[00:00] hatch: Are you gone for the day? [00:04] hatch: Are you gone for the day? [00:04] something's not right [00:05] im still here [00:06] hatch: If I override the init method on the tokens then it won't render the tokens again when I close and open the machine view [00:07] ok looking at the code [00:09] when you close and open the machine view does it destroy all the tokens? [00:09] it looks like it [00:10] yeah [00:10] so that doesn't make any sense [00:10] so it only works once? [00:11] hatch: Actually, that might not be entirely correct. [00:11] hatch: If you drag a charm to the canvas, deploy and then open the machine view they don't appear [00:12] hatch: Even though the machine token's render function is called. [00:14] not sure [00:14] I can look at it later on [00:14] but atm I've got a bunch of stuff to do heh [00:20] no problems [00:28] hatch: it does pass! Woot woot! [00:28] I'm going to squash again, but I fully expect that to pass as well. [00:34] Oh, heh, the version that passed doesn't have the async fix yet. Never mind. [01:15] huwshimi: I'm knocking off for the night, but if you want to QA my PR, it's ready to be broken :-) https://github.com/juju/juju-gui/pull/393 [01:16] kadams54: No problems, I'll take a look! [01:16] Have a good one… [01:16] kadams54: Night [01:38] evening [02:46] huwshimi hey how goes it? [02:46] hatch: ugh [02:48] that good eh? anything I can help with? [02:49] hatch: Well, I think I know that the container is not being set correctly, I just don't know why [02:50] although now it's not working at all, even if I don't overwrite the init [07:06] mornin' all [07:11] rogpeppe1: Morning [07:13] huwshimi: hiya [08:34] rogpeppe1: morning [08:34] frankban: hiya [08:34] frankban: hangout? [08:34] rogpeppe1: yes [11:09] morning all [12:26] morning, all. [12:27] rick_h_: you back today? [12:28] jcsackett: yep, back and emailing wheee [12:28] ah, the email excavation, always fun. [12:31] * frankban lunches [12:39] rick_h_: thanks for your email to ben. i was left scratching my head too. [12:43] bac: heh, maybe I was just tired. We'll see [12:43] rogpeppe1: https://github.com/juju/juju/pull/124 is part of what I was talking about stealing from core. I'm not sure if we can use it but just something to be aware of [13:01] rick_h_: that seems largely overkill for what we need to do, i think (though i reserve the right to change my mind later :-]) [13:01] rick_h_: in particular i hope that we can avoid using the txn package in store [13:02] rogpeppe1: ok, cool. Just want to make sure we don't reinvent the wheela a couple times [13:03] rick_h_: i suspect this wheel will be a different enough shape that there wouldn't be much saved by trying to reuse the code. [13:03] rogpeppe1: sounds good thanks [13:04] rogpeppe1: you also see the point on the individual file blogs inside the charms/bundles? [13:04] rick_h_: i'm not sure. we definitely want to expose *some* data/metadata from bundles, but i don't know about individual file extraction. [13:05] rogpeppe1: well during publish time we want to store then apart [13:05] we need this for things like icons and such, without extraction on demand [13:05] rick_h_: because you can't create zip files in JS? [13:06] rick_h_: icons, certainly [13:06] rogpeppe1: no, more because when we load api results we need to shot icons, allow readmes, etc [13:06] s/shot/show [13:06] rick_h_: anything else? [13:06] and we allow things like browser the hooks and such [13:07] rogpeppe1: the list of whitelisted files are in charmworld second [13:08] rogpeppe1: http://bazaar.launchpad.net/~juju-gui-bot/charmworld/trunk/view/head:/charmworld/models.py#L914 [13:08] are what we currently load/allow. I suspect we'll want to add actions to that soon [13:09] rick_h_: yeah, actions should definitely be there [13:10] rogpeppe1: and we expose those whitelisted files over http://charmworld.readthedocs.org/en/latest/api.html#retrieving-the-files-that-make-up-a-charm [13:11] rick_h_: does the GUI know how to parse those files? or does it just show the raw data? [13:11] rogpeppe1: raw data but it needs to know how to parse the actions to build the UI for them [13:11] so that's not always going to be true (that it doesn't need to parse) [13:11] rogpeppe1: oh, well I guess it renders .md file for README so it parses some, plain dumps others [13:11] rick_h_: so you've got a decent YAML parser in JS? [13:12] rogpeppe1: nope, we get json :) [13:12] rogpeppe1: that's why we've pushed so hard for jsonschema and limiting the yaml in there to stuff that's valid json as well [13:12] rick_h_: how will you get json from a YAML file? (i think the actions file is yaml) [13:12] the api will save us? :) [13:12] it's what we've been working on with the actions guys from the start. Making sure it's targeted to json/gui use [13:12] rick_h_: so that means it's not just extracting files from the zip, but applying appropriate transformations to them too? [13:13] but that will come from core, not from the store [13:13] from the store we just want to be able to show users what's in the charm [13:13] rick_h_: so you won't be able to parse actions unless you've got a live environment? [13:14] rogpeppe1: hmm, not sure. On the one hand we NEED parsed for a live env so we can show you the UI generated based on the actions file [13:14] rick_h_: BTW i'm not necessarily opposed to doing some store-side transformations - i'm just trying to work out whether we're talking about a general file retrieval mechanism or tailored metadata [13:14] rogpeppe1: but in the store side, we want to show users 'this charm supports these actions with these docs/notes' [13:14] rogpeppe1: to date, it's just been file retrieval [13:15] rogpeppe1: with the only store side transformations being around icons so that we don't show icons for non-recommended charms [13:15] rick_h_: from what i can see, it'd be quite straightforward to provide access to individual files from the zip if necessary, while still storing the files in zip format in the db [13:16] rick_h_: writing is a bit different though [13:16] rick_h_: and if you can't create zip files client-side, that may be awkward [13:16] rogpeppe1: yea we can't do that reliably [13:17] rogpeppe1: otp can chat about it in a bit [13:17] rick_h_: a mime multipart archive might be possible [13:17] rick_h_: ok, enjoy! [13:23] juju-gui: anyone have the story on the "activeID in charm results" card in Project 1? i'd like to start in on it, but i need more context, like what part of charm results? (and is that 'precice' a typo or something we're meant to support) [13:29] oops. jujugui^ [13:33] jcsackett: I can fill you in after the call if no one else gets to you [13:34] rogpeppe1: I am back in the hangout [13:34] rick_h_: after the call i'll have ~40min of work left. might as well leave it till monday. [13:38] jcsackett: I mean my current call sorry [13:38] 30ish min [13:38] rick_h_: oh, ok. [13:46] hi rick_h_ -- i've tried quickstart on utopic with mixed results. it works great against ec2 but on local provider i cannot get services to fully start. [13:46] rick_h_: so there may be a juju issue there but not a quickstart one, afaict [13:48] bac: yea, we're waiting on the upcoming release [13:50] rick_h_: i've got 1.19.3. the fix for bug 1314686 was targeted to 1.19.2 so i thought it would be in [13:50] <_mup_> Bug #1314686: Please add support for utopic

[14:03] jcsackett: free [14:03] jcastro: do you know who works with sputnick stuff on our end? [14:04] yeah [14:04] bac: there was some mongo issue that I think needs the upcoming 1.19.4 [14:04] jcastro: I wanted to buy one but they won't let me and say stuff changing and want to make sure if I wait there is a new one vs it's ended or something [14:04] rick_h_, it was kamal mostafa, no idea if he still does it [14:04] jcastro: cool thanks [14:04] oh you mean business wise [14:04] jcsackett: meet you in the standup hangout [14:07] rick_h_: dig. [14:08] rick_h_: be a moment. it's doing the "you don't have the plugin installed" nonsense. [14:09] rick_h_: today's standup hangout? b/c i'm there, but you're not. and i can't imagine i beat you. [14:12] rick_h_: any suggestions on what i should do next? [14:14] bac: sec, after this call can chat. [14:16] hatch: I'm not seeing any changes on develop - did huw's branches not land yet? [14:16] kadams54 nope https://github.com/juju/juju-gui/pull/391 [14:17] I'm doing another review and qa atm [14:17] then it will [14:17] if all goes well [14:17] I can review and QA as well [14:18] It seems like there are quite a few things coming out of the machine view work that warrant further discussion with the design team. [14:18] oh? [14:19] There are consistency