=== Ursinha is now known as Ursinha-afk === mrevell__ is now known as mrevell === Ursinha-afk is now known as Ursinha [15:00] #startmeeting [15:00] Meeting started at 09:00. The chair is bac. [15:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [15:00] hi, who is here today? [15:00] me [15:00] me [15:00] for the launchpad reviewers meeting [15:01] me [15:01] me [15:01] me [15:01] (mars will probably be absent) [15:01] bigjools ping [15:02] EdwinGrubbs: ping [15:02] me [15:02] me [15:02] \o [15:02] me [15:02] me [15:02] me [15:03] team leads would you ping your absent folks [15:03] me [15:03] (foundations is as here as we're going to get) [15:03] gary_poster: thanks [15:03] me [15:03] np [15:03] hi intellectronica, jelmer [15:04] hi bac ;) [15:04] hi Brad [15:04] ok, let's get started and hopefully some stragglers will appear [15:04] [topic] agenda [15:04] New Topic: agenda [15:04] * Roll call [15:04] * Agenda [15:04] * Outstanding actions [15:04] * Mentoring update [15:04] * New topics [15:04] * ec2 failures update - bac [15:04] * Reviewer move to Friday? - bac [15:04] * Peanut gallery [15:05] [topic] outstanding actions [15:05] New Topic: outstanding actions [15:05] there are only two outstanding actions...both assigned to me. [15:05] but i'm batting .500 [15:05] * bac to write community reviewer and contributor policy and announce it on the list. -- done [15:06] i got that out, as promised, before leaving for UDS. [15:06] but remain stalled on: [15:06] * bac to define new doctest policy regarding what is "testable documentation". [15:06] i'll try to get to that next week [15:06] [topic] * ec2 failures update - bac [15:06] New Topic: * ec2 failures update - bac [15:07] actually, i was hoping to have mars give us an update but he's not here today [15:07] I can summarize maybe [15:07] sinzui reported in our standup that work has been done but it is still very much a problem [15:07] yes, gary_poster? [15:08] mars has been investigating the problems on ec2 test. As a workaround, he has disabled the Windmill tests on ec2. This is obviously very suboptimal, but considered to be less sub-optimal than ec2 break one out of four times, which I gather is the rough stat. [15:08] He will continue to try to figure out and address the underlying cauase so he can reinstate the windmill tests in ec2 [15:09] * bac saw about 60% failure last time he tried [15:09] note that windmill tests still run in buildbot [15:09] so this is a testfix risk [15:09] hopefully I didn't get anything wrong; that's what I understand. done. [15:10] gary_poster: when tests hang, and the script kills itself after timing out, does it send a farewell email? i think that would be a useful addition if not. [15:10] bac, that is on mars' backlog, and in fact will be one of the next things he does to address, yes. [15:11] ok, great [15:11] gary_poster: when were the windmill tests disabled? [15:12] bac, IIRC, Monday, with issues; issues were resolved yesterday [15:12] sinzui: did you see your tests disappear after that or before? [15:12] One disappeared yesterday afternoon [15:13] sinzui, bac, we'll need to bring this up to mars. Would you like me to take that action ite,? [15:13] item [15:13] we should all keep an eye on our ec2 tests and report to mars instances of them disappearing. [15:13] +1 [15:13] gary_poster: that would be great. [15:13] ok [15:13] [action] gary to follow up with mars on ec2 problems. [15:13] ACTION received: gary to follow up with mars on ec2 problems. [15:14] gary_poster, I think my branch was within a few hours of the mailmanlayer shutoff. the issue may be fixed [15:14] sinzui: ack wll pass along, thanks [15:14] [topic] * Reviewer move to Friday? - bac [15:14] New Topic: * Reviewer move to Friday? - bac [15:14] our pool of reviewers has been pretty hard hit lately with personel changes [15:14] https://dev.launchpad.net/ReviewerSchedule [15:15] right now, on fridays we only have adeuring working [15:15] would gmb or henninge entertain the idea of moving to friday? [15:16] bac, I believe adeuring is a Friday reviewer; I wonder if it's the best thing for hte bugs team to lose two devs on the same day... Not sure it matters though. [15:16] not that I would mind sharing reviews with gmb or henninge, but I think if we have somebody in a US timezone would be better [15:16] adeuring: yes it would be better [15:16] adeuring: