[00:33] <StevenK> wallyworld__: *DB-*
[00:33] <StevenK> You said stable, which is not right
[00:33] <wallyworld__> shit
[00:34] <StevenK> We can't fix it without reverting, so it's not worth it
[00:34] <wallyworld__> it merged the right file though
[00:35] <StevenK> Oh sure, it's just the commit message that's the problem
[00:36] <wallyworld__> hopefully no one else will notice
[00:36] <StevenK> wgrant will. Especially if I tell him first.
[00:42] <wgrant> Heh
[00:45] <StevenK> steven@undermined:~/launchpad/lp-branches% ls -1 | wc -l
[00:45] <StevenK> 356
[00:45] <StevenK> Hmmmm
[00:45] <wgrant> I think you may have a problem.
[00:46] <StevenK> /dev/mapper/sys-home  150G  105G   38G  74% /home
[00:46] <StevenK> Not yet :-P
[01:25] <wallyworld__> StevenK: ian@wallyworld:~/projects/lp-branches$  ls -1 | wc -l
[01:25] <wallyworld__> 575
[01:27] <StevenK> Haha
[01:28] <wallyworld__> mine is bigger than yours :-P
[02:43] <wallyworld__> StevenK: do you know where can look to see the preferred way to capture log messages or otherwise inspect what is logged during a job run so that i can write a test assertion?
[02:44] <StevenK> Why are you asserting on log output, anyway?
[02:44] <wallyworld__> the required job behaviour is to do a no-op and log a message
[02:44] <wallyworld__> and i want to check tthat that occurs
[02:46] <StevenK> wallyworld__: You probably want to create a logger with a Handler that backs onto StringIO
[02:46] <StevenK> And then pass that in
[02:46] <wallyworld__> yeah, i was hoping for an example to save me writing it :-)
[02:47] <StevenK> BufferLogger from lp.services.log.logger
[02:47] <lifeless> wallyworld__: StevenK: or the one in fixtures
[02:47] <lifeless> you don't need to write it at all
[02:47] <wallyworld__> thanks, will look
[02:47] <lifeless> and the one in fixtures knows how to install itself and restore state afterwards
[02:48] <wallyworld__> lifeless: so, BranchScan job does a "from lp.services.scripts import log" in it's run method. i guess i need to change that too?
[02:48] <lifeless> shouldn't, no.
[02:48] <wallyworld__> right, i'll read the code
[02:49] <wallyworld__> thanks
[03:16] <wallyworld__> StevenK: can haz quickie? https://code.launchpad.net/~wallyworld/launchpad/branchscan-deleted-branch-602323/+merge/133175
[03:18] <StevenK> wallyworld__: Your test doesn't actually test anything
[03:18] <StevenK> Oh
[03:18] <StevenK> It's in the with, sneaky
[03:18] <wallyworld__> yeah
[03:18] <wallyworld__> i didn't even now that context manager existed
[03:18] <wallyworld__> know
[03:19] <StevenK> expected_message = 'Skipping branch %s because it has been deleted.' % (
[03:19] <StevenK>     db_branch.unique_name,)
[03:19] <lifeless> wallyworld__: you can use 'with foo as bar:'
[03:19] <StevenK> Might work in terms of line length, too
[03:19] <lifeless> so with try_advisory_lock(...) as lock:
[03:20] <wallyworld__> can do
[03:20] <wallyworld__> it will only have 1 or 2 loc, and i was copying an existing test
[03:20] <StevenK> Refactor the other tests?
[03:20] <StevenK> As a thought
[03:21] <wallyworld__> for what purpose? they read fine as written
[03:21] <StevenK> Save a few LoC?
[03:21] <wallyworld__> a few? for what benefit?
[03:21] <wallyworld__> if it were many, then sure
[03:22] <wallyworld__> there's no readability issue or anything concrete that will benefit from changing thing here i don't think
[03:25] <StevenK> wallyworld__: That's one difference between you and me -- I will refactor to save a few LoC, and you will not.
