/srv/irclogs.ubuntu.com/2015/02/06/#juju-gui.txt

=== kadams54 is now known as kadams54-away
sebas5384hey! o05:33
sebas5384o/05:33
sebas5384somebody know what happened with the service's settings in the juju-gui ?05:33
rick_h_sebas5384: what do you mean?05:41
sebas5384rick_h_: it isn't appearing anymore05:41
rick_h_sebas5384: umm, no but thakns for mentioning it. We'll get folks right on it05:43
rick_h_huwshimi: any time left in the day? ^05:43
sebas5384http://awesomescreenshot.com/0154cnfkad05:43
huwshimirick_h_: Yep, let me take a look05:43
sebas5384thanks rick_h_ !05:44
rick_h_sebas5384: yep, can replicate on demo.jujucharms.com as well05:44
rick_h_sebas5384: ty, will file a bug and get the guys on it right away05:44
rick_h_huwshimi: added a card marked critical with a bug on it. I assigned it to hatch assuming you'll not have time in your day to complete it but plese leave notes/etc either in the bug or the card05:47
huwshimirick_h_: Yep, sure thing. I can see there's an error there so I'll try and track that down and see how I go.05:48
rick_h_huwshimi: review done?05:48
huwshimirick_h_: Sure is :)05:48
rick_h_huwshimi: <305:48
sebas5384rick_h_ and huwshimi thanks for the fast response :)05:51
rick_h_sebas5384: np, not sure how that one slipped by but hopefully it's a small fix do to some new apis we're using and should be fast. We'll try to get a new rev'd charm out by EOD05:52
rick_h_US EOD that is05:52
sebas5384I'm going to demonstrate juju gui at a conference tuesday05:52
rick_h_sebas5384: you'll definitely have it by then05:53
rick_h_sebas5384: fyi, the icon thing is a bug in chrom v40 we're tracking upstream05:53
rick_h_sebas5384: so FF is currently the best demo browser unfortunately05:53
rick_h_sebas5384: until chrome fixes the svg issue and get it into a release05:53
rick_h_https://bugs.launchpad.net/juju-gui/+bug/1415550 for tracking05:55
mupBug #1415550: Charm's SVG icons with wrong size in Chrome <icon> <svg> <juju-gui:Confirmed> <https://launchpad.net/bugs/1415550>05:55
sebas5384yeah definitively rick_h_ i'm going to use it with firefox for now05:56
rick_h_sebas5384: looking at the upstream bug though looks like v41 will include a fix yay!05:57
sebas5384rick_h_: uhuu!05:58
sebas5384:)05:58
huwshimirick_h_: So it looks like hatch landed a branch which fixed this a couple of days ago: https://github.com/juju/juju-gui/commit/9dabbda4597c53545fc9ed912165a583d3fd133705:59
huwshimirick_h_: I've rolled back to the commit before that and confirmed that the bug was there before that commit.06:00
huwshimirick_h_: devel should be fine now.06:00
rick_h_huwshimi: ok, awesome. We should get a release rolling then. I'll make sure to bug hatch about it06:02
rick_h_huwshimi: can you email -peeps with your findings and I'll reply asking for the release today please?06:02
huwshimirick_h_: Sure thing.06:02
rick_h_ty kindly06:03
huwshimirick_h_: Sent!06:09
rick_h_huwshimi: awesome06:09
hatchfrankban: wouldyou be able to point me to the correct place to inspect the path in guiserver? I was thinking the on_message method15:27
frankbanhatch: looking at the code15:27
hatchtornado is very confusing15:28
frankbanI've found it less confusing than other async frameworks15:30
hatchquite possible - it feels to me like it's shoehorning async into a language that isnt15:31
hatch(I wrote an async framework for php back in the day) 15:31
hatch:)15:31
hatchbut php didn't have yeild so I actually had an event loop lol15:31
frankbanhatch: yield is just a way to avoid the callbacks hell15:32
frankbanhatch: so, I'd presume you'll have to extend url routing in server/apps.py15:34
frankbanhatch: right now only /ws$ is routed to the websocket handler15:34
hatchand what's the path supposed to be now?15:35
hatchwon't it still be /ws/uuid/15:35
frankbanhatch: to avoid confusion, I'd make them reflect the juju API server ones. so we need to keep handling /ws$ and add support for /ws/environment/:env_uuid:/api15:36
frankbanhatch: so same as juju but in the ws/ namespace15:37
hatchoh so they actually added the "environment" keyword in there15:37
hatchI thought that was just a placeholder :)15:37
frankbanhatch: heh, no, I think that's the real endpoint path15:37
hatchalright, is there a way I can test the api using curl or the like to see if that is indeed the endpoint?15:38
frankbanhatch: I am pretty sure that's the real endpoint. however, for these kind of tests I usually use ipython + https://pypi.python.org/pypi/websocket-client/ (it's really quick with the latter to establish a websocket connection)15:40
hatchahh cool15:41
hatchok well first is gui release then I'll continue on your reivew comments and this15:41
frankbanhatch: for instance, I use this script sometimes to watch env changes: http://pastebin.ubuntu.com/10093532/15:41
hatchoh yeah...just tiny :P15:41
frankbanheh but easy to change to make other calls15:42
=== kadams54 is now known as kadams54-away
=== kadams54-away is now known as kadams54
hatchmakyo_: didn't you do a guicharm release in 2015?17:09
hatchhttps://jujucharms.com/juju-gui/trusty/17 see the revisions list17:09
hatchor was that updated but never released?17:10
=== kadams54 is now known as kadams54-away
hatchmakyo_: whenever you return I'm wondering if something borked up becuase there is no git tag to match 1.3.017:59
hatchwait there is on trunk...hmm17:59
makyo_https://github.com/juju/juju-gui/releases/tag/1.3.0 ?18:00
hatchmight help if I used the --tags18:00
hatch*facepalm*18:00
makyo_Whew18:01
makyo_Had me worried18:01
hatchso that's one problem down - now the charm release :)18:01
=== makyo_ is now known as Makyo
hatchMakyo: so did you push a new charm release and it maybe not make it onto jujucharms?18:01
hatchI was pretty sure you did one rather recently18:02
hatchin fact http://bazaar.launchpad.net/~juju-gui-charmers/charms/trusty/juju-gui/trunk/files18:02
hatchyou did18:02
MakyoI did, yeah.18:02
Makyohttps://jujucharms.com/trusty/juju-gui in the releases dir is juju-gui-1.3.0.xz18:03
MakyoSo maybe the changelog is borked?18:03
hatchhmm yeah that's odd18:05
hatchok well I'll just go on the assumption that it's the UI that's a little busted because the code all looks good18:06
hatchMakyo: was sphinx supposed to be in the sysdeps?18:34
hatchoh it is....ok then...hmm18:35
hatchcode release done - commits can resume to the gui19:12
hatchstill working on the charm side 19:13
=== kadams54 is now known as kadams54-away
=== kadams54 is now known as kadams54-away
=== kadams54-away is now known as kadams54
=== kadams54 is now known as kadams54-away
=== kadams54-away is now known as kadams54
=== kadams54 is now known as kadams54-away

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!