but of the active reviewers in the americas we're spread pretty thin [15:17] right [15:17] bac, I'm happy to do it if deryck's okay with the change. [15:17] i could move to friday but that leaves an americas hole on tuesday [15:17] I don't mind moving to Fridays [15:17] I don't know if I want gmb and adeuring sidelined fridays [15:17] deryck: ok. henninge do you mind moving again? [15:18] bac: no problem at all [15:18] but, as was said, the problem is America [15:18] * henninge wonders if he should move west ... [15:18] henninge: let's not bring politics into the meeting [15:18] lol [15:19] But I was talking about the whole continent anyway ... [15:19] henninge: why don't you try friday and if seems we have a problem i'll then switch [15:19] It's in the way of any girl that wants to sail around the world... [15:19] [action] henninge to move to friday review EU review slot [15:19] ACTION received: henninge to move to friday review EU review slot [15:19] bac: on the same subject, note that there's a slot freein on monday EU. when i'm gone there will be no reviewer until abentley gets workin, and judging by the last few mondays there's a lot of activity around that time. [15:20] intellectronica: i know. i was hoping you'd change your mind before then. :) [15:20] intellectronica: where are you going, what did I miss? [15:20] intellectronica: you can always help out on mondays as a community reviewer! [15:20] wow, i managed to omit all those Gs. i'm not lolcat speaking, it was as accident [15:20] :) [15:21] intellectronica: what is the date for you leaving? [15:21] henninge: i'm leaving launchpad at the end of next week. [15:22] intellectronica: oh, too bad :( [15:22] ;) [15:22] in light of that, perhaps it would make more sense for henninge to move to eu-monday and i move to friday [15:22] henninge: what can i do, all those monday review shifts finally got to me [15:22] :) [15:23] henninge: let's talk about it off-line. [15:23] bac: ok [15:23] [topic] peanut gallery [15:23] New Topic: peanut gallery [15:23] any new topics? [15:23] orphaned commits [15:24] This is the what we have so far: https://wiki.canonical.com/Launchpad/QATeam/OrphanedCommits/10.05-devel [15:25] Ursinha, noted that this class of landing was not tracked with bug tags last release [15:25] hard to have a bug tag with no bug. [15:25] FWIW, If we talk about this here, I'd be +1 on this simply leading to someone taking an action item to raise this on the mailing list [15:26] In the examples I saw related to the registry team they were contributor branches that were not linked to bugs [15:26] * mars is guilty of submitting build system cleanups without bugs. [15:26] Do we want contributor work to always have a link to a bug/spec [15:27] i think reviewers should be on the look out for any MP that isn't linked to a bug. not saying it isn't sometimes appropriate but it should raise a flag. [15:27] sinzui: with contributor work it's even more important, because you want to help contributors plan and track the work before they start [15:27] intellectronica: agreed. so sinzui's suggestion is a good one, IMO. [15:28] Or in my case, remove part of the branch a week later because the feature is causing errors [15:28] but only 2 of the 10.05 orphans are from community [15:29] yes, there is another issue, but in the lp engineer case, flacoste assumes these are on the kanban board [15:30] sinzui: well, that does help. but if we're going to track qa with bug tags then it would make sense to *require* a bug or we lose all QA testing...unless we employ a separate tracking mechanism. [15:31] bac: if all we want is to track qa, filing a bug automatically sounds like a good idea [15:31] BjornT: at what point? [15:33] * gary_poster idly thinks of a flag for branches that says "addresses" rather than "fixes" [15:33] The tags are hard to visualise at the moment since the page is missing: http://people.canonical.com/~lpqateam/test-plan-report-10.05.html [15:33] bac: the script that scans commits and add qa tags probably [15:33] where addresses means that it is an incremental step towards a bug [15:33] but does not actually resolve a bug [15:34] * gary_poster suspects this is a can of worms [15:34] BjornT: so it should create a bug if none is linked to the commit? [15:35] henninge: if all we want is to track QA, yes, why not? [15:35] gary_poster, that's why I'm not