/srv/irclogs.ubuntu.com/2013/04/25/#juju-gui.txt

hatchrick_h_: are you around?01:32
rick_h_hatch: yea, what's up?01:35
hatchquestions about your proposed branch01:35
hatchis that for the little tab to open the store?01:35
rick_h_hatch: shoot01:35
rick_h_right01:35
rick_h_that opens/closes01:35
hatchso why couldn't you have a simple css based tab with a open/close method on the subapp ?01:35
hatchit looks like that's all it's doing but with way more code01:36
rick_h_hatch: no, because it needs to be state, back button enabled, urlable, etc01:36
hatchwell doesn't it only have two states? shown and not?01:37
hatchand it's shown if the other two are closed01:37
rick_h_hatch: maybe I'm not following. So there's three 'states' to the browser. minmized, fullscreen, or sidebar view. And tomorrow I have ot add a hidden state as well01:38
hatchwana guichat it?01:38
rick_h_hatch: and I was going to ask to pair up on that because whenever we're on the juju-gui interior views, we have to get the sidebar to go to a hidden state and make sure the back button pops it back to the old state/etc01:38
rick_h_hatch: can we setup first thing tomorrow? I've only got 20min left here at the coffee shop01:38
hatchoh yeah sure np01:39
rick_h_hatch: cool, let me konw when you get in tomorrow morning and we can chat as I wanted to grab your time on the hidden state stuff as well.01:39
hatchgoing out tonight so it definitely wont' be before my usual time (8am) :)01:39
rick_h_hatch: rgr, have fun01:40
hatchcurrently 7:40pm01:40
rick_h_yea, add 2hrs here01:40
huwshimirick_h_: In fact the minimised sidebar needs to remember its state so that when you open it again it remains the same as it was before you minimised it01:43
rick_h_huwshimi: right, so it should. It pushes a new state and doesn't change anything like charmids, search input, etc01:45
rick_h_huwshimi: so when you open it, it reloads. When we get the cache into place it'll avoid that reload as well01:45
rick_h_huwshimi: but that cache state stuff will be later on. For insance we can cache the interesting charm list and search results from fullscreen/sidebar and such01:45
huwshimirick_h_: Ah I see your branch actually does that, but shouldn't it just be the real sidebar there that is just moved over?01:46
huwshimi(Although I don't know how that should be handled when you were in full-screen mode)01:46
rick_h_huwshimi: so yes/no. Yes, but that'll be harder since we'll have this state change but moving it over while this took me 2hrs tonight to add01:46
rick_h_huwshimi: so going with this for a first pass that does what it should and we can look closer at how to do it with sidebar/fullscreen/animation/etc as a later improvement01:47
huwshimirick_h_: Yeah, fair enough, that makes sense01:47
hatchrick_h_: btw- 3.10 doesn't make any apreciable difference in speed in the tests :(01:48
hatchappreciable01:48
rick_h_hatch: :(01:49
hatchyeah....oh well01:49
hatchhere's hoping 3.1101:49
hatch:)01:49
* hatch out!01:49
rick_h_hatch: we'll just have to fix our crap and make it faster :P01:50
hatchhaha01:50
* hatch really out!01:50
=== matsubara is now known as matsubara-afk
=== rogpeppe1 is now known as rogpeppe
gary_posteruh-oh, new version of nodejs....12:08
rick_h_lol, run for the hills!12:09
gary_poster:-)12:09
gary_posterugh, that's a lot of fonts to download.  Did you do a comparison on the download size rick_h_ ?  I'm fine with that landing, but I may want to negotiate with UX if this is a big increase12:10
rick_h_gary_poster: yea, I'm looking at it qa'ing it now12:11
gary_postercool thanks12:11
rick_h_gary_poster: 482B gzip12:11
rick_h_gary_poster: well that's the css12:11
gary_posterB?12:11
rick_h_gary_poster: there we go, 164KB in fonts12:11
gary_posterheh12:11
rick_h_for the woff font in chrome12:12
gary_posterwhat was it before do you know?12:12
rick_h_I'm checking aganist uistage12:12
gary_postercool12:12
rick_h_gary_poster: 164? what it says but wtf12:13
gary_posteruh12:13
gary_posterooooh kay12:13
gary_poster:-)12:13
gary_postermaybe the different weights are done programmatically?12:14
rick_h_checking in FF. Havnig a hard time buying it12:14
gary_posterthx12:14
bacoops, i forgot to wear my raring t-shirt today.  /me changes12:19
gary_poster:-)12:19
rick_h_gary_poster: oh, so that's the old value. my local instance didn't pick up the change to index.html so it's still loading the old size12:20
