[16:54] right, so that's all the tests able to be built again [16:54] just got to fix a few more test failures [16:54] and then I can push the commit of doom! [16:55] 18 files changed, 2271 insertions(+), 3012 deletions(-) [17:39] :-) [17:41] being the commit that splits Job into Job and JobConfig [17:43] though I might try and turn it into lots of bi-sectable commits [17:43] not sure how likely that is [17:48] WOO! [17:48] ALL TESTS PASS! [17:48] I cannot tell you how happy this makes me! :D [17:48] 18 files changed, 2549 insertions(+), 3105 deletions(-) [17:49] Whee [17:51] * Keybuk checks the build with optimisations [17:57] (and fixes the usual gcc "I'm stupid" issues) [18:11] ok, now the coverage check [19:04] it seems that the link on http://upstart.ubuntu.com/download.html to the readme file for 0.3.9 is broken... [19:06] phsdv: thanks, fixed [19:06] thanks, starting to read it [21:36] hello [21:47] hello [21:47] do you happen to know how to make upstart listen to kernel boot options (like "3" to make init/upstart boot into runlevel 3)? [21:48] It's my understanding that you just have to put a "3" as a kernel boot parameter, but this doesn't seem to work with upstart [21:49] right, it doesn't work [21:49] it's a bug [21:49] oh, ok [21:49] thanks [21:49] https://bugs.edge.launchpad.net/upstart/+bug/74664 [21:50] (and also) [21:50] https://bugs.edge.launchpad.net/upstart/+bug/85014 [21:50] depending which side of the bug you care about ;) [21:50] thanks for the links [22:56] http://codebrowse.launchpad.net/~keybuk/upstart/main/revision/scott%40netsplit.com-20071019214909-iyq8ua9epn5wsmfw?start_revid=scott%40netsplit.com-20071019214909-iyq8ua9epn5wsmfw [23:21] wow [23:23] Hehe [23:39] ba [23:39] I hate it when I find bugs just by looking [23:39] wing-commander init% ./test [23:39] test (#0) goal changed from stop to start [23:39] test (#0) state changed from waiting to starting [23:39] event_new: Pending starting event [23:39] test (#0) goal changed from start to stop [23:39] Handling starting event [23:39] event_finished: Finished starting event [23:39] test (#0) state changed from starting to waiting [23:39] event_new: Pending stopped event [23:39] Handling stopped event [23:39] event_finished: Finished stopped event [23:39] ... [23:39] if the goal changes to stop while the starting event is blocked, you never get a stopping event [23:39] just a stopped one [23:54] bugger [23:54] need two different next states from waiting [23:55] from starting, I mean