[00:27] hatch: ugh, ok let's chat in the morning and get a plan together [00:28] +1 [00:28] our brains need to rest a bit tonight :) [00:29] hah rick_h_ I think the CI is stuck [00:29] looking [00:29] I'm not entirely sure what it's even running [00:30] all good, killed it. It's done that once/twice before [00:30] just login and you can stop the in progress job [00:30] ohh ok cool thanks === TheRealMue is now known as TheMue [01:11] so long everybody! I'll miss you. :-) === gary_poster is now known as gary_poster|away [11:19] morning [11:31] rick_h__, g'morning [11:40] hazmat: still an early bird? little ones not sleeping in for you yet? [11:42] rick_h__, today was all me.. up at 2 sore throat, and lots to do. [11:42] ugh, /me sends some meds down your way [13:27] morning hatch [13:27] morning [13:49] * benji_ takes the car to the shop. [14:01] rick_h__ so thoughts on upping the timeout and putting that chrome bug link? [14:01] hatch: yea, let's chat. I've got a few notes to run through [14:02] ok coffee is on, 2 mins [14:02] :0 [14:02] :) [14:02] doh [14:02] ok, shoot me a link [14:03] bleh it's taking forever https://plus.google.com/hangouts/_/76cpjp39uok6fu53clukspafeg?hl=en [14:27] hatch: I can dupe :( [14:27] that seems like a blocker, no? [14:27] :( [14:27] * hatch was really hoping it was his branch [14:28] hatch: it means it goes to 'feature flagged' :/ [14:28] so we can still release pretty relation lines for MWC demos [14:28] and they can demo in chrome ok [14:28] do we want to 1.0 without local charms? [14:28] no, it'll have it, just not public. We won't blog/etc it [14:28] but the MWC people can demo it [14:28] I am not sure pretty relation lines are 1.0 worthy :) [14:29] yea, true I guess [14:29] so maybe it's just another rev, but we need a release for MWC [14:29] no offence to the pretty lines of course :D [14:29] oh yeah I totally agree [14:30] hatch: any chance you can look at that today and work on the upgrade afterwards? [14:31] hatch: if it's a quick fix maybe we can get it in a release today for monday? [14:31] yep I can, although frankban wrote it, he might be faster if he has the bandwidth [14:31] if not, I can get on it [14:31] we know the upgrade story will take on [14:31] but created #1283067 [14:31] <_mup_> Bug #1283067: dropping a local charm in Firefox does not deploy it into the sandbox [14:32] frankban: do you have the time to peek at this bug and see if there's a quick fix we can get today for release? ^ [14:32] * hatch grabs his coffee [14:33] rick_h__: yes I'll take a look [14:33] thanks frankban [14:33] cool, [14:33] rick_h__: I had hoped sleeping on it would provide more insight, but I can't get the multi-field indexing approach to work (well it doesn't break anything but it doesn't do what we want), so I'm afraid doing this quickly is a bust. We need to learn more about ES. === benji_ is now known as benji [14:34] benji: ok, did you and bac get a chance to catch up yesterday? [14:34] benji: he's going to be out Mon/Tues so if we can get that doctype search working for Monday deploy that'd still be a partial win [14:35] since he was going to be out early next week I figured if I got this working it would have to be someone else so I didn't talk to him. I wasn't aware that I would be taking his branch. [14:35] yea, understood. 0/2 for monday [14:37] rick_h__: I bet I can figure out his branch though; should I pick that up next? [14:37] benji: I was just looking to see if he pushed up anything before he left [14:37] I'm not seeing anything other than his https://code.launchpad.net/~bac/charmworld/improve-bundle-search/+merge/207290 [14:38] * benji looks [14:38] benji: if you can run with it and get that up and going I'd appreciate it. I can get a charmworld deploy done monday and that would be nice to have [14:40] benji: yea, see how far you can get. I might be able to run with it some this weekend so let's make sure to catch up before your EOD and hangoff to me. [14:40] rick_h__: that doesn't look like the latest code. He asked me to review that branch on Wednesday and I noted that it didn't actually do what the card was about. He then began working on it some more. I can take it from that branch (or perhaps start over, since it