=== barry-away is now known as barry [02:58] barry: were we going to have that meeting? [02:58] thumper: i think we should [02:58] * thumper nods [02:58] thumper: 2 minutes [02:59] ok [02:59] hello [03:00] #startmeeting [03:00] Meeting started at 21:03. The chair is barry. [03:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [03:00] hello everybody and welcome to this week's asiapac reviewers meeting [03:00] hello [03:00] i see my clock is off by 3 minutes? [03:00] who's here today [03:00] i am here [03:01] I am [03:01] thumper is too i know [03:01] me [03:02] i can't update the meeting page because i guess the machine updates are still running [03:02] apologies for last night btw [03:02] == Agenda == [03:02] * Roll call [03:02] * Next meeting [03:02] * Action items [03:02] * Queue status [03:02] * Mentoring update [03:02] * Review process [03:02] * Module alternatives - do we really want them? [03:02] shall we go ahead and do our regular time next week? [03:02] yep [03:03] okay [03:03] cool [03:03] * Action items [03:03] yes please [03:03] [TOPIC] * Action items [03:03] New Topic: * Action items [03:03] nothing outstanding [03:03] * Queue status [03:03] [TOPIC] * Queue status [03:03] New Topic: * Queue status [03:04] does it seem like we have a bajillion branches that need reviewed this cycle? [03:04] yes. [03:04] * jml hasn't checked his queue today. [03:05] gee, I wonder why? [03:05] i don't know that it's going to be any more intense than usual for week 3 though [03:05] It seems like the bzr team have a bajillion reviews to atm too :) [03:05] i thought that was just one humungous review :) [03:06] i'm a little worried that ocrs are getting overloaded or burned out. do any of y'all feel that way? [03:07] rocs? [03:07] orcs? [03:07] it's much easier for us 'round here [03:07] barry: I don't. [03:07] on call reviewers :) [03:07] ah [03:07] barry: there's a minor problem where I get requests for reviews on Friday afternoons [03:08] barry: that's resulted in a couple of Europeans having to wait 24hrs between review cycles. [03:08] jml: you get them while on call or in your queue? [03:08] jml: is there someone to pass the reviews onto? [03:08] barry: while on call. [03:08] thumper: there wasn't this time. [03:09] who is the european on on Friday? [03:09] at other times in the cycle I probably would have just said no :) [03:09] david murphy [03:09] then in the us it's superman, er, sinzui [03:09] barry: where does the schedule live? [03:09] https://launchpad.canonical.com/OnCallReviewers [03:10] i still would love to fill wednesday more [03:10] but we lost statik as a reviewer, which is too bad, because he was good [03:10] otoh, bigjools is going to graduate [03:10] cool. [03:11] how many non-reviewers do we have currently? [03:11] mwhudson: i was just looking that up [03:11] I'd like to get abentley onto the team shortly [03:11] except i can't get into the directory :( [03:11] thumper: good idea [03:11] aaron, mars, leonard ? [03:12] I'll follow it up with him first [03:12] thumper: sounds good [03:12] to make sure he is up for taking it on now [03:12] (or after 2.0) [03:12] * mars stirs [03:12] mwhudson: those are the next logical choices i think [03:13] oh, al-maisan isn't a review either i guess [03:13] reviewer [03:13] and rockstar [03:13] rockstar is still very new though [03:13] so is al-maisan? i haven't seen many of either's branches. have you guys seen any of them? [03:14] al-maisan is still pretty new [03:14] I just reviewed one of his branches. [03:14] i doubt al-maisan or rockstar are ready yet [03:14] * jml agrees. [03:14] Yes, I am not ready... [03:14] but yeah, i guess we should bully the three i first mentioned i guess [03:15] [TOPIC] mentoring update :) [03:15] New Topic: mentoring update :) [03:15] mwhudson: all we need to do is find mentors for them [03:15] with bigjools graduating, i can take someone else on [03:16] al-maisan is euro? rockstar, you're us right? [03:16] barry: al-maisan is in .de [03:16] Yup, US [03:16] i know mars is ca, eastern [03:16] jml: cool [03:16] leonard is us eastern too [03:17] barry: abentley is .ca -- my geography doesn't extend that far north though [03:17] :) [03:17] i _guess_ one of us could do one of the new guys, but the difference is a bit much [03:17] yeah. [03:17] (until the dst thing happens again) [03:17] better to get someone from atlantis [03:17] yeah it's tough when there's so little overlap [03:17] I mean AMEU [03:17] jml: unless you can convince someone to move! [03:18] barry: I would like Australia to move :) [03:18] :-D [03:18] I'm not sure what the real estate market in Atlantis is like... [03:18] spiv: you can get great stuff cheap these days [03:18] * thumper seemed to miss 5 mintutes [03:19] jml: http://www.satirewire.com/news/jan02/australia.shtml [03:19] barry: I'd worry about home owners going under. [03:19] spiv: you haven't heard about our real-estate crash? it's only going to bring down the entire world economy [03:19] mwhudson: seen it :) [03:19] jml: good :) [03:19] maybe we need some reviewers in hawaii [03:19] mwhudson: :) [03:19] it's a great time to /buy/ a house, a horrible time to sell one [03:20] anyway, all good suggestions, thanks [03:20] I checked out the cost of a house in hawaii [03:20] it isn't cheap [03:20] * spiv would want to bouy a house if he lived in Atlantis [03:21] we seem to be somewhat off topic [03:21] :) [03:21] spiv: there are people who will sell one to you [03:21] [TOPIC] * Review process [03:21] New Topic: * Review process [03:21] * Module alternatives - do we really want them? [03:21] barry: what's a module alternative? [03:21] lets see if i can swap this one in [03:22] what is this exactly? [03:22] intellectronica brought this up [03:22] whether we want to wrap imports of say, from cFoo import Foo in try/excepts [03:22] or just do the import [03:22] and expect them to work [03:22] just import [03:23] oh just import [03:23] * jml seconds jamesh [03:23] just import [03:23] have you found a Python installation without cStringIO or cPickle? [03:23] we control our environment [03:23] We know what dependencies we have, don't we? So just import. [03:23] and what mwhudson said [03:23] s/have/have installed/ [03:23] that's basically what we decided. i think there was some question about forward-proofing etree, elementree and the like [03:24] mmm [03:24] the one place I think a try/except might be useful is for module name changes going 2.4 -> 2.5 [03:24] mwhudson: wasn't there something that you wanted to bring up at the review meeting? [03:24] e.g. ElementTree [03:24] thumper: it's on the agenda :) [03:24] * jamesh is too slow [03:24] jamesh: some talk about having a canonical.alternatives package for hiding the 2.4/2.5 differences [03:24] mwhudson: I didn't see it on the agenda [03:24] well i added it [03:25] anyway, we're all agreed, just import it [03:25] and upgrade to 2.5 when we do :) [03:25] * (mwhudson) What are page tests for? [03:25] mwhudson: the floor is yours [03:25] oh, it's just a proposed item [03:26] this came from some conversations i had with jml and thumper [03:26] basically, writing and maintaining pagetests is a pain [03:26] and i think we are a bit muddled as a team as to what the purpose of a page test is [03:27] some of them are 'customer stories' [03:27] and some of it is just testing that the templates work [03:27] (and most are some mixture of both) [03:27] so i'd like to encourage a wider conversation on the topic [03:27] My first reaction is that both are useful, but keeping them separate would be a good idea. [03:27] some try to do exhaustive testing of alternatives [03:28] i guess i should mail the list about this, but let's have a short argument here first :) [03:28] I'd prefer to see template testing move to unitests [03:28] spiv: that's my first reaction too [03:28] customer stories are fine [03:28] thumper: +1 [03:28] spiv: +1 [03:28] if they are customer tests, we should consider not running them on every landing [03:28] one question i've often had is how much should we be sneaking around the web u/i in a pagetest? [03:28] Basically: for tests the key question iswhat is the intent of the test, and for docs the key question is what is the audience of the document. [03:28] mwhudson: not sure about that [03:29] (And for doctests, there's also the question of "is this a doc or a test?") [03:29] thumper: i agree, i'd hate to stop running some of our important tests [03:29] I think being clear about the answers to those will help us figure out how to organise our pagetests. [03:29] mwhudson: for a while, it wasn't possible to do testbrowser stuff outside of the pagetests [03:29] spiv: very good point. i love doctests for everything, but some are documentation and some are not [03:30] thumper: this pre-supposes that all the important things are tested in non customer tests [03:30] I think that trying to do exhaustive tests of alternatives should never be in a doctest or pagetest [03:30] due to some required setup/teardown being done in the FunctionalDocFileSUite [03:30] jamesh: oh, interesting [03:30] that is fixed now though, so there isn't any technical reason not to do testbrowser in unittest code now [03:30] barry: I *will* persuade you that doctests for everything is bad at some point, I promise :) [03:30] jamesh: i have written unittests that use testbrowser :) [03:30] spiv: :) [03:30] spiv: I'm sure doctests are good somewhere [03:30] mwhudson: and that worked because I moved the test setup to the layers [03:31] (it's still a pain because you have to log in and out all over) [03:31] jamesh: i shall buy you a beer for that at some point [03:31] mwhudson: what about setting the unitest in the pagetestlayer? [03:31] thumper: i have this theory that we can use the doctest format for everything (and that it's better than python tests) [03:31] barry: ew, NO! [03:31] this is a separate punch-up [03:31] :) [03:32] yeah, let's not get distracted :) [03:32] these days, there isn't any special setup done for doc tests, other than default globals, iirc [03:32] jamesh: right, and we're trying to discourage magical globals in doctests anyway [03:32] barry, +1 :) [03:32] yep [03:32] are we? [03:32] I just added one [03:32] * mwhudson shall have to buy xUnit Test Patterns for everyone on the launchpad team, it seems [03:33] i've had exactly one case where i think it's justified, but other than that, the imports are good to put in the doctest! [03:33] at least it makes it clear where things are coming from [03:33] so, should i mail the launchpad list about the page test thing? [03:33] mwhudson: ideally with a short proposal about what to do about it [03:33] mwhudson: I think that's a good idea. [03:34] thumper: yeah, i hate having to go hunting for the harness that sets up a doctest just to figure out what some global actually is [03:34] mwhudson: just to try to limit the inevitable bike-shedding a little [03:34] and say that launchpad/pagetests should be stories [03:34] mwhudson: +1, spiv +1 [03:34] and that doc/ should be documentation [03:34] mwhudson: +1 [03:34] and tests (whether .py or .txt) should be in a tests directory? [03:34] mwhudson: +1 [03:34] mwhudson: +1 +1 +1 [03:34] ok [03:35] and can we for gawd's sake please do something about tests, testing, and ftests? [03:35] Rename ftests to f'ing tests? [03:35] well, testing is for scaffolding [03:35] barry: tests & testing is a sensible split [03:35] stuff in ftests can go to tests [03:35] ddaa tried to land a branch that moved everything in ftests into tests [03:35] jml: yes, maybe just get rid of ftests. i would be so much happier [03:35] testing is infrastructure [03:35] the tests / testing split is good [03:35] jamesh: sometimes it needs to go to testing :) [03:35] jamesh: +1 [03:36] jml: right [03:36] jamesh: +1 [03:36] but it's a conflict nightmare [03:36] ok, i can add this to my mail too :) [03:36] each team should move its own tests [03:36] from ftests to tests in the appropriate