[14:02] oi === bigjools-afk is now known as bigjools_ [14:02] me [14:02] me [14:02] I think some peoples' clocks are running a touch on the slow side... [14:02] me [14:03] kiko, SteveA, Rinchen: Meeting time? [14:03] Time is an illusion. [14:03] Lunchtime doubly so. [14:03] time is a corportist plot [14:03] hello hello [14:03] welcome to the launchpad meeting [14:03] me [14:03] and hmm how does one drive this bot! [14:03] hang on first ;) [14:04] okay, got some docs up [14:04] #startmeeting [14:04] Meeting started at 14:04. The chair is kiko. [14:04] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [14:04] me [14:04] so who is here this week, apart from leonardr? [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me === jt1 is now known as jtv [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me [14:04] me === EdwinGrub is now known as EdwinGrubbs [14:05] me [14:05] me [14:05] me [14:05] me [14:05] me [14:05] me [14:05] * kiko looks for danilos, cprov [14:05] SteveA, and stub? [14:05] danilos is off today [14:05] me [14:05] hey mthaddon, glad you could make it [14:06] hey kiko [14:06] very well [14:06] Rinchen is out because he pulled an all-nighter [14:06] and that means you get me as your most charming host today [14:06] so let's move on. that was the roll call! [14:06] kiko: I already said "me" [14:06] * [14:06] Roll call [14:06] * [14:06] Agenda [14:06] * [14:06] Next meeting [14:06] * [14:06] Actions from last meeting [14:07] * [14:07] Oops report (Matsubara) [14:07] * [14:07] Critical Bugs (Rinchen) [14:07] * [14:07] Bug tags [14:07] * [14:07] Operations report (mthaddon) [14:07] * [14:07] DBA report (stub) [14:07] * [14:07] Sysadmin requests (Rinchen) [14:07] * [14:07] A top user-affecting issue (mrevell) [14:07] SteveA, sure, but stub didn't. [14:07] and to wrap that up, Blockers [14:07] TOPIC next meeting [14:07] I recommend pasting from the raw text of the wiki page. [14:07] oh bar [14:07] ahem [14:07] [TOPIC] next meeting [14:07] me [14:07] New Topic: next meeting [14:07] hello there [14:07] sorry for being late [14:07] SteveA, I'm not very good at this :-) [14:08] so for next meeting, same time, same place? who's here, who's not? [14:08] I'll be here. [14:08] I'll be at a sprint in cocoa beach, along with the foundations and lpcomm teams [14:08] I expect we'll still make the meeting [14:08] foundations and commercialization will be sprinting [14:09] so confirming, sprinters will make the meeting? [14:09] yeah [14:09] great [14:09] anyone on vacation? jtv -- danilos? SteveA -- stub? [14:09] kiko: yes, danilos is off today (see above) [14:09] I'm in another meeting right now and I'm attending this meeting [14:09] jtv, I meant next week, sorry. [14:09] stub should be here [14:09] so attending a sprint is no excuse :-) [14:09] I'll give him a buzz [14:10] [AGREED] meeting next week, same time, thursday. sprinters should make it. no other absentees forseen. [14:10] AGREED received: meeting next week, same time, thursday. sprinters should make it. no other absentees forseen. [14:10] [TOPIC] Actions from last meeting [14:10] New Topic: Actions from last meeting [14:10] mthaddon, first topic is whether https://staging.launchpad.net/successful-updates.txt has been linked to from somewhere. [14:11] stub is having network problems [14:11] he's working on fixing it [14:11] second topic is whether launchpad-dependencies was updated by salgado [14:11] kiko, not currently - I have a script on devpad that checks it and emails me if it's not completed in the last 24 hours, which I will now change to email the launchpad list [14:11] third topic is whether salgado investigated putting launchpad-dependencies in /sourcecode -- or if at least a bug was filed. [14:12] what's it about the update? [14:12] mthaddon, okay, great. can we link to successful-updates.txt at least from LaunchpadProductionStatus, perhaps prominently? [14:12] kiko, sure [14:12] my box is half-crashed and I didn't see the beginning of the meeting [14:12] (can't launch firefox either. am running xchat remotely) [14:12] salgado, whether it was done or not -- I believe the answer is yes, to include lxml and that other one [14:12] salgado: using hardy? [14:13] kiko, yes, even mhonarc has been included [14:13] SteveA, no, it's diskless gutsy, long story. [14:13] salgado, good job. [14:13] salgado, I noticed that today you noticed a branch using a package that isn't required [14:13] I think we need a process to have people preemptively request packages for inclusion [14:13] that's actually from zope [14:13] because otherwise it's always a mad scramble [14:13] I see [14:13] kiko, it's a problem with sys.path, it seems [14:14] * SteveA quotes Ghostbusters: "Everything was fine until the grid was shut down by diskless here." [14:14] me [14:14] quick poll: are people aware of the fact that they need to do something special if you want to require a new python (or otherwise) package? say yes if you know that you need to. [14:14] yes [14:14] kiko: I think you have a good point there about packages, I've noticed that I have so many dev packages on my machine I don't feel confident I will notice if something new needs to be installed - this is why the lxml thing didn't affect me [14:14] "yes it is true. This man has no disk" [14:14] yes [14:14] yes [14:14] yes [14:14] yes [14:14] yes [14:14] yes [14:14] yes [14:14] yes [14:14] yes [14:15] yes [14:15] yes [14:15] yes [14:15] yes [14:15] yes [14:15] yes [14:15] yes [14:15] yse [14:15] n/a [14:15] yes [14:15] yes [14:15] yes [14:15] okay. so gmb and allenap don't know. [14:15] In theory, if we miss a required package branches that depend on it won't get through pqm [14:15] Yes. [14:15] I know. [14:15] I'm lagged. [14:16] while it's true that the branch won't get through PQM [14:16] it's also true that that's too late and then we need to rush IS into installing stuff [14:16] which might need to be backported -- see abel's issue with lxml one release ago [14:16] kiko: are we still running dapper on production? [14:17] carlos, yes. [14:17] ok [14:17] so how about we add this as a weekly topic, next to the "pending IS requests" topic? at least then people can think about it once a week. [14:17] anyone opposed? [14:17] is it common enough? [14:18] kiko: does that differ from pending RT requests? [14:18] The reason people don't think about it is that it is uncommon I think. [14:18] yeah, it's uncommon [14:18] jamesh, there might not be an RT request for the package -- I want to make sure people are aware of it [14:18] it's just that three new dependencies were introduced in the last vcycle [14:18] well, it's uncommon but still not that rare -- twice this cycle, for instance [14:18] or 3 times [14:19] * bigjools_ will have one too [14:19] 4 new deps, but I actually did only two updates to the package [14:19] Spread over all our devs, that comes to once every 2-4 months an individual dev will do it. [14:19] * sinzui suspected bigjools_ will have one [14:19] it's harmless to have the topic and just say "no." [14:19] So a reminder such as in the meeting would be good. [14:19] agreed. [14:19] I'll amend the template. thanks! moving on.. [14:20] the last topic was about Rinchen and tim figuring out who can attend the meeting from down under [14:20] I think it's really a big deal that australians/new-zealanders can't make the meeting [14:21] and Joey and I have talked about changing the time to make it less depressing for mthaddon and more humane for australians [14:21] cycle the meeting times? [14:21] however, that leaves us with the problem of being inhumane with jtv and stub and possibly danilo and BjornT [14:21] kiko: it's tough. we have two reviewers meetings for this reason, with me attending the asiapac one at 10pm localtime [14:22] barry, I know -- but we can't really have two meetings for launchpad, so rotation might be the best approach. [14:22] I'm happy to not turn up to meetings if it makes things easier :-) [14:22] we're still talking about it [14:22] thanks stub [14:22] okay, that's all on this topic -- we'll talk about it again next week when Rinchen is awake [14:22] so [14:23] [ACTION] kiko to add the lp-dependencies question to the MeetingAgenda template [14:23] ACTION received: kiko to add the lp-dependencies question to the MeetingAgenda template [14:23] [ACTION] Rinchen