=== Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [00:33] wallyworld__: *DB-* [00:33] You said stable, which is not right [00:33] shit [00:34] We can't fix it without reverting, so it's not worth it [00:34] it merged the right file though [00:35] Oh sure, it's just the commit message that's the problem [00:36] hopefully no one else will notice [00:36] wgrant will. Especially if I tell him first. [00:42] Heh [00:45] steven@undermined:~/launchpad/lp-branches% ls -1 | wc -l [00:45] 356 [00:45] Hmmmm [00:45] I think you may have a problem. [00:46] /dev/mapper/sys-home 150G 105G 38G 74% /home [00:46] Not yet :-P [01:25] StevenK: ian@wallyworld:~/projects/lp-branches$ ls -1 | wc -l [01:25] 575 [01:27] Haha [01:28] mine is bigger than yours :-P [02:43] 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] Why are you asserting on log output, anyway? [02:44] the required job behaviour is to do a no-op and log a message [02:44] and i want to check tthat that occurs [02:46] wallyworld__: You probably want to create a logger with a Handler that backs onto StringIO [02:46] And then pass that in [02:46] yeah, i was hoping for an example to save me writing it :-) [02:47] BufferLogger from lp.services.log.logger [02:47] wallyworld__: StevenK: or the one in fixtures [02:47] you don't need to write it at all [02:47] thanks, will look [02:47] and the one in fixtures knows how to install itself and restore state afterwards [02:48] 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] shouldn't, no. [02:48] right, i'll read the code [02:49] thanks [03:16] StevenK: can haz quickie? https://code.launchpad.net/~wallyworld/launchpad/branchscan-deleted-branch-602323/+merge/133175 [03:18] wallyworld__: Your test doesn't actually test anything [03:18] Oh [03:18] It's in the with, sneaky [03:18] yeah [03:18] i didn't even now that context manager existed [03:18] know [03:19] expected_message = 'Skipping branch %s because it has been deleted.' % ( [03:19] db_branch.unique_name,) [03:19] wallyworld__: you can use 'with foo as bar:' [03:19] Might work in terms of line length, too [03:19] so with try_advisory_lock(...) as lock: [03:20] can do [03:20] it will only have 1 or 2 loc, and i was copying an existing test [03:20] Refactor the other tests? [03:20] As a thought [03:21] for what purpose? they read fine as written [03:21] Save a few LoC? [03:21] a few? for what benefit? [03:21] if it were many, then sure [03:22] there's no readability issue or anything concrete that will benefit from changing thing here i don't think [03:25] wallyworld__: That's one difference between you and me -- I will refactor to save a few LoC, and you will not. [03:25] 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] wallyworld__: I don't think lines 62 and on don't need to be in the outer with [03:27] yes, good pickup [03:27] i saved 1 loc in the job method \o/ [03:27] lp is now fixed [03:27] wallyworld__: This means you can probably reflow expected_message [03:28] sorry, too sarcastic [03:28] wallyworld__: Your review is now 'Needs Attitude Fixing' [03:28] :-P [03:28] hah [03:29] at least i leave no ambiguity as to my philosophy [03:30] sadly expected_message still neews 2 lines [03:30] shoot on sight ? [03:32] wallyworld__: Better than 3 [03:32] it was only 2 [03:32] No, it was three [03:32] 62 + expected_message = ( [03:32] 63 + 'Skipping branch %s because it has been deleted.' [03:32] lifeless: shoot what? [03:32] 64 + % db_branch.unique_name) [03:32] ah, sorry, still 2 [03:32] 3 [03:32] *3* [03:32] wallyworld__: Shoot the messenger [03:33] StevenK: https://pastebin.canonical.com/77799/ [03:34] i changed the existing "lock = foo..." to "with foo():" to save 1 line [03:34] wallyworld__: your philosophy, is it 'shoot on sight' ? [03:35] wallyworld__: Yes, I got that :-) [03:35] lifeless: shoot the messenger? [03:35] wallyworld__: forget it [03:35] sorry, slow today [03:35] wallyworld__: I guess the expectedLog really only needs to be around the job.run [03:35] Today? [03:36] What's your excuse the rest of the time? [03:36] well, always [03:36] i