hatchI'd just hate for you to go down the wrong hole 00:00
rick_h_got it :) now to make it all actually work out00:53
hatchgot a diff?00:54
hatchoh, on gh00:55
rick_h_working through it, I know I just blew up a TON of tests00:55
rick_h_so let that roll and generate a big giant number for mw00:55
rick_h_for me that is00:55
rick_h_  ✖ 27 of 1694 tests failed:00:56
rick_h_not bad actually00:56
hatchahh you moved the search view out of the sidebar00:56
hatchgood move00:56
rick_h_right, we're working around the sidebar having the single search UI stuff00:56
rick_h_so this kicks the workaround and just makes it work *right*00:57
hatchyup good call00:57
rick_h_then the card fix is the one line   ✖ 27 of 1694 tests failed:00:57
rick_h_        viewCfg.filters.text = metadata.search;00:57
rick_h_that line ^00:57
rick_h_is the actual card issue00:57
hatchmuch nicer without all of the workarounds to keep it in the sidebar00:58
rick_h_and seems to work ok here in a quick dev/test run of it00:58
rick_h_makes the home button work out fine, etc00:58
hatchsidebar is doing less and less :) pretty soon it'll just be a div haha00:58
rick_h_hey, I always said it should be00:58
rick_h_just a single container to style and clear/etc00:59
rick_h_anyway, now on to problem two for the night00:59
hatchgood so looks like you fixed that the proper way so I'm happy01:00
rick_h_I try to do things right :)01:00
hatchlol, I mean, instead of introducing another workaround01:00
hatchso does the state workflow make sense now? 01:01
rick_h_yea, little bit. The "it ONLY goes through routeDefault" was really easy to miss01:01
rick_h_it'll need some comments and some others to grok it and it'll be cool01:02
rick_h_we had talked about it in principle, but I'd missed the actually talking to code01:02
rick_h_and it calling the same sidebar() threw me as it does a ton of work that clearly wasn't meant to be done in the routeDefault01:02
hatchahh yes yes, there is going to be so much code that will be removed01:03
hatchI'm sure it'll be sad to see all that code you wrote go away :)01:03
rick_h_not at all01:03
rick_h_this is stuff I always knew/hoped would get cleaned up01:03
rick_h_it's not like fullscreen where I STILL think it's the only decent way to search charms/bundles01:03
rick_h_oh bah bah bah, stupid breakout panels01:04
hatchlol - yeah they will add it back in01:04
hatchI know it01:04
rick_h_no, it's going into the new project so not our problem01:04
hatchohh gotcha01:05
rick_h_but it will be a fullscreen experience, so at least I can feel like I was right01:05
hatchsounds good01:05
hatchwell I'm going to pop off, see some of you in DEN, the others in LAS 01:06
hatchhave a good night01:06
rick_h_have a good night and safe travels01:06
hatchthanks you too01:06
huwshimihatch: See you in a few days!01:12
rick_h_huwshimi: make sure to block out thurs during sprints for hte team dinner01:15
huwshimirick_h_: Sure thing.01:16
* frankban lunches12:03
jcsackettrick_h_: my parents dropped in to stay with us last night. i can come to the call at 9 but i may have to hop off in the middle for a few moments to see them off.12:29
rick_h_jcsackett: all good thanks for the heads up12:30
rick_h_jcsackett: I can run through it early/later if it works better12:30
rick_h_just wanted to get the word out12:30
jcsacketthow long a call do you figure?12:30
jcsackettyou want to push it to 9:30? they claim they'll be gone then, so no interruptions. :p12:31
rick_h_jcsackett: ok, will see if kadams can12:31
bacrick_h_: update on charmworld deploy.  it has been bumpy due to obsolete bundles on production that never got updated.  data on the new staging was fresh as it came from ingest with current proof rules.  some bundles on production do not pass proof rules so it took a few iterations to get the migration right.12:40
