[00:19] anyone? https://github.com/juju/juju/pull/247 [00:20] * thumper remembers that axw and wallyworld are in the states this week [00:44] thumper: https://github.com/juju/juju/pull/248/files [00:44] waigani: https://github.com/juju/juju/pull/248 [00:46] davecheney: LGTM [00:47] waigani: thank [00:57] davecheney, waigani: https://github.com/juju/juju/pull/249 [01:08] FUCK [01:08] + exit 0 [01:08] All passed, sending merge [01:08] /tmp/hudson6791049747065303228.sh: line 44: /home/ubuntu/jenkins-github-lander/bin/lander-merge-result: No such file or directory [01:08] /tmp/hudson6791049747065303228.sh: line 44: /home/ubuntu/jenkins-github-lander/bin/lander-merge-result: No such file or directory [01:08] Build step 'Execute shell' marked build as failure [01:08] Description set: davecheney 128-update-ssh-terminal [01:08] build bot is broken [01:09] thanks gz [01:09] thanks for testing that [01:17] :( [01:26] davecheney: is the bot completely stuffed? [01:27] like... nothing will land? [01:27] because that would kinda suck [01:31] yes [01:31] that [01:37] thumper: also, the bot isn't reporting the build failure [01:38] double :( [01:38] +1 to making changes on a friday night [03:09] hm [03:09] i wonder why juju thinks these mustang board only have 1 cpu each [03:09] "hardware: arch=arm64 cpu-cores=1 mem=16061M" [03:31] davecheney: is this fixed from your crypto update: https://bugs.launchpad.net/juju-core/+bug/1261780 [03:31] <_mup_> Bug #1261780: go 1.1.2 TLS-enabled client does not accept our CACert

[03:31] davecheney: or is it likely to be fixed? [03:48] davecheney: merged \o/ [03:49] davecheney: well, yours was... waiting for mine === vladk|offline is now known as vladk [06:14] vladk: you're cutting out a bit, do you want to drop your video? === urulama_ is now known as urulama [06:59] dimitern: I have to go help my wife for a sec, back in a couple of minutes [07:00] jam, that's ok, i'll be late 15m as well, sorry === vladk is now known as vladk|offline [07:01] dimitern: k, I'm there now, but I'll just sit around until you get here [07:04] dimitern: also a reminder that you're OCR today === vladk|offline is now known as vladk [07:11] morning === rogpeppe1 is now known as rogpeppe === rogpeppe1 is now known as rogpeppe [07:54] dimitern: so I set up the 360 reviews for yourself and Frank. I can't actually set one up for myself, but you should have something that you can interact with. [07:54] they have a specific step-by-step of questions to answer. [07:55] jam, thanks, will have a look later [08:05] dimitern: as OCR one quick PR https://github.com/juju/juju/pull/252 , only about renaming a dir inside /doc [08:07] TheMue, looking [08:07] dimitern: is a bit in context of the 250 PR by Jesse ;) [08:07] TheMue, why contribution and not contrib? [08:08] TheMue, it might mean contribution or contributions [08:08] dimitern: couter question, why abbreviating if there’s no need? [08:09] TheMue, i'd prefer the plural form unless this will be the only document in that directory [08:09] dimitern: renaming it to contributions sounds reasonable, yes [08:09] TheMue, LGTM, thanks [08:10] dimitern: ok, will change to plural, convinced :) [08:13] :) [08:50] win14 === vds` is now known as vds [09:26] dimitern, TheMue, vladk, voidspace: we should also remember that today is the start of a cycle, so we want to be thinking about how we're going to break down the next 2 weeks of work. [09:27] jam, right, IPv6 is still a priority? [09:27] dimitern: yeah, AFAIK that's still the plan is to just focus on IPv6 support and push that through [09:28] jam, all right then [09:39] anyone able to rubberstamp this PR, please? it just updates the charm store to use the new package path, gopkg.in/juju/charm.v2 [09:39] dimitern, jam: ^ [09:40] rogpeppe, which one? [09:40] dimitern: https://github.com/juju/charmstore/pull/12 [09:40] dimitern: ta! [09:40] rogpeppe: I'm actually not confident in using gopkg.in, I've heard some reservations as it adds a