=== salgado-afk is now known as salgado === cprov is now known as cprov-away === mrevell_ is now known as mrevell-lunch === bac_lunch is now known as bac === mrevell-lunch is now known as mrevell [15:00] #startmeeting [15:00] Meeting started at 09:00. The chair is barry. [15:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [15:00] hello everybody and welcome to this week's ameu reviewers meeting. who's here today? [15:00] me [15:00] mö [15:01] me [15:01] me [15:02] me [15:02] wow, light attendance today [15:02] me [15:03] allenap, BjornT, EdwinGrubbs, gmb, rockstar ping [15:03] me [15:03] me [15:04] me [15:04] danilos: ping [15:04] abel? [15:04] [TOPIC] agenda [15:04] New Topic: agenda [15:05] me [15:05] * Roll call [15:05] * Graduations [15:05] * Is OCR still relevant in a world with merge proposals? (barry) [15:05] * Sabbaticals? (barry) [15:05] * Do we need a standard cover letter template for merge proposals? (barry) [15:05] * https://pastebin.canonical.com/11767/ [15:05] * If there's time, the old boring script [15:05] * Next meeting [15:05] * Action items [15:05] * Mentoring update [15:05] * Queue status [15:05] [TOPIC] * Graduations [15:05] New Topic: * Graduations [15:05] so bac recommends graduation for rockstar, but since he blew us off today... :) [15:05] me [15:06] naw. so rockstar graduates. congratulations rockstar-in-abstentia [15:06] go rockstar! [15:06] whoo! [15:07] yeah [15:07] it should be a rule [15:07] rockstar's reviews have been very attentive and he asks lots of good questions. he's ready to be graduated. [15:07] if you miss the reviewer meeting of your graduation [15:07] you get to be on probation for one more month :-) [15:07] he needs to work on his attendance, though! [15:07] that might seems unfair for the mentor [15:08] flacoste: i'm ok with another month of probation if you'll pick up the mentoring [15:08] but mentoring is also about instigating reliability in the mentee :-p [15:08] flacoste: right! or we can just publicly mock him relentlessly at his graduation announcement [15:08] barry: that's probably funnier [15:08] let the mocking begin [15:08] me [15:09] late but still me [15:09] maybe he should attend meetings even on days when we don't, for a month [15:09] * barry thinks he should /run/ the meetings (including asiapac) for a month [15:10] moving on, but don't let that stop you if you come up with a good zinger... [15:10] [TOPIC] * Is OCR still relevant in a world with merge proposals? (barry) [15:10] New Topic: * Is OCR still relevant in a world with merge proposals? (barry) [15:11] so, i've had a few people mention that ocr seems "weird" now that we have merge-proposals. so i'll open the floor to discussion: do we want to keep ocr now and if so, in its current form or change it? [15:11] weird, how? [15:11] in that it's less about getting a review RIGHT NOW than it is spending time popping things off the top of the queue [15:11] i don't see how it's weird. you still want to encourage high velocity, and you still need some mechanism for allocation of reviews [15:11] I say yes, they are relevant. I know who will be taking the review. I have someone I can discuss the matter with *before* the merge proposal it made [15:12] I agree with... well, all of the above. [15:12] OCR is still very relevant for velocity [15:12] barry: Otherwise MP just becomes the new PR. [15:12] gmb, agreed [15:12] It encourages people to *seek* a review rather than letting branches sit around. [15:12] without it we'd go back to the days of 2+ days reviews, chasing reviewers and getting frustrated [15:13] and who would allocate? [15:13] i do see people creating MPs but then not showing up to ask for OC reviews. so i tend to pop things off the MP stack and then hope they are around for questions. [15:13] * mars didn't know there was a MP stack [15:13] bac: that shouldn't happen. it's a coder's responsibility to make sure that they're branch gets reviewed [15:13] * bac wishes there were a real stack instead of an unordered list [15:13] I sometimes create an MP so that I can make sure I ask for a review *tomorrow*, but I don't expect someone to JFDI if I don't ask. [15:14] just like when using PR if no one asks for an OCR we would pull things off the general queue. same, same. [15:15] btw, i'm not saying /i/ think we should get rid of ocr. i'm just bringing up some things i've heard so we have a chance to discuss them [15:15] okay, so does anybody disagree with keeping ocr the way we have it now? [15:16] 5 [15:16] bac: yes, but that should be the exeception. the review process is so much more powerful when you can chat about it in real time [15:16] 4 [15:16] 3 [15:16] 2 [15:16] intellectronica: agreed. just sharing what often happens on tuesdays [15:16] 1 [15:16] cool, thanks [15:16] [TOPIC] * Sabbaticals? (barry) [15:16] New Topic: * Sabbaticals? (barry) [15:17] +1 for sabbaticals [15:17] on a related note, there's been some requests for sabbaticals and/or half day review slots. we have pretty good coverage for euro and america these days, so perhaps we could institute some kind of rotation [15:17] everyone needs a break from routine once in a while [15:17] thoughts? [15:17] i think i originally introduced the idea then promptly forgot about it. [15:18] i was thinking about it a bit since i made the request [15:18] i think it's a good idea so reviewers don't get burned out. [15:18] +1 [15:18] I recall the backlogs we have when foundations or registry team are sprinting. [15:18] i think that regulating this too much will not work well, because reviewers know best when they really need a break [15:18] perhaps instead, each reviewer should have an allowance [15:18] Sabbaticals are fine so long as someone is sill available to do the reviews. [15:18] much like leave days from work [15:18] sinzui, right [15:18] sinzui: But we have some overlap now. Those backlogs occurred because foundations/registry were splinting and we didn't do anything to deal with the hole. [15:19] sinzui: +1 sensible [15:19] splinting? [15:19] Whatevr. [15:19] sinzui: i don't think that's a problem these days. we have many reviewers [15:19] gmb: you're turning japanese? [15:19] intellectronica: except in exceptional circumstances perhaps, but we can deal with those [15:19] intellectronica: Appalentry so. [15:20] * gmb wonders who we should offend at next week's meeting [15:20] i suspect if we have the concept of sabbaticals they will rarely be used. but it's good to know there is a socially-approved mechanism for declaring you need a break. [15:20] so, not counting mentats, we have 2 more reviewers than we have ocr slots, so there is room for breaks [15:20] intellectronica: correct, yet if all the sabbatical takers abandon Wednesday, there is a problem. I am suggesting the we need to move who is working the slots now that we can afford to give reviewers a break from reviewing [15:20] what if every reviewer had, say, 10 leave shifts a year, which they are responsible to coordinate, make sure someone is there to replace them, etc? [15:21] sinzui: We need floating reviewers. [15:21] sinzui: yes, just like normal leave. you don't allow everyone to take time off at the same time. you coordinate [15:21] gmb: i don't mind floating (i think :) [15:21] barry: Likewise, once al-maisan has graduated. [15:22] sounds too formal for me. i'd prefer people just announce they need a month off, whatever, and barry busts anyone that seems to be abusive. [15:22] how long should a sabbatical be? 1 week, 1 cycle, up to the reviewer? [15:22] gmb: cool [15:22] we're all consenting adults - please, leave it to common sense [15:22] i think that 2-8 weeks [15:22] don;t wrap it up in more red tape [15:22] bigjools: +1 [15:22] bac: i am all for more opportunities to wield my power [15:22] +1 [15:23] +1. my suggestion was just because i didn't like the idea of organised rotation === salgado is now known as salgado-lunch [15:24] salgado takes a sabbatical from this discussion [15:24] okay, sounds good. so if you need a break, please come see me and i'll coordinate to make sure we have proper coverage. please be mindful that your teammates may also need breaks now and then [15:24] okay, thanks everyone. good discussion [15:24] [TOPIC] * Do we need a standard cover letter template for merge proposals? (barry) [15:24] New Topic: * Do we need a standard cover letter template for merge proposals? (barry) [15:25] i've been using this: https://pastebin.canonical.com/11767/ [15:25] everything that's standard, should be part of the form [15:25] and experimenting with a == diff == section at the end so the whole thing is in one cover letter [15:26] 1) do we want a standard form; 2) is this one a good start; 3) if so, what's missing? [15:26] lint output? [15:26] the bug *should* be redundant as the MP should be linked [15:26] well [15:26] bug report are often cryptic [15:26] so having a summary of the issue at the top of the cover letter is usually good [15:27] adeuring: lint output, good addition [15:27] plus it's easier on the Thailand/Australia/NZ folks [15:27] save a trip to Launchpad [15:27] flacoste: They chose to live on the wrong side of the internet. [15:27] Bugger 'em. [15:28] * gmb jokes, because thumper is bigger than he is [15:28] I often have a == rules == section that defines the scope of the fix. [15:28] flacoste: i agree. i've found it's good to have a summary there [15:28] sinzui: what does "the scope of the fix" mean? [15:29] sinzui: i think that's what my "proposed fix" section is for iiuc [15:29] ok, ok, my point is we agreed earlier that we should use --fixes or another mechanism to link the bug. i've been requesting it when not done. [15:30] bac: yes, definitely --fixes is a good thing, but under the = Bug XXXXX = section, i usually put a summary of what the bug is about [15:30] right, and the scope of the fix should usually be the bug [15:30] barry: yes, good idea. [15:30] does anybody disagree that we should have a standard form? [15:31] intellectronica: I write down what the crucial changes that need to happen. There are lots of other things that can complicate the fix, or introduce feature creep. [15:31] i agree, but i think that we should aim at making it part of launchpad itself, not some kind of template [15:31] sinzui: makes sense. that's something that is