2014/11/17/#juju-gui.txt

=== 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
=== rogpeppe3 is now known as rogpeppe
[13:34] <rick_h_> morning
[14:26] <rick_h_> kadams54: morning, did a QA on your branch and linked some screenshots
[14:26] <kadams54> k
[14:26] <rick_h_> kadams54: be cool if you and hatch could look into it please
[14:29] <kadams54> rick_h_: my understanding is that highlighting a service that's not related to anything should hide everything else.
[14:29] <rick_h_> kadams54: ok, it seemed different
[14:29] <rick_h_> kadams54: so I thought it was odd
[14:30] <rick_h_> kadams54: if that's expected then qa ok
[14:30] <kadams54> Well, QA is not OK :-\
[14:30] <hatch> rick_h_: where there any errors in the console?
[14:30] <kadams54> The behavior in the other screenshots is not good
[14:30] <kadams54> Highlighting should hide everything not directly related
[14:30] <kadams54> Including services that aren't related at all.
[14:31] <kadams54> So I'm investigating that.
[14:31] <kadams54> Getting a real AWS env setup
[14:31] <hatch> kadams54: does lxc not work for you? It's usually way faster than aws :)
[14:32] <kadams54> hatch: I work in OS X, so I'd have to fire up a VM and then run LXC in the VM.
[14:33] <hatch> ohhh 
[14:33] <kadams54> I've tried before but stuff like networking setup gets tricky
[14:34] <hatch> I found that running the parallels vm with ubuntu is considerably less buggy than using vagrant 
[14:34] <kadams54> vagrant or virtualbox?
[14:34] <hatch> the vagrant bugs are very odd too....like phantomjs crashing with no error
[14:35] <hatch> I didn'tlike virtual box, parallels does a much better job 
[14:36] <kadams54> vagrant supports multiple providers, seems like you ought to be able to run it over parallels
[14:37] <hatch> well if you have parallels there isn't much point in running vagrant heh, just run the whole os
[14:37] <kadams54> https://github.com/Parallels/vagrant-parallels
[14:37] <hatch> (although parallels does support vagrant images)
[14:38] <kadams54> I'd much rather use vagrant's stripped down OS and easy provision/reprovision.
[14:38] <hatch> yeah I tried that but I couldn't resolve the bugs
[14:39] <kadams54> rick_h_, hatch: there was a bug fix that had been merged to develop in a separate PR. I've rebased that into this branch.
[14:40] <hatch> cool
[14:49] <hatch> kadams54: I've added a few review comments
[14:51] <rick_h_> kadams54: ok
=== tvansteenburgh1 is now known as tvansteenburgh
[14:57] <kadams54> rick_h_: I think the bug fix I merged in addressed the problem you were seeing
[14:57] <kadams54> https://ec2-54-148-167-86.us-west-2.compute.amazonaws.com/
[15:04] <rick_h_> kadams54: ooh ok cool
[15:05] <hatch> kadams54: so you don't get phantom crashing with no error when running in vagrant?
[15:05] <kadams54> yes, I do
[15:05] <hatch> oh ok good - glad it's not just me heh
[15:05] <hatch> well not good
[15:05] <hatch> but
[15:05] <hatch> you know
[15:05] <hatch> )
[15:05] <hatch> :)
[15:05] <kadams54> I mostly test in browser and use lots of .onlys
[15:06] <hatch> heh painful
[15:06] <hatch> on that front
[15:06] <hatch> jasmine 2.1 will have fit and fdescribe 
[15:06] <hatch> which is like only
[15:20] <hatch> rick_h_: kadams54 I filed the agent_state_data bug https://bugs.launchpad.net/juju-core/+bug/1393452
[15:20] <mup> Bug #1393452: agent_state_data isn't populated when service units fail on start <juju-core:New> <https://launchpad.net/bugs/1393452>
[15:21] <kadams54> rick_h_: Looks like I just need your OK on QA to land 658.
[15:21] <rick_h_> kadams54: ok, go ahead. If the bug fixed it then I'll trust that.
[15:22] <rick_h_> kadams54: let me know when that lands and I'll start the release process
[15:23] <kadams54> Will do
[15:23] <kadams54> hatch: thanks for getting that bug files
[15:23] <kadams54> *filed
[15:27] <hatch> it'll probably get marked lower than low since I think we added it in the first place lol
[15:28] <hatch> but at least there is a record somewhere :D
[15:28] <rick_h_> lol
[15:29] <hatch> kadams54: this but in your branch was the second one that had been fixed in develop but not in the branch - maybe we should make a 'rule' to rebase develop in before pushing?
