ruote log_2010-09-29

2010-09-29 06:05:38 utc jmettraux tosch_le: hello
2010-09-29 06:05:54 utc tosch_le hello!
2010-09-29 06:06:47 utc jmettraux not sure if I can do a 2.1.11 today, I'm stuck with a ruote-couch issue
2010-09-29 06:07:40 utc tosch_le keep cool, no complains about 2.1.10 <> 2.1.11 clashes since saturday ;-)
2010-09-29 06:13:15 utc jmettraux well, I can still do it, wednesday ain't over yet, especially for you
2010-09-29 06:13:47 utc tosch_le :-)
2010-09-29 06:13:59 utc tosch_le i'll help were i can
2010-09-29 06:14:04 utc tosch_le just drop a line
2010-09-29 06:14:19 utc jmettraux thanks !
2010-09-29 08:12:51 utc jmettraux ruote-couch looking good
2010-09-29 08:13:51 utc tosch_le_ great
2010-09-29 08:27:44 utc jmettraux lbt: hello, is there a way for me to ask questions about http://wiki.meego.com/Release_Infrastructure/BOSS/Performance/Results ?
2010-09-29 08:28:19 utc lbt I'll find a way
2010-09-29 08:42:31 utc jmettraux well, I shouldn't, I'm not a meego employee
2010-09-29 09:58:50 utc lbt jmettraux: as in I'll make sure we all join a public/open mailing list that's on-topic so you can ask there :) And meego is a normal OSS project; nothing to do with "employee"
2010-09-29 09:59:29 utc jmettraux so if you stop being paid, you'll go on contributing ?
2010-09-29 09:59:46 utc lbt I actually do most of my MeeGo community work on my own time
2010-09-29 09:59:55 utc lbt the entire community OBS is on my time
2010-09-29 10:00:05 utc jmettraux cool
2010-09-29 10:00:26 utc jmettraux I don't want to join the meego community
2010-09-29 10:00:43 utc lbt no, I don't expect you to need to
2010-09-29 10:11:01 utc lbt jmettraux: I've asked our team to all join http://lists.meego.com/listinfo/meego-distribution-tools ... I think they've set it up as subscriber only; would you mind joining and mailing them there. And I agree they should cross post this information to the openweferu list
2010-09-29 10:11:33 utc lbt Maybe leave it until tomorrow as they're probably leaving work nowish
2010-09-29 10:11:54 utc jmettraux yes, standard procedure
2010-09-29 10:12:07 utc jmettraux thanks for pushig
2010-09-29 10:12:10 utc jmettraux pushing
2010-09-29 10:12:23 utc jmettraux I'd love to fix those issues
2010-09-29 11:47:56 utc jmettraux tosch_le: hello, do you think ruote-kit itself is ready for 2.1.11 ?
2010-09-29 11:48:38 utc jmettraux tosch_le_: hello, do you think ruote-kit itself is ready for 2.1.11 ?
2010-09-29 11:50:13 utc kennethkalmer jmettraux: hi
2010-09-29 11:50:23 utc jmettraux kennethkalmer: hello
2010-09-29 11:51:13 utc kennethkalmer doing well today ?
2010-09-29 11:51:36 utc tosch_le_ hi kenneth!
2010-09-29 11:51:43 utc tosch_le_ doing well, thanks. and you?
2010-09-29 11:51:59 utc jmettraux fine
2010-09-29 11:52:12 utc tosch_le_ jmettraux: will wait for the commit kenneth proposed yesterday and then, i think, rk is ready for release
2010-09-29 11:52:16 utc kennethkalmer good thanks
2010-09-29 11:52:29 utc kennethkalmer i'm getting failures on running rk's tests with 1.9.2
2010-09-29 11:52:52 utc kennethkalmer and a spec that is hanging
2010-09-29 11:53:02 utc tosch_le_ you're not the only one, iirc
2010-09-29 11:53:35 utc kennethkalmer getting the same thing on tosch/master too
2010-09-29 11:53:40 utc kennethkalmer any ideas ?
2010-09-29 11:54:20 utc jmettraux any gist ?
2010-09-29 11:54:23 utc jmettraux ;-)
2010-09-29 11:57:17 utc tosch_le_ did you run "bundle install" before running the specs?
2010-09-29 11:57:46 utc tosch_le_ updating the bundled ruote to the latest version is a good idea
2010-09-29 11:59:29 utc tosch_le_ errors_spec: http://gist.github.com/602631
2010-09-29 12:01:11 utc jmettraux tosch_le_: the rest is fine ?