bacrick_h_: thedac worked out the procedure to copy the prod database to staging and the latest code drop did the exodus cleanly.  so now we just have to apply it to production.  waiting on him to handle it so i don't have to bring the vanguard up to speed.  he's vancouver-based so it'll be a few hours.12:41
rick_h_bac: since bundles are one revision and reimportable can we blow them away and then migrate?12:43
rick_h_bac: do we have the code in place to convert the strings on reingest?12:44
bacrick_h_: as i said, the exodus works cleanly now on a copy of the production db.  so it should apply to production ok.  i'd rather let it run when thedac gets in.12:44
rick_h_bac: sorry, read that as it applies on a clean db12:45
bacrick_h_: yes, ingest does the convert12:45
rick_h_bac: ok cool12:45
rick_h_bac: thanks for the heads up. Let me know what we need in case you need to head out before things complete12:45
bacok.  RT 69701 (noted in card) has the details12:46
hatchmorning all12:50
bacugh.  just realized i have to pack for cold/rainy and hot/dry.12:50
bachi hatch12:50
hatchbac cold/rainy? Where are you going?12:51
rick_h_bac: heh, I hate that. 12:51
rick_h_packing for montreal sucked, but looking forward to just vegas12:51
hatchflight delayed 3h, sitting at airport....awww yeah12:51
* hatch puts on sunglasses like CSI12:52
hatchfree wifi at the airport though12:52
hatchslow....so hotspotting instead12:52
jcsacketthotspots always seem better than airport wifi.12:53
hatchLTE hotspot will definitely be better than airport wifi lol12:53
hatchand I don't have to worry about anyone wiresharking me12:54
jcsackettrick_h_: i'm guessing 9 is still go time, which works for me.12:58
rick_h_jcsackett: ok12:58
rick_h_jcsackett: yea, not seen kadams yet this morning12:58
jcsackettrick_h_: yeah. i'm guessing his hours start at 9 like mine, but he has less of an email addiction. :P12:59
rick_h_heh, he accepted late last night :P12:59
hatchand kids12:59
jcsacketti take it all back. we'll just blame the kids.12:59
hatchour new airport has plugins at each seat.....everyone here is on a laptop13:00
jcsackettrick_h_: i think, given you assigned me to a card i thought i was already assigned to, that i wasn't connected to the internet when i was manipulating kanban yesterday evening. and now i find that to move my card goes over lane limits. should i do so? don't think we have time for me to do slack while waiting.13:01
rick_h_jcsackett: over limit is ok. we'll clear it out today and priority is prepping for release13:02
rick_h_jcsackett: if you were really motivated you could review/land huw's branch to make room13:02
rick_h_but I can try to look at that after our call 13:02
jcsackettif you don't have time for that i can certainly review it, but also post call for me.13:03
rick_h_but Makyo's card should be landing, it was in review and commented on yesterday13:03
hatchrick_h_ http://www.engadget.com/2014/04/22/air-berlin-pebble/?ncid=rss_truncated :-)13:03
rick_h_hatch: nice13:03
hatchthought you'd like that....being a pebble user and all13:03
rick_h_the ios part sucks but oh well :P13:04
hatchhaha yeah13:06
rick_h_hatch: heh you wanted deleting https://github.com/juju/juju-gui/pull/255/files13:15
hatchnoice!! almost 1k lines13:15
hatchwhere we're going, we don't need code!13:15
=== BradCrittenden is now known as bac
rick_h_jcsackett: if you're available let's do this sans kadams then. Not sure where he is13:30
jcsackettrick_h_: there he is, and i'm in the hangout.13:34
rick_h_jcsackett: in https://plus.google.com/hangouts/_/canonical.com/go-over-search?authuser=113:34
jcsackettrick_h_: it was attached to the appt, i'm already there.13:34
rick_h_jcsackett: try again, that's the one attached to the calendar 13:34
rick_h_kadams54: and I are there13:34
jcastrohey guys, did the gui stop showing icons for local deployments recently?13:36
jcastroI moved a demo over to a new box and no icons13:36
rick_h_jcastro: no13:36
rick_h_jcastro: nothing changed recently, no releases done though hoping for one tomorrow13:37
