[00:32] davecheney: not yet... mostly knowledge transfer for cherylj, although I do have one problem I'm putting off today that would be good to do [00:34] yessss, give me ALL THE KNOWLEDGE! [00:38] * perrito666 wants to know new zealand and thinks thumper should adopt him into his team temporarily [00:38] perrito666: happy to :) [00:39] perrito666: and... Air NZ now doing direct flights to Argentina [00:39] probably has something to do with rugby and the pumas :) [00:39] thumper: price in bars of gold? [00:40] last time I checked direct flight was somewhere in the 20k usd :p [00:40] probably going to get cheaper :) [00:40] sweet [00:40] * perrito666 packs cooking gear [00:43] perrito666: cooking gear?.. [00:45] anastasiamac: I am sure there is something nice to cook in new zealand and I dislike cooking without my own knife :p [00:45] so who is OCR? http://reviews.vapour.ws/r/736/ this inlocks the critical bug [00:46] * perrito666 installs go cross compile to avoid having to go for a windows every time [00:50] waigani: ? [00:52] davecheney: actually, wifi works, it just took an unusually long time connect after suspend. possibly nothing to do with the latest kernel. [00:53] menn0: I have heard more people complain of things like that with latest kernel [00:58] thumper: i plan to monopolise mwhudson 's time and grind through as many bugs as we can on our port [01:02] perrito666: reviewed [01:02] fix it, then ship it [01:02] davecheney: td [01:02] tx [01:03] aghh I keep forgetting we are in 2015 [01:04] 2015, all year [01:04] perrito666: aah... that would explain high prices on travel to nz [01:05] perrito666: if everyone form argentine brings their ***cough*** cooking knives [01:05] perrito666: u do know that Maories r peaceful pppl? not easily provoked? nor known for their aggression? .. :P [01:06] anastasiamac: I dont know if you noticed this but there is no call where thumper doesn't have a bruise [01:06] perrito666: ha [01:08] perrito666: rofl [01:09] can we make the bot oblivious to the diff between - and _ ?? [01:12] perrito666: I gave you a review on that restore fix [01:12] perrito666: is there a reason why you can't just add a restore_windows.go? [01:13] right now the pressing time [01:14] ericsnow: anyway, it was faster to stub Restore than to stub all of the methods that fail [01:16] anastasiamac: standup? [01:16] ericsnow: to be honest I am a bit curious on why do we build any of that in windows [01:22] thumper: i'm not having anyluck getting someone on #is to look at that RT [01:34] ericsnow: can you please close/delete this review http://reviews.vapour.ws/r/621/ [01:34] it's dead [01:34] davecheney: sure [01:35] davecheney: the PR should be closed too [01:35] davecheney: did as much poking as I could... [01:35] menn0: I have a bad feeling about this [01:35] thumper: about what? [01:36] menn0: someone elses upgrade step [01:36] that I don't think it does what they thought it did [01:36] I cribbed someone else's setp [01:36] step [01:36] and did things slightly differently [01:36] particularly about the assertions [01:36] perhaps it is working by side-effect [01:36] * thumper shrugs [01:38] thumper: which one? [01:38] menn0: the one in agentconfig.go [01:38] I'm also changing agent config [01:38] but I'm trying to see where it actually gets written to disk [01:38] and it isn't during the step [01:40] hmm... only have a ConfigSetter [01:40] there must be a reason why we don't write in the upgrade steps [01:40] thumper: not sure... looks like it's Dimiter's work [01:46] menn0: ah ha [01:46] menn0: the changes to the agent config are saved at the end of the upgrade steps [01:46] * thumper fixes his tests [01:47] thumper: of course. [01:47] thumper: the steps are run inside a ChangeConfig [01:48] yep... [02:03] thumper: core meeting? [02:03] menn0: also? [02:03] davecheney: ditto [02:21] anastasiamac: I trust you noticed the coloring in thumper 's right eye [02:21] perrito666: you are imagining things [02:21] not been hit for ages [02:21] as I've had to let my shoulder heal [02:21] thumper: you live in a dangerous place, you have orcs for god sake :p [02:22] I have senn it on tv [02:22] perrito666: orcs r seasonal [02:23] lol [02:23] perrito666: only from Boxing Day onwards [02:55] menn0: fyi https://github.com/juju/juju/pull/1417/files [02:56] anyone game? http://reviews.vapour.ws/r/738/ [02:56] wow, three pages of diff [02:57] $ git diff master | wc -l [02:57] 1792 [02:57] sorry reviewer [02:57] most of that is real necessary [02:57] some of it isn't... [02:57] but most of it is [03:46] thumper: i'm having a look [04:06] menn0: how did you test upgrades before when the upgrade is