2010-09-29 12:01:28 utc tosch_le_ no, workitems_spec throws errors, too
2010-09-29 12:02:01 utc tosch_le_ updated gist
2010-09-29 12:02:44 utc jmettraux OK
2010-09-29 12:04:00 utc kennethkalmer running bundle update instead of install
2010-09-29 12:04:37 utc jmettraux I'm finishing a ruote-couch test here and then I switch to rk
2010-09-29 12:05:00 utc tosch_le_ "spec spec/ruote-kit_configure_spec.rb" hangs
2010-09-29 12:06:04 utc jmettraux this one is maybe completely off
2010-09-29 12:15:38 utc tosch_le_ this one hangs when calling "RuoteKit.run_worker(Ruote::HashStorage.new)"
2010-09-29 12:16:19 utc tosch_le_ and that's no surprise, isn't it?
2010-09-29 12:18:16 utc tosch_le_ jmettraux: is it ok for you if we disable the 'run_worker' spec?
2010-09-29 12:18:30 utc jmettraux please fix it as you will
2010-09-29 12:18:38 utc jmettraux I'm working on the workitem spec right now
2010-09-29 12:20:29 utc jmettraux skip => -100, limit => 100 is disorienting the storage
2010-09-29 12:21:11 utc tosch_le_ you wanted corner cases? you get corner cases ;-)
2010-09-29 12:21:55 utc jmettraux cornered indeed
2010-09-29 12:24:29 utc jmettraux @skip = @count - @limit if @skip >= @count
2010-09-29 12:25:21 utc tosch_le_ '=' is too much :-(
2010-09-29 12:25:44 utc tosch_le_ i should think before coding…
2010-09-29 12:27:21 utc jmettraux the tests caught it
2010-09-29 12:31:34 utc tosch_le_ will you fix it or shall i?
2010-09-29 12:32:53 utc jmettraux I'm fixing it
2010-09-29 12:33:09 utc jmettraux please fix the configure issue
2010-09-29 12:34:51 utc jmettraux if you push before me, please tell me
2010-09-29 12:35:29 utc tosch_le_ fixing will be difficult, that one has to do with the engine joining the thread in RuoteKit.run_worker
2010-09-29 12:35:53 utc tosch_le_ i'm really tempted to throw that example group out
2010-09-29 12:36:16 utc jmettraux please do it
2010-09-29 12:36:54 utc jmettraux just pushed
2010-09-29 12:39:42 utc tosch_le_ re-running specs now
2010-09-29 12:40:09 utc tosch_le_ pushed
2010-09-29 12:40:34 utc tosch_le_ specs green with ruby enterprise edition
2010-09-29 12:40:52 utc tosch_le_ 1.9.2 is still compiling
2010-09-29 12:41:00 utc jmettraux ;-)
2010-09-29 12:41:13 utc jmettraux ruby 1.8.7 (2010-01-10 patchlevel 249) [i686-darwin10.2.0] green
2010-09-29 12:41:34 utc tosch_le_ kenneth? waiting for your commit ;-)
2010-09-29 12:41:50 utc kennethkalmer let me just run again on 1.8.7
2010-09-29 12:49:09 utc tosch_le_ i'd like to have a faster computer. still: "info: ruby-1.9.2-p0 - #compiling"
2010-09-29 12:50:03 utc jmettraux we should find a way to sponsor a new machine for you
2010-09-29 12:51:09 utc jmettraux ruby 1.9.2p0 (2010-08-18 revision 29036) [x86_64-darwin10.4.0] green
2010-09-29 12:52:09 utc kennethkalmer man, ZA bandwidth sucks... bundler is doing its thing
2010-09-29 12:53:26 utc tosch_le_ sponsoring more bandwidth for ZA will be more a challenge :-D
2010-09-29 12:54:04 utc kennethkalmer oh yeah ;)
2010-09-29 12:54:06 utc jmettraux and sponsoring a ZA machine is like sponsoring 2 machines form anywhere else ;-)
2010-09-29 12:54:35 utc jmettraux you guys should host zarubygems.org
2010-09-29 12:58:11 utc tosch_le_ ACTION celebrates: "info: ruby-1.9.2-p0 - #installing"
2010-09-29 13:03:27 utc kennethkalmer pushed
2010-09-29 13:03:36 utc kennethkalmer been thinking about local rubygems mirror
2010-09-29 13:04:54 utc kennethkalmer yep, same failures as on 1.9.2, nothing new after the patch
2010-09-29 13:07:51 utc tosch_le_ ruby 1.9.2p0 (2010-08-18 revision 29036) [i686-linux]
2010-09-29 13:07:54 utc tosch_le_ green
2010-09-29 13:08:08 utc tosch_le_ including kenneth' commit
2010-09-29 13:09:19 utc jmettraux tosch_le: you merged kenneth's changes into your repo ?
2010-09-29 13:09:40 utc tosch_le_ yes
2010-09-29 13:10:12 utc jmettraux great
2010-09-29 13:11:27 utc tosch_le_ kenneth: next time, could you please rebase your commits so that yours are on top of the older ones?
