[01:09] rick_h__: Thanks a lot! [01:12] huwshimi: np, best of luck [01:12] huwshimi: make sure you doc all this stuff down somewhere to ease us when/if we pull something new [01:12] rick_h__: OK will do. [04:58] Oh no, a positive lines of code branch! [11:16] jujugui, I'm reviewing huwshimi's branch unless someone tells me they are already doing it [11:18] gary_poster: thanks! [11:18] :-) welcome [11:42] rick_h__, hey. did you and antdillon discuss what to do about the cookie footer now that the main footer is gone?it hides behind the charm browser sidebar now, which does not seem good. The best easy option I see is to have it pop over with a z index, and then have the charm browser add a bottom margin that makes it possible to see all content above the cookie warning. wdyt? [11:42] 'course, then it will also be above the inspector, which won't look right. :-( [11:42] the z-index sounds good, I'd not worry about the margin as it's a "warning" type dialog that goes away [11:43] I guess so [11:43] "hey, here's this thing in your way, click it and decide so you can move on" [11:43] imo [11:43] :-) ok [11:43] thanks [11:43] but no, antdillon and I didn't talk about it. Didn't notice it. [11:43] UX might have different feelings [11:43] rick_h__, gary_poster I agree popping it over the top is good [11:44] cool thanks antdillon. [11:44] rick_h__, antdillon I'll make a card for that in urgent lane. should be super fast. [11:44] gary_poster, I'll run it by Luca when hes next in and have a "final" answer shortly [11:44] cool perfect [11:45] gary_poster: rgr [11:46] jcastro: looks like your bundle made it http://manage.jujucharms.com/search?search_text=wordpress&op= [11:49] and it exposed a GUI bug :-P [11:49] yea, recent commit error [11:49] filing now [11:50] thanks [11:51] bac, oh you're not here for me to tell you I assigned the bug so...you're assigned [11:52] :-P [11:52] I'm super helpful like that :) [11:53] heh [12:10] I have performed the all-important job of adding a new cute icon for Juju Quickstart to the kanban board. [12:10] Go me. [12:11] lol [12:11] come on, they should all be halloween themed for today [12:11] heh, if only I had that option [12:16] frankban does not have IRC pings on his current computer. :-) [12:43] hi evilnickveitch [13:14] rick_h__, ah nice, I love causing bugs! [13:15] should bundles show up in http://manage.jujucharms.com/recently-changed ? [13:15] rick_h__, ok I'll commit the other ones, now that I have them I'll make the READMEs more spiffy too [13:15] they're just placeholders right now [13:59] huh. bac *was* here [13:59] * gary_poster goes to get some water [14:04] * rick_h__ runs for coffee [14:07] gary_poster: ping [14:19] rick_h__, ok all my bundles are pushed, poke me if there's an error [14:27] jcastro: rgr [14:42] weeeee https://twitter.com/FromAnEgg/status/395923814279696384 [14:43] hatch: heh [14:43] :-) [14:44] just don't forget to keep that block of code up to date if you add a new widget [14:44] :/ [14:44] hopefully we don't have to! [14:44] die widgets die [14:44] psh, shush. I thought we were done trying to rewrite all the things [14:45] I'm never done, I'm just tired from the last go around [14:45] lol [14:45] * rick_h__ pushes hatch off the edge and into the pool to cool off [14:45] ahh a pool [14:45] https://github.com/incompl/cloak [14:46] not really related to us, but a cool project none the less [14:47] the awesome charity truck build that I'm involved with http://wheelsofdreams.ca/ had it's private unveiling last night. It's awesome, I can't wait until the embargo is off so I can share pictures of the completed truck [14:50] jujugui call in 10 [14:55] rick_h__, from sinzui: "/me tries on staging [14:55] charmworld does not support featuring bundles. not in models, not in UI [14:55] * gary_poster just tried" [14:56] that was sinzui, not me [14:56] gary_poster: has to be promulgated first [14:56] ah thanks [14:56] gary_poster: at least that was the discussion yesterday, not