dependency on another 3rd party website that may have some uptime issues. [09:40] rogpeppe, liiking [09:40] jam: we've discussed this a fair amount, and there seemed to be general agreement it was ok [09:41] rogpeppe: if so, then you already have LGTM from frankban [09:41] jam: we did it together, so that's kinda not really sufficient :-) [09:41] rogpeppe, LGTM [09:41] dimitern: thanks [09:43] dimitern: TheMue: and michael's gone this week for katco's induction sprint, right? [09:43] jam, yes, fwereade is gone this week as well [09:43] jam: yep [09:54] jam, will you be available tomorrow @9 UTC for a g+ with the some of the maas and server teams re juju+maas networking? [09:55] dimitern: I can try, I might have to run an errand at that time, but if I don't I'll happily be there. [09:55] I think I'll be back by then [09:56] jam, great, i'll send the invite [10:01] TheMue: i accidentally deleted our meeting for today, but obviously we already had it :) [10:08] jam: hehe, I had the alert in my mailbox [10:45] TheMue: dimitern: vladk: standup === vladk is now known as vladk|offline === vladk|offline is now known as vladk [11:54] dimitern, urulama, jam: large-but-mechanical branch that updates juju-core to use the new charm package API: https://github.com/juju/juju/pull/253 [11:54] reviews appreciated before the branch bit-rots... [11:58] rogpeppe, will take a look in a bit [12:02] rogpeppe: why are all bundles turned into CharmArchive, but in juju/conn.go in 207 Bundle is just an Archive? [12:05] rogpeppe: also in state/api/client.go in line 629 ... if this is because of BundleArchives in the future, is this "Archive" any type (bundle or charm) or must it be called BundleArchive? [13:05] morning all :) [13:09] morning [13:09] a lot of the US devs are in a room in Lexington, MA [13:12] urulama: ArchiveTo is a method that can apply equally to charms or bundles [13:12] urulama: and it does the same thing in both cases (write the contents as a zip file to the given file) [13:13] urulama: so I think it makes sense to call it the same thing (it's possible we might want to make an interface that includes that method in the future) [13:13] urulama: does that make sense to you? [13:15] rogpeppe, ping? [13:15] mattyw: pong [13:16] rogpeppe, quick one - I was searching the juju api for any command that will let me find it a machine id exists [13:16] from a client command [13:17] looks like status/ fullstatus is all there is for now [13:17] mattyw: one mo, let me look [13:17] hi sinzui, are you familiar with bug 1316174 ? [13:17] <_mup_> Bug #1316174: 'precise-updates/cloud-tools' is invalid for APT::Default-Release [13:18] mattyw: i think you're probably right [13:19] bac, a precise machine that will be localhost must add the ctools archive, or the juju stable archive. [13:20] sinzui: i don't think juju/stable is sufficient. [13:20] bac. When it was clear many precise users were getting modern juju, but not using modern deps from ctools, I copied the lxc and mongo deps to the juju stable ppa [13:20] bac, yes [13:20] bac, which juju are you using? [13:20] sinzui: 1.20 [13:21] bac? My rebuild? [13:21] which i got from juju/stable and still see the issue. i can bring up my precise vm and confirm [13:22] bac, "apt-cache madison juju-local" to be certain [13:22] bac For 3 hours, precise and saucy had a juju-local packages with the wron mongo deps [13:25] sinzui: i do not have access to the precise VM i used on friday so i cannot say for sure what i had installed. [13:26] okay [13:27] bac 1.20.0-0ubuntu1~12.04.2~juju1 (note the .2 after the series) is the one that was built to replace the broken version [13:29] bac: Your vm does need juju stable ppa to get the right mongo. The bug doesn't affect many people anymore, so I was wondering if your machine got unlucky and saw the old package from Thursday [13:30] sinzui: i don't think so. i have the juju-local you saw. http://paste.ubuntu.com/7760165/ [13:30] s/saw/referenced/ [13:51] sinzui: did