always good to have [15:32] intellectronica: agreed, but don't let the perfect be the enemy of the good :) [15:32] intellectronica: +1 [15:32] greasemonkey script? :) [15:32] huh? [15:32] intellectronica: I started doing it to make myself focus on the fix. I kept doing it because I could point to that when the branch is a CP candidate. [15:32] bigjools: api/bzr plugin? :) [15:33] sinzui: also, i'm trying hard to start filliing out the template as soon as i start the branch, and not to wait until the mp is posted [15:33] barry: We could start by repurposing bzr write-cover-letter... [15:33] barry: or that! [15:33] If this template is good it should go in the tree I think. [15:34] barry: I have a macro that prompts me for it [15:34] gmb: i'm thinking something like that [15:34] sinzui: gedit macro? [15:34] yes [15:34] cool [15:35] if there are no objections, i'll take this to the mailing list [15:36] and if anybody wants to work on a plugin to use it, i'll help out [15:36] * gmb wishes he had time [15:36] api support through lpreview would be awesome [15:36] gmb: forced vacation is coming up :) [15:36] gmb: time is left-brain illusion, you have it :-) [15:37] barry: Really? I thought it was UDS... [15:37] * bigjools concentrates hard like Hiro [15:37] anyway, that's everything on my list. does anybody have an item not on the agenda? [15:38] barry: I did add something [15:38] but it's not on your agenda here for some reason [15:38] * barry refreshes [15:38] [TOPIC] * Gotchas with new celebrities (Julian) [15:38] New Topic: * Gotchas with new celebrities (Julian) [15:38] bigjools: the floor is yours [15:39] We had a problem in the week with an oops being generated because Celso added a new celebrity and some bugs code on edge was trying to iterate over them [15:39] so it's just a heads up to say, if you add celebrities, remember to find a LOSA to patch the DB as well [15:39] EOF [15:40] surely we can find a better way?... [15:40] i won't remember this [15:40] don't iterate over celebs? :) [15:40] why does it need to do that anyway? [15:41] either that, or test on startup? [15:41] bigjools: which file? [15:41] gmb: I can't remember, let me take a look [15:41] ok [15:41] There's some code that checks if a bug tracker is a celebrity, in which case deletion is forbidden. [15:41] Hum. [15:41] It does it by iterating through all the celebs. [15:42] * rockstar mes [15:42] allenap: Let's discuss this outside the meeting [15:42] Yep. [15:42] doesn't adding a celebrity require a DB patch? I had to do that when working with the janitor [15:43] sinzui: Not necessarily. [15:43] And usually, no. [15:43] gmb: well kind oif [15:43] sinzui: just a new row [15:43] actually it does [15:43] sampledata change [15:43] yeah, right [15:43] but you should do that using a db patch [15:43] so that it's applied automatically on roll-out [15:43] flacoste: Unless you're adding something as a celeb that already exists in prod. [15:43] right [15:44] gmb: OOPS-1063EC95 [15:44] flacoste: Is that in our guidelines for adding a celebrity? [15:44] bigjools: Ta [15:44] gmb: using a db patch? [15:44] that makes sense then: if the celeb exists on prod, add sample data. if not, db patch is necessary to add it in both places [15:44] flacoste: Yes [15:44] does that make sense? [15:44] i'm not even sure we have formalized guidlines on this [15:45] but yes, barry's suggestion is sane [15:45] and by using a db patch, your code won't run on edge [15:45] until the DB has the celebrity [15:45] hmmm that's kinda restrictive [15:46] bigjools: well, that's the policy, there might be reason to do otherwise [15:46] flacoste: i agree [15:46] but in that case, you should know what you are doing [15:46] okay [15:46] and not trigger an OOPS on edge :-) [15:46] I'm entirely innocent (this time) [15:46] :) [15:47] bigjools and innocence together, i can't believe it [15:47] [ACTION] someone capture celeb addition policy in dev wiki, when the reviewer-tips have been moved from the old wiki [15:47] ACTION received: someone capture celeb addition policy in dev wiki, when the reviewer-tips have been moved from the old wiki [15:47] barry, that's an action that will never happen! [15:47] * bigjools considers kicking sand at flacoste at the TL sprint [15:47] flacoste: indeed :) [15:48] anyway, that's all the time we have today [15:48] #endmeeting [15:48] Meeting finished at 09:48. [15:48] thanks everyone! [15:48] thanks barry [15:48] thanks barry [15:48] see y'all back a the ranch [15:48] bigjools: sand will be nice change from snow! [15:50] flacoste: amen brother :) [15:51] it's warm here today, we're 4C [15:57] 0C here! [15:58] Hello there and sorry for being late [15:58] with a dull grey sky [15:58] ah well it was bright blue here but it's nearly dark now :) [15:59] bright blue is nice! [15:59] dull grey kills my productivity [16:00] I hear ya [16:01] * gmb is waiting for the snow to come === salgado-lunch is now known as salgado === bac is now known as bac_lunch === bac_lunch is now known as bac === salgado is now known as salgado-afk === bac is now known as bac_afk === bac_afk is now known as bac === spmcinern is now known as spm