[15:29] <hatch> I never do heh, but it might be a good idea going forward :)
[15:29] <rick_h_> hatch: well the merge thing does that
[15:29] <rick_h_> hatch: it's kind of rare that things are so tied on each other like that
[15:29] <rick_h_> hatch: I think it's just because we're at the release edge
[15:29] <hatch> right - but your qa wouldn't have failed if that had been done
[15:29] <hatch> yeah that's true
[15:29] <rick_h_> hatch: right, but it's ok. The work on that vs the work of always rebase :/
[15:29] <kadams54> Yeah, in this case there was no excuse.
[15:30] <rick_h_> I think it's ok as is, but something I should have kept in mind during QA
[15:30] <kadams54> I did the bug fix branch
[15:30] <kadams54> I did both branches.
[15:30] <kadams54> I should have realized that I needed to rebase one into the other before sending the second off to QA
[15:30] <kadams54> I blame it on being excited and relieved at finally being done :-)
[15:30] <hatch> haha 
[15:30] <hatch> we hope ;)
[15:31] <kadams54> Generally I try to sync my develop with remote and rebase onto latest before sending off to QA.
[15:31] <hatch> it is pretty awesome when you hve a large environent
[15:31] <hatch> highlight, fade, highlight, fade
[15:32] <rick_h_> kadams54: :) go enjoy some nice maint cards, no pressure, just little things without big picture pains
[15:32] <kadams54> lol
[15:32] <kadams54> Can't relax yet. The merge build is still in progress.
[15:33]  * rick_h_ is optomistic today
[15:33] <hatch> haha
[15:33] <hatch> it's snowing fluffy snow
[15:34] <hatch> that's a good sign
[15:34] <perrito666> rick_h_: you feel like things can go in eight different ways?
[15:34]  * hatch can read snowflakes 
[15:34] <perrito666> rick_h_: ah, no that would be octomistic :p
[15:34] <hatch> lol
[15:34] <rick_h_> hah
[15:34] <perrito666> yours is more like "things can go in a very optical way"
[15:35] <hatch> wouldn't that be opticalistic? 
[15:38] <hatch> kadams54: yay tests passed
[15:39] <rick_h_> lol
[15:39] <hatch> kadams54: one good thing came out of your branch - I verified that we can use Array.isArray() in all our supported browsers
[15:39] <hatch> one less YUI method :)
[15:39] <hatch> IE<9 pssshhhhttt
[15:40] <rick_h_> lol
[15:44] <hatch> this weekend I spent some time researching js module systems and loaders.....
[15:44] <hatch> seriously I have no idea what people have been doing since 09
[15:44] <hatch> nothing even comes close to the YUI loader
[15:45] <rick_h_> hatch: :( stop, I'm still trying to get past everyone saying how we don't need YUI any more
[15:45] <hatch> well there are few things in it we 'need' but I honestly can't find anything to replace the loader with
[15:46] <frankban> hatch: never looked at it, but what about require.js?
[15:47] <hatch> frankban: the problem with js module loaders is that they only support one type of module
[15:47] <hatch> AMD, CommonJS etc
[15:47] <hatch> the yui ones we can simply wrap the file in a YUI.add() and be done
[15:48] <hatch> not so with these others
[15:48] <hatch> and ES6 modules don't make it any better
=== kadams54 is now known as kadams54-away
[15:48] <hatch> luca__: hey
[15:49] <luca__> hey hatch 
[15:51] <hatch> frankban: although AMD Is the lesser of the 3 evils :)
[15:54] <rick_h_> uiteam call in 7 kanban please
=== kadams54-away is now known as kadams54
[15:59] <kadams54> rick_h_: 658 just landed so we can proceed with a juju-gui release
[16:00] <rick_h_> kadams54: on it
[16:39] <hatch> kadams54: the card that you're working on right now - we used to set attributes using d3 on the svg and then use the d3 js to alter those attributes to fade in/out
[16:39] <hatch> we should be able to add/remove a class to have it use css animations and be a lot easier
[16:39] <kadams54> Yeah, and you want to switch it to altering classes
[16:40] <kadams54> I remember :-)
[16:40] <hatch> css overrides the attribute styles which were causing issues
[16:40] <hatch> oh ok cool
[16:40] <hatch> I wasn't sure how much I had spewed out before :P
[16:41] <hatch> rick_h_: man I want your job https://twitter.com/iamdevloper/status/534385415860805633 :P
[16:41] <hatch> lol
[16:42] <hatch> Makyo: so jealous of your yard :)
[16:44] <Makyo> Just a bit of snow :)
[16:44] <Makyo> https://www.dropbox.com/s/tbtjgxek13fdxqs/IMG_20141117_093828.jpg?dl=0
[16:45] <hatch> yeah saw that on the twitters
[16:45] <hatch> james doesn't mind the drive in?