jcastrorick_h_, how can I force the icons on again?13:37
rick_h_jcastro: you can't at the moment13:38
jcastroany idea why icons would show up on some machines but not others?13:38
kadams54jcsackett, rick_h_: Google Hangouts is now crashing immediately on me… gonna restart the browser.13:43
jcastrorick_h_, is there a bug for linking to the service's page from the gui?13:43
rick_h_services page?13:43
jcastroso like I fire up say, hadoop13:44
jcastroand I want to see the hadoop page13:44
jcastroI have to URL mangle/guess13:44
rick_h_which hadoop page? 13:44
rick_h_I need more info, we've got a lot of hadoop pages13:44
jcastroI deploy hadoop in the gui13:44
jcastrothen I click on the box13:44
rick_h_ok, yes, you deployed it13:44
jcastrothen I find the unit13:44
jcastroclick, then I find the IP13:45
jcastrothen I click on it13:45
jcastrobut like, it takes me "x.x.x.x" and not "x.x.x.x:xxx/hadoop" or whatever it is supposed to be13:45
rick_h_ok, we don't know what hte url is in the charms13:45
rick_h_do they expose "available urls"?13:45
jcastroright, I winder if that's a good idea13:45
rick_h_that's not something we can do. 13:45
jcastrolike, it should be right, if you're in the gui13:45
jcastroyou should be able to hit any page on any service13:45
rick_h_ah ok, it's been talked about but we'd need the charms to expose it. I forget what it was talked about13:46
jcastrorick_h_, I mean "we" as in all of juju, not just you13:46
jcastrohmm, I'm going to ask13:46
rick_h_ok cool13:46
rick_h_yea, we can look at it13:46
jcastrodo any of you have time to help debug our gui?13:47
jcastrowe need the icons for mark's demo13:47
rick_h_jcastro: ok, otp give me 213:47
kadams54Apparently the browser restart was not sufficient13:47
rick_h_kadams54: all good, I think we're set13:48
kadams54I shall go forth and code… er… test… er… fix broken tests… by coding.13:48
jcsackettkadams54: only question left was do i need to sync with your other branch, or just the one rick has put up?13:48
rick_h_kadams54: jcsackett let me know if you hit any issues that delay. We're squeezing in for release tomorrow13:48
rick_h_jcastro: sure thihng13:48
kadams54jcsackett: my other branch is crap, so I suspect not :-)13:48
jcsackettkadams54: dig. :)13:48
rick_h_jcastro: is the gui up where I can hit it?13:48
jcastroone sec13:48
kadams54Too many rabbit holes with no results13:48
jcastrorick_h_, do you have access to the lab?13:49
rick_h_jcastro: not that I'm aware of13:49
kadams54rick_h_: hey, what would be the best way to take over your branch/PR?14:41
kadams54I could use the `git qa-pr` alias to pull down a local copy…14:41
rick_h_kadams54: you could use the qa-pr tool to pull it down and wokr on it, rebase it, and then push to your own as git push origin xxx:xxx14:42
Makyojujugui did we ever figure out why mocha sometimes fails with no difference between actual and expected?14:45
frankbanMakyo: because useless failure output is one of the mocha design goals?14:46
Makyofrankban, +100014:46
bacit is a wildly successful program, then14:46
Makyojujugui call in 1014:50
bacjcsackett: here are the logs from staging.  https://pastebin.canonical.com/108999/plain/14:54
bacjcsackett: the beaker exceptions in app-exception.log have been seen for a while and thought to be innocuous14:54
jcsackettbac: the same errors are seen on production?14:55
jcsackettthat is a damn shame.14:56
jcsackettyeah, i don't see anything in here that explains the login behavior.14:56
rick_h_jcsackett: hangout time15:01
rick_h_frankban: ^15:01
kadams54Bah, Google Hangouts is crashing on me again…15:01
MakyoSo.  assert.deepEqual(actual, expected) fails, but assert.equal(JSON.stringify(actual), JSON.stringify(result)) passes.  Nice.15:11
rick_h_I'm going to print this 100 times and ship it in a box to hatch's house http://hadihariri.com/2014/04/21/build-make-no-more/ :)15:15