listed against 1.23.0 and the version is 1.23-alpha1? [04:11] menn0: wondering because I have a running 1.22 environment and I thought I might as well test the upgrade [04:11] but since the upgrade steps are listed against 1.23.0... I don't think they'll run [04:20] axw: blake suggested in the maas doc using "/dev/disk/by-id/" as the block device identifier. I think this would work? Can you comment in the doc to confirm? [04:21] thumper: I changed things a while back so that upgrade steps for the final version are run when you upgrade to any alpha or beta of that version [04:21] thumper: the steps will run [04:21] menn0: oh nice [04:21] menn0: thanks [04:22] thumper: it was wallyworld's suggestion. that way we can test steps while working towards a final version. [04:22] makes sense [04:26] thumper: Fix it and ship it [04:28] menn0: confirmed that starting a new environment has all agents dialing the env uuid version [04:29] menn0: I wish we had a setting that could say "please restart after you have finished upgrading" [04:30] menn0: thanks for the timely review of this massive piece of work [04:30] menn0: but I think it should wait until monday to land [04:30] I still need to test upgrades and a few tweaks from the review, but running out of time today, need to pack [04:32] davecheney, you still around? [04:33] mattyw: ping [04:46] wallyworld: will do. [04:46] ty [06:42] jam: did we ever get a definitive yes/no on whether Openstack is in scope for storage? It's missing completely from the Provider Specific Tactics section of the spec, but is a line item in the "in scope" section [07:44] jam, axw, PTAL http://reviews.vapour.ws/r/740/ - migrating goamz imports from LP to GH [07:44] * dimitern needs to step out for ~1h [07:46] dimitern: lgtm [07:47] axw, thanks! [07:50] wallyworld_: can't read your pgpmail [07:50] oh, ffs, stupid thunderbird [08:02] wallyworld_: you have question, do you mean for 1:1 tomorrow? [08:02] questions* [08:49] wallyworld_: if you're around, https://github.com/juju/juju/pull/1420 [08:49] bot hasn't picked it up for some reason ... [09:15] * fwereade out for a bit [10:03] TheMue, standup? [10:58] dimitern: morning and sorry for missing the hangout [10:58] dimitern: third bad night in a row, had to change clothes several times [10:59] dimitern: at least in the morning hours I found some sleep [11:00] dimitern: so will now work on the couch reading all the networking information and ask you each time there's something unclear :) [11:02] TheMue, hey, morning! [11:02] TheMue, sorry to hear that :/ I hope you'd start feeling better! [11:03] TheMue, sure - sgtm, there's a lot to read :) [11:04] dimitern: I hoped so too, I seldom have something for more than one day. so I already canceled being at the Oldenburg English Club today :( [11:04] dimitern: yeah, I thought that's a good opportunity, have to make the best out of the situation [11:05] perrito666: I have reopened bug 1411024 because the mac client still does not build with your change from last night [11:05] Bug #1411024: Win client/agent cannot bug built because of backup deps

[11:08] TheMue, +1 [11:08] TheMue, I wish you a quick recovery in any case [11:09] mgz, ha!@ [11:09] dimitern: thanks, I wish it too. *lol* [11:09] mgz, I though it might be a problem as in the fix the "_linux" suffix was used, rather than _unix [11:10] dimitern: well, it should be an easy fix to the fix then :) [11:11] mgz, however I'm not sure whether we support backups on mac os? [11:13] dimitern: I think we clearly don't need to at the moment [11:13] there's probably no real reason it shouldn't work, but we just need the client to build for now :) [11:14] mgz, yeah, but do you know if it *ever* worked on mac? [11:14] mgz, because if it didn't the fix is a lot easier [11:14] dimitern: I suspect the old shell script hack would have, if anyone ever tried [11:15] it's certainly not something we signed up for supporting, or have been verifying works [11:15] (the backup/restore test jobs only use the client under ubuntu) [11:15] mgz, cool, that's all the confirmation I need for now :) [11:45] mgz, https://github.com/juju/juju/pull/1421 - that's the fix [11:46] mgz, can you re-run the failing jobs after it lands manually to speed-up the unblocking? [11:49] can anyone review this please? ^^ axw, fwereade, voidspace, ? [11:49] dimitern: looking [11:49] voidspace, ta! [11:50] dimitern: LGTM [11:50] for some reason a RB diff wasn't generated [11:50] dimitern, LGTM too, I'm just a bit too slow [11:50] voidspace, thank you, setting to merge to unblock trunk [11:50] dimitern: it should go straight through - the build jobs are early on as well [11:52] mgz, cool! [12:01] mgz: tx [12:02] dimitern: and tx to you too [12:03] perrito666, np, I'm opening a beer tab ;) [12:03] you won it :p will have to wait until apri though [12:04] hehe it's closer than you think ;) [12:05] dimitern: we can go have beer and sausages with sauerkraut [12:06] perrito666, great idea! we should definitely put it on the agenda [12:07] well, it is on mine :p [12:13] ci is building the new rev [12:25] mgz, it would be very useful if on pages like this http://reports.vapour.ws/releases/2232 "Last build" links *both* to the job in jenkins and the console log output (ideally not as application/octet-stream but text/plain, so it can be viewed in the browser directly) [12:26] dimitern: yeah, I'm working on tidying those last bits up now [12:26] mgz, you rock! [12:49] voidspace, please, take a look at the stub NetworkInterfaces - http://reviews.vapour.ws/r/742/ [13:02] wallyworld_, don't suppose you're around? wanted to ask about when we set StatusInstalling [13:03] fwereade: give me 5 [13:09] fwereade: when ModeInstalling is called and also when ModeConfigChanged is called and state is not Starting [13:10] wallyworld_, right, am confused about setting it at ModeConfigChanged time [13:10] maybe that's not necessary - i think it was a s/Installed/Installing thing [13:11] wallyworld_, ok cool [13:11] but if we do get config changed and we are not set up yet [13:11] then installing seems reasonable [13:11] maybe the check just needs to be expanded [13:12] wallyworld_, then we should surely still be in Installing state? ah-*ha*, we might have gone into hook error on install [13:12] yup [13:12] wallyworld_, I don't think it should be a config-changed-time check though, I think it's part of committing an installed state [13:12] wallyworld_, does that sound sane? [13:13] i think so, but am a little tired so will think properly tomorrow [13:13] wallyworld_, or, hmm [13:13] wallyworld_, no, actually, I think you're right [13:13] which bit? [13:14] wallyworld_, setting it at the start of config-changed [13:14] that did seem reasonable [13:14] at the time [13:14] wallyworld_, it was judicious and forward-thinking, I appreciate it [13:15] not sure if i put it there or adapted it or whatever [13:15] wallyworld_, fwiw I think you set StatusInstalling in the wrong part of ModeInstalling but I'm touching that as I go [13:15] could well do - i am a uniter novice [13:16] be gentle with me, it's my first time [13:16] doing stuff with uniter [13:16] wallyworld_, no worries, it's just that ModeInstalling isn't the mode func, it *returns* the mode func [13:17] yes, you are right. maybe i was thinkin that if it was at the point of setting up for install, it should be put in installing [13:17] ie transition from allocating to installing sooner [13:18] as by then clearly allocation had been done [13:18] wallyworld_, anyway you don't need to do anything, I just wanted to check up/rubber duck what was going on [13:18] np, ty [13:18] wallyworld_, I dunno, I think the moment we start looking up the charm is a fine time to set the status [13:18] wallyworld_, btw did we get an upgrading status? [13:19] hmmm, no [13:19] we should probs add that, after checking with john [13:20] wallyworld_, we can manage fine without it but I think it's cleaner [13:20] i think i agree [13:21] wallyworld_, cool, I will assume that'll be the approach if I bump up against it [13:21] ok [13:30] mgz, got a link to the win/mac build jobs? [13:33] mgz, ah, found them from the last curse message [13:40] dimitern: any luck? [13:42] perrito666, well - so far so good [13:42] if anyone's got time for a quick review, that'd be great... may have slipped by unnoticed since the RB bot didn't pick it up: https://github.com/juju/juju/pull/1420 [13:43] perrito666, only the "revision results" job has to succeed [13:44] perrito666, but afaics the win/mac builds passed [13:47] dimitern: looks good [13:47] dimitern: still reading [13:48] dimitern: why the ++ changed to -- [13:48] voidspace, great [13:48] dimitern: why the ++ changed to -- ? [13:49] voidspace, well, it seemed more appropriate - releasing an address should decrement the maxAddr as it could be useful for tests like Allocate, Release, Allocate [13:51] ah, ok [13:52] at least you haven't claimed to do live testing on this one... [13:54] Dummy implementation looks good/useful [13:58] dimitern: if the -- change made no difference to any tests, or any code anywhere, then I'm sceptical it's needed at all... [13:58] dimitern: other than that, LGTM [13:58] dimitern: well, let me look at the dummy implementation one more time before I sign off [13:58] lots of TODOs added as well [13:59] voidspace, I