[04:00] #startmeeting [04:00] Meeting started at 05:01. The chair is barry. [04:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [04:01] hi everybody and welcome to this week's asiapac reviewers meeting [04:01] who's here today? [04:01] me [04:01] me (kind of) [04:01] Well this is looking embarrassing. [04:01] I'm here. [04:01] sinzui: i was going to say, if it was just you and me, we should do it when the sun is up :) [04:02] hi [04:02] jml: hi [04:02] spiv, thumper hi [04:02] mwhudson: are you around? [04:02] thumper is unwell and is off [04:02] thumper: :( [04:03] [TOPIC] agenda [04:03] New Topic: agenda [04:03] mwhudson last said "afk for a little while" 4 minutes ago [04:03] jml: okay thanks, oh well [04:03] * Roll call [04:03] * Next meeting [04:03] * Action items [04:03] * Queue status [04:03] * Mentoring update [04:03] * Review process [04:03] * '''Pre-imp calls are falling by the wayside''' (gmb) [04:03] [TOPIC] next meeting [04:03] New Topic: next meeting [04:03] so, is this a good time for y'all? [04:03] for me, yes. [04:04] normally, but I'm somewhat uncommunicative right now [04:04] This is fine for me too (same timezone as jml). [04:04] thumper: no worries. anybody know if this is a good time for jamesh? [04:04] but I'd be equally with something up to four hours earlier or two hours later, if things need rescheduling. [04:04] barry: it's 11am in WA right now. [04:04] barry: it *ought* to be a good time for him :) [04:05] :) [04:05] (the WA that's on the west coast of Australia, not the one on the west coast) [04:05] jml, naw, i was thinking maybe 1 hour earlier, but this works just fine for me too. in daylight saving's its 11pm, and i fake-tivo the daily show so it's all good [04:06] :) [04:06] cool, let's keep 0300utc then [04:06] +1 [04:06] [TOPIC] action items [04:06] New Topic: action items [04:06] * (continued) thumper to report on pending-reviews killer in LP [04:06] barry: sorry, it's coming [04:07] but nothing else to report right now [04:07] thumper: no worries, we'll just keep continuing it if you want [04:07] yeah [04:07] thumper: just want to be sure you don't want to ditch the whole thing :) [04:07] jamesh: hi! [04:07] hi [04:07] barry: not just yet [04:08] boo! [04:08] cool [04:08] * barry is frightened [04:08] just for completeness, here are the ameu action items: [04:08] * gmb to hack review-submit to enforce 800 line limit. [04:08] * schwuk to work with mwhudson to get instructions for running loggerhead onto the wiki [04:09] any comments or should we just move on? [04:09] barry: I don't think enforcing the 800 limit is a good idea. [04:09] barry: unless that means "show a warning" [04:09] me neither [04:09] jml: 800 + shame [04:09] yeah, it's mostly so you have a warning [04:10] i really think the option to override should be --isuck [04:10] rather than say enforce [04:10] say 'warn' perhaps ? [04:10] jml: the 800 line flag really means you need to send with the -r switch (have a review lined up) [04:10] if you mean warn :> [04:10] * barry speaks as someone who sucks often [04:10] sinzui: perhaps. [04:10] yes, the point is you really need to arrange things with a reviewer if your >800 lines [04:11] I think that's fair enough. I still think you should be able to override whatever policy the tool has by default. [04:11] i'm happy to reword this action item to be clearer about it, but i think gmb knows what we mean [04:11] jml: definitely [04:12] anyway, review-submit will be an addon to send eventually. [04:12] barry: cool. now I know what gmb means, I'm happy :) [04:12] cool! [04:12] [TOPIC] queue status [04:12] New Topic: queue status [04:13] some bonehead's got two branch's in the general queue [04:13] :) [04:13] :-D [04:13] just to be clear [04:13] are we supposed to send emails *and* hack wiki? [04:14] jml: i don't know anymore. i do, but that's just me [04:14] hah! [04:14] ok :) [04:14] i fully intend to ask tomorrow's on-caller to review these branches though [04:14] someone should patch submit-review to edit the wiki [04:14] there's also at least one email sent that's not on the wiki page and has no assigned reviewer [04:15] lifeless: yes. it's getting close to that already. [04:15] jml: it's still the responsibility of the dev to get his branches reviewed [04:15] *nod* [04:15] * barry still likes the pending-reviews page tho [04:15] it tells me we have 5 pink branches [04:16] i know about stub's two [04:16] jml: what's up with your use-inmemory-proxies branch? [04:16] barry: shelved, for the time being. [04:16] jml: should we move that to wip? [04:17] barry: yeah, probably. [04:17] jml: i'll let you do that :) [04:17] ok [04:17] thanks [04:17] any other queue-related comments? [04:18] none from