[16:46] <Makyo> Not too much. He's got a nicer car than I do :)
[16:48] <hatch> haha 
[16:49] <hatch> I'd love to move out of town but sometimes the highways are a skating rink
[16:50] <Makyo> Yeah, we ran into some of that last week, but thankfully they were on top of getting them plowed.
[16:58] <rick_h_> hatch: see you in standup in 2?
[16:58] <hatch> sounds good
[17:11] <rick_h_> hatch: hah! I want that job :P
[17:11] <hatch> haha
[17:13] <rick_h_> hatch: http://www.kevindangoor.com/2014/11/simplifying-javascript-projects-with-react/ for the 'lunch reading' folder
[17:14] <hatch> cool I'll check it out
[17:14] <hatch> this industry is love hate, love that it's always changing, hate that it's always changing :D
[17:27] <kadams54> Soooooo
[17:27] <kadams54> I just spent my lunch hour on the phone with DISH.
[17:28] <kadams54> Because they signed up a Kathy Adams awhile back and messed up her e-mail address.
[17:28] <kadams54> I ended up getting all her e-mails, including signed copies of her service agreement and other paperwork.
[17:28] <kadams54> Which has home address and phone number on it.
[17:28] <hatch> what's DISH?
[17:28] <rick_h_> hatch: sat. tv
[17:28] <kadams54> Satellite TV
[17:28] <hatch> ohh
[17:28] <hatch> :)
[17:29] <hatch> Just because we don't have enough compile to js languages http://spiderlang.org/
[17:29] <kadams54> DISH's first attempt at a fix was to change e-mail prefs so she (I) didn't receive any more marketing e-mails.
[17:29] <hatch> lol
[17:29] <kadams54> DISH's second attempt was to file a report on my (non-existent) account.
[17:29] <hatch> haha why didn't they just call her?
[17:30] <kadams54> Third attempt I got someone who removed my e-mail from her account, but because they can't make outgoing phone calls, they couldn't get the right address.
[17:30] <kadams54> So I just called her and left her a very odd voicemail.
[17:30] <hatch> rofl
[17:30] <kadams54> Way to go DISH.
[17:30]  * hatch hopes she calls you back and you become best buds
[17:30] <kadams54> She might be related.
[17:31] <kadams54> I grew up in Indiana and have family scattered throughout small towns and farm country
[17:31] <kadams54> She happens to live in a small southern IN town that's close to where one of the family branches settled.
[17:31] <hatch> ohh look at that
[17:32] <kadams54> It's a common enough last name that it's not likely, but it's still kinda funny that she's from Indiana, since DISH is nation-wide in the States.
[17:32] <kadams54> I was actually kinda hoping to talk with her.
[17:33] <kadams54> But her VM was one of the generic "You've reached this phone number" messages that doesn't give away any personal information.
[17:33] <kadams54> Which is kinda funny considering.
[17:34] <hatch> haha
[17:34] <hatch> "we take your privacy very serious...unless we enter in the wrong email address, at which case...well you know"
[17:47] <kadams54> Note to self: make sure e-mail verification is a part of new user/customer sign up in all future endeavors. Gracefully handle failures.
[17:53] <hatch> haha
[18:31] <kadams54> Changing locations, back in 10 or so.
=== kadams54 is now known as kadams54-away
[19:22] <hatch> well there is no question react is scalable now: https://twitter.com/remibecheras/status/534391509190459392 cc rick_h_
[19:46] <hatch> rick_h_: Makyo what's happening with local charm icons with the new api?
[19:47] <hatch> re we still returning the default icon?
[19:47] <rick_h_> hatch: so the deal is we'll use the new api
[19:47] <rick_h_> hatch: and we'll need to filter non-promulgated client side
[19:47] <rick_h_> hatch: but we want a flag that you can turn off in the gui config ui
[19:49] <hatch> right but the charm path is local
[19:49] <hatch> not in the store
[19:50] <rick_h_> hatch: where are we looking?