issues with the create container form, such as the "Move/Create/Add" label on the same button. Ditto for some labels within the form. [14:19] kadams54 ok if you could do the QA then I could keep going on my remove relation stuff [14:20] be very thorough there was a lot of hidden bugs the last time heh [14:20] jcsackett: frozen fail [14:20] There are questions about how to handle situations like dropping a unit on one machine while you have another machine selected [14:20] rick_h_: i've been knocked off hangouts; google claims i have no internet, which is funny b/c other things open. [14:20] lol [14:20] it hates you man. You've got more hangout issues than anyone I know [14:21] i know dude. [14:21] and i'm not doing anything unusual. [14:21] i have three google accounts--i know people with 5 or 6 who have fewer issues. [14:21] rick_h_: ok, it's loading again. [14:21] Now that machine view is really starting to shape up, we're driving out all the normal UX issues that happen when you take static visuals and turn them into a working, changing app. [14:21] should be there in a second. [14:22] kadams54: wheeeeee [14:25] antdillon: jcsackett hatch http://techcrunch.com/2014/06/19/google-launches-web-starter-kit-a-boilerplate-for-multi-screen-web-development/ kind of cool [14:25] yeah I've got it open in a tab [14:25] heh [14:25] haven't tried it yet [14:26] bac: ok got a sec? [14:26] rick_h_: sure [14:26] bac: @ standup hangout? [14:27] yep, joining [14:27] rick_h_: looks like it is occupied [14:27] bac: it should be free [14:27] I'm in it [14:28] rick_h_: url? normal daily standup has fb and rp in it [14:28] https://plus.google.com/hangouts/_/calendar/cmljay5oYXJkaW5nQGNhbm9uaWNhbC5jb20.t3m5giuddiv9epub48d9skdaso?authuser=1 [14:28] bac: ah I'm in the friday url [14:29] blast the friday url [14:32] makyo_ is there a secret 'get relation by endpoints' method hidden somewhere? [14:34] ahah I found it... [14:37] rick_h_ whenever you get a second we should chat about the ghost inspector and the ecs change stuff [14:41] rick_h_: that is pretty cool. [14:41] hatch: sure thing, sec [14:43] rick_h_, Looks good, I've been using browsersync.io, but will give it ago [14:43] antdillon: yea, all good. Just something to keep an eye on [14:44] jujugui: is /charmName a valid url? I thought we only wanted /series/charmName and /series/charmName-version [14:44] jcsackett it is [14:44] at least it used to be [14:44] hatch: i mean, it's allowed by the app. it's the reason we're having the login weirdness. [14:44] deal with it [14:44] ..... [14:45] * hatch snickers [14:45] and, if anyone creates a charm named login, is going to be an issue. [14:45] jcsackett: yea, it defaults to latest series [14:45] i mean, that's an *unlikley* charm name. [14:45] how about moving /login and /logout to a /somenamespace/loging etc [14:45] jcsackett I'm not sure I understand the problem...can't you just check if it's login/logout and call it a day? [14:46] hatch: what if someone makes a charm for a single sign on service and names it login? [14:46] i'm not saying it's likely, but it's possible. [14:46] we tell them in review that they can't [14:46] :) [14:46] it'll only matter on promoted charms [14:46] promulgated? [14:46] yea, only promulgated charms get that [14:46] approved [14:46] yeah....approved [14:46] :D [14:46] so it's a list we control [14:47] ok, if people are comfortable with that being a review req, i'm fine with it. just wanted to raise it as an issue. [14:47] hatch: ok, give me 2 and we can standup? [14:47] rick_h_ yep no rush [14:47] I'm reading some old code that has no comments and illegible var names [14:47] I feel like I'm reading some third party node.js code [14:47] rofl [14:48] ahh I crack myself up.... [14:48] hatch: k, loading [14:48] umm it's being broken... [14:49] just a sec, switching networks... [14:55] jujugui call in 5 === tvansteenburgh1 is now known as tvansteenburgh [14:59] jujugui call in 1 [15:01] jujugui: also, trusty->utopic via 'update-manager -d' hangs. 