sure if that happened/not [14:56] * hatch officially proposes changing promulgated to promoted :P [14:56] :-) [14:56] hatch: hey, I ran into promulgated in a book the other day and went "Oh oh I know what that means!" [14:57] hahaha [14:57] heh [14:57] rick_h__: did you see my comment yesterday about 22 Jump Street being filmed in Regina Saskatchewan? [14:58] we are taking over....I am tellin ya! [14:58] hatch: yea, just didn't mean anything to me. I thought 21 jump street is a 20+yr old tv show? [14:58] :P [14:58] then I hear they made a bad movie out of it [14:58] lol [14:59] jujugui call in 1 [15:19] Haloween - the only day of the year when it's acceptable to run door to door screaming at the home owners and taking candy from strangers [15:19] woot! [15:19] lol [15:19] I'm sure when I have kids I'll enjoy it again haha [15:19] hazmat: feature request, tags at release points please :) [15:20] right now it's firmly in the "Get off my lawn" category :D [15:20] rick_h__, bzr tags? ack [15:20] hazmat: yes please, tracking down a change from 1.1 to 2.5 and would <3 to have a tag to head back to [15:21] * hazmat does parse those rev numbers.. [15:21] rick_h__, i assume you mean deployer? [15:21] hazmat: yes [15:21] hazmat: come to find out we already had it in charmworld for a get_flattened_deployment function that no longer exists [15:22] rick_h__, that never got merged [15:22] hazmat: ah ok, good to know [15:23] rick_h__, fwiw https://code.launchpad.net/~abentley/juju-deployer/get-flattened-deployment/+merge/175689 [15:24] hazmat: cool, yea pulled it from the download-cache. Will get it updated. Thanks. [15:25] hazmat: do we support inheritence? /me thought it was not allowed, but maybe that's a *for now* thing [15:29] rick_h__, inheritance works now.. i'd suggest subclassing a deployment and overriding fetch_charms/get_charms with charm objs from mongodb [15:32] charmworld supports inheritence on ingest; we use the deployer as a library and perform the inheritance before storing the result [15:32] benji: yea, I'm looking at how best to update it for the updated deployer version [15:32] k [15:48] benji: ping, got a sec to chat? [15:49] rick_h__: sure [15:49] benji: https://plus.google.com/hangouts/_/72cpieoqhi2s1ddd7061crjaqg?hl=en [15:51] PSA for OS X Mavericks users: ssh only reads ~/.ssh/authorized_keys now. if you have authorized_keys2 it won't be used so just rename it [15:51] marcoceppi, hey. You available for a hangout to help me understand some of your concerns before the larger call? [15:52] gary_poster: yeah, I have a call now but in an hour I'll be available [15:52] ok thanks marcoceppi [15:53] Mostly, just need some clarifications :) [15:55] gary_poster: do you still have trunk with maartens file? I'm wondering if you get an error in utils.js with it [15:55] it appears to be the method which finds the center and pans to it [15:56] I can't seem to repro it before I changed the error reporting.... [15:56] which doesn't make any sense haha [15:56] hatch, I still have it. but it didn't deploy because of an error [15:57] I gave you the file because it was able to trigger an error [15:57] so... [15:57] I dunno, what do you wnat me to do? [15:57] :-) [15:58] ohh ok - trunk has that fix [15:58] I thought you were running on trunk [15:58] no problem I'll debug further [15:58] Uh [15:59] hatch why don't we have quick hangout? Don't want to be quite this lost ;-) [15:59] haha ok [15:59] hatch https://plus.google.com/hangouts/_/calendar/Z2FyeS5wb3N0ZXJAY2Fub25pY2FsLmNvbQ.t3m5giuddiv9epub48d9skdaso [16:07] there is something definitely wrong with the 'layout' stuff - if you put a debugger in the charm deploy code, they don't get placed in the proper spots [16:08] this one won't be fun to debug ^ Makyo [16:08] just FYI [16:09] I'm missing large chunks of information, but I'll trust your judgment that it won't be fun :) [16:11] haha - sorry that was definitely not a complete thought [16:11] lol [16:11] basically