MakyoOnce per day for a year?15:19
rick_h_that sounds good15:20
=== BradCrittenden is now known as bac
bacrick_h_: charmworld updated and seems happy15:43
rick_h_bac: woot thanks!15:45
rick_h_bac: have a good trip, see you on the other side15:45
* bac bye16:02
Makyojujugui ready for re-review  - I've confirmed that the tests pass in IE10, but will be looking at the saucelabs video just to be sure.16:10
jcastrorick_h_, icons still broken16:12
jcastroand also it seems that when we deploy it forgets where the boxes are supposed to go in the bundle16:12
Makyojujugui forgot the link: https://github.com/juju/juju-gui/pull/253 (confirmed it was a timeout on saucelabs' side, too)16:15
rick_h_jcastro: k, bringing up a test environment16:18
rick_h_jcastro: will figure out how to build the url manually16:18
MakyoHoly crap, they accepted our offer!16:19
frankbanguihelp: I need two reviews for https://codereview.appspot.com/90570044 (quickstart/python, high priority) with my apologies for the long diff. Anyone available? Thanks!16:19
rick_h_Makyo: woot! party in vegas to celebrate!16:22
rick_h_jcastro: frankban can help look into it. Can you pm him the info from juju status and the environment admin secret so he can look at how to build a url to check if the files are in juju or not?16:22
Makyohttps://www.flickr.com/photos/ranna/sets/72157644219100342/ 1.09 acres, 2362 sqft., hickory floors, quartz counters, glass backsplash tile...16:33
rick_h_go hickory, that's what I put in. <3 those floors16:36
rick_h_frankban: so if you're helping jcastro I'm going to go back to working on the inspector release blocker stuff. Let me know if you need a hand16:37
frankbanrick_h_: sure thanks16:37
jcastrorick_h_, all fixed, I owe frankban more beer16:59
rick_h_jcastro: cool, what was it frankban ?16:59
frankbanoutdated charm branch16:59
rick_h_oh lol17:00
rick_h_nice and easy then cool17:00
Makyojujugui running out to pick up hatch from the airport.  Will be back from the hotel.17:00
rick_h_Makyo: rgr, give him the grand tour and maybe find him some chicken and waffles :P17:01
rick_h_jujugui need a review and qa for a small branch please https://github.com/juju/juju-gui/pull/25717:38
rick_h_kadams54: jcsackett either of you have a chance to review https://github.com/juju/juju-gui/pull/257 please? Tests pass now :018:48
rick_h_kadams54: jcsackett going afk. I'll be back tonight. If you need anything important ping me on hangouts19:16
jcsackettrick_h_: i may have time to look at it later today.19:24
kadams54rick_h_: Taking a look…19:34
hazmatrick_h_, can we make gui use the same trick it does for local charms on store charms?20:05
hazmatrick_h_, re icons20:05
hazmatrick_h_, ie. always get charm icons from the backend on deployed charms20:06
hazmatguihelp /juju-gui ^ 20:14
rick_h_hazmat: https://plus.google.com/hangouts/_/72cpi460salmu56blakp9rmtes?authuser=1&hl=en20:58
rick_h_hazmat: app/views/utils.js getCharmIconUrl has the if local condition you want21:00
hatchhi all21:01
rick_h_hey hatch 21:01
rick_h_hazmat: testing it out on ec2 with https://github.com/mitechie/juju-gui/tree/always-store-icons21:02
hatchhazmat are you in DEN yet?21:03
rick_h_hatch: yea, he just ping'd from the hotel there. I think most folks are there based on earlier irc21:04
hatchoh cool, I got the sim card updated but had to pay the $10 for hotel wifi....ugh21:04
hatchcharging for internet must be a North American hotel thing21:04
rick_h_yea, it sucks21:05
rick_h_was $25 when I was in chicago once21:05
hatchholy crap! 21:05
hatchI think I'll hot spot for the rest of the time 21:05
hatchman npm issues galore in CI21:05
jcsackettkadams54: did your branch land yet?21:10
kadams54No, net yet21:11
kadams54hatch: for you: https://github.com/juju/juju-gui/pull/257/files#r1192412421:13
hatchlol ^521:13
* hatch ducks and covers21:13
jcsackettjujugui: can i get a review of https://github.com/juju/juju-gui/pull/258/files21:16
