[16:06] <stgraber> jodh: gosh, 15k long test file, there goes my plan to read through test_initctl.c ;)
[16:09] <jodh> stgraber: I'd look at test_show_config() and test_reexec() in that file.
[16:12] <stgraber> jodh: my plan for an easy test is to add code to test_upstart_open which tests that initctl connects to the bus set in UPSTART_SESSION if it's set and fails if UPSTART_SESSION is unset and user_mode is set
[16:13] <stgraber> jodh: that'll essentially test the code I added to initctl.c, the rest is identical to standard upstart, so I'm not sure it makes sense to add a specific test to each of the test function
[16:15] <jodh> stgraber: sounds good to me.
[16:26] <stgraber> looks like it's going to be one of those rare branches where adding tests will take less time than writing the initial code ;) that's assuming I don't get into any weird nih bug when writing the tests ;)
[16:34] <xnox> well.... how long have you already spent sieving through test_initctl.c?
[16:34] <xnox> =))))
[16:51] <stgraber> xnox: just long enough to notice vim telling me how long it was ;) then I decided to use grep to figure out where to add my tests ;)
[16:51] <stgraber> anyway, half the tests are done now
[16:51] <xnox> stgraber: i remember onces doing a grep -C 10 & using patch-mode to write a patch and path it in......
[16:52] <xnox> good old days of not enough ram and patching 10MB file =)))))
[16:54]  * xnox used to do crazy stuff, for little reason.
[16:55] <stgraber> ;)
[17:00]  * xnox doesn't like
[17:00] <xnox> precise-desktop-base login: <4>init: setvtrgb main process (258) terminated with status 1
[17:00] <xnox> upon starting lxc containers on raring.
[17:04] <stgraber> xnox: I agree and I had an action to add some "and not-container" to those upstart jobs but slangasek felt we should just let them fail ;)
[17:04] <xnox> stgraber: why does it complaint in my face and not somewhere else?
[17:04] <xnox> =)
[17:05] <stgraber> xnox: because it'd complain in your face on a real system, so we do that too in containers ;)
[17:05] <stgraber> or more likely, it'd do that on a real system if /dev/console didn't support those ioctls and your machine was booting fast enough to cause a race between setvtrgb and tty1 :)
[17:18] <stgraber> jodh: branch should be ready for another review. I added the two tests I mentioned earlier and they pass as expected.