isn't on-target) but it'll be rework. [14:41] benji: yes, understood. It'll be rework and can't be helped without delaying into Wed or later next week in which case we'll miss the MWC demo time this was asked specifically for. [14:41] will do [14:41] thanks [14:49] rick_h__ we are going to have to parse the zip contents - on a real env we would need to store the zip filename in the annotations otherwise [14:52] so unfortunately that part is kind of an all-or-nothing approach [14:58] hatch: what's wrong with the annotation? I'm confused [14:59] benji: one heads up, when brad and I talked the one thing that came out was that the api search and the website search are two code paths [14:59] benji: and that a "" search in the api returns all results [14:59] well they get sent along with every request, so it bulks up the requests and it's pretty clunky [14:59] at least imho [14:59] yep [14:59] while that did not seem true on the website search [14:59] benji: cool [15:00] hatch: ok, well it was on the roadmap for the feature anyway [15:00] which? [15:00] hatch: the zip to help find service by name not zip filename [15:00] ohh yeah - I'm hoping that the sandbox api can be repurposed [15:00] I'll look into that first [15:00] hatch: but let's make sure we shortcut, if we dont' have any local charms in the env we can skip it still [15:01] oh yeah for sure [15:01] k [15:01] friday is not liking us today is it [15:01] this week has just been frustrating all over :) [15:02] yea, suppose so [15:02] I need an idea for a Go lang project [15:04] get my head outa js for the weekend [15:04] heh, I'll send you my list over [15:04] haha I'm looking for small projects [15:04] maybe a week of personal time sort of things [15:05] benji what was wrong with your car? [15:05] hatch: yea, charm proof would be a small something to check out doing. I was thinking of doing my small readable web service in Go as an experiment and test scale [15:05] it has a coolent leak (and needs some brake work) [15:06] shitty, what kind of car? [15:06] rick_h__ oh Go can handle scale :) [15:07] charm proof in Go? Isn't it written in python already? [15:07] hatch: yea, just mean a good example project to 'port' and compare [15:07] ohh gotcha [15:07] that might not be a bad idea [15:10] maybe I'll write a Bookie competitor in Go :P [15:10] jk [15:13] hey, more power to you [15:14] it could be called Gookie [15:14] lol [15:16] rick_h__ http://nightwatchjs.org/ nodejs selenium driver api [15:16] it 'looks' simpler than our python one [15:19] hatch: hmm, but this is a selenium server and we use saucelabs for that [15:19] https://github.com/beatfactor/nightwatch/issues/53 [15:20] oh shoot I misread [15:20] I thought that it interacted with a selenium server [15:20] it IS the server [15:20] is how I read it [15:20] well it does...but right heh [15:22] teach me for not reading before sharing [15:39] benji hey when you had your git rebase issues did you try creating a patch and applying that on a new branch? [15:40] hatch: I tested things out. The key is that if you updated your branch to latest develop to do 'git rebase develop' and all your commits get layered nicely [15:40] rick_h__ following along in #yui? [15:40] hazmat: did that on https://github.com/juju/juju-gui/pull/139 [15:41] hatch: oh no, /me goes to look [15:41] hatch: yea, use rebase when push upstream changes into topic branch [15:41] that works nicely [15:41] hatch: I considered it, that may have even been the final solution; I can't remember exactly. [15:42] rick_h__ well remember benji's issue with a bunch of wips and merges from develop [15:42] I was hoping for a reasonable 'fix' and it looks like creating a patch file is the best [15:42] yea, I mean I've done both. Created a new branch from develop and cherry pick. [15:42] rick_h__ so then what are we talking about re the rebase develop if you weren't following along in yui :) [15:42] that went ok, but tedious [15:42] no, I wasn't at the time [15:43] I was just replying to your questoin to benji [15:43] as it was something I wanted to test out and just had cause with the safari CI branch since it hung for a week there [15:43] ohhh ok ok [15:43] so you had commit commit merge commit commit ? [15:43] and