[00:00] wallyworld_: but I think RB has versioned several times since then [00:00] menn0: yeah, it's not a valid command, yet the email says to use it, so i'll have to dig into the doc [00:00] wallyworld_: which version of the RB tools do you have installed? maybe you ended up with an older version somehow? [00:00] 0.6.2 [00:01] that the latest [00:02] the docs here don't mention a pull command: https://www.reviewboard.org/docs/rbtools/dev/ [00:02] I wonder if Eric meant "post"? [00:03] wallyworld_: that's the command I remember using the most. it's what creates and updates review requests. [00:03] could be, i was wondering the samw === meetingology` is now known as meetingology === jcsackett_ is now known as jcsackett [00:17] wallyworld_: cheers for the review [00:17] np === bigjools_ is now known as bigjools [01:23] wallyworld_: I'm back at limited capacity - whole house has had the flu === axw__ is now known as axw [01:23] oh no :-( [01:24] take it easy then [01:26] yuck, hope you feel better axw [01:27] wwitzel3: thanks, I'm mostly better, just a bit run down. [03:01] menn0: you're going to like my next PR [03:01] all the state package _test_'s have stopped importing the api [03:02] davecheney: ##\o/## [03:03] menn0: basically inside most of the tests we do somethingl ike [03:03] m := state.NewMachione [03:03] api.ChangeMAchine(m.Id() [03:03] assert m.Thing == expected change [03:03] what is clear after my change is that these two methods are indepdent [03:03] ie, you ask the api to do something [03:04] but the effect on the state is indepdented [03:04] independant [03:04] as identified by the fact that the data type of the state is not the same as the api [03:05] sounds good. I look forward to seeing it. [03:29] * menn0 is hating reviewboard right now [03:39] wallyworld_: when you have a chance: http://reviews.vapour.ws/r/20/ [03:40] menn0: sure, i just noticed it :-) [03:40] first go at using review board [03:41] wallyworld_: I had a lot of trouble creating that request. it turns out the "master" branch of your personal GH repo (i.e. not your local master) has to be up to date or else you end up creating reviews for stuff that's already committed. [03:41] wallyworld_: that's pretty inconvenient [03:41] wallyworld_: I'll reply to Eric's email [03:41] menn0: that sucks actually [03:42] a lot [03:42] wallyworld_: there must be a better way [03:42] yep, there had better be [03:42] wallyworld_: I don't recall this being a problem when I've used RB before but that was with git-svn, not pure git. [03:42] i just wish we used launchpad still :-( [03:43] it handles this stuff well [03:43] * menn0 nods [03:46] menn0: why are we using reviewboard again ? [03:47] davecheney: AFAIK, because Github reviews are bit sucky: e.g. as soon as you write a comment, that's it the whole world see it without a chance for you to delete/edit [03:47] menn0: is that the only reason ? [03:47] davecheney, wallyworld_: hopefully this issue I've found today can be sorted out. it could just be a setup thing. [03:48] hope so too [03:48] davecheney: I think there were others but I'm not sure what they are. I haven't been too involved with the decision. [03:48] davecheney: there's other reasons too [03:48] eg lack of decent review queue [03:48] so we can have depdenent branches ? [03:48] that too [03:49] oh well, we can't know these things til we try them [03:50] davecheney, wallyworld_: RB is pretty flexible about where it gets its diffs from so hopefully this issue I've run in to is fixable [03:50] menn0: it feels like the RB <> GH integration is wrong [03:50] the other systems 've used integrate into the GH review, not the other way around [03:50] i have no idea how they do this [03:50] they jsut do [03:51] cf. werker and travis [03:51] they use the GH API [03:51] travis was