you get a chance to look at the paste ^^ ? [14:05] bac, does the machines have the archive? In the pastebin example, Juju will only install from mongodb from that archive. [14:05] sudo add-apt-repository cloud-archive:tools [14:06] oh, 2.4.9 in now in the tools. Maybe I need to backport that to juju stable [14:08] sinzui: adding that archive produces the same output as i pasted. [14:11] interesting in a bad way bac [14:11] bac, does the machine have precise-updated enabled? [14:12] sinzui: yeah. fwiw, i discovered this doing QA for juju-quickstart. if we discover juju is not installed we install juju-core and juju-local from juju/stable. if local on precise requires other repos we need to document that. [14:14] bac: agreed [14:19] so perrito666, wwitzel3 did you guys have a place to sleep last night? [14:20] alexisb: yeah, they let perrito666 in his room and I just let them auth my card. [14:20] alexisb: email this morning says it should be all sorted out now :) [14:20] heh [14:21] alexisb: I slept on the street :p [14:21] nothing like spending 24+ hours in the airport just to sleep on the street [14:21] alexisb: nah, they authorized a part of the fee on my card until the card locked them and then they trusted someone from the company would call this AM [14:22] btw, non us people http://www.usatoday.com/story/news/nation/2014/07/06/airport-security-electronic-devices-dhs-tsa-johnson/12266639/ [14:22] you better charge your things before leaving [14:28] good morning juju land. looks like i'm reviewing today! :) [14:34] hmpf, my local provider doesn’t start anymore [14:44] cmars: yay; wanna look at https://github.com/juju/names/pull/12 ? [14:44] :) [14:44] jcw4, sure, next on my list [14:44] cmars: ta [14:46] anyone know if rick_h__ is working today? [14:57] jam: short info, endpoint error is also on local [15:02] TheMue, jam, vladk, https://github.com/juju/juju/pull/254 - dummy IPv6 support [15:05] why is coludinit log in /var/log, shoulnt it be /var/log/juju/cloud-init-output.log instead? [15:05] dimitern: oh, cool, will take a look [15:05] TheMue, ta! [15:09] perrito666: cloud-init isn't juju specific [15:10] mgz: rick_h__ is out today, back tomorrow [15:13] mgz: tx [15:15] jrwren: thanks! [15:20] dimitern: you’ve got a review :) [15:20] TheMue, thanks! [15:21] TheMue, yeah, that net.JoinHostPort taking a string port sucks :) [15:35] dimitern: vladk and I still had comments. while answering to his comment I found a possible problem. [15:36] TheMue, yes? [15:36] dimitern: IMHO a missing slash when building a URL. [15:37] dimitern: storage.go line 157 [15:37] TheMue, the slash is part of s.path [15:37] TheMue, I haven't changed that and the tests still work [15:37] dimitern: ah, good to know, thanks [15:38] TheMue, np :) [15:38] dimitern: and the squeare brackets come with JHP() [15:38] TheMue, yep [15:38] * dimitern needs to step out === urulama is now known as uru-afk [16:01] Thanks cmars ; I have this pending change in juju https://github.com/johnweldon/juju/commit/50dc10fdc4575e0eee0743545541c2a2a174d42f [16:02] cmars: which might address your questions on https://github.com/juju/names/pull/12 [16:03] jcw4, thanks, taking a look to see how it will all fit together [16:03] tx [16:38] axw: https://github.com/juju/jenkins-github-lander/pulls [16:41] here's a PR to add some external verification to charm bundles. anyone care to take a look? https://github.com/juju/charm/pull/11 [16:42] mgz, natefinch, wwitzel3, voidspace: ^ [17:17] anyone know why we have build-essential in our Makefile dependencies? Just for gccgo? [17:24] sinzui: I thought I saw a LP regression get linked the other day, something that I had worked on. [17:27] mgz, to get make...ha ha [17:28] mgz, build-essential predates gccgo. CI installs it separately to get make [17:31] sinzui: okay, I'll just remove it then I think, really don't want to be installing x11-common and such like when we don't need to [17:31] wwitzel3, what do you think you worked