could move it down a few lines [03:38] StevenK: i made that change and pushed [03:40] Ah ha, I think I've worked out how to refactor this lot. [03:43] wallyworld__: r=me [03:44] StevenK: thanks. [04:19] wallyworld__: You have broken buildbot. [04:19] yeah - fixing. stupi celery [04:20] apparently it is illegal to get the branch unique name in the job init when running with celery [05:19] wallyworld__: https://code.launchpad.net/~stevenk/launchpad/local-codeimports-bad/+merge/133183 [05:20] wallyworld__: "[testfix][r=wallyworld][no-qa] Test fix for r16242 - rever" Huh? [06:01] StevenK: looking now [06:02] And here I was, about to reboot so I can play HL2. [06:08] StevenK: r=me [06:08] now you can play :-) [06:08] my fingers sure messed up that commit message [06:09] Haha [06:10] wallyworld__: I can't, I have to pick up my wife soon :-) [06:10] hmmm. HL2 or wife [06:10] tough choice [06:10] I didn't know you'd played HL2? [06:11] i haven't but i am not ignorant of its existance [08:58] good morning === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === jtv1 is now known as jtv === mpt_ is now known as mpt === dimitern is now known as dimitern_lunch === vds` is now known as vds === dimitern_lunch is now known as dimitern === gary_poster|away is now known as gary_poster === mpt_ is now known as mpt === yofel_ is now known as yofel === BradCrittenden is now known as bac === matsubara is now known as matsubara-lunch === deryck is now known as deryck[lunch] === matsubara-lunch is now known as matsubara === deryck[lunch] is now known as deryck === Ursinha-afk is now known as Ursinha === Ursinha-afk is now known as Ursinha === Ursinha_ is now known as Ursinha-afk === Ursinha_ is now known as Ursinha-afk === Ursinha_ is now known as Ursinha-afk [20:43] flacoste: hi, i'll be on leave to go to the cricket on friday, can we do it today? [20:43] wallyworld__: are you available now? [20:43] wallyworld__: or in the next 1h15? [20:43] wallyworld__: it's my wife's birthday today, so coming back in the evening is not an option ;-) [20:43] 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] ah ok [20:44] wallyworld__: do you have any time between the school run and your stand-up? [20:44] what about sat morning my time? [20:44] i have 30 minutes [20:45] wallyworld__: that should be plenty [20:45] in about 45 minute's time [20:45] ok [20:45] wallyworld__: ping me when you're available [20:45] * wallyworld__ nods [20:45] 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] flacoste: yes but with hockey you get really serious injuries! [20:47] cricket they just get sun burnt! [20:49] lol [20:50] 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] czajkowski: yeah, hockey is a violent game [20:50] especially in North America [21:18] 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] deryck: just as an fyi spread the knowledge. [21:18] hello folks [21:18] howdy bigjools [21:18] 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] rick_h: LEDs in the ceiling [21:19] ah, less impressed now :( [21:19] picture didn't do it justice though [21:19] very well done [21:19] first plane I've been on with orange lighting too! [21:21] rick_h, interesting. I guess setTimeout is unavoidable. But I guess that makes sense when we're dynamically constructing nodes. === Ursinha_ is now known as Ursinha-afk [21:21] deryck: that last comment about forcing a repaint by resetting display on the DOM node works [21:21] so it's ugly, but not set timeout interval and should work peachy [21:21] bigjools: how long was the flight ? [21:22] see in the jsfiddle: dom_node._node.style.display = document.defaultView.getComputedStyle(dom_node._node)['display']; [21:22] czajkowski: "too" [21:24] rick_h, yeah, that's better than the setTimeout. [21:31] 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] wallyworld__: want to use skype? [21:32] ok [21:57] wallyworld__, StevenK, jcsackett, I will be 30 minutes late to our meeting. Feel free to post pone it. [21:57] i have no problem starting 30 minutes late [21:59] fine by me. [22:00] Okay, that's fine.