[03:25] <wallyworld__> i will if there's a decent gain. but otherwise it make the mp diff harder to read and achieves very little benefit
[03:26] <StevenK> wallyworld__: I don't think lines 62 and on don't need to be in the outer with
[03:27] <wallyworld__> yes, good pickup
[03:27] <wallyworld__> i saved 1 loc in the job method \o/
[03:27] <wallyworld__> lp is now fixed
[03:27] <StevenK> wallyworld__: This means you can probably reflow expected_message
[03:28] <wallyworld__> sorry, too sarcastic
[03:28] <StevenK> wallyworld__: Your review is now 'Needs Attitude Fixing'
[03:28] <StevenK> :-P
[03:28] <wallyworld__> hah
[03:29] <wallyworld__> at least i leave no ambiguity as to my philosophy
[03:30] <wallyworld__> sadly expected_message still neews 2 lines
[03:30] <lifeless> shoot on sight ?
[03:32] <StevenK> wallyworld__: Better than 3
[03:32] <wallyworld__> it was only 2
[03:32] <StevenK> No, it was three
[03:32] <StevenK> 62	+ expected_message = (
[03:32] <StevenK> 63	+ 'Skipping branch %s because it has been deleted.'
[03:32] <wallyworld__> lifeless: shoot what?
[03:32] <StevenK> 64	+ % db_branch.unique_name)
[03:32] <wallyworld__> ah, sorry, still 2
[03:32] <wallyworld__> 3
[03:32] <wallyworld__> *3*
[03:32] <StevenK> wallyworld__: Shoot the messenger
[03:33] <wallyworld__> StevenK: https://pastebin.canonical.com/77799/
[03:34] <wallyworld__> i changed the existing "lock = foo..." to "with foo():" to save 1 line
[03:34] <lifeless> wallyworld__: your philosophy, is it 'shoot on sight' ?
[03:35] <StevenK> wallyworld__: Yes, I got that :-)
[03:35] <wallyworld__> lifeless:  shoot the messenger?
[03:35] <lifeless> wallyworld__: forget it
[03:35] <wallyworld__> sorry, slow today
[03:35] <StevenK> wallyworld__: I guess the expectedLog really only needs to be around the job.run
[03:35] <StevenK> Today?
[03:36] <StevenK> What's your excuse the rest of the time?
[03:36] <wallyworld__> well, always
[03:36] <wallyworld__> i could move it down a few lines
[03:38] <wallyworld__> StevenK: i made that change and pushed
[03:40] <StevenK> Ah ha, I think I've worked out how to refactor this lot.
[03:43] <StevenK> wallyworld__: r=me
[03:44] <wallyworld__> StevenK: thanks.
[04:19] <StevenK> wallyworld__: You have broken buildbot.
[04:19] <wallyworld__> yeah - fixing. stupi celery
[04:20] <wallyworld__> apparently it is illegal to get the branch unique name in the job init when running with celery
[05:19] <StevenK> wallyworld__: https://code.launchpad.net/~stevenk/launchpad/local-codeimports-bad/+merge/133183
[05:20] <StevenK> wallyworld__: "[testfix][r=wallyworld][no-qa] Test fix for r16242 - rever" Huh?
[06:01] <wallyworld__> StevenK: looking now
[06:02] <StevenK> And here I was, about to reboot so I can play HL2.
[06:08] <wallyworld__> StevenK: r=me
[06:08] <wallyworld__> now you can play :-)
[06:08] <wallyworld__> my fingers sure messed up that commit message
[06:09] <StevenK> Haha
[06:10] <StevenK> wallyworld__: I can't, I have to pick up my wife soon :-)
[06:10] <wallyworld__> hmmm. HL2 or wife
[06:10] <wallyworld__> tough choice
[06:10] <StevenK> I didn't know you'd played HL2?
