[01:13] rick_h_: back by any chance? [01:17] jcw4: what's up? [01:18] quick question [01:18] (and if not it can wait) [01:18] shoot [01:18] running juju-gui against local dev version of juju [01:18] HACKING.rst is bit-rotted [01:18] jc :( [01:18] is there a quick howto somewhere? [01:19] jcw4: that's it. If it's bitrotten we need to fix it [01:19] I have juju local running [01:19] the instructions aren't clear how "make devel" hooks up with the local env [01:19] jcw4: oh it doesn't :) [01:19] it talks about api-port but doesn't say where exactly [01:19] jcw4: it uses a fake backend (all JS) in devel/etc [01:19] have to deploy a charm? [01:19] jcw4: rgr [01:20] best way to go [01:20] so any dev changes on the gui side have to be packaged and redeployed? [01:20] i.e. no runtime code changes? [01:20] on the juju-gui side [01:20] jcw4: so what we do is there's a config on the charm for 'juju-gui-source' [01:20] okay [01:20] that takes a branch name you can set [01:20] and it'll download/load from github/etc [01:20] jcw4: or you can use a tarball at a url, etc [01:21] i see [01:21] jcw4: so normally we dev against our fake provider, and then when we QA on live env we'll deploy, set the source to our branch, and test stuff out [01:21] hmm, in this case I want to play on the -core side and tweak on the -gui side simultaneously... [01:21] okay [01:21] that helps a lot [01:21] jcw4: yea, will be interesting. [01:22] if I can improve the docs I will [01:22] jcw4: there are configs on the charm to not use the minimized JS and such [01:22] oh [01:22] but it'll take some poking to get at the source in the build dir/etc [01:22] okay I'll be sure to check that too [01:22] jcw4: and the normal "make devel" won't work since you won't have a bunch of dev tools [01:22] so it's kind of hacky [01:22] hehe [01:22] no worries [01:22] but it's enough to get decent tracebacks/etc just to not have it minified [01:22] now that I know it's not supposed to 'just work' I'll get my boots on [01:23] jcw4: but yea, config optinos for the source, for enabling the console (console.log/error), and for debug mode should come in handy [01:23] (config options on the charm) [01:23] in summary - I'll build the charm with options to keep source un minified [01:23] deploy to local juju [01:23] and then hack around in the charm dir to tweak code [01:24] deploy to local juju, then once it's up change the config options (or set them on deploy using a config.yaml file) [01:24] k [01:24] just like a normal charm, they're just set via "juju set ..." [01:24] k [01:24] see https://jujucharms.com/trusty/juju-gui-9/#configuration [01:25] I saw that but now it makes more sense [01:25] tx rick_h_ [01:25] jcw4: cool, np [01:48] there, get annoyed, write some code :P https://github.com/juju/juju-gui/pull/601 [01:48] huwshimi: if you're up for a GUI review ^ appreciate it [01:48] huwshimi: but if you're slammed for luca, that comes first [01:49] rick_h_: OK, I'll take a look if I get a chance [01:49] and booya, managers coding ftw1 [01:49] err ! [01:54] rick_h_: A couple of notes about the styles, I'll do a proper review later if I don't run out of time today. [01:55] huwshimi: thanks [01:55] huwshimi: I knew you'd call me out on that :P [01:55] rick_h_: There's no excuse :) [01:56] huwshimi: but but but ... yea [01:56] hehe [01:59] ok, time to head home night all [02:00] rick_h_: Night [03:53] rick_h_: jcw4 you can hack on the deployed charms code by ssh'ing into the juju-gui instance [04:04] rick_h_: just figured that out a little bit ago [04:04] thanks! === uru_ is now known as urulama [07:41] Luca__: Hey are you around yet? === rogpeppe is now known as rogpeppe1 === rogpeppe1 is now known as rogpeppe [10:58] morning [11:32] morning rick_h_: do you need a review for your card in maintenance? link? [11:35] frankban: yes please [11:42] antdillon: what are you doing around here? [11:42] rick_h_, ha couldnt stay away [11:43] antdillon: I'm going to tell your boss to send you home [11:43] antdillon: did you have your kid? [11:44] rick_h_, yeah little girl Chloe [11:44] antdillon: woot! congrats! [11:45] rick_h_, thank you very much [11:45] rick_h_, dont tell my boss ... [11:52] frankban: can we join our call early please? e.g. now? I want to squeeze it in before I leave for the doc please [11:52] rick_h_: now is ok [11:52] ty [12:07] jujugui heading off to the morning doc apt. I'll be back in a bit. [12:17] * frankban lunches [13:32] * rick_h_ is back [13:34] jujugui