive in this case
20:24:06brixenrememeber, we use Ruby because if its utility, not because of some mythical properties
20:24:26mjijacksonbrixen: just to be sure this is necessary, i'm loading rubygems via export RUBYOPT="rubygems" in my .bashrc
20:24:30mjijacksonthat should work, right?
20:24:39brixenmjijackson: there was an issue fixed for that
20:24:48brixenso try on master and let me know if it works
20:42:12evanno
20:42:17evanRUBYOPT didn't work in rc4
20:42:23evanit's in rc5
20:42:43mjijacksonevan: awesome! thx
20:43:00evanwell, master/rc5
21:40:08evancould someone update and see if the top commit they have is 65fd26b
21:40:18evangit says it's in the romet
21:40:19evanremote
21:40:23brixenyep, sec..
21:40:24evanbut github isn't saying so an the UI
21:40:40Defileryep
21:40:48Defilermay 5 14:35:59
21:41:10evank
21:41:12brixenyep, me too
21:41:15evanjust github being wacky then.
21:41:29evanArray::Frozen is gone now
21:41:33evanthere actually weren't a lot of methods
21:41:37brixenyeah
21:52:20brixensomeday, we will have nice things
21:52:38brixenfor example, it is ridiculous to have to start a subprocess to spec command line option parsing
21:52:51brixenwhen you have such a lovely OO language
21:55:56brixenevan: I just push the command line spec stuff
21:56:19brixenevan: do you want to do the rc release or shall I continue pushing the spec fixes?
21:58:04evango ahead and finish
21:58:08brixenk
21:58:16evanbetter to do a rc release at the beginning of a day
21:58:37brixenwell, not sure how long the c-api stuff will take, but I'll keep working on it
22:01:52boyscoutConvert Array to use Ruby.check_frozen - 65fd26b - Evan Phoenix
22:02:53evanwell, i don't want to go in and do a bunch of work right after rc5
22:03:00boyscoutRemoved full.mspec config file. Use bin/mspec ci to run all CI specs. - 3952b50 - Brian Ford
22:03:00boyscoutDeleted obsolete Task related specs. - b0527c1 - Brian Ford
22:03:00boyscoutFixes for command line specs. - 503beaf - Brian Ford
22:03:00boyscoutFixed parsing options with optional arguments. - 4f914f2 - Brian Ford
22:03:00boyscoutFixed command line parser per current specs. - 4018f24 - Brian Ford
22:03:00boyscoutEnabled command line specs. - 5e47491 - Brian Ford
22:03:04evanso either the work is done now, or post 1.0
22:03:13evanhow much work is it?
22:04:07brixenwell, there's some capi specs excluded, I'm looking into them now
22:04:32evanok, check them out and advise me on them before you start in on them
22:04:39evanso we can decide if we can postpone or not
22:04:42brixenk
22:12:21brixenevan: thoughts on rb_define_hooked_variable ?
22:12:31evanpost 1.0
22:12:33brixenlooking at codesearch, doesn't seem like we need to impl it
22:12:37evanno one uses it.
22:12:44evanand it's going to be a pain to implement
22:13:22brixenI guess I need a way to omit stuff from C extension
22:13:36brixenC-API specs that is
22:27:14boyscoutCI: Commit 5e47491 failed. http://github.com/evanphx/rubinius/commit/5e474917b21a9211b152fa016e370a08530fee09
22:30:58brixenthat is the randomm frozen error I get
22:31:01brixenrandomly
22:35:58evan:/
22:35:59evanyeah
22:48:02Defilerhuh, boyscout doesn't link to the build output anymore?
22:49:22evancijoe doesn't
22:49:26evanI should change that.
23:35:05boyscoutAdded rb_struct_define. - 92c6a15 - Brian Ford