=== urulama___ is now known as urulama [14:09] kadams54: hey get my email this morning? [14:09] Yeah, looking things over right now [14:09] great thanks [14:10] the fix ended up being rather extensive lol [14:11] kadams54: I'm going to take this branch and work on the services side now [14:11] OK [14:25] hatch: when you have a few minutes, I'd like to talk over the changes. [14:31] yeah I'm good [14:31] standup? [14:32] ^ kadams54 [14:32] yup [14:50] jujugui call in 10 please kanban [14:50] jujugui hi mom [15:37] hatch: all changes are now pushed to the PR === alexpilotti_ is now known as alexpilotti [15:37] Will get the tests working and then (hopefully) land this puppy. [15:38] thanks [15:58] kadams54: did you want to rebase your branch to get it down to less commits? [15:59] That'll be the last step before landing. [15:59] Once I get the tests passing. [15:59] ok I'll make sure that my stuff can be added on after [15:59] just lemme know if you make any code changes :) [16:08] hatch: just changes to test code [16:09] Which should be all fixed now. *fingers crossed* [16:12] haha nice [16:12] kadams54: without any changes the 'highlight' and 'fade' work for the source service but any related stuff doesn't work [16:13] darn heh [16:13] was hoping for better results by default [16:13] I'm actually surprised anything works, so it's glass half full for me. [16:16] Just pushed test fixes, so hopefully it'll have a green light in about 20 minutes. [16:18] great [16:18] kadams54: there is also a css bug in the tooltip when deploying a bundle, it's shifted to the left [16:19] this likely happened when I did the header changes to make the header tooltip notificatiosn line up [16:19] maybe that'll be a simple fix you can do as a driveby with the other css stuff [16:23] hatch: Yeah, I can take a look as I'm updating the icon. [16:23] guihelp: Need another QA on https://github.com/juju/juju-gui/pull/631 [16:24] thanks [16:24] hatch: you did the first pass can you make sure the issues are all better? [16:24] yeah I already have locally and we did a shared qa [16:24] but we both have been in here a while :) [16:25] This is complex enough that I want a fresh set of eyes on it. [16:25] kadams54: hatch rgr [16:26] Makyo: you were a test away from your branch up for review, can you qa while you get reviews? [16:26] rick_h_, yep [16:26] ty [16:27] fabrice: rather than the cookie one can you look at the clicking on deploy card next? [16:28] fabrice: updated the board to reflect the priority/new card there if that's ok? [16:31] kadams54: I created the cards for your next tasks [16:31] Thanks [16:31] I THINK this is everythinig [16:31] (famous last words) [16:31] lol [16:43] luca__: any updates on the highlight icon SVG -> PNG? [16:43] kadams54: not yet, sorry. Give me a sec. [16:44] luca__: YO [16:44] I have a problem with your service spec [16:44] :( [16:44] wait no I don't [16:44] ignore me [16:44] I just figured it out [16:44] too many names for things lol [16:45] you keep doin good things :) [16:48] luca__: kadams54 is asking about the highlight icon? ^ [16:48] otp [17:14] :/ [17:15] having multiple hidden/shown in the canvas is seriously messing things up :) [17:21] kadams54: got a few minutes to be a sounding board? [17:24] hatch: yeah [17:24] ok joining standup [17:28] Makyo: you able to hop into standup? [17:29] rick_h_: maybe you will know, able to hop into standup? [17:30] nm Makyo joined [18:05] Makyo: were you able to qa 631? [18:09] hatch, no, not yet, I'm still struggling through mine. [18:14] ahh ok np [18:41] jujugui: need reviews and QA on https://github.com/juju/juju-gui/pull/633 [18:41] on it [18:41] hatch, rick_h_, Makyo: that's the toggling for highlight buttons, so that only one can be active at a time. [18:58] kadams54: review and qa done - just need one more for both brancheds [19:13] jujugui - I know everyone is busy but could we get another QA on 631 and a review on 633? [19:14] On it [19:14] I need out of test hell. [19:14] haha u rock [19:19] hatch, kadams54 just to confirm, 631 doesn't affect service view? [19:19] orrect it's very broken [19:19] :) [19:19] Assuming that's fixed later? [19:19] yup in my current local branch [19:20] just need to get 633 and 631 landed so I can finalize my branches code === kadams54_ is now known as kadams54 [19:41] kadams54: so looks like we can ship 631 now [19:41] Yay [19:41] odly enough I no longer get user icons in github [19:41] ahh had to clear the cache [19:42] Shipped [19:43] awesome [19:45] Gotta switch locations, be back shortly. [19:56] jujugui one more review required - pretty small, qa already done https://github.com/juju/juju-gui/pull/633 [19:58] hatch, kadams54 +1 [19:58] awesommmmmme [19:58] thanks for clearing these Makyo [19:59] can you +1 on gh? [19:59] Woot woot. [19:59] awesome so.... [20:01] kadams54: so I think the only thing else for this (besides my cards) is the tooltip thing I mentioned [20:02] Yeah, looking into that. [20:36] why does quickstart try to bootstrap a running environment? [20:36] hmm I've never seen that [20:37] but it has been a while [20:38] looks like it always trys to bootstrap which is inane [20:39] hmm odd maybe I've just never noticed :) [20:39] * hazmat files a bug https://bugs.launchpad.net/juju-quickstart/+bug/1387867 [20:40] Bug #1387867: always tries to bootstrap .. check the api endpoint instead [20:42] thanks [20:49] kadams54: I found another bug [20:49] Gah [20:49] maybe