=== gary_poster|away is now known as gary_poster === gary_poster is now known as gary_poster|away [01:49] StevenK: wallyworld mentioned on the call that you probably forgot to remove the feature flags from sampledata [01:49] Drag [01:49] *Drat [01:53] wgrant: with the change to give all new projects default sharing policies, setPrivateBugs() become irrelevant and broken. so i think i should remove it [01:53] or wil that break apport? [01:53] wallyworld_: Things that test private_bugs behaviour (which is still important, not for apport but for all unmigrated projects) will have to setBugSharingPolicy(None) first [01:54] there's a garbo job which has migrated all projects except a few commercial ones [01:54] private_bugs/setPrivateBugs will die with BVP [01:54] Sure [01:54] And the commercial ones are the ones that use private_bugs :) [01:55] hmm. ok. will set to None [01:55] If it was killable already, it would be dead already :) [01:55] But sadly not quite yet [01:55] yeah, me got a lot of test failures related to all this to fix :-( [02:03] wallyworld_: Not tooo many things should use setPrivateBugs [02:04] it's more than just that [02:04] wallyworld_: Most will probably be fixed by fixing makeProduct, I suspect [02:04] bvp stuff fails too etc [02:04] Ah, true [02:14] StevenK: Hm, why does lib/deb822.py still exist? [02:14] It's a broken symlink now [02:14] Did you miss it in your cleanup? [02:16] Hmm [02:16] Seems I did, indeed [02:19] Sigh, I bet https://code.launchpad.net/~stevenk/launchpad/feature-flags-sampledata is cursed. [02:21] [2012-08-29 02:03:46,105: INFO/PoolWorker-2] Job resulted in OOPS: OOPS-f59d1eb05281222ebbe8ce921e02b2d8 [02:21] d [02:21] cursed indeex [02:31] StevenK: Well [02:31] StevenK: I was going to say that was a trivial self-review [02:31] But you made a mistake :P [02:31] :-( [02:31] I was like "Green? There's not supposed to be any green in this branch ... Oh." [02:32] Ah [02:32] I see the next rev [02:32] heh [02:32] That's better, +0, -3 [02:35] wgrant: So now you've made fun of me for it, are you going to review it, or shall I self-review? [02:36] That's a self-review if I ever saw one [03:44] jam: what tz are you now? [03:55] lifeless: he's +4 [03:56] noice [03:56] halfway to a good place :) [04:59] lifeless: +4 [04:59] morning, btw [05:00] StevenK: auditor broke db_devel again :-( i need to land a branch to devel but have to wait till db devel is restarted with a testfix or completes fully? [05:02] wallyworld_: Force, wait 10 minutes, land [05:02] No need for a testfix or a green run [05:03] ok, will do, ta [05:03] wallyworld_: The new auditor wasn't in that db-devel run [05:03] ah ok, explains it. [05:03] were you going to land a testfix for it? [05:03] devel finally passed, so it will be in the next build after buildbot-poller actually merges in stable. [05:04] which will be when do we know? [05:04] PQM probably rejected it due to db-devel being broken [05:04] Force a build of db-devel [05:04] ok [05:05] Bleh, new auditorfixture [05:06] r15871 has hit stable at least. [05:07] Now to wait for the 8 revisions to be merged into db-devel [05:07] Hmmmmm [05:08] bigjools: /wave for being online [05:09] hi jam [05:09] gimme 5 mins and I'll grab you [05:09] bigjools: Hahaha. Grab him where? [05:13] jam: o/ :) [05:15] Come *ON* buildbot-poller [05:16] jam: its probably faster to do remote hands (tuolumne) [05:20] jam: ok, hangout? [05:28] bigjools: sure on the hangout, let me grab the camera [05:28] RARGH [05:28] buildbot-poller only merges if both builds worked [05:30] Yes [05:30] bigjools: https://plus.google.com/hangouts/_/70125daf3bc2bab87f46d925ed9e5fc229cdf07e?authuser=0&hl=en# [05:31] StevenK: so pqm is ignoring my lp-land and not even sending back hate mail. is that expected? [05:33] Haven't seen anything in the pqm logs [05:33] But archvsync is still chewing on my request [05:36] but lp-land should be expected to work right now i think? [05:36] yes [05:37] Which branch? [05:37] new-project-sharing-policies-1040989 [05:37] ? [05:38] no, branch-infotype-portlet2-1040999 [05:39] Hm [05:39] No conflicts [05:39] How many times have you tried to land it? [05:39] 3 maybe [05:39] :( [05:39] One more time for good luck? [05:39] once when it complained about a testfix, 2 since after forcing bb [05:39] ok, here goes.... [05:39] How long after forcing buildbot? [05:40] It should have been fine after the first */10 after the forcing [05:41] wgrant: https://code.launchpad.net/~stevenk/launchpad/shift-get_distroseries_pofiles/+merge/121751 [05:42] wgrant: yeah, i waited maybe 20 minutes [05:42] jam: you're frozen [05:42] wallyworld_: :( [05:42] wgrant: and lestest attempt, nothing [05:42] i gotta go get the kid, back soon [05:43] wgrant@carob:~$ rsync prasé:: [05:43] rsync: getaddrinfo: pras\#303\#251 873: Name or service not known [05:43] fail [05:44] wallyworld_: Hm, it's either hung or a wallyworld-denier [05:45] I see nothing in its logs from you in more than 50 minutes [05:45] wallyworld_: Ah [05:45] wallyworld_: Landed just now [05:46] Ah ha! It was waiting for wallyworld_ to be AFK. === jam1 is now known as jam [06:02] wgrant: No review for me? :-( [06:03] StevenK: Where's the cleanup? [06:03] Adding more methods to DistroSeries is probably not removing tech debt. [06:05] wgrant: Destroying half of a translations module doesn't do it for you? [06:07] wgrant: I'd argue that they probably belonged their to begin with [06:08] well, next time something fails, i'm going away again and it wil work [06:08] StevenK: Adding app-specific stuff to registry modules when it can be avoided makes me a very sad panda [06:09] Obviously, -60 doesn't make you happy either [06:09] model/distroseries.py is already 1800 lines long [06:10] -60 is nice [06:10] But bloating (I)DistroSeries is less so [06:10] wgrant: Then where? [06:11] IVPOExportSet? :P [06:12] Bleh [06:12] * StevenK deletes the MP [06:12] Code reduction is nice, but Distribution/DistroSeries/Archive/Person are already huge [06:12] We need a better solution than just adding every method in the application to them [06:13] stop using subclassing ? [06:13] lazr.restful [06:16] bigjools: it seems to be fairly specific to g+, do you want to try the hangout without video, or maybe mumble? [06:16] StevenK: wgrant: did you guys find any more email addresses on dogfood? === jam1 is now known as jam [06:17] jam1: looks like your whole internets given the disconnection from irc! [06:17] bigjools: well, hangouts triggers the disconnect. [06:17] I don't have mumble set up on here. I think we're pretty much done for now, perhaps you can try out the code and that'll generate more questions [06:18] yeah I guess they don't like all the bandwidth it seems to eat [06:19] wallyworld_: Wasn't there only one missing? [06:20] wgrant: not sure. i thought there were a few but could be wrong === jam1 is now known as jam [07:47] good morning === almaisan-away is now known as al-maisan [08:29] mgz: done now [08:31] ta. [08:57] Given a base url like http://host/path and an additional path/to/thing, what would you use in Python to join them? Yes, this is a trick question. [08:59] bigjools: urlparse.urljoin, as long as the first URL has a trailing / [09:01] bzzzzzzzzzt [09:01] wrong [09:02] Oh? [09:02] >>> urljoin("http://10.0.0.9/MAAS/", "/api/thing") [09:02] 'http://10.0.0.9/api/thing' [09:02] Well, yes, if the path is absolute :) [09:02] this is the trick part :) [09:02] But yours wasn't [09:02] wait, what is the trick part? [09:02] aha stripping the leading / from the path works [09:02] >>> os.path.join('where/did/this/go', '/hello/there') [09:02] '/hello/there' [09:04] nigelb: the trick is that if you have a leading slash on the path part it doesn't work [09:05] It trips everyone up, but it is consistent with other path handling and sort of makes sense. [09:05] oh wow. I totally didn't notice that :) [09:06] it's a nasty trap for URLs [09:06] makes total sense for is.path though [09:06] os.path, even === al-maisan is now known as almaisan-away [10:36] what does this buildbot failure http://lpbuildbot.canonical.com/builders/lucid_db_lp/builds/2252 mean? [10:41] adeuring: There's a new version of auditorfixture that will hopefully fix that spurious failure [10:41] wgrant: thanks [10:41] It should be in the next db-devel build [10:41] how do I unbreak my copy of launchpad today? [10:42] bin/buildout [10:42] mgz: ^^ [10:42] Or just make [10:42] ta. [10:45] so, basically I need to run make after bzr switch before running bin/test? I will get used to this in the end... [10:46] If you're changing between trees of significantly different versions, yes. [10:46] You do need to rebuild :) [10:48] too used to either python, where you don't, or C, where the way you run tests builds as a side-effect anyway [10:48] You do in Python unless you have modules installed system-wide. [10:48] Which is becoming a rarity [10:48] buildout and pip have this same problem [10:48] and virtualenv [11:04] mgz: any luck with the ppa