and kiko to figure out how to manage lp-bzr's absence from the meeting; consider rotation or time-shift [14:23] ACTION received: Rinchen and kiko to figure out how to manage lp-bzr's absence from the meeting; consider rotation or time-shift [14:23] salgado, what about putting the package in sourcecode/ ? [14:23] isn't aaron on that team? [14:24] statik, he is, and the fact that tim hasn't told him to come to the meeting kinda underlines my point. ;) [14:24] kiko, as I said, I'm opposed to that, so I may not be the best person to move this forward [14:25] tim doesn't even think about the meeting -- which means he misses out on OOPSes, policy changes, critical bug discussions, etc [14:25] aaron is in Sydney at the moment [14:25] anyway [14:25] topic for next week [14:25] kiko, I didn't even notice this had been assigned to me during the last meeting [14:25] hmph. okay, flacoste, can you figure that out later with salgado then. [14:25] ? [14:25] kiko: yes [14:26] moving along [14:26] point of information [14:26] [TOPIC] Oops report (Matsubara) [14:26] New Topic: Oops report (Matsubara) [14:26] I spoke with lifeless about putting our dependencies packages' source in sourcecode/ and he's very in favour of the idea [14:26] we discussed it at some length [14:26] end of POV [14:26] thanks SteveA! [14:26] um, POI [14:26] I've setup a oops report for today updated every hour. The url is in #lp-code topic. I'm checking that often and following up with people on IRC about oops related to the rollout. [14:26] that's all from me. back to you kiko [14:27] matsubara, that's a great change. I love it. [14:27] kiko: 1am is a perfect time for a meeting! [14:27] matsubara: that rocks [14:27] matsubara, I reloaded today and was able to complain already about my favorite OOPSes. [14:27] matsubara: YES! \0/ [14:27] Hobbsee: That depends on how drunk you are allowed to be [14:27] without even having to complain about there not being an OOPS report!! [14:27] stub: *grin*. there is that [14:27] matsubara, any OOPSes outstanding you'd like us to nitpick? [14:28] kiko: not really an OOPS, but a critical one that I found while testing +editemails. I can tell you details on #lp-code [14:28] On this topic, it would have been nice if we noticed a significant drop in timeout oopses after turning on the BBWC but I couldn't see anything particularly noticeable. [14:28] matsubara, yeah. so all OOPSes from today's report are being addressed already? give us at least a list so we don't double-check. [14:29] stub: are we sure it's turned on? [14:29] kiko: I'm still checking the +editemails one. [14:29] SteveA: I have no way of confirming except to ask the admins. [14:29] kiko: the soyuz queue one is on bigjools plate [14:29] stub, that's a pity - I feel like the site is a little quicker to render pages, but I may be kidding myself [14:29] IME, bbwc only improves latency, not throughput [14:29] and jamesh will prepare a patch for the openid one [14:29] SteveA, we're sure it's turned on [14:30] okay, thanks! [14:30] bigjools_: I hate hardware [14:30] mthaddon: is it something that is turned on for all disks/partitions, or just ones we choose? [14:30] mrevell, do you have input for critical bugs? or should I skip that topic? [14:30] stub: aye [14:31] SteveA, just for the ones we choose (well, that ones that have it available) [14:31] kiko: I think matsubara does [14:31] mrevell's pinging out, so I'm skipping that. [14:31] oh, really? [14:31] [TOPIC] critical bugs (matsubara?) [14:31] New Topic: critical bugs (matsubara?) [14:31] kiko: no critical ones from Rinchen. The only one is the private one I told you earlier [14:31] thanks matsubara [14:31] [TOPIC] bug tags [14:31] New Topic: bug tags [14:31] and the soyuz pubkey which bigjools fixed [14:32] I need r-c for that [14:32] no proposed tags this week, so moving on [14:32] my favorite topic now! [14:32] [TOPIC] Operations report (mthaddon) [14:32] New Topic: Operations report (mthaddon) [14:32] 1.2.1 released [14:32] Have some documentation updates following the release, and expecting a "second" rollout today or tomorrow [14:33] Er, not much more to say from me (except for