doing a rebase against develop removed the merge commit? [15:43] so I updated my safari branch with latest develop first `git co develop; git juju-sync; git co $mybranch; git rebase develop [15:44] it didn't remove it, it avoided it [15:44] a merge commit is a merge commit. It's now part of the history [15:45] hatch: so https://github.com/juju/juju-gui/pull/139 had a rebase on develop in there, two rebases for lint/etc during review. looks clean :) [15:45] rick_h__, hatch: quick fix for bug 1283067 ready for review -> https://github.com/juju/juju-gui/pull/140 [15:45] <_mup_> Bug #1283067: dropping a local charm in Firefox does not deploy it into the sandbox [15:45] frankban: looking [15:45] thanks [15:46] rick_h__ so how did you get rid of the merge commit there then? the rebase on develop got rid of it? [15:46] s/on/from [15:46] hatch: you're saying to get rid of something taht exists [15:46] I'm saying that you need to avoid the merge commit to start with by using rebase vs merge [15:46] well in #139 it looks like you did [15:46] there is no merge commit in there [15:46] right [15:47] I avoided it by doing a git rebase develop [15:47] that's how I updated my long running feature branch [15:47] hatch: follow? or am I still talking nuts? [15:48] ok so you didn't 'merge' develop in...you rebased it in [15:48] hatch: rgr [15:48] me likey [15:49] ok so from what I understand the 'best' way to fix benji 's issue is to use a patch and git 'should' ignore the parts of the diff that are unchanged in upstream [15:49] but the proper way to do it is to rebase upstream in [15:49] upstream in this case being develop [15:49] hatch: right [15:50] hatch: correct, you should be able to git rebase develop over and over in a long running branch safely, fixing conflicts between trunk and your WIP as you go [15:50] jujugui call in 10 [15:51] frankban: yay works. Want me to :shipit:? [15:51] rick_h__: yes thanks [15:52] stupid browser differences [15:52] frankban how did you figure out that that was necessary? [15:53] hatch: there's a bug in zip.js called out in the code [15:53] hatch: ff debugger + feeling lucky with google [15:53] lol "I feel lucky, oh so lucky" [15:54] nice diff to use that whitespace flag feature on as well <3 [15:54] yes indeed [15:55] frankban ok cool :) I was hoping that there might have been some cool technique or something :P [16:00] jujugui call now [16:18] frankban have a quick second now to run through my questions? [16:18] https://plus.google.com/hangouts/_/calendar/cmljay5oYXJkaW5nQGNhbm9uaWNhbC5jb20.t3m5giuddiv9epub48d9skdaso?authuser=1 [16:19] hatch: sure, be there in a minute [16:26] hatch: I am there [16:34] rick_h__, you mentioned copying me on an email, but having a hard time finding it. [16:34] What was the subject? [16:35] Makyo: sec, it was a document. Looking [16:36] rick_h__, oh, oops. Will check. [16:36] There it is. [16:36] Sorry, got it :) [16:36] Makyo: https://docs.google.com/a/canonical.com/document/d/1ma0U1ZxILTh5s3NoHiuwKiclRSxJLIY-Q2MEMzBO7Dk/edit [16:36] Makyo: cool, adding the link to the card as well [16:40] Cool, thanks. [16:40] * Makyo restart for updates. [16:54] http://vimeo.com/58200103 [17:25] anyone else excited to see pictures of Pluto? [17:25] like 4 serious! [17:25] :) [17:33] we're getting pics of pluto? [17:36] well in a year probably [17:37] http://pluto.jhuapl.edu/index.php [17:37] 9.5 year trip or something hah [17:47] jujugui looking for a really quick review on https://github.com/juju/juju-gui/pull/141 no qa needed [17:48] hatch: I'll take a look [17:48] thanks [18:07] rick_h__ do you know where I can find the aforementioned charm-proof repo? [18:30] hatch: it's in charm-tools [18:30] ohh duh [18:31] I'll take a look at the Go conversion....if you don't mind [18:31] woot, got the ok for the 1.0 release go go [18:31] hatch: yea, definitely. [18:32] cool, I have no idea what will come out of it [18:33] yea, just something small to tinker with [19:06] so I just looked at it [19:06] it's actually pretty sizeable [19:06] once you include the scripts for actually parsing the stuff [19:06] good couple thousand lines [19:07] which is fine [19:07] I had assumed it would be smaller for whatever reason [19:38] now to play "run