[00:05] kadams54: Welcome, by the way! [00:05] hatch: replied to your branch [00:05] rick_h_ thanks, looking [00:06] Thanks huwshimi [00:06] kadams54: How's it all going? Settling in? [00:06] Makyo: thanks for looking at that PR. Didn't have a chance to post it here. [00:06] hatch: so we'll need the inspector trigger point as well in there and then something for machine view right? [00:06] huwshimi: yup. I participated in my first architectural discussion with Jeff and Rick today. [00:07] Whee! [00:07] he's still here yay! [00:07] rick_h_ yep [00:07] kadams54 lol! That one was pretty calm [00:07] we are getting better [00:07] :D [00:07] haha [00:07] hatch: but yea, that's the general pain in the @$#@$ idea for the first pass [00:07] we don't have gary to be peacemaker any more :( [00:07] kadams54: Got caught in the middle, huh. [00:08] he just took his headphones off and walked away [00:08] I agree 100% with rick_h_ [00:08] rofl [00:08] kiss a$$ [00:08] Except when I agree with hatch [00:08] haha [00:08] lol [00:09] rick_h_ I spent a bunch of time spinning my wheels on this today so the full inspector and machine view should be faster [00:10] hatch: cool, appreciate it. I know it's now fun :) [00:10] hatch: but, we'll hopefully get out of here with the inspector and browser.js cleaned up which will be good [00:10] just have to make sure we keep the rest of MV going and not get bogged down too much [00:11] yeah definitely - it's interesting how a poor architectural choice over a year and a half ago has so many implications today [00:11] re double dispatch and the like [00:11] huwshimi: did you grab a card? If so please make sure to update kanban so we don't dupe work [00:12] well, that's a big one, but I want everyone to walk away with the fact that it's more than that [00:12] there were still the render race conditions/etc. [00:13] oh right...I think those could have been mitigated without state but state was definitely the best tool for the job for the big picture [00:13] rick_h_: Oh, I didn't, I was working on the full height inspector. [00:13] it was funny how Nick's presentation was all stuff I hate. Fat views, promises all the way, etc [00:13] rick_h_: Is that this card: "update new inspectors UI to fit within space of the sidebar as a subordinate View"? [00:13] haha yeah I was laughing inside about that [00:13] huwshimi: ok cool, just want to make sure we don't dupe things and waste time [00:13] I liked the dependency injection though, that was unique [00:14] I am not sure I'd use it [00:14] but cool none the less [00:14] huwshimi: it was meant to be, but I think that card got moved over when you did the ghost inpsector [00:14] huwshimi: so feel free to setup a new card or something [00:14] huwshimi: just so we can tell what's up in the morning [00:14] rick_h_: Yep, will do. [00:14] hatch: yea, I mean it's something I'd have thought of, but we've been in the battlefield. [00:15] hatch: and that changed opinions somewhat. Testing, etc [00:15] yeah - I think in Java land people are used to stubbing out more complex structures than we do [00:15] although you would only need to stub out one large one and pass it around [00:15] which might be nice [00:16] well I got the impression their testing was less than thorough [00:16] which helps when you can just say "this is hard to test, carry on" [00:16] yeah he complains about that in #yui [00:16] vs revisiting the design and looking at *why* it's hard to test [00:16] I guess people would just let things through [00:16] hatch: yea, it was one of his issues with current employer which is good [00:17] hah yeah [00:17] ugh this cold is kicking my @$##@$# [00:17] heh you did sound a little out of it [00:17] like you were talking through your nose a bit [00:17] I have a hard time thinking when I'm sick. Brain slows down [00:17] yeah, same for all [00:17] so you got me at half strength today lol [00:18] resources are being used for killing baddies [00:18] haha, good easy pickins [00:22] huwshimi: oh sorry, missed the card for it in the ready to code [00:22] yes, that's the one [00:22] huwshimi: blind tonight [00:22] rick_h_: Oh, I just added that :) [00:22] huwshimi: yea, got it reset [00:23] rick_h_: Oh, you're playing with things :) [00:23] huwshimi: yea, updated it right [00:23] rick_h_: Thanks [00:23] trying to see if i can find something feaure related for kadams54 next :) [00:25] a grunt task for building yui modules - maybe we can use this instead of our loader (which doesn't work very well) https://gist.github.com/nhusher/9902519 [00:25] ^ written by Nick [00:25] as soon as we can ditch the grunt part of it [00:25] :P [00:25] Hah! Not sure if I want to see what sort of fun rick_h_ finds for me when his mind is a cold-medicated haze. [00:25] slack task! [00:26] kadams54: hey, I've avoided medication all day [00:26] that just makes me useless, so I'll just stick with miserable [00:26] lol [00:26] * huwshimi heads home from coffee shop [00:28] rick_h_: feel free to send any additional state cards my way… I see several more out there, but not sure if I should hold off until hatch's work lands. [00:28] kadams54: so would love to see you pick up either the deployed unit token or the listing of machines in to the MV panel [00:28] kadams54: we can pre-imp to go over them in the morning if you're up for them [00:28] kadams54: yea, I think the inspector is blocked until hatch is done, no pressure hatch [00:28] Sure, listing of machines sounds interesting [00:29] It looks like those two cards are related? [00:29] kadams54: yea, should be light but get you touching the environment, db, etc. [00:29] kadams54: yea, so one column is a list of machine, the other is the containers in that machine [00:30] so they're close, but slightly different bits of data so two chunks of work [00:30] one is the nested data of the other [00:30] OK [00:30] Let's do it [00:31] Do those require the architecture chat about the machine view view organization? [00:32] hatch: I don't think so, it's just expanding the current View with some data from the db [00:32] ahh ok cool [00:32] we should maybe block out some time this week for that chat [00:32] hatch: nothing about workflow, etc. [00:32] yea, but need to unblock inspector/deployer first [00:32] machine view is the last thing to get done as it needs everything else [00:32] oh yeah for sure [00:33] so it might be a bit before we get full into MV arch [00:33] so many moving parts at once atm wheeee [00:34] haha yeah after this is all done i'll be booring [00:34] heh, except then we'll have the big daddy of work coming down the pipe [00:34] on top of a bunch of smaller things we've not gotten done yet [00:34] haha truth truth [00:35] if we can hit boring by vegas I'll be a happy techie. It won't be afterwards [00:35] lol [00:35] Yo Dawg We Heard You Like Webkit..... https://github.com/trevorlinton/webkit.js [11:59] * frankban lunches [13:10] jujugui afk for a bit making sure the boy goes to tumbling class [13:19] guihelp: I need two reviews for https://github.com/juju/juju-gui/pull/215 . Anyone available? [13:22] frankban I can take one in about 40mins [13:23] hatch: thanks === rogpeppe2 is now known as rogpeppe [13:33] rick_h_: let me know when you want to chat about machine view [13:58] frankban so the deployer now reports back to the gui the status of the bundle deployments? [13:58] Is this a new deployer feature? [13:58] hatch: no, that's an old guiserver API we never used [13:59] kadams54: rgr, shoot me a link [13:59] oh, so did we do it incorrectly when we did the old GUI implementation? [13:59] curious as to why we didn't do it this way to begin with [14:00] hatch: just didn't think about that when we added it [14:00] hatch: previously we just watched the progress of deployment started by the GUI itself [14:01] ohh ok cool [14:01] I was just curious [14:01] rick_h_: https://plus.google.com/hangouts/_/7ecpikj49vjg2en7dif8j4dvuc [14:01] hatch: fyi, added some more notes and such to the state urls google doc for the refactor work [14:03] kadams54: sec, google hating on my dual account nature [14:05] rick_h_ ok I see - so much $ [14:05] ;) [14:06] jcsackett I couldn't get wordpress to balance using haproxy so I think I might be using haproxy wrong... [14:06] do u have any experience with it? [14:21] hatch: heh, yea well trying to break down what the new bits of data we'll need. Simplify it down some [14:21] hatch: i haven't played with haproxy at all, tbh, but i can try to look through the docs some evening this week. [14:21] :) [14:22] jcsackett well I create the relation...then what? I figured you would access the related service on the haproxy ip [14:22] but no such luck [14:22] thanks - I'm trying to find time when I can [14:31] * hatch is trying juju quickstart lxc in a saucy vagrant [14:31] * hatch crosses fingers [14:31] hatch: http://the.taoofmac.com/space/HOWTO/Vagrant [14:31] Pretty detailed instructions on setting up LXC containers in Vagrant on a Mac [14:32] kadams54 I'm leaving that to juju [14:32] I've gotten about half way through them - running into problems installing Guest Additions from the commandline. Not seeing the CD mount anywhere. [14:32] it's apparently provisioning the machine [14:33] Let me know how it goes [14:34] The instructions above get you to a point where LXC containers have direct access to files on the Mac file system and connecting in is "ssh