me [04:18] [TOPIC] mentoring update [04:18] New Topic: mentoring update [04:18] i don't think there are any asiapac mentors or mentorees [04:18] are there any of you guys who /aren't/ already reviewers? [04:19] That was to requirement for mwhudson to emigrate. [04:19] barry: all australasian LP hackers are reviewers [04:19] we even have a non-LP hacker who's a reviewer :) [04:20] jml: excellent! [04:20] a veritable superabundance of reviewers. [04:20] you need to hire more grunts [04:20] [TOPIC] review process [04:20] New Topic: review process [04:20] * '''Pre-imp calls are falling by the wayside''' (gmb) [04:21] What does that mean? [04:21] people 'round here slacking on pre-impl calls [04:21] have you noticed that? is it a problem? and if so what do we do about it? [04:21] here's specifically what gmb says: [04:21] * '''Pre-imp calls are falling by the wayside''' (gmb). Since we've adopted the on-call review process I've noticed that less and less people (including myself) have been having pre-imp calls before they start work. Partly, I figure that this is because the on-call process allows the developer to talk about their implementation decisions with the reviewer and so helps to substitute for a pre-imp in a lot of ways. However, I've had one [04:21] branch land this cycle that happily had r= from the reviewer but which, had there been a pre-imp, might not have landed in the form it did (which would have avoided me needing to ask for an RC). Should we be enforcing the requirement to have pre-imp calls, or at least be more scrutinous of those branches that don't? [04:22] oopsie [04:22] I've observed that people aren't rigorous about calling someone before they begin work on a branch. [04:22] I don't think that's a problem. [04:23] I think I've done only two pre-impl calls with people in the last month (and only one was actually voice) [04:23] let me ask: how have the quality of branches been lately when first submitted for review? has it been getting better, worse, or about the same? [04:23] the way we'd measure it being a problem is if reviewers were asking for branches to be re-implemented :) [04:24] barry: that's hard to discern. The code I get to review is generally very clear. [04:24] i'm curious about the rate of needs-reply branches. [04:24] the sample is limited, because we are increasing the size of the review team. [04:24] jml: many of the branches i've reviewed lately have been merge-* right off the bat (i.e. very good quality) [04:25] something that's definitely happening is that i'm basically only reviewing things from the code team [04:25] needs-reply doesn't always mean lower quality [04:25] and occasionally jamesh [04:25] jml: no, you're right, but it's an indicator that the branch maybe wasn't quite ready to code yet [04:25] I often flag things needs-reply because I don't understand a part of the problem domain. [04:25] mwhudson: interesting [04:26] yeah, I'd agree w/ mwhudson on that. [04:26] is that basically geography? [04:26] I also almost always get code or bzr ppl to do my pre-impl calls [04:26] well, it isn't that surprising that NZ reviewers mainly have overlap with NZ and AU coders [04:26] indeed [04:27] jamesh: yep [04:27] for pre-implementation calls, it is even more of an issue [04:27] since that's real time when reviewing might not be [04:27] cool, so basically you don't feel that the current rate of pre-impl's is a problem? [04:27] no :) [04:28] i don't know how much of a problem this is, it reduces the cross-team spread of knowledge a bit [04:28] barry: no [04:28] mwhudson: good point. [04:28] while there might not be as many formal pre-impl calls as there once were, people do consult others on IRC about implementation issues [04:28] (at least in this time zone) [04:29] if it's not a problem, i'm definitely not going to try to "fix" it then :) [04:29] thanks, this has been good feedback [04:29] so... that's it from me. we have 15 minutes if anybody has anything not on the agenda, or messages to convey to ameu [04:30] * jml has nothing on topic [04:30] cool. time for the colbert report then :) [04:30] :) [04:30] #endmeeting [04:30] Meeting finished at 05:31. [04:30] thanks everyone! g'night :) [04:31] thanks barry [04:31] thanks barry [04:31] (and sorry for being late) [04:31] mwhudson: no worries. talk to you guys soon... === cprov is now known as cprov-afk === cprov-afk is now known as cprov === mrevell is now known as mrevell-lunch === mrevell-lunch is now known as mrevell === vednis is now known as mars === salgado is now known as salgado-lunch === salgado-lunch is now known as salgado === Seveaz is now known as Seveas === mwhudson_ is now known as mwhudson === mwhudson_ is now known as mwhudson