[00:18] hatch: or maybe not, damn I'm beat. Going to call it good for now and catch back up in the morning [00:29] rick_h_: i've been mojo-ing using ec2 as the host (not target) and i'm getting the same juju-deploy failures i see everywhere [00:29] i forgot to capture the output, so i'm running again to get it for analysis tomorrow === kadams54 is now known as kadams54-away === kadams54-away is now known as kadams54 [01:33] bac: rgr ok thanks for the heads up === kadams54 is now known as kadams54-away [03:41] #jujuskunkworks [03:41] grh [03:41] fat-fingered the /join somehow === kadams54 is now known as kadams54-away [14:39] hatch: How're things looking with the services view? [14:40] hey aren't you supposed to be gone? :) [14:40] hatch: Yeah, we have parent-teachers conferences today and I'm waiting for the next to begin. [14:41] ahhh - yeah it's going [14:41] just working on moving of that mvvisibility method [14:41] Up to the DB? [14:43] down to the DB [14:43] :P [14:45] hatch: http://paste.ubuntu.com/8761147/ - is this expected behavior? take a look at the postgres defaults.. this is an export from jujucharms.com [14:45] it predisposed them with "nulls" and .nan [14:47] loooking [14:47] lazyPower: no, I thought this was fixed. This is latest revision of GUI? [14:47] no that's definitly not correct [14:47] rick_h_: its whatever is deployed to jujucharms.com [14:47] :/ [14:47] lazyPower: oh well then heh [14:47] i did the drag/drop thing, relate, and export for a user in #juju [14:48] lazyPower: drag/drop your bundle or one in the gui? [14:48] e.g. 'drag/drop thing' ? [14:48] oh you grabbed the services [14:48] not a bundle, ok [14:49] rick_h_: no i built it on jujucharms.com and exported it [14:49] sorry, i should have been more specific in how i phrased that [14:49] lazyPower: anyway, no. I thought we didn't include the config items that weren't the default and looks like we're failing to diff null vs default [14:49] hatch: ^ make sense? [14:51] yeah this is all broken - can someone try comingsoon to see if it exhibits the same issue? [14:52] ok finished what I was doing [14:52] now I can test it [14:53] hatch: yea, does it on comingsoon as well [14:53] glad we caught it internally [14:54] rick_h_: want a bug for follow up on this? [14:54] lazyPower: got a card on the board for it and we'll look into it thanks [14:55] rick_h_: ack. Thanks for taking a look. [14:55] jujugui call in 6 kanban please [14:55] rick_h_: bug I found while testing that https://bugs.launchpad.net/juju-gui/+bug/1388119 [14:55] Bug #1388119: Relation selector shows ghost model id instead of display name [14:56] ugh [14:56] and yes I was able to reproduce that issue [14:57] there are some postgres config values which are "null" there might be an inforrect conditional in there [14:57] incorrect even [15:24] lazyPower: I'm starting on that bug fix now - was there a bug created on LP? [15:24] hatch: negative [15:24] ok np [15:38] lazyPower: rick_h_ I have found the issue with the export - it's actually on deploying the charm just fyi [15:38] * hatch stepping out to grab a coffee [16:04] ugh haloween is the worst [16:04] glad noone on my block has decorations for it this year [17:04] bahaha https://twitter.com/rob_pike/status/525533749677932545 [17:15] ahh today is just one of THOSE days [17:19] rick_h_: lazyPower so postgres actually defines fields as null [17:20] aws_access_key_id: [17:20] type: string [17:20] default: null [17:20] description: EXPERIMENTAL. Amazon AWS access key id. [17:20] which acts all sorts of stupid in js [17:20] hatch: which charm rev is that? [17:20] is this valid juju? [17:20] umm [17:20] http://bazaar.launchpad.net/~charmers/charms/precise/postgresql/trunk/view/head:/config.yaml [17:20] it should be either default:, default: "", or default: "null" depending on how its being checked in the code [17:21] ok so should this be supported in the GUI or should proof have failed? [17:22] dealing with types across 3 languages is....intersting :) [17:22] 4 languages I guess [17:22] YAML > Python (proof) > Go > JS [17:22] lol [17:24] rick_h_: lazyPower I'm going to bench this one and move