place [03:36] * barry has a secret plan for montreal :) [03:36] we file a bug [03:36] create tasks for each team [03:36] and JFDI! [03:37] thumper: right! i want a special pqm tag for that [03:37] :) [03:37] barry: why a pqm tag? [03:37] thumper: I think you'll find that one team has already done it :P [03:37] jml: not quite [03:37] thumper: well, when i want to jf land a branch :) [03:37] jml: there are some older tests still in c.l.ftests [03:37] Some people claim "Jedi" as their religion. thumper's religion is "JFDI". [03:38] thumper: let us continue this team-specific discussion outside of the meeting. [03:38] 2007's motto was "canonical fuck yeah!". 2008's should be JFDI [03:38] barry: can you add a mootbot task for that? [03:38] [ACTION] mwhudson to start discussion on page test purpose [03:38] ACTION received: mwhudson to start discussion on page test purpose [03:38] barry: I was meaning a bug for ftests -> tests [03:39] [ACTION] thumper to submit a bug for moving ftests contents to tests [03:39] ACTION received: thumper to submit a bug for moving ftests contents to tests [03:39] anything else on this topic mwhudson? [03:39] not for today i think [03:40] i can hear jml's stomach rumbling from here :) [03:40] :) [03:40] :D [03:40] okay, one quick last one [03:40] The "D" represents my gaping maw. [03:40] * (bigjools) On-call reviewers to remind devs to be available during the review. [03:40] that's a good idea. [03:40] +1 [03:40] I'll do that in future. [03:40] bigjools brought up to me that some people are using on-call requests to jump ahead of the review queue [03:40] we've got way too much obsolete test infrastructure still in tree [03:41] i.e. they request an on-call review then disappear [03:41] jamesh: yes [03:41] barry: yeah, this has happened to me. [03:41] i told bigjools that if that happens, you are allowed to reject the ocr request and move on to someone who is present [03:41] the whole point of ocr is to have a dialog with the dev [03:41] i think i've done this [03:42] This makes good sense to me. [03:42] at least, if i have two reviews to do and the developer is there for one of them, he gets done first [03:43] anybody disagree? [03:43] nope [03:43] nope [03:43] great. that's it for me. shall we end 2 minutes early and let jml get some food? [03:44] I think we should discuss it first :) [03:44] jml: but only for 1 more minute :) [03:44] #endmeeting [03:44] Meeting finished at 21:47. [03:44] thanks everyone! [03:44] barry: thanks! [03:44] thanks barry [03:44] barry: thanks === barry is now known as barry-away === Moot2 is now known as MootBot === mwhudson_ is now known as mwhudson === mpt_ is now known as mpt === mwhudson_ is now known as mwhudson === salgado-afk is now known as salgado === mrevell_ is now known as mrevell-lunch === mrevell-lunch is now known as mrevell [15:00] #startmeeting [15:00] Meeting started at 09:03. The chair is barry. [15:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [15:00] hi everybody. welcome to this week's ameu reviewers meeting [15:00] who's here today? [15:00] me [15:00] me [15:00] me [15:00] me [15:01] me [15:01] me [15:01] danilos, BjornT ping [15:01] flacoste: ping [15:01] me [15:01] me [15:01] me [15:01] barry: thanks [15:02] me [15:02] salgado: ping [15:02] [TOPIC] agenda [15:02] New Topic: agenda [15:02] me [15:02] == Agenda == [15:02] * Roll call [15:02] * Next meeting [15:02] * Action items [15:02] * Queue status [15:02] * Mentoring update [15:02] * bye to statik [15:02] * bigjools graduates [15:02] * new recruits? (leonardr, abentley, mars) [15:02] * Review process [15:02] * (bigjools) On-call reviewers to remind devs to be available during the review. [15:02] * (gmb) Wrapping long argument lists: What's the agreed style? [15:02] * (mwhudson) What are page tests for? [15:02] * thumper moving ftests contents to tests [15:02] [TOPIC] next meeting [15:02] New Topic: next meeting [15:02] * barry kicks mootbot [15:03] me [15:03] me [15:03] jtv: hi! statik