rick_h_kadams54: hatch what's the goal of the preventDefault vs halt?21:16
hatchrick_h_ halt stops the propagation of the event whereas preventDefault only prevents the default event for that handler21:16
hatchso you won't be able to listen on a container for that event and there will be no way to debug it21:17
kadams54I view halt as a bit of a nuclear option21:17
hatchit burnt me hard earlier in the gui and in past projects21:17
hatchso I really try to avoid it21:17
rick_h_hatch: ok, but it's used on a close button. It's already on a container, and it's own html/template of control. No one else should be listening to that things events. 21:17
rick_h_it's like having code across the page looking in case someone does some raw JS event on some temporary node used in a widget21:18
rick_h_you bind to events on the widget, not click events in the widgets html21:18
rick_h_otherwise it's fragile as can be21:18
hatchwell say you wanted to capture all click events in a element for whatever reason21:18
hatchthis click event would never reach your handler if halt is used21:18
rick_h_hatch: right, I'd scream to high heaven you can't do that :)21:18
hatchthat's how pjax works21:19
rick_h_so no, I prefer my nuclear option in my widget :)21:19
rick_h_ok, will look at the scope but not sure this is an 'always good' rule and that it's appropriate here21:19
hatchthe only reason halt is 'required' here is because of pjax, else prevent Default would work just fine21:20
hatchjcsackett I can review your branch21:20
jcsacketthatch: cool, thanks.21:20
hatchoh crud, I can't QA21:21
hatchI can't connect to my vagrant for whatever reason21:21
hatchprobably this network21:22
hatchjcsackett done21:30
hatchyou just need someone to qa21:30
rick_h_I can qa tonight at the coffee shop21:30
jcsacketthatch: how strongly do you feel about that personal preference noted? b/c it runs entirely counter to my personal preference. :p21:30
jcsackettrick_h_: thanks.21:30
hatchjcsackett haha, tbh I'll be voted down on it so up to you21:30
hatchI think it looks more like python style indentation21:31
hatchso I figured you python guys would prefer it21:31
hatchbut doesn't look like it :)21:31
rick_h_hatch: multiple lines is per our code convention docs21:31
jcsacketton short constructs, yeah. on large blocks like that it makes it way harder to keep track of what's where, imo.21:31
rick_h_hatch: with any complex objects as their own var = line21:31
rick_h_hatch: we had a friday retro on this21:31
hatchjcsackett but in python you don't have that problem.... :)21:32
rick_h_we can revisit, but I'm with jcsackett. Seeing vars vs hiding vars. I hate we don't use the rule on out tests and don't force them to be alphabetical21:32
hatchrick_h_ yeah - I'm talking about the curlies21:32
rick_h_oh, that one21:33
hatchI think smushing all the curlies up on the end of the last line is cleaner21:33
rick_h_meh, I'm still with jcsackett. Lines up and easy to see. 21:33
rick_h_the fact that you call it "smushing" is hurting your case :P21:33
rick_h_I don't ever think "boy that's some great smuching code. Much more readable smushed like that"21:34
hatchI really wish github had syntax highlighting in diffs21:34
jcsackettdoes anyone have that?21:34
hatchjcsackett I dont' think so, I think they all just colour the diffs21:35
jcsacketthm. i find myself missing `bzr shelve`.21:35
hatchgit stash :)21:36
hatchand it's cousin, git pop21:36
hatchhmm I really have no idea how to debug/fix this network issue....21:37
hatchguess I won't be fixing it today :)21:37
jcsackettgit stash doesn't do quite the same thing.21:39
hatchoh? 21:39
hatchI thought they were almost identical21:39
hatchwhat is the difference?21:39
jcsackettif i have several chunks of changes in one file, bzr shelve lets me shelve them selectively.21:39
jcsackettstash is just "what's the WIP? let's stash it" -- or at least appears to be that.21:40