onto the other critical bug until we have an answer on where the problem actually si [17:25] hatch: good question [17:25] default: null is valid [17:25] it just depends on how the charm is readng the vaules [17:25] *values [17:26] ok so does null === "null" ? [17:26] or is null truley the null value [17:26] if it is "null" then it's working properly [17:27] well - at least properly enough :) [17:27] it will evaluate to string [17:28] ok so it's current functionality is acceptable then? [17:29] ahh there is an issue then - because we should show 'null' in the pre-deployment inspector [17:30] lazyPower: so `null` should be treated as "null" ? because in JS/JSON null is special [17:30] hatch: field type is string [17:30] thats why ti would be "null" [17:30] can it be anything else? [17:30] if you want the null character, you omit a value [17:30] thats the *only* way to get a null character to return from a yaml field of type string. [17:30] ahh [17:31] ok there is a bug here then - but aiui it shouldn't break [17:31] your bundle export should still be valid [17:31] no? [17:32] hatch: i suppose, i'm going to flying jumpkick a bug against postgres for follow up [17:32] lol [17:32] ok I'll bench this one then for now - put it in tracking until we can confirm [17:34] sounds good hatch - if its not a bug in gui sorry for the run around. [17:35] lazyPower: well no there is a bug - we don't show the null value in the pre-deployment inspector - and then we send "null" strings thinking that it had been changed [17:35] BUT [17:35] ah combination of wonky behavior [17:35] yeah [17:35] so yes the export is 'broken' but I don't think it's critical :) [17:36] I'd like to get a little more input on the real desired outcome before implementing a fix [17:41] lazyPower: rick_h_ I updated the card in tracking on the kanban so we can discuss [17:42] Makyo: didn't you have some drupal sites? http://www.eweek.com/security/drupal-users-had-seven-hours-to-patch-or-be-hacked.html [17:43] hatch, wordpress. Both are awful, each in its own way. [17:43] oh haha ok :) [17:43] I can honestly say I've never worked with either [17:43] well I think I deployed a wordpress site once to play with it [17:45] hatch: ok, thanks for that [17:45] hatch: are you going to look at the relation card then? [17:45] already started, yup [17:45] going to grab lunch soon though [17:47] rofl nodejs was forked [17:47] could there be any more drama in the js world [18:41] Oh man, another spaceflight accident :/ http://www.breakingnews.com/topic/virgin-galactic-spaceship-incident-oct-31-2014/ [18:45] :/ [18:59] :( === alexpilotti_ is now known as alexpilotti [19:21] so there is a competition for Ubuntu scopes...hmmmm === kadams54 is now known as kadams54-away [21:18] nice, bug fixed, time to write tests [21:20] ahh all the old tests still pass - that means alls good right? ;) [21:25] :P [21:25] well I suppose it would if this was ever tested :/ [21:45] writing these tests are going to take a while [21:46] there are 0 tests for the addambiguousrelation method [21:46] fertile ground :) [21:46] lol yeah this code has only been in here for....2 years? [21:46] lol [21:50] we got a roomba yesterday [21:50] it's cleaning algorithm seams very inefficient [21:50] heh [21:51] yea, everyone thinks that the first time they see one in action [21:51] but after it's done the carpet is remarkably clean [21:51] it has also never gone back to its base station [21:51] lol [21:52] I don't know if it learns or not [21:52] I don't think so because it runs into the same thing multiple times [21:52] no, I don't thihnk it does. It can't tell that you're running it in the same spot vs a new room/etc [21:53] there is a sticker on it which says that they encourage people to reprogram it :) [21:53] maybe I can develop a smarter algo :P [21:53] what do you mean? I we have the dock in one spot and it just goes all over the house [21:53] so it should be able to build a map [21:57] * rick_h_ goes to get ready for trick-or-treating [22:01] enjoy === lazyPower is now known as lazyWeekend === kadams54 is now known as kadams54-away === kadams54-away is now known as kadams54