on? [17:37] wwitzel3, This is the short list of open regressions https://bugs.launchpad.net/juju-core/+bugs?search=Search&field.importance=Critical [17:37] sinzui: thanks [17:38] sinzui: none of that is anything I worked on, so it might of already been resolved .. William pinged me on something right before I left for the long weekend and I've forgotten what it was. === vladk is now known as vladk|offline [17:38] sinzui: I just remember it was a regression. (or I thought it was) [17:39] wwitzel3, https://launchpad.net/juju-core/+milestone/1.19.4 lists the closed bugs from a few weeks ago === wwitzel3_ is now known as wwitzel3 [17:40] wwitzel3, All open regressions https://bugs.launchpad.net/juju-core/+bugs?field.tag=regression [17:41] wwitzel3, voidspace: the answer to the question, [17:42] sinzui: can we throw in --no-install-recommends as well... [17:43] mgz, interesting...that is why I keep removing tk [17:45] could anyone ptal https://github.com/juju/juju/pull/256 ? [17:45] * sinzui wont let tk, java, or erlang back on to his computer [17:46] sinzui: thanks for the links [17:46] sinzui: so, my instinct is to s/build-essential/make/ in the job, and add the --no-install-recommends and remove build-essential in the Makefile [17:46] * perrito666 imagines sinzui with a religious icon shouting "out you devil" to his computer [17:47] mgz, +1 [17:48] mgz, you may want to look at run-unit-tests and fix the other makefile issues for i386 [17:48] yeah, I see some dodgy seds :) [17:52] cmars: I addressed your comment on https://github.com/juju/names/pull/12 [17:52] jcw4, got it. about to approve, with some comments on the WIP branch [17:52] thanks [17:53] cmars: perfect... if you like I can open a WIP PR so you can more easily comment [17:53] jcw4, that'd be great, thanks [17:54] cmars: https://github.com/juju/juju/pull/257 [17:54] wwitzel3: Do you know about http://irclogs.ubuntu.com/2014/07/03/%23juju-dev.html ? [17:55] wwitzel3: going to Friday: http://irclogs.ubuntu.com/2014/07/04/%23juju-dev.html [17:55] wwitzel3: wwit [17:55] wwitzel3:https://launchpad.net/bugs/1289316%3E [17:55] <_mup_> Bug #1289316: lxc not installed from ubuntu-cloud.archive on precise

[17:57] jam: wwitzel3 is currently afk working with Ian [17:58] jam: you can add anchors for the timestamp: http://irclogs.ubuntu.com/2014/07/03/%23juju-dev.html#t02:24 === alexisb is now known as alexisb_lunch [18:12] jcw4, commented, I wonder if some of the parsing functions in state/action.go should go in names. what do you think? [18:13] cmars: let me look; In my mind I've been trying to keep the names and the state package conceptually separate, and make sure that names doesn't need to know anything about the internals of state [18:14] jcw4, ah, that makes sense. if you're just working with a naming convention in these methods though, maybe that is general enough to go in names? [18:14] s/methods/functions/ [18:17] so I have backup and restore packages and there are a bunch of things that both share, what is recommended, do I create a utils package for those? or do I define them in one of the 2 and then import them from the other? [18:17] natefinch: voidspace mgz thoughts? [18:19] cmars: yeah; maybe I can abstract it enough that it's not state specific [18:20] jcw4, see if it makes sense & feels right. if not, i'm ok with landing https://github.com/juju/names/pull/12 as is [18:20] tx cmars [18:20] thank you jam [18:31] sinzui: [18:32] * sinzui looks [18:33] I have just push --overwrittedn a tyop in the lp branch [18:37] mgz, r=me and merged...check that tip is as you expect [18:38] cmars: on the one hand none of the other Tags make an effort to keep the internal representation of _id separate from the public id [18:38] cmars: but in this case fwereade also made it clear that the internal representation of the action id shouldn't be exposed [18:39] cmars: I've treated those duplicated functions as coincidentally the same rather than fundamentally the same [18:39] jcw4, fair