[00:33] Morning [00:34] morning huwshimi [00:34] good flight? [00:34] rick_h__: Yeah, not bad. Get home ok? [00:35] oh yea [00:35] nice direct single flight :) [00:37] Nice! [00:50] rick_h__: It's nice to return to see that change to connections :) [00:59] huwshimi: :) [00:59] huwshimi: yea, I ping'd tim and said "Hey, I want to give you the FYI that we're going to make some waves about this up through Mark Ramm and up to Mark S if necessary" [00:59] and his reply "already gone" [01:00] hehe === uru_ is now known as urulama [07:44] urulama: mornin' [07:44] rogpeppe1: morning [07:45] urulama: how did your trip back go? [07:45] rogpeppe1: fine, thanks, came back to a rainy and foggy home :D [07:46] urulama: still quite nice up here :-) [07:46] urulama: here's a review to brighten your day :-) https://github.com/juju/charmstore/pull/34 [07:47] rogpeppe1: looking at it [07:47] urulama: ta [07:48] rogpeppe1: so you are not using your swap day? [07:48] urulama: i think i'll take it tomorrow [07:48] urulama: or... perhaps i might take it in two halves - this afternoon and tomorrow morning [07:49] urulama: i did this PR on the train north [07:49] rogpeppe1: great :D [07:50] rogpeppe1: my kids are at home this week (kindergarden closed for their vacations) - so I'll probably work in the morning till 3PM and than continue at 8PM (our time), just to let you know [07:50] urulama: (and i also have nearly done a signficant refactoring of the meta include stuff; i decided that the GetItem stuff was going to be too complex and inefficient. [07:51] urulama: ok, good to know [11:05] morning party people [12:35] morning [12:35] morning jrwren_, good trip? [12:36] very good. [12:36] awesome [12:36] and you? [12:36] tired, great sprint though. [12:36] rick_h__: do you have a minute to chat about bug 1348192 ? [12:36] <_mup_> Bug #1348192: series should not be determined to create machine until a charm is placed upon it [12:36] frankban: sure thing [12:37] frankban: invite me in and I'll join in [12:37] rick_h__: https://plus.google.com/hangouts/_/canonical.com/gui [12:38] hi there jrwren_, welcome back [12:45] thanks urulama. [12:47] jrwren_: so catch up on emails and we'll get a sprint summary email out in a bit [12:47] jrwren_: once you're caught up let's chat and see where to head this week? [12:49] sounds great. [13:03] luca: call today? [13:03] or skipping? [13:03] rick_h__: oh! [13:03] rick_h__: I really wasn’t watching my calendar [13:03] rick_h__: I’ll jump on, sorry [13:14] hi there hatch [13:15] hey urulama hows it going? [13:15] hatch: great, thanks ... had a good way back? [13:15] flight was delayed out of LHR so I had to spend the night in YYZ before getting back home [13:16] at which point I rolled my ankle [13:16] lo [13:16] l [13:16] so no [13:16] haha [13:16] auch [13:16] how was yours? [13:17] nothing as drastic as yours ... taxi, bus, writing some docs at the airport while eating "great" food, landing in a rainy and foggy Slovenia ... and all the English complaining that they should of stayed at home :D [13:18] haha [13:19] hope your ankle is better now, otherwise your 1% of life in your work-life balance could be endangered! [13:20] haha right [13:36] hatch: are you on the search bug this morning? [13:36] rick_h__ yeah I will be [13:36] hatch: ok cool thanks [13:37] rick_h__: it seems the API allows for hulk-smashing two different units in the same top level machine [13:37] frankban: it works? I talked with someone on core about it and everything. I hope I'm not crazy [13:38] I can't seem to be able to make g+ turn my London photo album into a story :( [13:38] hatch: sometimes takes a few days [13:38] oh ok I'll give it a wait [13:46] over 400 emails down to 280 go go go [13:50] rick_h__: ctrl+a, mark as read :P [13:50] :P [13:50] frankban: sorry for the bug mail span [13:50] I fail at tab mgt today [13:51] rick_h__: The summary is all good [13:51] luca ty much [13:51] rick_h__: no probs [14:16] rick_h__: I confirm juju stable allows co-locating two units. Do we still want to enforce one unit per container? [14:16] (tried also on ec2) [14:18] rick_h__: I'd suggest not to. And we can just add those checks: 1) on an existing machine/container, no services with a different series are allowed. 