when deploying a bundle from a yaml file, if you put a debugger in the 'loop' somewhere, the positioning gets broken [16:12] so it's not a real bug, but probably a symptom of a larger issue.....whatever that issue may be [16:14] gary_poster: I've got time now [16:14] marcoceppi, awesome, thank you https://plus.google.com/hangouts/_/calendar/Z2FyeS5wb3N0ZXJAY2Fub25pY2FsLmNvbQ.t3m5giuddiv9epub48d9skdaso [16:16] so odly enough my phone just dinged telling me that gary and marco were in a hangout.....should I be concerned? lol [16:16] lol [16:19] I thought hangouts were private by default, I wonder why it invited me [16:19] because I reused [16:20] ohhh [16:20] makes sense [16:25] * hatch shakes mocha/phantom "why you crash?????" [16:25] it hates you :P [16:26] clearly! [16:30] bac: no codereview emails came from your merge proposal FYI [16:31] hatch: it is still thinking [16:31] ohh odd :) [16:31] it's not sure if it wants to review your code? :P [16:31] 'lbox propose' is running, has printed the LP MP url but hasn't proceeded [16:32] 'make check' run by lbox failed for me multiple times. :( [16:32] does it just crash for you too? [16:33] hatch: no, there is a spurious test failure that only occurs when make check is run by lbox [16:33] oh odd mine is the exact opposite - it crashes when it's NOT run by lbox [16:33] lol [16:34] so that leads me to believe it has something to do with all of the data being pushed through the console which causes phantom to crash [16:39] gary_poster: when you have a moment could you qa/review https://codereview.appspot.com/17540044/ since you have the yaml files :) [16:52] on it, hatch [16:53] thanks [16:56] frankban: are you going to need to pass your branch off to someone? [16:58] hatch, I *think* the bad layout issue might happen when you drag in a deployment file too soon after a reload? so far I have not been able to disprove that guess [16:59] I was thinking similar but the same issue happens with debuggers in the deploy 'loop' so my guess there is some underlying issue that these are symptoms of [16:59] huh [16:59] ok [17:08] https://bugs.launchpad.net/juju-gui/+bug/1246462 rick_h__ gary_poster do we have any ideas as the best place to put this? I was thinking beside the developer name [17:08] <_mup_> Bug #1246462: bundle tokens need to be basket:bundle for the name/title [17:09] hatch I think they need to be part of the big na,e [17:09] that's going to be tough because we only really have room for....say 15 characters [17:09] hatch, where now they say wiki, it would say "envExport" they would say "wiki:envExport" [17:09] hatch: the api provides a basket_name attribute that needs to be in the model and then part of the bundle token title [17:10] hatch: well, but let's see it and we can toss an example to UX from there? [17:10] yeah well it could be SuperCharged:TestBundle at which point it would show SuperCharged:... [17:10] :) [17:10] https://manage.jujucharms.com/api/3/bundle/~hatch/wiki/7/TestBundle example api data [17:10] hatch agree with rick_h__ . If we show only one thing at the top it should be SuperCharged, and then TestBundle would be beneath [17:11] hatch: yea, so maybe the token needs a XXX: SuperCharged, or the title becomes the basket name and the bundle name is the sub attribute [17:11] hatch, make syre that mouseovers show full struing? [17:11] ok so the bundle name beside the developer and the basket at the top [17:11] that sounds good [17:11] mm [17:12] uh [17:12] hey luca__ ! :-) [17:12] gary_poster: Heya [17:12] hehe [17:12] gary_poster: hows it going? [17:12] you should have ran [17:12] :P [17:12] lol [17:12] lol [17:12] hey. pretty good. :-) quick question for you, I hope [17:12] go ahead [17:12] the big thing is which is the more important bit of data to the user? the basket name or the bundle name? (I'm thinking basket name myself) [17:13] rick_h__: I'm going to tend to agree but I haven't fully wrapped my head around the basket and deployer relation yet [17:13] so, we need to show the "basket: (deployer file/bundle collecton) name and the bundle name. concrete example might be that we have a bundle collection named mediawiki, and we have three bundles named small, medium and large [17:14] so we can show that it two ways that we see in the token [17:14] (1) the main name becomes mediawiki: small [17:14] (and mediawiki: medium and so on [17:14] ) [17:14] (2) we show mediawiki as the main name [17:15] and then put the bundle name--which is kind of like a flavor, let's say, in a subordinate line, like with the maintainer [17:16] luca__, they both are names, but "mediawiki" is kind of like the family name and "small" or so on is the bundle's given name [17:16] only together are they really best distinguished [17:17] http://localhost:8888/sidebar/search/bundle/~hatch/wiki/6/envExport/:flags:/charmworldv3/?text=hatch is a way to look at the current state if that helps [17:17] uh [17:17] http://comingsoon.jujucharms.com/sidebar/search/bundle/~hatch/wiki/6/envExport/:flags:/charmworldv3/?text=hatch [17:17] :-D [17:18] I lean towards #1, but hatch's concern is that the name might get too big. I think it is the name anyway, and we should just make sure that mouseover tells the whole story [17:18] If we do #2, I think it is harder to convey the idea that the secondary bit is the bundle name [17:19] luca__, finished. what do you think? [17:19] also to note about #1, nothing else does the mouseover-for-more, so disoverability would suffer....maybe? [17:19] maybe we can use a smaller font :D [17:20] I think ellipsis is a pretty common indicator of mouseover for more. [17:21] yeah you're probably right [17:21] (of course luca probably has totally different plans) haha [17:21] :-) [17:21] I think #1 sounds the better option [17:21] I wouldn't do a hover over [17:21] ok [17:21] just have the ellipses to show that this name is longer [17:21] just click on it for details, as usual? [17:22] yeah [17:22] yeah ok [17:22] It's a difficult thing to solve [17:22] luca__, you should be used to it ;-) [17:22] would we allow bundles to be named Mediawiki? [17:22] people can call them anything they want [17:23] in promulgated bundles I think we would encourage that family name/given name style [17:23] That seems a little odd, but potentially its ok [17:24] ok. hatch, you ok with that? should be easy to do at least :-) [17:24] yep for sure - I'm not sure that solves rick_h__'s original concern though. That you will just see X number of bundles and have to click through to see why they are different [17:25] if the names are big yes [17:25] but not if small [17:25] similar to charms [17:25] let's see if it proves to be a problem [17:25] hatch: well primary is I think that basket_name isn't shown at all, that's primary as I think it's the most important bit [17:25] and consider it once we have real-world data [17:25] right [17:25] yeah that's fine [17:25] +1 interation ftw [17:25] man we need more data [17:25] :D [17:25] :-) [17:26] one review/qa needed for https://codereview.appspot.com/20170044/ [17:27] * bac beat up lbox! [17:27] we need to add try/catch with retry on loadFixture. may do that as a slack task soon. [17:27] bac: looking [17:28] bac: I'd actually propose a loadFixture cache instead [17:28] get rid of all those darn GET requests [17:28] hatch: the load fixture is being called by charm cache filling [17:28] loadFixture shouldn't be getting. It's a fixture. [17:28] you give it json and it loads it [17:28] * rick_h__ is confused [17:28] can't cache the cache before you fill the cache [17:29] loadFixture doesn't make an IO request? [17:29] does too [17:29] right - so if you request the same fixture twice, it shouldn't make a second IO request [17:30] hatch lots of small things in my review but LGTM and QAOK [17:30] _cached_charms is where the failure occurs. loadFixture throws a 404. [17:30] so loadFixture would need to be some persistant class in the tests for that I guess [17:31] bac: ohh