can I get a second review of https://github.com/juju/juju-gui/pull/601 please and I'd appreciate a second QA as I'm unable to dupe the issue frankban had and curious if others can replicate [13:34] rick_h_: I had a second QA comment, fixing my first one FYI [13:35] frankban: oh oh oh I see now. You're right, they're not checked based on current value on load. That's bad on my part I'll update in this branch [13:35] should be a simple read/set checked=checked thing in the template [13:35] * rick_h_ misread it at first [13:35] rick_h_: cool thanks [13:45] jcsackett: around to chat on the feature flag branch? [13:46] rick_h_: rick_h_ ping you in a few? trying to work my way through a mountain of PRs [13:46] jcsackett: rgr [13:46] all small, should be done in just a bit. [13:46] all good, just wnat to make sure I unblock as soon as I can [13:46] unblock you that is === kadams54_ is now known as kadams54 [14:13] rick_h_: that took a bit longer than i thought. you free? [14:13] jcsackett: otp atm [14:13] natch. :p [14:13] well, i'll start reviewing your branch, then. [14:18] jcsackett: free now [14:18] jcsackett: get in while the 10min last :) [14:18] rick_h_: coming to the standup hangout now. [14:25] running to grab a coffee [14:57] jujugui call in 4 kanban please === kadams54_ is now known as kadams54 [15:00] frankban: call [15:04] frankban: did you get a response from homebrew? after the initial acceptance updates seemed to be automatic in the past. [15:05] bac: I know, it's like this: https://github.com/Homebrew/homebrew/pull/32866 [15:29] rick_h_: jcsackett fixed the manual provider issue - just need tests now [15:33] DO is so darn fast compared to ec2 lol [15:45] hatch: whoo! [15:56] *sigh* the go.js test file is over 2k lines [16:13] rick_h_: I’ve assigned a card to Huw to build the new footer design, is that ok? [16:20] jujugui I need a review and qa for a manual provider bug [16:20] https://github.com/juju/juju-gui/pull/602 [16:21] we have a new footer design? luca__ what did I tell you about changing the design?? HUH HUH?? :P [16:24] rick_h_: next card of choice for me? [16:27] * hatch started a new js framework last night [16:33] hmm I sure wish apple had a 4k thunderbolt display - having a hub in the monitor would sure be nice! [16:38] frankban: thanks for the review [16:38] frankban: did you qa? [16:38] hatch: I am bootstrapping a manual env for the first time [16:38] Ooo :) [16:38] on Digital Ocean? [16:39] no, on a vmware machine locally ;-) [16:40] oh haha I never even thought of doing that :) [16:40] hatch: it worked pretty fast [16:40] now switching to your branch [16:42] hatch: can you see if kadams54 or Makyo need help. Priority is to look at getting those up for landing so release is possible tomorrow [16:42] hmm, luca is gone [16:42] hatch: I'll need QA shortly ;-) [16:42] kadams54: lemme know! [16:43] hatch: code changes are minor, so it's mostly QA. [16:43] I'm a boss at qa [16:43] I can break anything [16:43] :P [16:45] jujugui has anyone else been able to add the calendar in Sarah's email? [16:47] hatch: yeah, just did it. [16:47] hmm - you used the url and went to Add Url? [16:47] hatch: I've got it from before and can't share it it seems [16:47] it says that the url is invalid for me [16:48] I got it by pasting the email address in 'Add a coworker's calendar" [16:48] ahhhh [16:48] there it goes [16:48] thanks peeps [16:48] http://cl.ly/image/132H0Y3R0l3V [16:49] kadams54: haha that's cool how did you do that? [16:49] cloud.app [16:49] https://www.getcloudapp.com [16:50] I've started sending animated GIF screenshares instead of typing up 6 page tech support e-mails to friends and family. Very handy. [16:50] darn osx only [16:51] we need some of thse for Ubuntu [16:54] hatch: done [16:54] awesome thanks [17:02] kadams54: do you know if this cloudapp will work in fullscreen vm? So I could use it from osx in my ubuntu vm? [17:03] I don't, sorry. [17:03] I have screencloud for Ubuntu but it doesn't do 'video' [17:03] which would be helpful for some things [17:04] Yeah, definitely. Cloudapp just added the video feature in the last release; I was pretty excited to see it. [17:19] jujugui reviews/qa please! https://github.com/juju/juju-gui/pull/603 [17:19] I'm on it [17:19] BOOM, just like that [17:19] Thanks. [17:19] but now you have to review mine :P [17:19] Fiiiiine [17:21] hatch: you need a second in addition to Makyo? [17:21] nope frankban got the review and qa [17:22] I'm running make