Index

Show enters and exits. Hide enters and exits.

17:31:18brixenwell, I got the redmin unit tests to complete
17:31:19brixen692 tests, 2600 assertions, 12 failures, 2 errors
17:31:35brixenthere appears to be a thread related hang in one of the specs
17:32:04brixenfunctionals and integration complete with 6 and 1 failures respectively
17:37:28dbussinkbrixen: looks pretty nice :)
17:38:24dbussinkbrixen: any weirdness?
17:39:22dbussinkbrixen: you could run rubyspec.org on rbx if the tests pass ;)
17:44:59brixen:)
17:45:09brixenno significant craziness
17:45:26brixen14 f/e out of 2600 assertions is pretty good
17:45:51brixenintegration: 272 tests, 961 assertions, 1 failures, 0 errors
17:46:09brixenfunctionals: 529 tests, 1636 assertions, 6 failures, 1 errors
17:55:40dbussinkbrixen: going through them now then?
17:57:10brixendbussink: will be shortly, on a call atm
18:17:26dbussinkbrixen: does this make sense for you: https://gist.github.com/e0b972b021048b13e4df
18:17:35dbussinkbrixen: if the registered variable is an immediate
18:28:26brixendbussink: yeah, makes sense
18:28:48dbussinkbrixen: didn't want to take on that todo right away :P
18:29:54brixenheh
18:31:44brixentest unit runner is so gross :(
18:31:59brixenI want my nice mspec tools for disecting problems
20:26:32dbussinkbrixen: ok, quick question
20:27:31dbussinkspec/capi/spec_helper.rb:11
20:28:01dbussinkConfig::CONFIG['rubyhdrdir'] doesn't exist for mri, so the capi specs break here for (only if running them for the second time after each other though)
20:28:06dbussinkbecause it can't find ruby.h
20:34:44brixenhmm, well wth...
20:38:45brixendbussink: how are you running the specs?
20:39:23dbussinkbrixen: ./bin/mspec -tr spec/capi/array_spec.rb
20:39:38dbussinkit works the first run, fails the second one directly after
20:39:41brixenand why do you think it's failing because it can't find ruby.h?
20:39:56brixenit couldn't run the first time if it wasn't finding it
20:41:04dbussinkbrixen: i have no idea, it's really weird
20:41:11brixenit's not getting to the File.mtime code the first time
20:41:48brixenanyway, let's see if I can find the correct config item
20:41:50dbussinkbrixen: this is my output: https://gist.github.com/0795a88916eb534f3c94
20:45:14brixenyeah
20:45:47brixenso, would ruby.h be in Config::CONFIG['includedir'] ?
20:45:51brixenoh no, it wouldn't
20:45:59brixenit'd be in Config::CONFIG['archdir']
20:46:08brixen<rolleyes>
20:46:35brixenbut checking that in a non-Apple install
20:47:04brixenseems to be
20:50:11brixendbussink: sec, pushing you a fix
20:50:35dbussinkbrixen: archdir doesn't seem to exist for rbx yet though, right?
20:50:39brixenredo'ing compilation of capi specs is on my list
20:50:51brixenno, it doesn't really
20:51:04brixenit's a stupid dir for an include file anyway
20:51:22brixenI don't know where rubyhdrdir came from, but it sounds like a good idea :)
20:52:39boyscoutPurged tags for non-existent specs. - 63d66e7 - Brian Ford
20:52:40boyscoutSearch for ruby.h in correct dir. - d52dbe7 - Brian Ford
20:54:12brixenoh man, that file is such a mess
20:54:27dbussinkthe capi spec_helper?
20:55:28brixenit's going to get rewritten
20:55:31brixenyes, that file
20:55:42brixenit has, um, "evolved" over time :)
20:56:40brixenonce I get to write the new build system, it will be trivial to build C extensions
20:56:56brixenfor some values of trivial
20:57:34dbussinki always have visions for how some ideal things could work
20:57:43dbussinkthen you build it and it turns into a pile of crap
20:58:11brixenheh
20:58:52brixenyeah, handling edge cases can turn any elegant design into a nightmare
20:59:27brixenthere are 0b10 kinds of programs in the world, those that deal with edge cases and those that don't
21:00:06boyscoutBetter fix for finding ruby.h. - 8eda754 - Brian Ford
21:01:04boyscoutCI: rubinius: d52dbe7 successful: 3447 files, 13468 examples, 41016 expectations, 0 failures, 0 errors
21:03:55boyscoutBe sure to handle immediates when registering globals - 6286137 - Dirkjan Bussink
21:03:55boyscoutRemove tags for passing Hash specs - bbaf54c - Dirkjan Bussink
21:03:55boyscoutAdd some basic specs for rb_register_address - ec96326 - Dirkjan Bussink
21:12:13boyscoutCI: rubinius: ec96326 successful: 3448 files, 13480 examples, 41038 expectations, 0 failures, 0 errors
21:24:39dbussinkbrixen: http://www.jruby.org/2010/04/26/jruby-1-4-1-xss-vulnerability.html
21:24:46dbussinkbrixen: looks like rbx has the same issue
21:43:59brixenthat description is hard to understand
21:50:06brixendbussink: I'm not sure what the expected behavior should be
21:50:14brixenbut it doesn't look like rbx behaves like jruby
21:50:15brixenhttp://gist.github.com/379980
21:50:42brixenotoh, would have been nice to hear about this
21:50:54brixenmaybe evan got an email, I sure didn't