known PPA issue) [14:33] hoping we can do that late today [14:33] not too late I hope... [14:33] but that depends mostly on getting the OOPSes fixed today [14:33] matsubara, are they going to get fixed today? [14:34] sorry, I meant the slave_scanner issue [14:34] sure. [14:34] thanks! [14:34] [TOPIC] DBA report (stub) [14:34] New Topic: DBA report (stub) [14:34] Do we know that we will need to do a second rollout the following day before the OOPS reports come in? If so, should we consider letting rollout times slip by a day or three to improve overall reliability? [14:34] (oops.... old topic. Must type fasteer) [14:35] Priority of replication has been bumped in favour of making LP more useful for auth by Landscape and other systems with similar use cases. We still need to proceed on replication, but the timeframe isn't as critical now given we are getting hardware upgrades. [14:35] Nothing else to report. [14:35] on the oops issue -- are these oopses ones we could have discovered on edge? [14:35] or by some other means, before the roll-out? [14:35] SteveA, in the soyuz case, yes on at least one count. [14:35] matsubara, what do you say of the others? [14:36] the openid one is quite old, it's not specific to the rollout [14:36] we couldn't find the soyuz config one because it's a specific lpnet config issue [14:36] SteveA, the portlet-details one is on bounties, so... moving back on topic, though -- stub, okay. have you and SteveA re-discussed using carbon for the staging DB? [14:37] not sure yet about the +editemails one, but I think it's an old one as well [14:37] there are various discussions going on about shifting around hardware [14:37] Carbon for staging has been discussed via email. Carbon for staging won't be an option for long term as I have it earmarked for the first replica lp database. [14:37] however, we could use it in the mid term [14:38] particularly as you're shifting down the immediate priority of replication [14:38] right, that's what I'm asking if we talked about. [14:38] kiko: this depends on various other discussions about hardware [14:38] kiko: so, it's still going on. [14:38] SteveA, okay, thanks. please update us on this next week. [14:38] I expect a resolution in a week or so [14:38] It also depends on how soon landscape need two databases. [14:39] [ACTION] SteveA and stub to discuss options for staging DB and report back [14:39] ACTION received: SteveA and stub to discuss options for staging DB and report back [14:39] [TOPIC] Sysadmin requests (kiko standing in for joey) [14:39] New Topic: Sysadmin requests (kiko standing in for joey) [14:39] anyone have any ones that they want done asap? [14:40] I have RT #29797 which is blocked itself on packaging work [14:40] any others? [14:40] bigjools_, please ping me today so we talk that over [14:40] sure [14:40] [AGREED] RT 29797 priority for bigjools [14:40] AGREED received: RT 29797 priority for bigjools [14:41] [TOPIC] A top user-affecting issue (mrevell) [14:41] New Topic: A top user-affecting issue (mrevell) [14:41] howdy [14:41] This week's issue deals with merging accounts. At present, it's not easy to find how to merge two accounts. [14:41] The fix for bug 30439 gave us the "Claim this account" link on profile pages. [14:41] Launchpad bug 30439 in launchpad "Link to page for merging accounts is well-hidden" [Medium,Fix released] https://launchpad.net/bugs/30439 [14:41] However, for someone who is logged-in there's still no link to merge with another account. [14:41] At least, no link on the profile page. [14:41] Support requests suggest that we could make this easier to find. [14:41] I've reported bug 185486 in response to the most recent help request on this subject. [14:41] Launchpad bug 185486 in launchpad "Merge account option should be on profile page" [Undecided,New] https://launchpad.net/bugs/185486 [14:42] The Actions menu is somewhat busy already. Where else in someone's profile pages do you think we could place a link to merge accounts? [14:42] Thanks kiko [14:42] mrevell, the 'Claim this account' is a link to +requestmerge when the person is logged in [14:42] mrevell, so if I'm at ~kiko/ you'd like the