you can work on this one [20:50] K [20:50] well, can you try and reproduce it first heh [20:50] it is... [20:50] load up the gui so it lands on the service view and charmbrowser on initial load [20:50] deploy [20:50] the mongodb bungle [20:50] click to highlight mongosvr [20:50] configsvr* [20:50] switch to mb [20:50] mv* [20:50] no machines are hidden [20:51] you will need to add this code to models.js:1424 [20:51] if (!machine.units) { [20:51] // If the machine view hasn't been opened yet then there won't be a [20:51] // units collection on the machines. [20:51] return; [20:51] } [20:51] K [20:51] to avoid it from falling over [20:54] I'll make a card and get on it. [20:55] thanks [20:55] I really hope this isn't yet another oddity [21:03] Makyo: hey - we don't have a 'hide relation line' method? [21:03] we have show and fade but no hide? [21:04] I can just copy the other methods but I thought we had something to do this already [21:05] is there a config option yet to display icons for non official charms? [21:05] hatch, no, nothing in place [21:07] allllllllrighty [21:19] Makyo: just to confirm - in a relation topology d.source is the 'near' and d.target is the 'far' ? There is a bug where it's showing relation lines where the 'far' is hidden so I'm just wondering if this terminology is consistant [21:19] it's around line 770 in topology/relation.js [21:19] it appears to be so [21:20] It shooould be? [21:20] heh ok I'll asume so as well :) [21:23] d3 is pretty awesome even if it's api is horrible ;) [21:33] ahhhhh just rebased in develop and it broke it! [21:33] blargggg [21:35] kadams54: so the branch which did the toggle thing actually has these events propogate everywhere [21:35] we need to use method calls instead of events [21:35] Why? [21:37] kadams54: actually I was able to fix it reordering the events [21:37] it was highlighting then unhighlighting [21:37] hah [21:37] by switching them around it works [21:37] I can make it less spammy [21:37] Great [21:38] phew [21:44] Makyo: FYI target is not always 'far' [21:44] :'( [21:45] :/ [21:55] sweet I got all the functionalities down [21:55] kadams54: any luck tracking down that bug? [21:56] Not yet. My work time's been sporadic with dinner time, but the family is off to dance class now so I now have some quality bug squashing time. [21:56] ohh right it'a past your EOD [21:56] well you could just leave it for me [21:57] Now that I have this functionality done I need to write tests [21:58] rick_h_: are you around tonight? === kadams54_ is now known as kadams54 [22:00] hatch: It's no problem. [22:02] jujugui if anyone wants to do some preemptive qa work on this added services stuff it would be appreciated https://github.com/hatched/juju-gui/tree/canvas-fade [22:03] On it [22:03] Makyo: thanks - the only currently known bug is the one I mentioned to kadams54 at 14:50 in the scrollback [22:04] and now to write tests....OOOOBOY!!! [22:05] Makyo: also I seem to have re-introduced the error when trying to hide/fade ghost services :/ [22:05] will look at that now [22:10] ok that's being done as a follow-up [22:10] moving the card back [22:16] hatch: OK, I think I figured out what the bug is. [22:16] ok shoot [22:16] It's actually the line you had me add at 1424 in models.js :-) [22:16] if (!machine.units) [22:16] machine.units isn't set in the DB [22:16] We set it by making a call to db.units as part of render() [22:18] sooooooooo howdowefixit? [22:18] _updateMachineWithUnitData, machine-view-panel.js, line 1409 [22:19] I think setMVVisibility needs to be moved out to the DB or a util method [22:19] so on render it should populate that unit data with the visibility statuses [22:19] It will need to make a similar query to db.units first and then do the unit check [22:19] well setmv is in the db [22:19] It's in MachineList [22:19] I think? [22:20] Yeah. [22:20] oh yeah [22:20] but that's ok [22:20] if we have the db we can access the machone list [22:20] but you're suggesting updating the machines with the unit data before we open the mv? [22:20] No [22:21] I'm saying setMVVisibility just needs to do this: [22:21] var units = db.units.filterByMachine(machine.id, true); [22:21] if (!units) { [22:21] ... [22:21] Instead of just: [22:21] if (!machine.units) { [22:22] That make sense? [22:22] nope [22:22] but I can try it [22:22] lol [22:22] sec [22:23] We can't count on the internal list (machine.units) being initialized when setMVVisibility is called [22:23] We need to go back to the DB (db.units.filterByMachine()) instead [22:23] oh wait I can't, now I see why you want it on the db [22:23] heh [22:23] ohhh [22:24] so you're saying we need to update the 'real' units [22:24] Yes [22:24] Those "real" units will get populated into machine.units at a later point [22:25] so does it make sense to do that OR do we run the update on mv load? [22:25] I think your idea is better [22:25] Also: lines 1427-1432 in models.js is some crazy shit [22:25] yeah don't go there [22:25] nested some() calls [22:25] it's dangerous [22:25] therebedragons [22:26] # [22:26] #therebedragons [22:26] truth [22:26] Can you throw a comment on there? [22:26] no I don't remember what it does [22:26] lol [22:26] lol!! [22:26] jk yeah I can [22:27] kk, I gotta run again, but will check in again in about an hour [22:27] thanks! [22:27] cya === kadams54 is now known as kadams54-away === Guest89108 is now known as bodie_ [23:48] hatch: around what's up?