check right now for my PR :-) === urulama is now known as urulama-afk [17:25] Makyo: I think you missed the 'clear all' functionality [17:25] wait nm' [17:25] I am bling [17:25] blind even [17:25] Oh, just forgot that in the QA instructions [17:27] code looks good - qa'ing [17:30] Makyo: now that we have this functionality it's quite odd that you can't expose in the ghost inspector [17:30] thoughts? [17:32] and done [17:32] hatch: https://github.com/juju/juju-gui/pull/604 is ready for QA [17:32] ;-) [17:32] hatch, it's a little weird, yeah. Was going to make that a followup. [17:33] ...which you said. So yeah. [17:33] :) [17:34] man doing work like this is such a relief post MV release :) [17:34] frequent releases ftw [17:34] Continuous deployment, baby. [17:34] lol I wouldn't quite go that far :) we do ship a packaged app afterall [17:37] kadams54: lol got to love the branches where the test loc's FAR exceed the code changes [17:37] actual code < test code < time spent on manual QA [17:37] Whee! [17:38] Plus I had a comparison error yesterday that manifested by breaking other, unrelated parts of the code, even though the QA stuff worked fine. Also whee! [17:40] kadams54: just a request for the future - if you move code could you add a reviewer comment to point that out so reviewers don't have to stare at the diff for changes when none happened [17:41] Ah yeah, good point, will do. [17:41] The tricky thing is that some of that did change, so I'll annotate with comments [17:41] ahh yes plz do :) [17:41] Some tests moved into a describe block. New tests were also added to that same describe block. [17:45] OK, new tests annotated. The rest of the test deltas is just reorg. [17:48] thanks [17:59] jujugui stepping out over lunch to fedex some things. Back after. [18:21] jujugui kind of cool, have blog stats from google in image form on the mv blog post http://uploads.mitechie.com/lp/mv-blog-stats.png [18:23] rick_h_: it was so good, 700 people came back to read it again1 [18:23] s/1/!/ [18:23] lol [18:23] or that was you 700 times [18:23] well you know me and my refresh button issues [18:41] man google maps 'create map' is missing a huge feature "add points from starred" :/ [18:52] hi all [18:53] hatch: I have a YUI question for you. We have an issue on the Landscape project involving clicking on links handled by event handlers before all of the JS has run, causing either nothing to happen or the wrong thing to happen. Is there a standard YUI approach to that problem? [18:57] wb benji [18:57] long time :) [18:57] :) [18:57] no there isn't a standard problem but there is a common technique to solve it though [18:58] the first step of course is to try and reduce that time :) [18:58] the second is a raw js handler which catches all clicks on A's and then queues them up [18:58] once YUI comes online it re-fires them [18:59] s/problem/solution [19:00] benji: but with that said you should be able to get a running YUI instance within a second [19:00] hopefully less than 500ms tbh [19:00] so you might have an architecture problem there [19:00] hatch: good stuff, thanks! [19:01] anytime [19:01] benji: using a combo loader? [19:02] rick_h_: I honestly don't know. I've tried to stay away from the JS stuff for a while, but I think I've built my tolerance back up enough to look at it again. ;) [19:02] lol [19:04] benji: really try to reduce the load time to be a non-issue first - implementing this can be a hassle. I've done it before :) [19:04] granted that was 4ish years ago hah [19:05] jujugui: heading out, should be back online in two hours. [19:05] hatch: yeah, I'm a little surprised that it can take as long as it does for everything to get into a working state [19:06] you may have a big blocker in there somewhere [19:21] ahh machine view is so bawler [19:21] * hatch feels like he has gold teeth and a rap album when he uses it [19:24] hatch: you don't own any rap albums? [19:25] maybe some but none with my name on them [19:25] has anyone gotten the 'cloud track' calendar to show up on their freedom-hating phone? it eludes me. [19:28] bac when you go to "Calendars to display' do you have a button 'Calendars to sync' ? [19:28] if so - go there and you can add it to the list [19:30] how do I get my version of the gui charm up on my local provider? I'm also going to tweak a simple charm to play with a new backend feature. I assume this info is in HACKING, but I must be missing something [19:30] there's a "local" charmstore you