'do-release-upgrade' is the way to go. === hatch__ is now known as hatch [15:17] hi hazmat, i see you're out next week. could you have a look at my merge proposal for jujuclient before you head out? [15:19] jujugui: US-based people, reminder there is a 401-K meeting at 12EDT [15:19] bac, yes, there another on the same topic [15:19] there is [15:19] Oooo I'll take some 401K's [15:19] by noon i mean 1pm sorry [15:19] yeah.. curious what that's about re 401k [15:19] are plan options kinda blow [15:20] our [15:20] hazmat: yeah, i think they are looking at new selections. i've been pressing for some ETFs or similar [15:21] juju-gui: i could use a review on https://github.com/juju/juju-gui/pull/397 [15:28] bac: thanks for the reminder [15:31] that was such a dumb thing to do! [15:31] rogpeppe1 lol [15:31] keybd now sigificantly broken [15:32] uh oh, well now you have the weekend to take it apart and clean it [15:32] Way to keep on the sunny side hatch [15:38] well it's been pouring here for a week [15:39] so something's gota be sunny! [15:48] just in case everyone misses GOTO https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/label :P === makyo_ is now known as Makyo [15:59] hatch: i've just cleaned it ok, but problem is that in taking some of the keys off, some of the little plastic pieces under the keys have broken [15:59] damn lenovo junk :P [15:59] hatch: so now i'm missing an x, an r and half the space bar [15:59] I'm going to bet you can buy those parts....I hope at least [16:00] hatch: when the fan broke, they replaced the keyboard and i'm sure it wasn't so vulnerable before [16:00] hatch: i hope so too [16:00] I bet that plastic gets fatigued over time [16:00] being banged on day-in-day-out [16:00] hatch: it's less than 6 mo old [16:01] well what the heck... [16:01] IMO, keys should not break if you pull them off [16:02] I'm with you on that one [16:02] when I moved some keys around on my MBP they flew apart but didn't break [16:02] hatch: i spilled a glass of gin & tonic all over the old keyboard on the plane once, and it ended up fine [16:03] clearly it can hold its booze [16:03] ... [16:03] :D [16:09] have a good weekend all. [16:09] good. laptopkey.com seems to be the biz [16:11] hatch: Finished QAing huw's branch and left a comment. I really only found one problem and that might be better to address in a separate branch. [16:12] looking [16:12] rogpeppe1 great! [16:12] three new keys ordered... [16:12] $17.60 [16:12] ahh I love the internet, obscure businesses for everything :D [16:13] they even had pictures of the three possible keyboards this laptop could have [16:13] anyway, time for me to vamos [16:13] happy weekends all! [16:14] cya rogpeppe1 have a good one [16:14] kadams54 thanks for the qa - I agree, follow-up card [16:14] I only had one small note which maybe you can address in your branch [16:14] I'll shipit [16:15] No. [16:15] no don't shipit? [16:15] I categorically refuse to address anything else in my branch. [16:15] oh lol [16:15] ;-) [16:15] Alright, let's land this puppy so I can move from test hell to merge heaven… [16:15] ok shippppped - so kadams54 there is the one trivial comment I made [16:15] Oh, wait. [16:15] too late [16:16] :P well not really, but what's up? [16:16] OK, I need to quit making jokes that can be misinterpreted as "stop what you're doing" [16:16] can you create a bug for your first qa comment [16:16] lol [16:16] My "oh wait" was a follow up to my comment about "merge heaven" [16:16] ohhhh I get it [16:17] maybe it won't be so bad [16:17] hatch: will do on the bug [16:17] kadams54 so it's still going to be a good 30mins until it's merged just fyi [16:18] Yeah, I'm going to grab lunch [16:28] *sigh* turns out the method I thought we had, doesn't work like I thought it did [16:38] * rick_h_ runs to get lunch [16:39] rick_h_ saw your tweet, what are you getting? [16:39] hatch: ordered the xps13 developer edition with ubuntu on it [16:40] need a laptop with working mic/camera for all my darn calls [16:40] woah..... [16:40] don't those run 12.04? [16:41] yea, but not like I'd ever run the OS delivered anyway [16:42] ok, off to lunch now biaf [17:04] joining that conference call is the hardest thing i've done today [17:06] lol [17:40] rick_h_ heh true - so does 14.04 have all the drivers necessary? [17:41] hatch: I assume, we'll find out [17:41] at least the camera should work :) [17:41] haha - you hope... [17:41] :P [17:41] others have them in the company so I feel pretty good [17:41] I'm not going to love it, but there's such a lack of 'good' laptops out there right now [17:46] it's true [17:47] I'm digging this MBP but I really need to get some support from