ruled out because it doesn't test properly [03:51] we can improve the integration as Eric indicated in his email [03:51] by taking advantage of GH's and RB's APIs [03:52] I'm guessing we're going to end up with something like "lbox propose" that ties together or GH and RB usage [03:52] s/or/our/ [03:53] and the circle is complete [03:53] all roads lead to lbox [03:57] I was kinda hoping we'd just do a pull request, and that would cause a RB review to be created [03:57] is the reviewboard tool manditory ? [04:02] % rbt pull [04:02] usage: rbt [--version] [options] [] [04:02] rbt: error: 'pull' is not a command [04:02] what did I do wrong [04:02] this is step 1 [04:04] axw: yeah, me too [04:04] davecheney: there's a typo in the email - it should be post [04:04] % rbt post [04:04] ERROR: There don't seem to be any diffs! [04:04] is this because I need to keep pushing maste3r ? [04:04] not sure, it just worked for me [04:04] you need a gh pr first [04:05] done that [04:05] https://github.com/juju/juju/pull/752 [04:05] i tried to create ie from the gh diff [04:05] https://github.com/juju/juju/pull/752.patch [04:05] menn0: first review done using rb, did you get an email? [04:05] but reviewboard doesn't like it [04:05] i've not tried using a patch [04:06] i just used the rbt tool and it worked [04:06] wallyworld_: no email :( [04:06] wallyworld_: I did configure an email address for my account earlier today [04:06] menn0: well, that sucks, i just finished, so maybe it will get there eventually [04:06] error: The following untracked working tree files would be overwritten by merge: .reviewboardrc [04:06] davecheney: that file was added to master i believe [04:07] so step 2 of hte instructinos is also wrong [04:07] so if you pull master, you'll get it [04:07] could be, i just created the rc file in my home directory [04:07] ahh, so that file needs to be in every branch [04:07] that makes sense [04:08] it seemed to find it in my home dir, but there's also a copy checked in which i don't quite understand [04:08] still can't make it work [04:09] i can't help much - i know very little about rb [04:13] http://reviews.vapour.ws/r/21/diff/# [04:13] oh that is brilliant [04:13] if you have a large diff, rb gives up [04:13] and only shows the diff if you ask explicitly [04:18] wallyworld_: https://code.launchpad.net/~axwalk/gomaasapi/maas-testserver-bootimages/+merge/234614 [04:18] lookng [04:18] davecheney: send an email to eric and the list so it can get fixed [04:18] will do [04:27] menn0: I think I've got services migrated. just doing a final make test ... so far so good! [04:29] waigani: sweet. I'm keen to see the PR. [04:29] * menn0 has to deal with his youngest for a bit [04:29] menn0: just give the tests a day or two to run... [04:31] axw: code looks good, but we need a test or two [04:32] ok [04:32] menn0: nop, rain check on that [04:32] tests for tests [04:33] forgot about that :) [05:11] wallyworld_: I've replied to some of your comments here: http://reviews.vapour.ws/r/20/ [05:11] ok [05:15] menn0: i saw one reply which i replied to - are there others? [05:18] wallyworld_: it was just that one. I wasn't sure if you're getting emails from RB. [05:21] menn0: no, not yet. i just wanted to check that since "some" implies > 1 :-) [05:21] wallyworld_: regarding that issue, I don't like --resolve-previous-error because it's not clear that the error is about a previous upgrade attempt. [05:21] maybe reset-previous-upgrade ? [05:21] wallyworld_: yeah sorry I thought there was more. I have more planned in my head regarding the other thing you raised :) (I agree with everything) [05:22] wallyworld_: bingo. reset-previous-upgrade is the best one yet. [05:22] \o/ [05:22] naming things and cache invalidation :-) [05:22] 2 of the hardest problems [05:22] or cache