going to touch it :) [15:35] :-) [15:35] sinzui: would you take an item to talk to QA and then make a proposal on the list, taking into account BjornT's suggestion? [15:35] BjornT: sounds like a good idea. Let's see if Ursinha likes it, too. [15:35] yes [15:36] thanks curtis [15:36] [action] sinzui to talk to QA about our QA tracking problem and create a proposal on the mailing list [15:36] ACTION received: sinzui to talk to QA about our QA tracking problem and create a proposal on the mailing list [15:36] any other topic? [15:36] 3 [15:36] 2 [15:37] 1 [15:37] thanks for coming everyone. [15:37] #endmeeting [15:37] Meeting finished at 09:37. [15:37] thanks bac [15:37] so, intellectronica, which team won you over and how? ;-) [15:38] henninge: the summer won me over. in a couple of weeks i'll be out of employment and ready to rock'n'roll :D [15:39] intellectronica: oh, because you said "leaving launchpad", I thought it was within the company. [15:40] intellectronica: so you have nothing new lined up yet? [15:40] henninge: nothing at all (other than some fun-related plans) [15:41] intellectronica: well, have fun, then! ;-D [15:41] henninge: thanks! [16:12] * Ursinha reads the backlog [16:12] bac: if all we want is to track qa, filing a bug automatically sounds like a good idea [16:12] we dropped out that idea because some of the commits aren't really bugs [16:13] henninge, ^ [16:13] BjornT: ^ ;) [16:13] Ursinha: I think that is the point. [16:13] Ursinha: To track even those commits that aren't really bugs. [16:14] henninge, that's how the script would initially work, but I dropped the idea after chatting with francis [16:14] henninge, and why would we need that? [16:14] Ursinha: we were talking about orphaned commits. [16:16] henninge, I see now... but the point is that the orphaned commits aren't all the commits the don't mention bugs, but only those which should and don't [16:16] s/the don't/that dont/ [16:19] henninge, I guess that not a lot of those supposedly orphaned commits are really orphaned, maybe we could fix that by adding the bug tag to the commit message? [16:19] henninge, making that a pqm requirement, maybe [16:21] Ursinha: I am not sure they have bugs. It did not sound like that. [16:22] henninge, and they should? [16:22] * henninge has little own experience, he always files bugs... [16:22] henninge, that's why I love you :P [16:23] * henninge blushes [16:23] Ursinha: well, jtv trained me that way ... [16:23] henninge, I love him too :) [16:24] Ursinha: The question was "Should all commits have a bug so they are qa'ble". [16:24] henninge, my personal opinion is no [16:24] I was not aware of the definition for orphaned commits that you just gave. [16:24] "should have bug bug don't" [16:25] henninge, the idea of having a bug attached to a commit is to track the qa you have to do on it [16:25] Ursinha: you are the authority on QA here, so if that is your opinion, we are fine, I guess ... ;) [16:25] on the bug, you mean [16:25] of course having a bug for real fixes is needed to track problems and future reference [16:25] but now we really need to care about fixes that need qa, those need bugs [16:26] so the rule should be "If the fix will need qa, attach a bug. If it doesn't a bug is optional?" [16:27] Ursinha: don't all fixes need QA? [16:28] henninge, I thought so, but it turns out some are just adjustments that not necessarily can be qaed [16:28] Ursinha: I'll put that on the agenda for next week's meeting. [16:29] henninge, well, speaking roughly in practical qa terms, yes, but I'd like to have bugs to all problems we have, even the ones that result in commits we can't test [16:29] but that's my wish :) [16:30] or the ones that don't even need commits to lp tree to be fixed :) === salgado is now known as salgado-lunch === salgado-lunch is now known as salgado === gary_poster is now known as gary-afk === gary-afk is now known as gary_posrer === gary_posrer is now known as gary_poster === salgado is now known as salgado-afk [22:33] thumper, rockstar: up for a reviewers meeting? [22:33] bac, havin' our standup right now. [22:33] I might have to skip it this week [22:33] rockstar: why do you schedule the standup during our meeting time? :) [22:34] thumper: ok. not much said but talk about broken ec2 and rearranging of OCR schedules due to all the people moving around. [22:34] next week...