option of initiating a merge from there? [14:42] salgado, sure, but it only appears on the pages of people who are inactive [14:42] mpt have an opinion? [14:42] kiko: I think so. it seems like the place I'd look for it and, it seems from talking to users, that its current location isn't obvious to them [14:43] mrevell, the only place I know to look for is on /people/ [14:43] yeah [14:43] yeah, I agree that's a problem. but what to do with the link is a bother [14:43] mrevell, you chase mpt down to talk it over? [14:43] Well, I'll discuss the with mpt after the meeting and anyone else who is interested [14:43] kiko: will do [14:43] kiko, mrevell discussed it with me yesterday or so [14:43] [ACTION] mrevell and mpt to discuss https://launchpad.net/bugs/185486 [14:43] ACTION received: mrevell and mpt to discuss https://launchpad.net/bugs/185486 [14:43] Launchpad bug 185486 in launchpad "Merge account option should be on profile page" [Undecided,New] [14:44] * stub notices his battery is low and has lost his power adapter [14:44] The person page is one of the pages for which the Actions menu problem will be solved [14:44] so, I'll have to find *some* way of fixing it :-) [14:44] [TOPIC] Beginning the UI review mentoring (mpt) [14:44] New Topic: Beginning the UI review mentoring (mpt) [14:44] thanks kiko [14:44] A couple of weeks ago I proposed mentoring reviewers on reviewing UI stuff [14:44] and the response was positive, which is great to see [14:45] So, I'm looking for one or two volunteers to start this next week :-) [14:45] volunteer reviewers, that is [14:45] come on! [14:45] mpt: i'm interested, but probably not starting next week [14:46] mpt: same as barry, as we'll be sprinting [14:46] we'll be focusing on non GUI stuff at the sprint [14:46] "I don't bite ... hard ..." -- Austin Powers [14:46] mpt: ok, i volunteer [14:46] * kiko looks at BjornT's team! [14:46] intellectronica! [14:46] thanks intellectronica [14:46] mpt: I'll be interested once I'm graduated as a reviewer. [14:47] Probably best to do only graduates [14:47] But I think >1 mentor would slow things down somewhat. [14:47] gmb: you don't need to be graduated for this [14:47] mpt: why? [14:47] exactly [14:47] because 2 levels of mentoring would slow down reviews too much [14:47] Especially when you consider on-call work. [14:47] makes sense [14:47] One other? [14:47] i thought it was an hands on training [14:47] salgado? [14:47] not a new level of mentoring [14:47] Considering that not all reviews will touch the UI [14:48] salgado: will be sprinting [14:48] kiko, sprint [14:48] ah, right. [14:48] mpt, let's do with one for now since we seem short of reviewers who are free. [14:48] ok [14:48] next week we'll get another [14:48] very well [14:48] also [14:48] flacoste, you mentioned training [14:48] UI design training is a separate thing, which I'm scheduled to organize next month. [14:48] I had a topic to discuss, but I'll leave it for next-week -- but will talk to mpt about it meanwhile. topic was Especially when you consider on-call work. [14:48] gar. [14:49] What should we do about pillar +portlet-details (kiko) [14:49] thanks kiko [14:49] that was the topic. [14:49] very good [14:49] do we need VNC or something [14:49] like, how do we effectively practice UI review mentoring? [14:49] how would VNC help, though? [14:49] screen sharing [14:49] I get the feeling it's more about having conversations over mockups [14:50] rather than say "the thing at the top, yeah, there, that thing" [14:50] you just say "see there" [14:50] SteveA, I think discussing diffs over the phone (+ IRC for URLs) will be sufficient in most cases [14:50] diffs? [14:50] not screens? [14:50] or screenshots [14:50] maybe there are tools to make getting screenshots more efficiently [14:50] we've learned that a fast turn-around is important [14:50] SteveA, this is mentoring of branches. Branches generate reproducible behavior at URLs that the reviewer and the mentor can look at. [14:51] and UI discussions are dynamic [14:51] mentoring of branch reviews, I mean. [14:51] so, please consider using technology to help have richer communication about the UI [14:51]