someone who knows how to write drivers :) [17:47] yea, the camera not working is a killer for me. [17:47] too many calls to have to do [17:47] and rebooting into osx to do calls is a pita [17:48] and using osx is an even bigger pita [17:48] yop! [17:53] Makyo: you're going off of https://drive.google.com/a/canonical.com/?pli=1#folders/0B7XG_QBXNwY1cWQ4TmRmRXJBc0E for the change version? [17:57] rick_h_, https://drive.google.com/a/canonical.com/#folders/0B7XG_QBXNwY1Vmowb25PejhJOTg [17:58] Which looks similar. [17:58] Makyo: oh cool I missed it on there when I loaded it up :/ [17:58] blindy friday [17:59] rick_h_ so on sprints are you going to bring three laptops now? lol [17:59] hatch: no, I'll be down to one [17:59] I'm tired of carrying two [18:00] and two power cords etc [18:00] This weekend I think I'll hop back into Ubuntu and see if I can get the issues fixed up [18:01] would really love to run it on metal [18:40] I'm going to run. Everyone have a good weekend. [18:40] cya rick_h_ enjoy! [20:47] kadams54 hey did the rebase go ok? [20:48] Yeah, it's pushed and the CI build is churning away on it right now. [20:48] oh cool, so it's ready for a review? [20:48] Yup. [20:56] bye jujugui. have a nice weekend everyone. [20:56] You too bac [20:56] you too, cya bac [21:02] kadams54 in the _displayCreateMachine method you're grabbing either the create-machine or create-container element for the container....but won't there also be a create-xxx in the unplaced units column when the user clicks the little button? [21:03] serviceunit-token.js handles its form a little differently [21:03] Since it needs to populate the dropdowns dynamically, including existing machines and containers [21:03] so we have two identical forms with two different code bases? :) [21:04] Fortunately no [21:04] oh ok good :) [21:04] It grabs the existing form and manipulates it into what it needs [21:05] I don't understand the comment about the kludge [21:06] Huw took the approach of showing/hiding various sections in the form by placing a "state class" at the form level… so if

, hide the constraints. If , show constraints. [21:06] I only discovered that after using the "hidden" class directly on the various sections within the form [21:07] In the other places the form is used [21:07] So in that bit of code we're both setting the state class AND having to remove the "hidden" class [21:07] It should only be one or the other and the approach needs to be consistent throughout the code [21:08] hmm ok thx [21:08] TBH, I meant to revisit that once everything was working and happy, but before it went out for review [21:08] Sorry. [21:08] np, we should get this landed [21:08] almost done review [21:13] kadams54 ok review is done [21:14] I'll do qa no [21:14] q [21:14] now [21:16] Ah yes… the async issue in the tests… [21:16] I tried wrapping them in after [21:16] And failed miserably. [21:16] Having chunks of the code executed as callbacks messed up stubbed methods [21:16] env.createMachine.lastArguments() was undefined [21:17] ok after QA I'll show you a diff [21:19] Here's what I tried… https://gist.github.com/kadams54/bc7cfc5ddb722bd42cc8 [21:19] and env was undefined? [21:24] One moment, let me check… [21:28] No, env is not undefined [21:28] And addMachines is still a stub method [21:28] It's just that lastArguments() now returns undefined [21:29] Something (f._allArguments?) seems to be dropping out of scope once the assertion happens in a closure [21:30] ok sorry I can look now, I was just sending an email re a qa issue [21:30] ok looking at the code [21:31] hmm that's very odd [21:32] kadams54 I"m ok with leaving it I suppose, it looks like it's causing more issues than it's solving....no idea why those issues are happening though [21:33] hatch: OK. FYI, I found a problem when QAing… now that I have Huw's stuff, the dropAction is not bubbling up through the events properly. [21:33] really? It's working here [21:33] what's your issue? [21:34] I'll try and repro it [21:34] When I drop an unplaced unit on the "Create Container" target in the container column header, the form shows up over in the machine column [21:35] The action === 'container' conditional is failing, so the container gets set to the form in the machine column. [21:35] OK, gotta run [21:36] ok be sure to add that comment to the PR so noone lands with it