user@containers.local" (yay Bonjour) [14:36] that's pretty cool, I'll give that a go, I haven't thought about running lxc's in vagrant for anything other than juju [14:36] it's still provisioning...maybe it's hung hehe [14:40] hey look at that, it's working [14:40] .... [14:40] famous last words possibly... [14:40] lol [14:41] frankban your review is done I'm just qa'ing it [14:43] looks like I'll have to create a ssh tunnel to the gui instance in the vagrant [14:43] it's given a 10.x.x.x ip [14:45] hatch: cool thanks [14:46] kadams54: so in poking around there is an app.db.machines list that's a YUI model list [14:46] Makyo do you know what the pw is for the vagrant images? [14:46] kadams54: and you should be able to monitor that model list ofr events to keep things up to date I'd think [14:46] you bet [14:46] (finally something that won't rely on dbupdate) [14:47] yea, I'd hate to have to do that for these Y.Views like that [14:48] Great [14:49] kadams54 http://yuilibrary.com/yui/docs/api/classes/ModelList.html look for the 'events' section [14:50] so did y'all hear about the world's biggest ferris wheel that opened yesterday in vegas? about 25% taller than london eye. [14:50] jujugui call in 10, please kanban [14:51] bac yeah? April Fools? [14:51] nope [14:51] linky [14:52] hatch: not handy. it is called 'high roller'. [14:52] LMGTFY [14:53] bac I see it now [14:53] http://en.wikipedia.org/wiki/High_Roller_(Ferris_wheel) [14:53] too bad it's view is of a big parking lot lol [14:53] $35 or so [14:53] yeah, its right behind our hotel [14:54] It needs to be on top of a casino [14:54] Like that roller coaster [14:54] Which was also called High Roller, but is now closed :-( [14:54] not to outdone they are planning a bigger one in dubai [14:55] s/to o/to be o/ [14:55] it'll probably have decent views at night [14:55] during the day vegas is just dirty [14:55] :) [14:55] and at night it is dirty and lit by neon? [14:55] Dubai: the Samsung Mobile of countries. "That's awesome. Now we'll copy and make it bigger." [14:55] glow in the dar dirt woo [14:56] bac the dirty is harder to see lol [14:57] hatch, sorry. What do you need a password for? [14:57] Makyo I found it, it's vagrant:vagrant u/p [14:57] Okay. [14:57] Makyo create a ssh tunnel to the juju-gui lxc in vagrant [14:57] I've not yet needed it. good find. [15:00] Makyo yeah - it's pretty cool, quickstart and everything works in vagrant [15:00] it's a little slow spinning up...but whichever [15:01] kadams54: ping for stand up [15:01] Hah, sorry [15:01] Too involved in this MV stuff [15:03] kadams54: :) [15:12] frankban I don't think the QA worked, do you have a second for a call? [15:12] oh wait... [15:12] hatch: sure [15:12] * frankban waits [15:13] sorry I just want to finish the qa :) [15:13] np [15:14] frankban in the standup room [15:21] hatch: if you need to double check bundles status, you can also point your browser to https:///gui-server-info [15:21] ohh ok, too late [15:21] :) [15:21] but It was working as expected other than that one 'issue' [15:21] heh [15:22] cool [15:38] kadams54 xscope looks pretty cool [15:39] Didn't buy it for a long time because $30 seemed a lot for a glorified screen ruler. I was wrong. [15:41] ccccccdilrkrudgkrilnrbkvbvvhbclkjfkbtugulikd [15:41] bah [15:41] rick_h_ is trying out a new vim-irc plugin [15:41] jujugui going afk for a little bit. If you need me feel free to call, text, or etc [15:42] get betta [15:42] kadams54 lol [15:43] OK, so app architecture question: this came up in Nick's interview, but how do we handle passing the DB into views? [15:46] kadams54: passing it as a view attr, so that the view can do "this.get('db')" is the option used when instantiating the environment view [15:47] kadams54 new Y.juju.myView({db: this.db}) [15:47] for example [16:02] OK, I follow that, but one thing that puzzles me is that I don't see db listed in the ATTRS bit in environment.js [16:09] kadams54 yeah sometimes we forget to add them [16:09] feel free to add it and document it :) [16:09] if you pass it in, it's there, weather you define it or not [16:09] kadams54 when I started almost no attributes were listed/documented [16:09] :( [16:09] :-( [16:11] Another question: it looks like there's a JujuBaseView in utils that has functions for binding a model to a view - should I use that mixin with MachineViewPanelView? [16:11] I'm guessing no [16:11] but let me see [16:11] Seems more useful when a view is tightly tied to a particular model (i.e. a detail view) [16:12] Less so when your view be handling multiple models [16:12] it's only used in the login view and notifications view [16:12] kadams54: yeah for a model list it does not seem to make much