2010-09-29 13:12:01 utc tosch_le_ ACTION doesn't like merge commits
2010-09-29 13:12:40 utc jmettraux git pull --rebase (ftw)
2010-09-29 13:13:37 utc jmettraux 1.8.7 and 1.9.2 green
2010-09-29 13:14:06 utc tosch_le_ does anybody like to test on 1.9.1?
2010-09-29 13:14:15 utc jmettraux I have it
2010-09-29 13:14:25 utc kennethkalmer i think 1.9.1 is dead already
2010-09-29 13:14:34 utc kennethkalmer 1.8.7 and 1.9.2 should be targets imho
2010-09-29 13:14:42 utc tosch_le_ +1
2010-09-29 13:14:45 utc tosch_le_ :-)
2010-09-29 13:15:10 utc jmettraux OK great
2010-09-29 13:19:46 utc tosch_le_ ah, exciting, it's release time :-)
2010-09-29 13:20:40 utc jmettraux well, 2.1.11 still not out :-(
2010-09-29 13:21:52 utc tosch_le_ what blocks the road?
2010-09-29 13:23:40 utc jmettraux net-http-persistent, working on closing the connections
2010-09-29 13:23:49 utc jmettraux (ruote-couch)
2010-09-29 13:24:08 utc tosch_le_ ouch
2010-09-29 13:36:07 utc jmettraux ruote-ouch
2010-09-29 13:36:31 utc tosch_le_ lol
2010-09-29 13:57:56 utc jmettraux starting to look god
2010-09-29 13:57:59 utc jmettraux good
2010-09-29 13:58:34 utc tosch_le_ how did you look before?
2010-09-29 13:58:37 utc tosch_le_ ;-)
2010-09-29 13:59:04 utc jmettraux :-)
2010-09-29 13:59:13 utc jmettraux human, all along
2010-09-29 14:00:06 utc tosch_le_ anyway, good news!
2010-09-29 14:02:55 utc jmettraux damn, "pagingation" http://github.com/tosch/ruote-kit/commit/14a0ea27914633d507fa52f345e0f0f352a9e259
2010-09-29 14:06:07 utc jmettraux still failing :-(
2010-09-29 14:06:16 utc jmettraux argh, that will have to wait then
2010-09-29 14:06:32 utc tosch_le_ is it failing only with couch?
2010-09-29 14:06:38 utc tosch_le_ or generally?
2010-09-29 14:06:52 utc jmettraux only with couch
2010-09-29 14:07:23 utc jmettraux and only with net-http-persistent
2010-09-29 14:07:49 utc tosch_le_ strange
2010-09-29 14:08:46 utc jmettraux the number of open files touches the ceiling and the test suite EEEEEEEEs
2010-09-29 14:09:05 utc jmettraux I implemented a fix, but it doesn't work
2010-09-29 14:09:10 utc jmettraux I will go on tomorrow
2010-09-29 14:10:35 utc tosch_le_ :-)
2010-09-29 14:10:51 utc tosch_le_ smells like a test specific problem to me
2010-09-29 14:11:21 utc jmettraux it's always test specific :(
2010-09-29 14:11:45 utc jmettraux how to run 399 functional tests with 399 different engines
2010-09-29 14:12:46 utc jmettraux for ruote-couch, the suite runs in 23 minutes ~
2010-09-29 14:12:50 utc tosch_le_ that needs a biiiiig machine, that needs.
2010-09-29 14:13:17 utc jmettraux half of that time is spent in setting up and tearing down the dbs
2010-09-29 14:13:21 utc jmettraux :-(
2010-09-29 14:13:22 utc tosch_le_ or better: More power!
2010-09-29 14:13:32 utc jmettraux 399 machines
2010-09-29 14:13:54 utc tosch_le_ s/machine/computer/
2010-09-29 14:14:01 utc jmettraux 2 months ago I set a CI server up
2010-09-29 14:14:44 utc tosch_le_ anyway, is your not working fix anywhere to see?
2010-09-29 14:14:59 utc jmettraux let me push
2010-09-29 14:17:06 utc jmettraux I suspect http://github.com/jmettraux/rufus-jig/blob/master/lib/rufus/jig/adapters/net_persistent.rb#L49 to not close the connections
2010-09-29 14:17:27 utc jmettraux my task tomorrow is to run the tests with "lsof watchdog" on
2010-09-29 14:20:15 utc tosch_le_ from Net::HTTP::Persistent's docs: "If you’ve used Net::HTTP::Persistent across multiple threads you must call this in each thread."
2010-09-29 14:20:37 utc tosch_le_ but i suppose that's not the problem
2010-09-29 14:21:15 utc jmettraux that could be it
2010-09-29 14:21:46 utc jmettraux tomorrow's lsof will tell me exactly what's going on
2010-09-29 14:25:47 utc jmettraux have a good evening !
2010-09-29 14:25:58 utc tosch_le_ you too!
2010-09-29 14:26:04 utc tosch_le_ bye!
2010-09-29 14:26:26 utc jmettraux many thanks !