you're getting the same 404 as Makyo - yeah I have no idea whats up with that, I've never seeen that one [17:31] gary_poster: thanks will look now [17:31] hatch: yes. [17:31] bac, hatch I can fix it if I swap test-debug and test-prod in make check [17:32] Makyo: that makes no sense but i'm all for it [17:32] bac, I know :P You're welcome to try and see if that helps? [17:32] gary_poster: ISTM? International Ski Travel Market or Institute of Secretariat Training & Management [17:32] ....lol [17:33] :-P [17:33] it seems to me [17:33] ohhhh [17:33] that makes more sense [17:33] It has to do with one of the tests changing the URL, then the next test requests a relative path [17:33] haha [17:35] right [17:35] let's fix it right and blow up all the tests to their own modules :) /me runs away [17:38] haha [17:38] I'm pretty sure that would dramatically slow down our test suite [17:38] but I agree we need to find some way to do that [17:39] I don't know, the whole mocha "Load all the things, parse for any .only, .skip, etc" step could go away [17:39] more disk IO, but w/SSD might be faster :) [17:39] and much easier to parallel-ize if we wanted/needed to [17:39] well it needs to load the index.html for phantom so it needs to load/parse all of the js each time then [17:40] so we would need an index.html for each test suite [17:40] which is fine [17:40] but I'm pretty sure it would slow it down [17:40] although the parallel story sounds interesting [17:40] prove it ;P [17:41] any slow down that 'may' be caused by parsing and loading new assets for every suite would certainly be offset by running 4+ tests at a time [17:41] well....2 for me ;) [17:41] "chrome, open these 40 tabs now!" *boom* [17:43] thanks bac, LGTM and comments inbound [17:44] lol - you're computer has 40 cores? :D [17:44] LUCKY!!!! [17:51] guihelp: I need one review + QA for https://codereview.appspot.com/20370043 thanks! [17:53] gary_poster: I need to go in 10 minutes, I'll take care of landing the branch (if it looks good to the reviewer) later today or tomorrow. [17:55] ok thank you frankban! looking at branch asap... [17:56] ugh I was hoping I'd never have to look at these endpoint functions again......THANKS gary_poster :/ haha [17:57] hatch, you don't hafta. You can leave those out. :-) [17:58] phew! Ok I'll just console log the data [17:58] thanks gary_poster [17:58] :) [17:59] enjoy your time off frankban [18:00] hatch: thanks have a great we [18:00] I will :) [18:08] jcastro: is there a better way for this guy to do what he wants than what I have said? http://askubuntu.com/questions/369127/juju-wordpress-charm-doesnt-update-from-git/369148#369148 [18:08] <_mup_> Bug #369148: E: linux-image-2.6.27-11-generic: subprocess post-installation script returned error exit status 10

[18:12] hatch, marco knows best, his charm [18:12] but that seems sane to me [18:12] marcoceppi: ^ [18:55] * hatch spazzes out on widget parent/child [18:58] Finally disentangled myself from insurance crap, ugh. Sorry for the wasted time :( [18:59] Makyo: live on the edge....go no insurance [18:59] That's...basically what's happening. [19:00] I....was just kidding [19:00] Three appointments, four prescriptions, about $1000, all claims denied. [19:00] :/ [19:01] All out-of-network visits are 100% subscriber's responsibility, and there are no in-network providers listed, which makes the plan non-ACA-compliant. [19:01] [19:02] move to Canada :D [19:02] Hahah! [19:02] I might, at that! [19:02] sure our income tax sucks, but at least I don't know of anything called in-network-providers wrt health care [19:02] Still not convinced Saskatoon is a real place, though. [19:02] Yeah, we actually live in the ether, only coming to your dimension for sprints [19:03] That was my best guess, yeah. [19:03] that's why it's so important I have a job that allows remote work [19:03] Hahaha [19:03] lol [19:05] jujugui lf a really quick review/qa https://codereview.appspot.com/20410043/ [19:05] hatch: there are nice parts of canada... [19:05] hatch, ont it [19:05] on it [19:05] Makyo: thanks [19:05] er, that was meant for makyo [19:05] makyo: there are nice parts of canada [19:05] bac, Canada sounds nice! Saskatoon just sounds made up :) [19:06] Makyo: i rearranged those make check targets. if it works it'll get submitted. :) [19:06] \o/ [19:06] feels a bit dirty though [19:06] look it's really there! http://goo.gl/maps/kza65 [19:07] oh, heck, it failed even with 'make check' from the command line, not run by lbox! [19:07] re-vert [19:07] Boo [19:07] We need to fix that test :| [19:08] * Makyo peers at map. [19:08] and the largest identifiable part of the city is named after a hockey player [19:08] hatch, code's good, QAing [19:08] bac: lol which part is that? [19:09] gordon howe campground [19:09] oh lol I dont' even know where that is [19:10] learn something new every day [19:12] hatch, any reason why we should't increase max-width of #subapp-browser #bws-sidebar .bws-content .yui3-token .title to 170px in your branch? Seems like it would fit and looks a lot better [19:14] gary_poster: looks a-lot better :) [19:14] I have nothing against it but I'll qa across ff and ie [19:14] thanks hatch. gave LGTM & QAOK with that. [19:15] I can't wait for SASS and source maps [19:15] * gary_poster runs away for halloween and such. back briefly later. bye all! [19:15] ooooooo boy that'll be awesome [19:15] cyaz [19:18] alright I'm going to grab some lunch now [19:19] ping if ya need me [20:10] of course the time I leave lbox running unsupervised it decides to stop working [20:10] haha [20:11] I think we have bonified ci failures in IE [20:11] as per the last email [20:13] gary_poster: not sure if you're around, but if so, do you have a card in mind for me? [20:14] I'm taking the cookie zindex for now [20:20] is there a nob in gui to turn on icon files for non promulgated charms? [20:20] that was fast...fixed [20:20] hazmat: negatory [20:20] it's a charmworld thing [20:20] it tells the gui what to display for it's icon [20:21] bummer [20:21] what's the usecase? [20:21] tbh I'd be ok with that knob :) [20:21] hatch, demoing a set of charms that haven't been promulgated, would be nice to have the icons.. [20:22] or if you just don't care if it's promulgated or not :) [20:22] rick_h_: ^ any input [20:22] he might not be around [20:23] hazmat: file a bug :) [20:23] k, thanks [20:24] jujugui anyone want to do a quick qa on ubuntu for me for this cookie zindex branch? https://codereview.appspot.com/20430043/ [20:30] hazmat: you'd have to have a fork of hte gui that set the model attribute 'shouldShowIcon' to true hard coded? or no, you can't do that any more [20:30] hazmat: so yea, the icons come from charmworld, so that would have to be branched/hacked [20:30] :) [20:31] rick_h_: what about a option in the gui to add something to the end of the data endpoint to include the real icons? [20:31] http://bazaar.launchpad.net/~juju-gui-bot/charmworld/trunk/view/head:/charmworld/views/api/__init__.py#L571 [20:31] doesn't sound difficult from the gui side but maybe charmworld can't handle it? [20:31] hatch: right, we'd talked about adding a url flag to charmworld api calls that you could set via a gui/etc config [20:31] hatch: but just not ever been done [20:32] hatch: you'd have to add a flag to that method ^^ that ignores all the rules and just returns the icon if it exists [20:32] I don't know how the routing works but that sounds simple on the face of it [20:32] of course more work than one would think because of the charm changes :) [20:32] hatch: yea, not saying it's hard at all [20:33] maybe we could have a 'prettify: true' option :D [20:33] hatch: just saying it's not there [20:33] hatch: heh, in the gui charm? [20:33] yeah [20:33] if (pretty === true) { unicorn(); } [20:34] hatch: heh, just a demo mode [20:34] does the funky scaling, turns on always-on icons, etc [20:34] but anyway, that's the low down on that [20:34] coolio [20:34]