can use, right? [19:31] bodie_: use the --repository flag when deploying via the CLI [19:31] https://juju.ubuntu.com/docs/charms-deploying.html#deploying-from-a-local-repository [19:31] sweet, thanks [19:31] rick_h_: updates to the kill-mv-env branch sent. [19:31] jcsackett: rgr [19:31] once you have the GUI up and running you can then drop local charm zip files onto the GUI canvas bodie_ [19:32] oh, that's handy. thanks hatch [19:32] no problem [19:32] what are you hacking on? [19:32] actions [19:33] ahh cool - I want those [19:33] :) [19:33] what about to hack on the gui charm itself? I guess you'd use upgrade? [19:33] yeah, we just got the worker / unit bits nailed down finally, and we have the skeleton api landed too [19:33] I'm not sure about upgrading the gui charm via the gui :) never tried that [19:33] should work though...heh [19:33] what could possibly go wrong? :P [19:34] I was thinking more about using the CLI for that, though [19:35] sure thing - if you want to hack on the actual GUI code in a live env you can do that too [19:35] without having to modify the charm [19:35] ah, juju upgrade-charm --repository [19:35] oh [19:35] hm [19:35] I guess that's what HACKING covers, right [19:36] something about that doc just didn't make things clear to me [19:36] yeah run `juju help upgrade-charm` for detailed instructions [19:37] ok. thanks again for the pointers [19:37] np I'll be here for a while yet [19:37] if you have any more [19:45] hmm in qa'ing kyles branch I'm having an issue with the gui charm not upgrading the ws [20:05] hatch: lemme know if you want me to test your branch on my env. [20:14] * rick_h_ steps out to throttle someone at a t-mobile store [20:18] rick_h_: when you're back from throttling if you can look at my mv-flag branch so i can move it to landing. :) [20:48] jcsackett: hey it has been merged [20:49] jcsackett: it would be awesome if you could qa kyles branch with me [20:49] https://bugs.launchpad.net/juju-gui/+bug/1376901 I found this bug while doing it so I'd like another set of eyes [20:49] Bug #1376901: Destroying a machine with a container on it causes machine view to break [20:54] Makyo: https://bugs.launchpad.net/juju-gui/+bug/1375251 requires you to deploy the env using the DO plugin [20:54] Bug #1375251: Juju gui malforms display with missing x/y coordinates on existing env w/ charm upgrade [20:55] hatch, yeah, I was just hunting for that. Have a link handy? [20:55] Makyo: from lazyPower http://blog.dasroot.net/juju-digital-ocean-awesome/ [20:55] awesome tut [20:56] * lazyPower hattips [20:56] thanks for the feedback [20:56] fyi, thats also in the documentation now [20:56] juju.ubuntu.com/docs/provider-digitalocean.html [20:56] oh cool [20:56] the video was really awesome but too bad the sound wasn't quite there [20:56] https://juju.ubuntu.com/docs/config-digitalocean.html rather [21:07] guihelp: Anyone around to provide a second QA and review on https://github.com/juju/juju-gui/pull/604 ? [21:07] yeah definitely needs another qa [21:14] We need more West Coasters ;-) [21:19] kadams54: will look in a bit [21:20] jcsackett: will look [21:22] hatch: i can take a look. be a bit to spin his branch up on ec2. [21:23] kadams54: haha we used to [21:23] jcsackett, rick_h_: thanks [21:29] Switching locations, will be back shortly. [21:38] hatch: so, i'm not going to be able to qa it. amazon is saying it has insufficient capacity. :p [21:38] lol [21:38] use DO [21:41] jcsackett: maas time! [21:41] :p [21:43] jcsackett: :+1: ty! [22:05] rick_h_: awesome, thanks. [22:05] rick_h_: the bug is an ec2 bug, alas. [22:06] jcsackett: rgr [22:20] my internet is super flaky mind if I skip on the aus call rick_h_? [22:21] hatch: np [22:22] hatch: thanks for the updates today to get ready for release <3 [22:25] yeah this fix release is going to be really nice [22:25] especially if we can keep up doing releases like this [22:25] really nice to do [22:26] hopefully [22:27] though new feature starts next week so it'll slow down a bit, but smaller features and more releases is good, out of MV's shadow [22:32] Morning [22:33] rick_h_ hatch: Sorry I'm late [22:33] huwshimi: all good [22:33] huwshimi: joining late myself [23:04] hi huwshimi [23:04] urulama-afk: i've submitted a pull request to your CScharmTools branch with working ngram searches. [23:05] bac: Hey! [23:17] huwshimi: have you not left yet? better hurry or you'll miss opening cocktails. [23:22] bac: haha, yeah, it's a bit like that...