2) on a ghost machine/container, dropped units must share the same series [14:20] juju-gui: are we aware of this bug? 1) go to http://comingsoon.jujucharms.com/ 2) drag/drop the mediawiki/single bundle 3) double click mediawiki to open the inspector 4) start zooming the canvas in/out and notice the mysql service box icon does not scale [14:22] looking [14:23] frankban working fine on OSX Chome 36.0 [14:24] hatch: chrome on linux here [14:25] frankban chrome has a number of svg rendering bugs, on linux even more so, I wouldn't doubt that it happens :) [14:25] hatch: works ok with firefox [14:27] frankban: I've seen problems with the SVG icons in Canary/Mac for awhile now. Like hatch says, seems to be something in Chrome/Canary. [14:27] hatch: I can dupe it with OSX chrome 36 :-/ [14:27] I don't know what, but it isn't fixed by switching Chrome versions or re-installing. [14:27] really? any errors? I really can't repro [14:29] hatch: no errors [14:30] frankban: yea that sounds like a plan I guess. [14:31] frankban: and yea, we'll need to make sure the UX makes it make sense and help guide the user [14:31] frankban: I hit that during my demo to mark and not sure what's up. Haven't investigated farther [14:31] frankban so the label for creating a relation doesn't scale (as intended) but the icons both scale without issue here... [14:32] frankban rick_h__ so we are allowing multiple units to be placed on the 'base container' via the GUI? [14:33] hatch: yea, something must have chnged since the demo and it seems it's possible now [14:33] hatch: so the reason we killed it was the api didn't let us do it, but seems that's changed or I had a dream about it all back in the demo days [14:33] hmm ok then - I am pretty sure there is still a bug somewhere in there because that's how kadams54 got into his broken state [14:33] hatch: oh definitely, this is a bit outside of that stuff [14:34] there's issues around series matching and container types allowed and more [14:34] gotcha ok cool, just trying to stay in the loop :) [14:34] heh, you and me both :) [14:34] haha [14:35] so it took me just over 1.5H to get from the office and through security in LHR and there wasn't a single delay [14:35] yea, I know huw and I planned on 2hrs to get through airport and be there 2hrs early so left at 9am for a 1pm flight [14:36] crazy how long it takes [14:42] hatch: jcsackett did we get anywhere with this 'major/minor uncomitted' stuff? [14:42] I feel like we brought it up but can't find any notes/bugs/updates on that [14:42] rick_h__ yeah the only 'minor' thing has been fixed [14:43] rick_h__: it was referenced, but i don't recall serious discussion about it. [14:43] rick_h__: i *think* hatch saw some stuff about it in one of the mockups. [14:43] it was just setting of a config moved into the other list [14:43] hatch: ok, I see your email to luca, you're set then? We have a bug/card to work on it? [14:43] e.g. I can delete this email :) [14:43] I did what needed to be done already - the new UI has a nicer implementation so we'll probably wait until that's done [14:43] ok cool [14:44] autocompleteDataFormatter is apparently undocumented everywhere wrt the autocomplete widget...heh [14:44] woot [14:44] do you remember where you found it? [14:44] the source probably [14:52] well it doesn't exist [14:52] heh [14:53] jujugui call in 8 [14:53] jujugui kanban please [14:53] hatch: oh, is that something I wrote then? [14:54] haha yup [14:54] bah clearly still jet lagged [15:00] jujugui call now jcsackett [15:02] rick_h__: whoops, coming now. [15:24] jrwren_: after you're done with reading docs, we could have a chat about that tomorrow morning (your time) ... rick_h__ want to join? [15:24] rick_h__, heya [15:24] rick_h__, so any progress on searching? [15:24] I still can't find anything [15:24] jcastro: yes and yes? [15:24] jcastro: what's up? [15:24] urulama: sure, happy to help [15:24] jcastro: hatch is working on the bug for the autocomplete that came out, but not sure what you mean 'I still can't find anything' [15:24] urulama: whenever you wish. I don't need to be done reading, I don't think :) [15:25] rick_h__, do you know if quickstart/deployer will work if bundles are in github? [15:25] REstart for updates, back in a sec [15:25] urulama: created the card: add a command to start the server... [15:25] jcastro: if you give it the raw bundle file (http://.....) it should work [15:25] frankban: thanks [15:25] jcastro: not sure on deployer, I know quickstart will use a url [15:25] at least I believe it will [15:26] frankban: i've moved it at the top, 1st card [15:27] jcastro what are you searching for that you're unable to find? [15:28] https://bugs.launchpad.net/charmworld/+bug/1348290 [15:28] <_mup_> Bug #1348290: charm searching returns odd results at times