updates? [11:05] jam: no, we need some major packaging changes [11:06] see the log from yesterday morning, maxb suggested a few things [11:10] mgz: "we need major changes" means we should bring up the idea of just going to P w/ flacoste [11:15] jam: I think so, yes, but lifeless wasn't keen [11:16] mgz: I'm not specifically keen, but if it is 'spend another 2 weeks sorting out py2.7 on lucid' vs, 'slowly migrate one appserver at a time to P' where the latter is where we want to be anyway... [11:18] mgz: I really like the idea of every step being something we can easily roll back, but remember my "installing py2.7 seems tractable" [11:18] if that part is invalidated === gary_poster|away is now known as gary_poster === gary_poster is now known as gary_poster|away [11:34] okay, lib/lp_sitecustomize.py is what made this so confusing [11:35] ...why is all that work being done pre-normal imports? [11:37] mgz: bin/py does not load 'site.py' the regular python site package, instead it uses a custom site.py so that it *only* loads eggs from the buildout location. [11:37] to promote isolation from the system. [11:38] sure, but that's site.py [11:38] this sitecustomize.py stuff is all just working around zope being weird and (failing to) silence various loggers [11:38] and importing a barrel of cruft in the process [11:41] deliberately importing bzrlib plugins before script execution seems to serve no purpose, and breaks our api version check stuff === gary_poster|away is now known as gary_poster === almaisan-away is now known as al-maisan [12:56] wgrant, just an FYI, the landscape project wasn't changed to proprietary by SQL [12:56] wgrant, I did it yesterday, so I think we hit a bug [12:56] (I just read the backlog of your conversation with fcorrea) [12:58] Beret: Around what time did you do it, do you recall? [12:58] 'cause when I set it away and back to the same setting, it worked fine. [12:58] Hmmm [12:59] I would guess around 22:00 GMT [12:59] Thanks [12:59] I'm not certain I just knowit was later in the day [12:59] after the email went out [13:04] Beret: Ah, found the bug [13:04] A cleanup job is being a tad aggressive [13:06] sweet [13:06] Only affects the initial setup, before there are any new bugs or branches [13:32] abentley, ping for standup [13:32] deryck: Sorry, having trouble with it on this machine. Can you please invite me and I'll answer on my phone? [13:32] abentley, sure [13:50] abentley: that makefile update looks good from here. Thanks for that. Did you want to submit that? === sagaci_ is now known as sagaci === abentley changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: abentley | Firefighting: - | Critical bugs: 4.0*10^2 [14:24] abentley: Hi, do you have a few minutes to review https://code.launchpad.net/~wgrant/launchpad/bug-1043319/+merge/121874 shortly? It's a pretty simple regression fix for the sharing migration that I'd like to get landed tonight. [14:25] wgrant: certainly. [14:25] Bah [14:25] Curtis beat you to it [14:25] Even though he's not here [14:25] this week [14:25] Bad sinzui [14:26] Thanks anyway :) [14:26] wgrant: np [14:30] rick_h_: Sure, I'll get that in. [15:27] rick_h_: Whenever you're ready for a review: https://code.launchpad.net/~abentley/launchpad/makefile-tweak/+merge/121885 [15:32] abentley: r=me thanks for the fix === al-maisan is now known as almaisan-away === deryck is now known as deryck[lunch] === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === deryck[lunch] is now known as deryck === abentley changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: - | Firefighting: - | Critical bugs: 4.0*10^2 === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === gary_poster is now known as gary_poster|away [23:13] StevenK: Morning [23:13] wgrant: O hai [23:13] StevenK: Can you QA your rev? [23:13] And I will QA my fix [23:13] And then we can deploy the bugfix [23:13] I am in the middle of doing so [23:14] thanks [23:16] StevenK: Did you see the garbo bug? [23:16] Quite amusing [23:16] wgrant: Not I recall [23:17] launchpad.net/bugs/1043319 [23:17] dammit firefox [23:17] <_mup_> Bug #1043319: UnusedSharingPolicyPruner doesn't preserve permitted access policies

< https://launchpad.net/bugs/1043319 > [23:17] stop pruning schemes [23:20] wallyworld_: Did you cast your eyes over qas sharing? [23:23] StevenK: yes, looks ok [23:33] wgrant: Just waiting for your QA and we can deploy [23:34] Great [23:35] And I have a row ready to paste into LPS. === Ursinha` is now known as Ursinha