[19:50] <rick_h_> oh local charm icons
[19:50] <rick_h_> hatch: yea so for now it's still got to be the local one
[19:50] <rick_h_> hatch: we can look to update that once we get things update down the road
[19:52] <hatch> ok sounds good
[19:54] <hatch> rick_h_: this is the bug that is referenced in the code around the local charms https://bugs.launchpad.net/juju-gui/+bug/1202703
[19:54] <mup> Bug #1202703: local charm icons are broken in GUI <charmworld:Triaged> <juju-gui:Fix Released by rharding> <https://launchpad.net/bugs/1202703>
[19:55] <hatch> maybe you want to close the charmworld one
[19:55] <hatch> or ammend or something
[19:55] <rick_h_> hatch: oh, those are pulled from the env now
[19:55] <rick_h_> yea, the charmworld one is nadda
[19:55] <rick_h_> ok, done
[19:55] <rick_h_> hatch: that's one case where that's not charmworld but the juju environment
[19:55] <rick_h_> hatch: so watch out for that one
[19:56] <rick_h_> man this added services feature rocks
[19:56] <hatch> yeah right now it's just defaulting to manage.......static/img/charm_160.svg
[19:56] <rick_h_> we need to do a screencast
[19:56] <hatch> so I'll just keep that functionality
[19:56] <rick_h_> hatch: right, that's the default but we added support for pulling it from the env
[19:56] <rick_h_> hatch: ok
[19:57] <hatch> oh did we? Was that ever implemented in the code?
[19:58] <rick_h_> hatch: ye
[19:58] <hatch> hmm odd I don't see that - sec I'll get you the link
[19:58] <hatch> https://github.com/juju/juju-gui/blob/develop/app/store/charmworld.js#L403
[19:58] <hatch> this is called from the handlebars helper
[19:58] <hatch> and others....but similar
[20:00] <hatch> so I'm not sure that it was ever hooked up to go to the env for local charm icons
[20:01] <hatch> at least not here
[20:01] <rick_h_> hatch: I know it works or worked at one point. It's probably from somewhere else though
[20:01] <hatch> ok I'll keep my eyes open but keep this methods functionality the same
[20:01] <rick_h_> hatch: k
[20:02] <hatch> also is ?demo=true supported? I don't see it in the api docs
[20:02] <rick_h_> ugh, we missed a git tag on the last release :(
[20:02] <rick_h_> hatch: no, it never worked
[20:02] <hatch> lol ok
[20:02] <rick_h_> hatch: it was meant to work but never completed, ignore/rip that out
[20:02] <hatch> consider it ripped
[20:05] <hatch> rick_h_: hmm ok so the syntax is now different heh https://api.jujucharms.com/v4/static/img/charm_160.svg is not found
[20:05] <rick_h_> hatch: yea, we'll have to use a local file I think.
[20:05] <rick_h_> in the gui, I think the icon is there. 
[20:06] <hatch> oh ok lemme check
[20:06] <hatch> it is
[20:06] <hatch> :)
=== urulama is now known as urulama___
[20:18] <hatch> rick_h_: bundle icons we are going to use a local file as well?
[20:19] <rick_h_> hatch: yes
[20:26] <rick_h_> uiteam new gui release uploaded, working on charm release now
[20:34] <hatch> awesome
[20:34] <hatch> now the question is - do we have the bundle icon locally 
[20:35] <rick_h_> hatch: if not, right-click save as :P
[20:35] <hatch> haha yeah just looking through the assets now
=== kadams54 is now known as kadams54-away
[20:37] <rick_h_> kadams54-away: or hatch can either of you take 5 and get me a few nice small screenshots of the added services bar stuff for a blog post once I ge the charm released and ingested
[20:37] <rick_h_> or any of you want to write up the blog post on the release youselves? I'm not greedy.
[20:37] <hatch> yeah I can do both
[20:38] <hatch> can I do it after this stuff? 
[20:38] <rick_h_> hatch: ok, thanks. One thing to remember is the charm has the new custom port as well to mention
[20:38] <hatch> like towards the eod
[20:38] <hatch> yeah definitely
[20:38] <rick_h_> hatch: well I'd like to read it over before release
[20:38] <rick_h_> hatch: so yea, but make sure to hit me up then before you hit publish?
[20:38] <hatch> ok sounds good - I just need to grab some lunch soon :)
[20:38] <rick_h_> and I'll try to look after hours tonight
[20:38] <rick_h_> hatch: ok
[20:38] <rick_h_> ty kind sir
[20:39] <rick_h_> https://github.com/juju/juju-gui/blob/develop/CHANGES.yaml#L24 changes
[20:48]  * hatch lunching
[21:36]  * rick_h_ grumbles about failing charm tests
[21:48] <hatch> boo
[21:49] <rick_h_> it works when I make deploy though wtf
[22:03] <huwshimi> Morning
[22:06] <hatch> morning huw
[22:59] <hatch> heh I just broke this code and all the tests pass
[22:59] <hatch> sumfing wong
[23:00] <hatch> gona eod, will bbl
[23:04] <rick_h_> hatch: night
=== kadams54-away is now known as kadams54
[23:47] <kadams54> rick_h_: how's the release going?
[23:48] <rick_h_> kadams sent an email to tge list with status
[23:48] <rick_h_> basically stuck and eod'ing
[23:48] <rick_h_> monday is my night with thge boy
[23:49] <kadams54> +1