consistency [05:23] isn't it: there are 2 really hard problem in computer science: [05:23] 1. naming things [05:23] 2. cache consistency [05:23] 3. off by one errors [05:23] ? [05:23] :) === uru_ is now known as urulama [05:38] morning all [06:19] wallyworld_: I'm going to change hte way I'm doing the MAAS fix. I'll just query the available nodes, and get the set of architectures from them [06:20] axw: instead of querying each cluster controller and asking for available images? [06:22] axw: bbiab, gotta go do school run [06:25] axw: don't do that, it's not scalable [06:26] bigjools: is there a theoretical limit on the number of nodes in MAAS? [06:26] axw: no [06:26] mk [06:26] did you have a problem querying clusters' images? [06:27] bigjools: I'm having trouble with /api/1.0/nodegroups//boot-images/ [06:27] I just get 404, despite feeding a UUID I got from nodegroups/?op=list [06:27] The requested URL /MAAS/api/1.0/nodegroups/356f23e1-5879-41fc-be81-e5f601428772/boot-images/ was not found on this server [06:28] umm [06:30] axw: do you have a new enough maas? [06:31] I can't remember exactly when this feature was added [06:31] but it's new-ish [06:31] bigjools: I have 1.4 [06:31] ah you need 1.5 [06:31] I rhink [06:31] think [06:32] it's in trusty [06:32] ah, doh [06:32] so then we need a fall back to listing nodes anyway. bleh [06:33] thanks bigjools [06:33] urgh [07:00] jam1, 1:1 ? [07:00] dimitern: yeah, I'm just trying to workout my firewall stuff, brt [07:00] jam1, np [07:02] jam1, if you're having trouble, we can try mumble instead? === tasdomas` is now known as tasdomas [07:38] wallyworld_: was a little more complicated than I expected: https://github.com/juju/juju/pull/754 [07:39] looking [07:46] axw: not sure if it's simpler to record boot images in a map and avoid the need for using the "seen" set. I guess a loop would then be needed at the end to extract the values so probably not a lot of difference [07:48] wallyworld_: yeah, much the same either way. [07:51] axw: nice, thank you :-) [07:52] wallyworld_: cheers [07:52] i'll backport when it lands [07:53] thanks [07:53] I'm going to test my charm storage changes live, and make the cleanups you suggested [07:53] hopefully be able to land later on [07:53] ok, after you were rudey distracted onto the maas thing [07:54] heh, it's fine :) [08:40] morning folks [08:48] TheMue: morning. I might be a little late to our 1:1 as I'll need to take my dog out, but I'll be close to on time. [08:49] jam1: just ping me then === wesleyma` is now known as wesleymason [09:28] TheMue: ping pong, walla walla ding dong [09:28] jam1: ouch, that hurts in my ears :D [09:28] jam1: omw === Spads_ is now known as Spads [10:47] dimitern: TheMue: standup? [10:48] jam1: ouch, omw [10:48] my browser just hung, so I might be a sec [10:49] jam1, sorry, omw === Spads_ is now known as Spads === jam1 is now known as jam === Spads_ is now known as Spads === Ursinha is now known as Ursinha-afk [12:30] * fwereade lunch === Ursinha-afk is now known as Ursinha === tdc_ is now known as tdc === makyo_ is now known as Makyo [15:38] ericsnow, ping? [15:38] mattyw: here [15:39] ericsnow, hey there - forgive my inexperiecne with review board - do I need to be running rbt setup-repo in each branch before I push? [15:40] mattyw: only if the branch isn't based on master [15:40] ericsnow, ok - yeah this one isn't [15:40] mattyw: I pushed a .reviewboardrc to our juju repo that should take care of it [15:40] ericsnow, I guess we should probably add .reviewboardrc to the .gitignore file as well? [15:40] mattyw: I haven't done so for our other repos [15:41] ericsnow, I'm playing with a branch that has a prereq [15:41] mattyw: not necessarily [15:41] mattyw: but that should still be based on master (ultimately), right? [15:41] ericsnow, yes === seelaman` is now known as seelaman === ev__ is now known as ev === gnuoy` is now known as gnuoy === jheroux_away is now known as jheroux [16:50] ericsnow: hey i'm doing something wrong with oath and rbt. username is "katco-@github"? and password is what? github pw? [16:50] s/oath/oauth/ === Ursinha is now known as Ursinha-afk [16:51] katco: try "katco-" for the username and "oauth:katco-@github" for the password [16:51] ericsnow: k sec [16:52] ericsnow: looks like that worked. thanks! [16:52] katco: cool === Ursinha-afk is now known as Ursinha === hazmat` is now known as hazmat [19:01] natefinch, on and ready when yo uare === Ursinha is now known as Ursinha-afk [19:02] alexisb: coming === Ursinha-afk is now known as Ursinha === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [21:02] mramm: morning [21:02] mramm: I got my physio time wrong [21:02] mramm: and I am around now [21:02] thumper: ok [21:02] sounds good [21:02] are you free in 30? === Guest31361 is now known as wallyworld [21:03] yep [21:03] mramm: as in minutes not seconds? [21:03] yea [21:03] * thumper nods [21:04] ok [21:04] * thumper stares in horror at his email queue, it was only a few days [21:07] but we missed you thumper [21:08] so you thought you'd send me heaps of email? [21:08] yay [21:08] =D [21:09] thumper: talk go well? [21:09] seemed to [21:09] nice [21:09] i ran into kit, he said he went up [21:12] hey, I ask this again in case new people is here, Do you know how is the path /var/lib/juju/agents/unit-- generated? [21:12] I have a unit name and I would like to do something nicer than just demi hard code it === Ursinha is now known as Ursinha-afk [21:14] thumper: meeting over, can we meet now? [21:15] mramm: yup === Ursinha-afk is now known as Ursinha [21:15] perrito666: how? yes it is based on the string value of the unit tag === urulama is now known as urulama-afk [21:34] TheMue: nice work on the API implementation guide. I wish that was about when I started! === jheroux is now known as jheroux_away [21:45] katco: meeting? [21:51] wallyworld: oh hey... sorry it wasn't on my calendar. one sec./ [21:52] wallyworld: curious... i can't find it in the past either? can you toss me a link [21:52] katco: https://plus.google.com/hangouts/_/canonical.com/ian-katherine [23:26] thumper: Params passed to UserManagerAPI.AddUsers [23:26] https://docs.google.com/a/canonical.com/spreadsheets/d/1cUmTCpsTdso5YJ7Dxvluz_E29Q2P4L__AeGZnP5Q9Ck [23:26] ta [23:26] thumper: [23:26] Params passed to UserManagerAPI.RemoveUser [23:26] https://docs.google.com/a/canonical.com/spreadsheets/d/1r-fKcMJw5U5dVqS2wpu5Z9xAB1yH7S2EQPQuqQsKCxw [23:26] thumper: Params passed to UserManagerAPI.SetPassword [23:26] https://docs.google.com/a/canonical.com/spreadsheets/d/1RS8bmU-g__8z7PnJtiekbl-ULG3arhi39KDPa9E-_lo [23:27] Params passed to UserManagerAPI.UserInfo [23:27] https://docs.google.com/a/canonical.com/spreadsheets/d/1caY03J8AalWXNXZ3Lh8DzE4LLYy99HYL1LaQ2UOSwh4 [23:37] thumper: https://docs.google.com/a/canonical.com/document/d/1wcu6ufHX_x7CM602VJn-3pgHzFMfuEz_DvxfvXX4Q3s/edit#heading=h.f5zva1iwyydv [23:38] arm status [23:40] thumper: WIP https://github.com/waigani/juju/compare/add-env-uuid-to-collections [23:41] thumper: I'm not clear on the scope of serviceGlobalKey (and all GlobalKeys for that matter) [23:42] thumper: I've appended the envuuid to serviceGlobalKey assuming global means global, not 'environmental' [23:44] davecheney: ta [23:44] waigani: thanks, I'll look shortly [23:49] * thumper -> physio [23:50] davecheney: hey, thanks for reviewing katherine's juju status branch but it was done on github not reviewboard so that make it a little hard to collaborate now === mup_ is now known as mup [23:54] wallyworld: sorry, just saw the email [23:54] np :-) === Ursinha is now known as Ursinha-afk