[15:28] if you search for "hadoop" you get a bunch of crap [15:28] jcastro: right, that's in progress right now [15:28] yeah I'm working on that one right now [15:28] jcastro: hopefully have a fix later today/tomorrow and we'll do a new release this week [15:28] jcastro, rick_h__, deployer supports a bundle url on the cli and also for internal imports [15:28] hazmat: awesome [15:29] ugh, this sidebar is killing me [15:29] I totally don't understand [15:29] jcastro: ? [15:29] can we get the fullscreen search back? [15:29] rick_h__, I still can't use the gui to search for bundles and charms I want [15:29] jcastro: rgr, went to PM [15:32] jujugui: going afk for 3h ... see you later === urulama is now known as urulama-afk [16:08] rick_h__ it seems that a number of people are using the autocomplete results instead of the real search results, maybe we should drop the autocomplete and put the autocomplete results in the sidebar...to avoid confusion [16:12] hatch: I think we can make the autocomplete better. I mean we're hearing a lot from a set of people using autocomplete because it comes up first. [16:12] I'm not sure the best answer is 'remove the feature' vs improving it [16:13] right - but it's super confusing, even with it working properly you could get three mysql's (for example) all with icons all called 'mysql' [16:13] hatch: otp but let's chat and see what we can do to make it better. [16:13] yeah sure [16:13] I'm just thinking ahead [16:13] ping whenever [16:14] hatch: nothing says we can't not do that. show one mysql named service and move up other options [16:14] maybe the issue is that we have never defined what AC was supposed to do, atm it's just a poor version of the real search results [16:14] rgr [17:02] * rick_h__ goes to get lunch [17:59] * rick_h__ is back and such [18:24] holy smokes qa'ing this bug is painful [18:24] I mean...it's AMAZING [18:27] ;) === urulama-afk is now known as urulama [18:32] painfully amazing or amazingly painful or both? [18:37] definitely amazingly painful [18:37] ;) [18:37] bbiab lunching [18:50] hatch: let me know when you're back [19:00] rick_h__ backattack [19:01] hatch: sorry, next call here. Will try to find a chance [19:01] haha np [19:31] jcsackett are you around? [19:47] hatch: what's up. [19:47] jcsackett didn't you land a branch which fixed the /inspector/serviceName bug which caused it to lock up the charmbrowser? [19:48] I know there was some issues but I thought we just ignored the /inspector urls for now [19:48] hatch: i dealt with a couple dispatch issues around inspector, but i don't remember anything about locking up the charmbrowser. [19:48] when I access a /inspector url the charmbrowser is blank and the charm details breakout is shown and blank [19:48] on a real env of course [19:50] this isn't ringing any bells? [19:51] hatch: I've got 10min [19:52] hatch: in the standup url [19:52] ok joining [19:52] it's taking a while to connect [19:53] hatch: no bells have rung, but if it's only in a real env my guess is the retry loop is blowing its stack. [20:00] jcsackett right - but there was some reason why you didn't fix it before [20:00] I remember you started to fix it then quit for some reason [20:00] but I wasn't sure where you quit at [20:01] because right now if you get handed a /inspector url the GUI is useless because it errors out [20:12] jujugui good night all [20:16] jcsackett well I wasn't sure if this existed or not so I created another bug report https://bugs.launchpad.net/juju-gui/+bug/1349565 [20:16] <_mup_> Bug #1349565: visiting a /inspector url in a real env causes the charmbrowser to be blank [20:51] jujugui: can someone take a look at https://github.com/juju/juju-gui/pull/459 [20:51] i guess two someone's per standup. [20:51] Looking… [21:06] jcsackett: comments added; let me know if I'm talking crazy talk. [21:10] jujugui when addressing review comments wrt this new 2-review policy we should not rebase until the final merge in so that we can keep the comment history alive [21:10] hatch: +1 [21:11] hatch: +1, though I thought that's what we were already doing? [21:11] kadams54: i think hatch means don't rebase between the two reviewers. [21:11] yeah that [21:11] jcsackett: Oh yeah… I tend not to rebase until the very last step before shipit. [21:12] It scales nicely with N+1 reviewers ;-) [21:12] kadams54: me too. [21:13] Definitely good to discuss though… that's one of the hard things as a newbie is the things that the team decided on in retros that's not documented anywhere. [21:15] we should probably update the hacking docs with this stuff. [21:16] * jcsackett calls "not it" [21:25] jcsackett: hatch friday card? [21:26] * rick_h__ runs away for to get the boy until later tonight [21:26] have a good day all [21:35] kadams54: you were not talking nonsense, for the most part--i'm going to look into the not using the DOM suggestion. see my response about the boolean parameter for finding things. [22:31] kadams54: pushed the change. [22:44] jcsackett: cool, looking... [23:08] Morning