hatchahh yeah, I didn't know bzr shelve did that. How does it determine where the chunks are?21:41
rick_h_bzr diff, grab each chunk21:41
hatchok so basically wherever there is a line which isn't changed is a divider 21:42
hatchI could see that being useful21:43
hatchthere has to be a way to do that with git21:43
hatch....you can do everything with git somehow :)21:43
hatchso you can git stash only a single file if you have edited multiple21:51
hatchbut I can't find any way to stash parts of a diff21:52
hatchoh wait21:52
hatch`git stash --patch --no-keep-index`21:53
hatch^ jcsackett  you can try that21:53
jcsacketthatch: alright, i'll give that a go.21:53
hatchyou might need to use `save --patch` 21:53
hatchbut maybe not21:53
hatchhttps://www.kernel.org/pub/software/scm/git/docs/git-stash.html grep for --patch for more info21:54
hatchI knew there had to be a way haha21:54
hatchapparently `git stash -p` works exactly like what you're asking....but I can't find any docs to back that up21:55
Makyojujugui quick code review for demo-able deployer button - https://github.com/juju/juju-gui/pull/25922:01
Makyojujugui also, confirmed the CI failure was an unrelated timeout on https://github.com/juju/juju-gui/pull/25322:02
hatchMakyo review done - I can't QA unfortunately22:04
hatchahah fixed the network issue22:08
hatchMakyo I can QA now22:08
rick_h_anyone recall the guiserver url?22:08
rick_h_guihelp ^22:08
hatchrick_h_ like you want to access the guiserver not the gui?22:09
rick_h_hatch: nvm, I think I'm close now22:10
hatchok cool22:10
rick_h_hazmat: I don't think it'll work. I don't know the core internals to tell but trying to force the icons to the store is easy, but the data isn't there in the api22:11
hatchhazmat Makyo I'm going to start walking to the meetup at about 4:3022:12
MakyoSounds good.22:12
rick_h_hazmat: pita, but think https://docs.google.com/a/canonical.com/document/d/1Y8Uhomr4_6L3nFXLXPZY2V-tFjl5KjOIknQLuXdzU_A/edit might be the best bet for the demo so that you can ingest the charms in that local store and provide the icons22:12
rick_h_hazmat: unless there's something in the core api for the files that will allow this to work for non-local charms22:13
rick_h_hazmat: https://ec2-54-87-142-1.compute-1.amazonaws.com/juju-core/charms?url=cs:precise/juju-gui-90&file= is the url for getting a list of the files22:13
hatchfixed the network issue now only to have other stuff break....bleh22:13
rick_h_hazmat: obviously change to your url/etc22:14
rick_h_Makyo: jcsackett so both of you need qa?22:17
hatchlooks like it22:17
MakyoYes please.22:17
hatchI'll try and get mine up and running when we get back22:17
hatchMakyo meet you in the lobby?22:18
MakyoYep, down in a few.22:18
hatchok cya all22:18
MakyoDucking out too.  Will catch up later, possibly while there.22:20
mbruzekHi rick_h_ Can I have a minute of your time?23:01
mbruzekI am doing a charm review, and Jose added defaults to the config.yaml for juju-gui.  https://code.launchpad.net/~jose/charms/precise/juju-gui/add-blank-defaults/+merge/21288523:06
mbruzekNormally this is not a problem, but I have encountered charm authors who counted on some parameters to be "unset" and I was wondering if anyone in the juju-gui would like to weigh in on this?23:07
mbruzekSpecifically the config options that he now defaults to "" are:  ssl-cert-contents, ssl-key-contents, login-help.  Does anyone know if the juju-gui depends on those 3 being unset rather than empty string?23:10
mbruzekalso password 23:10
rick_h_mbruzek: sure thing23:53
rick_h_mbruzek: we've got a task to look at unset itself. The defaults should be ok as "". I don't recall if we'll send them as "" or not23:54
rick_h_mbruzek: honestly, I'd try it out and see. I think the unset we need to do is to reset a config where they expect it to be unset23:55
rick_h_if he's defaulting to "" then he's expecting that?23:55

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