those functional charm tests!" [19:39] to the tune of Play That Funky Music [19:40] "get down and boogy and let functional tests pass" [19:40] lol [19:43] boom and fail [19:43] doh [19:46] lol [19:46] it seems anything with 'charm' in it blows up for us [19:46] charmworld [19:46] charm tests [19:46] :P [19:46] lets rename them to trinkets [19:47] ERROR empty image-stream in environment configuration [19:47] ummm, ok [19:47] uhh what [19:48] yea, that's my thought [19:49] Makyo new OneTab record here ....276 [19:49] and about 15 more tabs open lol [20:15] Makyo: did you have a card today? Were you looking at the video or something else? [20:15] * rick_h__ was just kanban gardening while waiting for live tests wheeeee [20:21] I found two really cool sublime plugins, gitgutter and git [20:21] gitgutter shows the changes in your file in the line gutter [20:21] and git gives you git commands in sublime [20:21] sublime > terminal is kind of a pita when you only want to see diffs and stuff [20:23] hatch: or Makyo able to help me test this please? [20:23] sure what do you need [20:23] preferably non trusty [20:23] I have 12.04 [20:24] hatch: I pull down lp:~rharding/charms/precise/juju-gui/1.0 and see if you can make test including the live ec2 tests [20:24] sure [20:25] bzr checkout --lightweight should be the default :) [20:25] bzr branch pssshhht [20:25] rick_h__: ready for handoff? I think all we need is review and QA and it'll be done. [20:25] benji: sure thing [20:26] rick_h__ running `make test` [20:26] benji: https://plus.google.com/hangouts/_/72cpiv9f3hglut66ffdl7mq950?hl=en [20:26] hatch: thanks [20:29] hah, he runs make test and then loses his internet connection [20:29] "thou shall not release!" [20:32] rick_h__ https://gist.github.com/hatched/cba361ff581bf4d51a6d this is the error I get [20:34] rick_h__, sorry, yeah, video. Have IRC open on the wrong laptop :( [20:34] hatch: did you run 'make sysdeps" [20:34] nope i'll try that [20:35] hatch: yea, run that first and then make ftest JUJU_ENV="ec2" or whatever your ec2 env is named [20:35] unable to locate libpython-dev [20:35] hatch: sudo apt-get install libpython-dev [20:36] oh, you don't have it on precise? [20:36] yeah that package doesn't exist [20:36] looking [20:36] what's the equivalent for precise? [20:36] pffft python portability issues [20:36] :P [20:36] try just 'python-dev' [20:37] but make sure you catch the rest of the sysdeps in the Makefile [20:38] already the newest version [20:38] will look at the other deps [20:38] k [20:38] sudo apt-get install build-essential bzr libapt-pkg-dev python-virtualenv rsync xvfb [20:39] libapt-pkg-dev looks like it was missing [20:39] will see if that was the issue [20:39] k [20:40] looks like that was the kicker [20:40] whatever it's for [20:40] hatch: cool, it'll run for a while hopefully [20:41] well it apparently doesn't like my JUJU_ENV value [20:41] hatch: right, it needs to be whatever you call your ec2 env in your environtments.yaml file [20:41] I call mine ec2 [20:42] https://gist.github.com/hatched/cd5196ac5b83cc4bf231 [20:42] ok [20:42] rick_h__: I did some basic QA and it looks good. Here's the MP: https://code.launchpad.net/~benji/charmworld/doctype-search/+merge/207738 [20:42] make ftest JUJU_ENV="amazon" [20:43] benji: awesome, will look when I get this deploy up and done. [20:43] cool [20:43] ok will try that [20:43] lol juju-test is missing now [20:43] I'm just going to read this hacking file to see what I am missing [20:44] lol ok [20:44] yea you need charm-tools [20:44] sorry, didn't realize you'd not messed with the charm [20:45] not on this vm I guess [20:45] best way to install charmtools? [20:45] sudo apt-get install charm-tools [20:45] installing from ppa now [20:45] do you have the stable ppa? [20:46] yay finally [20:46] looks like it's starting the env [20:46] yay [20:50] benji: does this require the : or is it optional? [20:51] rick_h__: required; I figured the structure was a better way to go, but it would be easy to make it optional [20:51] benji: ah, see in the tests that they both work cool [20:51] well the bare words work, but if you have terms after the word you have to have a colon [20:51] benji: well just 'bundle' works for all