sense [16:13] Pro tip: if your headset has controls that clip to your belt, it's always a good idea to remove it before walking away from your laptop. [16:16] lol [16:17] On a related note: USB: can haz magsafe connectorz? [16:18] hah! magsafe is pretty cool but I've found sometimes they dont' quite line up properly [16:18] although it has saved my laptop a few times :) [16:19] magsafe is magical if you have pets and/or kids [16:20] no kids!!! [16:21] I think dogs may be the worst [16:21] They get *SO* happy and hyper. [16:22] haha, we have 3 [16:22] OHMYGODSOMEONESATTHEDOORWAGTAILDESTROYEVERYTHING!!!! [16:22] That's my chocolate lab [16:23] My friend has pugs [16:23] haha it's true [16:24] I once watched them sweep a bowl of chips and a bowl of salsa off a coffee table by dragging all four PS2 controller cables across. And they didn't miss a beat, charging right on through and up to the door. [16:24] haha love it [16:26] Lunching [16:50] jcastro: do you have a link handy for the rails scalable bundle? I can't seem to find it at https://jujucharms.com/sidebar/search/bundle/~charmers/mongodb/5/cluster/?text=bundles [16:51] arosales, the scalable bundle doesn't show up in the store [16:51] rick_h_ is aware of it [16:51] http://manage.jujucharms.com/bundle/~charmers/rails/example-single is what we have right now [16:54] jcastro: for the flyer I was going to put the quick-start examples and we have rails scaleable there [16:54] arosales, yea it used to work and then something changed, I don't think it's broken forever broken [16:55] just normal broken [16:55] rick_h_, ^^^ [16:55] do we know what the quick-start command will look like, perhaps, 'juju-quickstart bundle:rails/scalable [16:55] juju-quickstart bundle:rails/example-scalable [16:56] jcastro arosales he is sick so probably will pop back in later on [16:56] ok [16:56] hatch: thanks [16:56] hatch, do you happen to know who was looking into our missing bundle? [16:57] do any folks know when 'juju-quickstart bundle:rails/example-scalable' will work again? [16:57] I would like to put this into print, but I don't want folks trying something that won't work either [16:57] jcastro umm one sec let me see if I can find the discussion [16:59] jcastro sorry I don't see it anywhere in the task list. bac do you know anything about the status of ^ [17:01] jcastro, arosales: "juju quickstart bundle:rails/example-scalable" seems to be working here. so my guess is it's present in charmworld but there is a problem with the search [17:02] http://manage.jujucharms.com/bundle/~charmers/rails/example-scalable [17:02] frankban: ah good to hear that the quick-start command is still working [17:03] frankban: the correct command though is "juju-quickstart bundle:" correct? [17:04] arosales: yes. quickstart accepts several ways to specify a bundle, for promulgated ones, bundle:basket/name is sufficient [17:05] frankban, oh cool, thanks for getting rid of ~charmers for that. <3 [17:05] :-) [17:06] I was talking to a guy the other day who said 'charmers' is so confusing [17:06] frankban: thanks for confirming the command still works I will insert that into our upcoming flyer and highlight the good works folks here are doing [17:06] he was wondering why you can't just promote someones charm, why you need to go through so many steps to find the owner of the charm [17:06] ^ jcastro just fyi [17:07] hatch, what do you mean by promote [17:07] promote as in promote in the store [17:07] or promote as in spread the word? [17:07] umm lemme find an example [17:07] arosales: cool thanks. simplified bundle URLs have been introduced in quickstart 1.2. "juju quickstart -h" includes some examples [17:09] jcastro so for example https://jujucharms.com/sidebar/search/~clint-fewbar/precise/haproxy-1/?text=haproxy vs https://jujucharms.com/sidebar/search/precise/haproxy-28/?text=haproxy the 'promoted' one is in an entirely different repository than the maintainers [17:09] oh you mean the ability to use a personal branch as the canonical store one? [17:10] yeah - he didn't understand why one just can't be 'promoted' and needs to be a copy losing all of the commit histories and suh [17:10] such [17:11] * arosales looking at "juju quickstart" and "juju-quickstart" [17:11] jcastro: it should be back in there now [17:11] rick_h_, not in the website [17:11] but we can quickstart it [17:12] jcastro: hmm, it was showing will look [17:12] otp [17:12] no worries [17:12] jcastro: we should sync up with curtis and have quick-start promoted bundles put into charm testing. [17:26] jcastro: bah, it was working. Now it's gone again. It exists but it's a search bug http://manage.jujucharms.com/bundle/~charmers/rails/example-scalable [17:27] * rick_h_ adds the bug back [17:57] yay my blog post finally came in handy for myself lol http://fromanegg.com/post/51864370819/xor-in-javascript [17:57] Heh [19:01] Holy crap forgot about appointment. Bakc in a bit! [19:34] somehow I broke half the routing [19:35] man this state stuff [19:35] .... [19:37] oh aync callstacks you rock [19:39] can anyone explain how i can clip my charm icon image to the rounded corners in the icon.svg template? [19:39] * tvansteenburgh is an inkscape newb [19:39] Hi tvansteenburgh there are some docs about that [19:40] one minute I'll see if I can find it [19:40] tvansteenburgh https://juju.ubuntu.com/docs/authors-charm-icon.html [19:40] it comes with a template [19:41] i'm embarrassed that i didn't notice the Charm Icons section of those docs [19:41] tyvm [19:41] haha np that is a long list :) [20:02] rick_h_ you around? [20:35] hatch: what's up? [20:35] I think I'm done this branch, just writing/fixing tests now and wanted to make sure I did what this card encompasses\ https://github.com/hatched/juju-gui/compare/juju:develop...hatched:new-states?expand=1 [20:36] looking [20:37] the viewstate needs to go it's quite the headache :) [20:37] but I think that's a breaking change [20:37] so maybe we want to do a release first [20:37] feeling better btw? [20:39] what's that? You don't care for viewstate? :P I'd never have guessed [20:41] lol - but THIS time I have science on my side! [20:41] * hatch considers programming a science [20:47] hatch: comments added [20:47] hatch: but yes, that's the card encompassing [20:48] hmm - I don't understand your last comment, hangout? [20:48] sure [20:48] https://plus.google.com/hangouts/_/7ecpjmkjhr157f7i1s2f9vdfb8?hl=en [21:11] hmm [21:29] hatch: can you jump back in that hangout? [21:30] https://plus.google.com/hangouts/_/7ecpiqb1951juiealdk2fq8r6k?hl=en [21:30] new link [21:42] kadams54 I thought you were taking off early? :) [21:42] "kid can wait" [21:42] I did. Now I'm back. [21:42] :P [21:42] Soon I'll be gone again :-) [21:42] haha [22:01] Morning [22:03] morning === hatch__ is now known as hatch [22:04] morning huwshimi [22:06] rick_h_ ok thingy moved....should be good enough to get others unblocked [22:06] now to fix tests and the like [22:07] hatch: make sure to use FF on the tests vs making them test broken things please [22:07] ideally the changes are behind FF and the existing tests 'just work' [22:08] FF? [22:08] flags? [22:08] yes feature flags [22:11] ohh, yeah, it's all under il and mv now [22:11] those should be merged soon [22:11] since both are required now [22:11] well not 'required' but might as well [22:11] well, yea they're different and helpful for QA for design atm [22:11] e.g. test inspector stuff, drag/drop vs the machine view panel covering everything [22:14] no longer an issue once this branch lands :) [22:18] hatch: close, but yea. After the refactor and and move the deployer bar stuff from IL to MV [22:34] hmm phantom keeps crashing again [22:35] I was wondering when this would start happening again [22:38] Makyo did your tests start to fail in juju-views-utils ? === hatch__ is now known as hatch [22:44] hatch: just tried out tests with updated trunk and after a clean-all to update the d3 stuff they passed [22:44] yeah these aren't failing or anything [22:44] phantom just stops [22:44] oh bummer [22:44] yeah this used to happen for the longest time then it just resolved itself [22:45] ....should have known [22:45] heh, cursed hatch computers [22:46] ExceptionHandler::GenerateDump waitpid failed:No child processes [22:46] ugh [22:46] PhantomJS has crashed. Please read the crash reporting guide at ... [22:47] hatch, erk, sorry, was on the couch with the other computer. It was something in there with a # has no method something something. [22:47] Makyo oh ok np, so it wasn't phantom just crashing ok [22:50] maybe a reboot will fix [22:50] :) [22:50] Hahaha [22:50] I feel like a windows developer again. [22:55] hmm nope [23:17] hatch - now crashing for me, too, same error, different location each time. Will try on ubuntu directly + browser in OS X [23:17] Er, I meant to leave that in the input buffer. [23:51] hatch - now crashing for me, too, same error, different location each time. Will try on ubuntu directly + browser in OS X [23:51] test-server works, though. [23:52] ohh interesting....ok well I'm glad that it wasn't just me [23:52] I was going to use test-server as well :/ [23:53] That will ensure that tests pass at least somewhere. [23:53] Will have to hunt that down. [23:53] yeah I tried to make it run in CI but GH was down [23:55] I hope that' CI isn't borked because it runs in phantom once too [23:55] at least I think..