hi! [15:03] barry, hi! [15:03] next week, same time and place? anybody know they will be sprinting or away? [15:04] cool [15:04] [TOPIC] action items [15:04] New Topic: action items [15:04] * intellectronica to file bug on lint issue regarding elementtree import [15:04] sorry, i didn't. let's carry this on [15:04] intellectronica: np [15:04] * mwhudson to start discussion on page test purpose [15:05] i will proxy mwhudson :) [15:05] he's going to start a ml thread on the purpose of page tests. more in a moment [15:05] * thumper to submit a bug for moving ftests contents to tests [15:05] more on this in a bit [15:05] [TOPIC] queue status [15:05] New Topic: queue status [15:06] any comments? i was freaking out on monday, but i've calmed down by now. thanks everyone for cranking away on reviews! [15:06] Stub has reviewed my two branches now; he has not updated the statuses [15:07] sinzui: are they merge-approved? [15:07] yes [15:07] sinzui: cool. i'll ignore those and you can just remove them when they land [15:07] my branch that kiko reviewed is still trying to land [15:08] * barry won't even mention testsuite2 :) [15:08] anything else on the queue? [15:08] We seem to have extra missing branches this week [15:09] sinzui: missing where? [15:09] yeah, I'm nagging stub about my oauth-context branch, which has a DB patch [15:09] barry: The red branches are missing. It implies they were removed from the repo, or renamed [15:10] sinzui: yeah, people are not updating PR [15:10] gmb and intellectronica have missing branches [15:10] and schwuk [15:10] everyone, remind your reviewees to clean up PR when their branch lands! [15:10] amen to that [15:10] * bigjools cleaned 4 of other people's out this morning [15:11] sinzui: oh right, i cleaned my repo but not PR, i guess. i'll sort it out [15:11] bigjools: yes, feel free to clean out any that you see [15:11] * barry is a culprit too [15:11] I don't like doing it though, it makes devs lazy! [15:11] bigjools: hmm, maybe refuse to review any new branches until they clean out their old ones :) [15:11] lol [15:12] anyway... [15:12] [TOPIC] * Mentoring update [15:12] New Topic: * Mentoring update [15:12] Or have PQM check for red branches and rejecting anything until they're gone... [15:13] so first, we say goodbye to statik who is moving off the lp team and won't be accepting any more reviews. i was tempted to go down to florida to camp on his doorstep until he changed his mind but it's a long train trip [15:13] statik: thanks and good luck! you're welcome to come do reviews any time you're bored :) [15:13] thanks! [15:14] next up: bigjools graduates. congratulations bigjools! [15:14] congrats bigjools! [15:14] * flacoste cheers bigjools [15:14] congrats bigjools! [15:14] hurrah [15:14] * bigjools takes a bow [15:14] Hurrah! [15:14] thanks guys [15:14] i will send an official email after this meeting [15:14] Barry is a great mentor - thanks to him also [15:14] remember, give all your hard branches to bigjools now :) [15:14] oof [15:15] :) [15:15] bigjools: thanks [15:15] but has it easy, being a soyuz developer [15:15] * bigjools considers retracting the last statement :) [15:15] too late [15:15] the worst branch are his :-p [15:15] flacoste: no kidding! [15:15] flacoste: lol [15:16] i think we can accept a few more recruits now. i'm certainly happy to mentor again. i think the next three potential candiates are leonardr, abentley, and mars [15:16] i talked to thumper and i'm thinking about mentoring abentley [15:16] we need mentors if we're going to accept anybody else, though i propose we do not accept until after 2.0 [15:17] so if you would like to volunteer to mentor leonardr or mars, please send me an email [15:17] i'm happy to mentor again, after 2.0 [15:17] intellectronica: awesome [15:17] I'll be happy to do the same [15:17] gmb: great, thanks! [15:17] btw, we mentioned rockstar and al-maisan but we want to give those guys a bit more time dev'ing for now [15:18] [TOPIC] * Review process [15:18]