enough [18:40] cmars: but I've also simplified the actionId on the state side now so that it's clean enough to use as the public Id [18:40] sinzui: thanks! tip looks good, pulled on the jenkins master. [18:41] cmars: If it seems okay with you I'll land the names change as it is and keep the possibility of sharing some of the parsing functions in the backlog? [18:41] https://github.com/juju/juju/pull/259 [18:41] katco: trade you, https://github.com/juju/juju/pull/258 [18:42] jcw4, perfect, thanks. i've lgtm'd [18:42] cmars: ta [18:47] mgz: LGTM, with reservations ;) [18:58] Does the instructions in juju-core's README.md on Github pull from -head? I'm trying to deploy juju, but when i run "juju generate-config" it gives a panic. [18:59] hasues: they do, can you pastebin the panic please? === vladk|offline is now known as vladk [19:01] perrito666: http://pastebin.com/fEA2Zp70 [19:02] hasues: is that linux? [19:06] PTAL https://github.com/juju/juju/pull/257 [19:08] here's that article on go that mgz recommended: https://plus.google.com/u/1/+JonathanLange/posts/Mv5V4r7m9nN [19:08] here's that article on go that mgz recommended: https://plus.google.com/u/1/+JonathanLange/posts/Mv5V4r7m9nN [19:11] voidspace: :P [19:11] mgz: I'm sorry [19:14] mgz: I thought voidspace was throwing you under the bus a bit with that link [19:14] :-) [19:14] jcw4: only deliberately.... [19:14] haha [19:29] hasues: ? === alexisb_lunch is now known as alexisb [19:59] mgz: howdy, this around the lander stuff? /me saw some pull requests but been in the wood without interwebs [20:03] rick_h__: yeah, I've put up pull requests for all the changes I make to the lander for juju, if you could find a mo to look at them that'd be ace [20:04] we do a few things differently to juju-gui (mostly around the tests not running till the merge is actually flagged), hopefully the proposals should make that clearer for you === vladk is now known as vladk|offline === vladk|offline is now known as vladk [20:11] wallyworld_: https://github.com/wwitzel3/juju/compare/019-environmentserver [20:15] perrito666: Yes, it was Linux. Sorry for the delay [20:18] hasues: may I know the version? [20:18] * perrito666 smells a bug [20:19] perrito666: Of linux or go? [20:19] hasues: linux [20:19] perrito666: Gentoo [20:23] perrito666: kernel 3.15.0 [20:23] perrito666: Go version 1.2 [20:24] hasues: that's your issue then... you'll need to teach our versions code about gentoo [20:25] mgz: cool will definitely give a look [20:25] mgz: noted. === vladk is now known as vladk|offline [20:29] hasues: see version/osversion.go [20:30] mgz: Thanks, I was browsing for something of that nature :) === mwhudson_ is now known as mwhudson [21:29] alexisb: morning [21:30] hey there waigani [21:30] on my way [21:31] waigani, are you on? [21:31] alexisb: i was... connection just dropped [21:31] alexisb: give me 2min [21:31] k [21:34] mramm: oh hai [21:34] mramm: weren't we going to have a talk earlier? [21:43] * thumper tries to distract himself with work [21:55] thumper: any idea why local provider bootstrap would exit cleanly on utopic, but the api state server wont stay started? It's not even creating the machine-0.log, so it never finishes the init phase before it bails. [21:56] lazyPower: not off the top of my head [21:56] lazyPower: but if mongo failed to start, the machine agent wouldn't [21:56] lazyPower: check the juju-db [21:56] juju-monogd is started [21:56] if you want to follow along - its happening in #juju on canonical irc [21:56] i'm stumped [21:57] lazyPower: I have a call shortly, but could help after === vladk|offline is now known as vladk === vladk is now known as vladk|offline [23:01] thumper: yo [23:03] davecheney: around? [23:03] wallyworld: hey [23:07] thumper: how goes? [23:33] rick_h__: hey [23:43] thumper: hey, back from the woods with no interwebs. Will do my best to get the emails/docs you need from me during work day tomorrow === Ursinha is now known as Ursinha-afk