agree we shouldn't plan too far ahead for what tests we *might* need [13:59] dimitern: I dislike/distrust code written for imaginery future use cases. [14:00] voidspace, that s/++/--/ seems like a good idea, but I don't mind reverting it if you disagree - it's just a detail anyway [14:00] voidspace, +1, although I need to get better at judging this :) [14:00] sure [14:00] * dimitern bbiab [14:01] balancing future planning with that is difficult [14:01] no easy answers [14:04] dimitern: yep, LGTM === lazyPower is now known as lp|BagelRun [14:04] dimitern: my mortgage guy is due here any minute and I'm going to make coffee [14:08] fwereade: o/ [14:08] anastasiamac, heyhey [14:09] fwereade: how r u? [14:09] anastasiamac, good thanks, and you? [14:09] fwereade: m good but weary [14:09] could I run smth by u? [14:09] anastasiamac, ofc [14:09] fwereade: about annotations... [14:09] anastasiamac, go on [14:10] fwereade: to meet kapil's requirments, we need to list charms [14:10] fwereade: I've added API call to current client [14:10] fwereade: bumping it's version but m not 100% sure it's good [14:11] fwereade: not sure if 1 call warrants a separate API either [14:11] anastasiamac, my gut would say to have a new facade [14:11] anastasiamac, hot on its heels will, I suspect, be requirements that we delete charms [14:11] anastasiamac, etc etc [14:11] fwereade: excellent! will change tomorrow :) [14:11] anastasiamac, and check the current api, there's a charm-info (or something) call [14:12] fwereade: shall i move it? [14:12] anastasiamac, which would be happy on that facade I think [14:12] anastasiamac, yes please [14:12] anastasiamac, which is to say don't *move* it, leave it where it is too, and then we don't need a version change on client [14:12] fwereade: purrrfect :) wallyworld_ and I were wondering but u've cleared it all up ;) [14:13] anastasiamac, cool [14:13] fwereade: thnx for ur time and insight :) [14:13] anastasiamac, np === lp|BagelRun is now known as lazyPower [14:50] anastasiamac, you've got a review on the ListCharms PR [14:52] wtf?! ERROR juju.cmd supercommand.go:411 failed to bootstrap environment: dial tcp 54.88.153.132:22: ConnectEx tcp: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond. [14:52] mgz, is the job itself flaky? http://juju-ci.vapour.ws:8080/job/win-client-deploy/1298/console [14:53] dimitern: sinzui has been wrangling with it [14:54] mgz, cool [14:54] dimitern, ssh on that machine sucks [14:55] sinzui, :/ [14:56] well, I hope there's still a chance to unblock master later today [14:57] dimitern: build-osx-client passed, so your fix was fine [14:57] voidspace, thanks for the review [14:58] * dimitern sings the happy dance ;) [14:58] or dances the happy song lol [14:58] eheheh [15:30] BLESS! [15:30] oh joy === ChanServ changed the topic of #juju-dev to: https://juju.ubuntu.com | On-call reviewer: see calendar | Open critical bugs: None [15:31] dimitern: yeay! unblocked [15:31] dimitern: thank youuuuu [15:32] perrito666, mgz, cheers guys! [16:52] wwitzel3: ericsnow: did either of you do a write-up on how you broke up your branch? i'm coming up on a large-ish change. [16:52] katco: we haven't gotten there yet :) [16:52] ericsnow: doh! [16:53] i suppose there must be something out there on google when the time comes [16:55] katco: but the idea I had is to remove everything I don't want in the review and commit that on the branch and then create the PR (and review request) [16:55] katco: then add back in the stuff for the next patch and commit in a new branch (and use rbt --parent) [16:55] katco: then repeat until it's back up to where it was before you started splitting things up [16:56] katco: that way you don't lose all your git history [16:56] ericsnow: i think my general hunch was the inverse: create a new branch and begin merging in certain files [16:56] katco: where to split things up is a different question [16:57] katco: though keeping each review request a manageable size is one key objective [16:57] ericsnow: yeah, that's the only reason i am doing it [16:57] katco: the problem with that is the loss of history [16:57] ericsnow: very true [16:58] katco: right now our branch has over 100 commits in it (I think) [16:58] ericsnow: wow that's awesome [16:58] katco: and most of that is valuable information [16:58] yeah, I'd like to avoid losing our commit history [16:58] katco: wwitzel3 and I have taken the approach of keeping commits focused [16:59] ericsnow: i aspire to be that diligent, but i always get trapped in the "gosh can this really be an atomic commit" [16:59] ericsnow: as in, could this functionality stand alone [16:59] wwitzel3: one alternative