[06:11] <wallyworld__> i haven't but i am not ignorant of its existance
[08:58] <adeuring> good morning
[20:43] <wallyworld__> flacoste: hi, i'll be on leave to go to the cricket on friday, can we do it today?
[20:43] <flacoste> wallyworld__: are you available now?
[20:43] <flacoste> wallyworld__: or in the next 1h15?
[20:43] <flacoste> wallyworld__: it's my wife's birthday today, so coming back in the evening is not an option ;-)
[20:43] <wallyworld__> i have to take lachie to school in 10 minutes and have my standup in 1 hour. can i ping you after that? say in about 2hrs?
[20:44] <wallyworld__> ah ok
[20:44] <flacoste> wallyworld__: do you have any time between the school run and your stand-up?
[20:44] <wallyworld__> what about sat morning my time?
[20:44] <wallyworld__> i have 30 minutes
[20:45] <flacoste> wallyworld__: that should be plenty
[20:45] <wallyworld__> in about 45 minute's time
[20:45] <wallyworld__> ok
[20:45] <flacoste> wallyworld__: ping me when you're available
[20:45]  * wallyworld__ nods
[20:45] <flacoste> wallyworld__: that's better than sat morning, will the cricket game even be finished by then ? ;-)
[20:46]  * flacoste likes hockey 2 hours and your done
[20:46] <czajkowski> flacoste: yes but with hockey you get really serious injuries!
[20:47] <czajkowski> cricket they just get sun burnt!
[20:49] <flacoste> lol
[20:50] <czajkowski> flacoste: I went to one game when I was over in Canada 3 years ago, and the amount of nose bleeds was unbelievable , not even in rugby is it that messy
[20:50] <flacoste> czajkowski: yeah, hockey is a violent game
[20:50] <flacoste> especially in North America
[21:18] <rick_h> deryck: http://www.mikechambers.com/blog/2011/07/20/timing-issues-when-animating-with-css3-transitions/ and the last comment got me going http://jsfiddle.net/jyjXN/7/
[21:18] <rick_h> deryck: just as an fyi spread the knowledge.
[21:18] <bigjools> hello folks
[21:18] <rick_h> howdy bigjools
[21:18] <rick_h> bigjools: so was that starts in the plane a sunroof glass kind of thing or was it projected/faked?
[21:19]  * rick_h goes completely off topic 
[21:19] <bigjools> rick_h: LEDs in the ceiling
[21:19] <rick_h> ah, less impressed now :(
[21:19] <bigjools> picture didn't do it justice though
[21:19] <bigjools> very well done
[21:19] <bigjools> first plane I've been on with orange lighting too!
[21:21] <deryck> rick_h, interesting.  I guess setTimeout is unavoidable.  But I guess that makes sense when we're dynamically constructing nodes.
[21:21] <rick_h> deryck: that last comment about forcing a repaint by resetting display on the DOM node works
[21:21] <rick_h> so it's ugly, but not set timeout interval and should work peachy
[21:21] <czajkowski> bigjools: how long was the flight ?
[21:22] <rick_h> see in the jsfiddle: dom_node._node.style.display = document.defaultView.getComputedStyle(dom_node._node)['display'];
[21:22] <bigjools> czajkowski: "too"
[21:24] <deryck> rick_h, yeah, that's better than the setTimeout.
[21:31] <wallyworld__> flacoste: google+ hangouts is screwing me over - it keeps wanting to install the google talk plugin over and over even though the hangout initially starts ok
[21:32] <flacoste> wallyworld__: want to use skype?
[21:32] <wallyworld__> ok
[21:57] <sinzui> wallyworld__, StevenK, jcsackett, I will be 30 minutes late to our meeting. Feel free to post pone it.
[21:57] <wallyworld__> i have no problem starting 30 minutes late
[21:59] <jcsackett> fine by me.
[22:00] <StevenK> Okay, that's fine.