bundles in the test [20:51] right [20:51] right, the bare words are what I was worried about. That's the main use case [20:51] cool [20:51] the : identifier is more advanced optional [21:01] hatch: still running? [21:08] yup [21:08] there has been an error though [21:09] :( [21:09] https://gist.github.com/hatched/9143581 [21:09] doesn't give me any information as to what failed yet though [21:09] yea, will have to wait for them to finish [21:10] sorry my internet is very slow right now for whatever reason [21:10] you had one 'ok' thuogh so that's more than me [21:10] it takes a while [21:10] lol I'm guessing that 1.0 is not being released today? :) [21:10] depends on what your errors say [21:14] rick_h__ https://gist.github.com/hatched/a9f7df762c14e797fbc1 [21:15] hatch: what's your juju version? juju --version [21:15] 1.17.1 [21:15] precise [21:16] it's `juju version` btw :P [21:16] heh [21:16] that's the Go way [21:16] lol [21:16] unless they fixed that [21:16] well --version works [21:16] it's what I've been using [21:16] hatch: can you test one more thing then please? [21:17] can you stick that directory inside a directoy named precise and try to install it and check the unit log? [21:18] oh ok they must have fixed it [21:18] install it in a real env? [21:18] yes please [21:19] sure thing [21:19] if the tests fail and the unit was in error it'd be helpful to find out what the error is [21:19] bootstrapping [21:19] I can't deploy right now, there's some issue core is looking into [21:19] thanks [21:19] lxc bootstrap is so much faster [21:19] lol [21:19] sorry to distract you. I'll start working on setting up a precise vm to work around the current issues [21:19] heh no doubt [21:19] no problem I needed a break from local charm anyways [21:19] hatch: you can test it there as well I guess. Maybe it does error in lxc [21:20] so you get me to deploy a local charm....damnnnn youuu [21:20] lol [21:20] lol [21:20] ahh it's ok the ec2 setup is already running [21:20] but if the tests says it's in error I need to know what error to try to see what's not happy in the charm [21:21] yup I'll keep you posted [21:21] thanks, /me goes to get a refill...come on EOD [21:21] going to have to start my birthday party drinking early after today [21:22] it's your bday? [21:22] well why didn't you say something! [21:22] tomorrow [21:22] well then! Happy early Bday [21:22] hah [21:22] thanks, where's the wine? [21:23] I'd say you could watch the USA Olympics hockey game tomorrow, but you guys lost :P [21:23] I hear it wasn't a very good game all around [21:24] I haven't heard much except my friends who care about hockey texted me lol [21:25] ERROR cannot repackage charm: symlink "tests/.venv/include/python2.7" is absolute: "/usr/include/python2.7" [21:25] have you seen that before? [21:25] running make clean [21:25] make clean on the new copy [21:26] ahh that appears to have fixed it [21:36] rick_h__ 2014-02-21 21:31:15 INFO install raise ValueError('Error: no releases found in the charm.') [21:36] that would be an issue :) [21:36] ok, so do me a favor. this is what I was thinking [21:36] favour [21:36] :P [21:36] in the releases folder change the version from 1.0 to 1.0.0 [21:36] ok changing [21:36] I wonder if that's what I'm hitting [21:37] wait....do you want me to do that in the deployed one and retry? or from ground 0 [21:37] in the local copy and redeploy the charm [21:37] alrighty [21:37] will ping [21:37] mv releases/juju-gui-1.0.xz releases/juju-gui-1.0.0.xz [21:38] juju-gui-1.0.xz ? [21:38] I've never seen that extension before [21:38] odd [21:38] its ok [21:38] it's our special super compression extension. It's rare [21:39] ohh ok :) what does it use? [21:39] xz :) [21:39] * Makyo : jerk. [21:39] http://tukaani.org/xz/format.html [21:40] lol SORRY [21:40] I'm doing three things at once here haha [21:40] It is 100% a Friday. [21:40] friday release day! [21:40] boom! [21:40] Woooo...ooo...oo~ [21:42] lol we said we would never do that! [21:42] well it was supposed to be easy. I think the issue is I left off a .0 [21:42] just hitting a juju bug causing me to not get passing tests [21:43] if this works we're back in business [21:47] rick_h__ started [21:48] hatch: woot [21:48]