is to manually generate the patches and create the review requests manually and then merge the whole branch when *all* the requests get a ship-it [16:59] wwitzel3: but that seems like more work [16:59] ericsnow: i think i need to shift my thinking on that somehow [17:00] katco: sometimes you can and sometimes you can't :) [17:00] katco: usually, though, you can :) [17:00] ericsnow: :) [17:00] katco: try using the "-p" option to git add [17:00] yeah -p for the win [17:01] katco: we've used it to great effect when we've made changes that we want to split into multiple commits [17:01] ericsnow: yeah, honestly, I think we could probably, in most cases just do file and file_test as a commit. [17:01] ericsnow: ah cool, so you're picking out bits of files? [17:02] ericsnow: not commit, review [17:02] katco: yeah, we just select the bits that make up an atomic commit [17:03] katco: I've found it much easier to do when you have all of the context, instead of trying to do with with forsight [17:04] ericsnow: anyeay, I agree it is more work, but I think we've done a good enough job of file seperate that it is probably the best approach [17:06] wwitzel3: i agree. i cannot with any reliability land slices of the stack without having first coded the entire stack. there's just too much back and forth as the design emerges. [17:50] does anyone know how to prevent google chrome from popping those notifications before meetings?? [17:51] perrito666, in the calendar - edit the event and uncheck reminders [17:51] dimitern: What I actually want is to stop chrome from breaking the 4th wall [17:51] perrito666, Notifications -> [ Pop-up ] [ 10] [minutes] (x) [17:52] perrito666, whaa? :) [17:52] dimitern: I dont want to be able to pop notifications into my window manager [17:52] I mean chrome [17:54] There is a small bell that shows in the indicator and a pop up thing tells me that I have a meeting, I dont want chrome to do that , it is invasive and also clashes with my ubuntu calendar [17:54] perrito666, aah! [17:55] perrito666, those .. I knew how at some point and managed to turn them off, but I forgot how [17:55] it is not obvious in the ui [18:07] rick_h_: there is a nice lady talking in yourbehalf in the call [18:15] dimitern: decorator fail : http://maas.ubuntu.com/docs/api.html [18:37] voidspace, I can't see anything unusual :) [18:39] dimitern: :-) [18:42] voidspace, all my branches merged, finally [18:43] dimitern: great [18:44] dimitern: I got sidelined into mortgage applications - didn't make as much progress as I should have [18:44] dimitern: but for once I have nothing planned this evening so might do some more in front of the TV [18:44] dimitern: I got the signatures changed, but not much on the implementation for MAAS / EC2 [18:44] voidspace, no worries, I wanted to unblock you before I go today [18:44] dimitern: easy stuff though [18:44] dimitern: yeah, thanks [18:45] ah, and the new goamz branch [18:45] cool [18:45] voidspace, yep - shiny and squeaky [18:45] Nice :-) [18:45] :) [18:46] dimitern: so, I already had booked a "geek dinner" in London of the Thursday that we're together in London [18:46] ok, I'll be off then - a long day that was [18:46] dimitern: you'd be welcome [18:46] dimitern: go have a break [18:46] voidspace, nice! [18:46] yeah, saves me a trip if I'll already be in London... [18:46] voidspace, awesome :) [18:47] I'm signing off for now [18:47] going jogging [18:47] g'night all [18:47] voidspace, enjoy [19:25] perrito666: who da what? [19:25] perrito666: nice lady? [19:28] rick_h_: on the call we where hearing an operator voice [19:29] ah lovely [19:29] what a mess [20:14] perrito666, do you have a few minutes to review http://reviews.vapour.ws/r/744/ [20:16] abentley, jog: I have my lull. I will review abentley's branch first [20:17] sinzui, that's fine, abentley, said he'd look at mine === kadams54 is now known as kadams54-away [21:15] katco: hey, what was the magic invocation to get different output formats in juju? [21:15] and which version is that in? [21:16] perrito666: juju status --format [21:16] perrito666: juju help status will give you all the options [21:16] perrito666: and i believe it's in >= 1.22 [21:24] didmitern: r u still around? [21:26] katco: tx a lot [21:27] perrito666: np, enjoy :) [21:29] perrito666, do you have a minute to review http://reviews.vapour.ws/r/744/ [21:30] sinzui: sure [21:30] anyone local provider savvy? === urulama is now known as urulama__ === TheRealMue is now known as TheMue === meetingology` is now known as meetingology === Tribaal_ is now known as Tribaal === mwhudson_ is now known as mwhudson === Viperz28 is now known as Guest82007