* bac much better12:20
rick_h_gary_poster: so the update isn't much bigger 205KB, but it's 5 files :/12:21
gary_poster:-/12:21
gary_posterrick_h_, was how many before? 1?12:22
rick_h_gary_poster: yea, one request out. Trying to see why it changed to 5. Not used the google webfonts much beyond "I want this font, load"12:22
gary_posteryeah, me too.12:22
rick_h_ah, css at least seems to have a diff url for each font/weight12:23
gary_posterah that makes sense12:23
rick_h_well, but had hoped google would combine. Almost wonder if it'll to all of the ubuntu font in one request but larger size that's cached12:24
rick_h_gary_poster: ok, so I dropped the Book 300. Not sure if we need that one. Got it down to 4 requests. but yea, I'm not sure how to drop it down more without getting UX to agree to limit fonts combo/italics12:32
rick_h_http://www.google.com/fonts/#UsePlace:use/Collection:Ubuntu+Mono:400,700,400italic,700italic|Ubuntu:400,500,700,400italic,500italic,700italic12:32
gary_posterrick_h_, ok, thank you.  I'd love to have your research saved someplace for later discussion...12:33
* gary_poster thinks12:33
gary_posterjujugui blog?  I can put it there or give you posting privs12:34
rick_h_gary_poster: yea, I'll put this up for review if that's ok and move forward, but file a bug/card to talk with UX on it. 12:34
gary_posterok perfect thanks rick_h_ 12:34
rick_h_gary_poster: sure, I can post something there. 12:34
gary_posteryour choice rick_h_ .  I'll go ahead and give you privs, and then you do whichever you prefer12:34
rick_h_gary_poster: ok, changed my mind. Did some more tweaking, posted to the blog. Thanks for access12:56
rick_h_gary_poster: if you get a sec, can you +1 the changes since you've peeked at it and we've had the chat?12:57
gary_posterrick_h_, will do.  thanks for blog post and for the further imporvements you talk about12:58
gary_posterrick_h_, LGTM done12:59
rick_h_ty much12:59
hatchmorning!13:01
rick_h_morn13:01
hatchrick_h_: looks like I made it an hour early13:01
hatchlol13:01
* hatch isn't here13:01
rick_h_hatch: :P yea, I need a sec to swap out hard drives so come back later 13:02
* gary_poster looks around, sure he just heard hatch somewhere,,,13:02
hatchhaha13:02
rick_h_my backup of my backup of my NAS finished after I started work so finally can swap the second drive out13:02
* benji sees a hatch-shaped ghost fly through the wall.13:02
* rick_h_ sees some green goo left behind on the wall and hopes he doesn't have to clean it up :)13:03
hatchdo we have any beer connoisseurs?13:03
rick_h_wine?13:03
hatchhttp://www.ratebeer.com/beer/westmalle-dubbel/2205/13:04
hatchtaists like cherry chocolate :)13:04
=== Makyo|out is now known as Makyo
gary_posterugh13:55
gary_posteranyone remember where the pyjuju logs are?13:55
gary_posteron the machines13:55
gary_posterjujugui ^^13:56
bacgary_poster: /var/lib/juju/units/.../charm.log ?13:57
baci'm not sure about the exact path but you should be able to follow from /var/lib/juju13:57
gary_posterbac that's it! thank you13:58
gary_posterhatch, probelm in CI is that make distfile is failing with this: http://pastebin.ubuntu.com/5601096/14:01
hatchoh awesome you were able to log in14:01
hatchlooking...14:01
gary_posterI assume we can dupe after a make clean-all14:01
hatchugh not that again14:01
gary_posterhatch, what is that?14:01
hatchit's a versioning issue14:02
hatchsomething between imagemagik and spritegen if I remember correctly14:03
gary_posterah14:04
hatchthis happened right after we switched node to 10.x right?14:04
gary_posterso what was the resolution before?  (and why is it failing only in Canonistack, I wonder?)14:04
gary_posteryes hatch14:04
hatchcan you do a version check on node, npm and the depdendencies so we can compare them to the working local versions14:05
hatchmaybe for some reason it's getting a different version?14:05
rick_h_hatch: ready to chat whenever you are14:05
hatchI can help look through the dependency list if you like as it's pretty large14:06
rick_h_luca_: so I'm using https://chrome.google.com/webstore/detail/whatfont/jabopobgcpjmedljpbcaablpmlmfcogm?hl=en to check the fonts and only the readme is incorrect. One suggestion would be to launch a private browsing session so that it should break appcache/js cache and view the page? http://uistage.jujucharms.com:8080/bws/fullscreen/precise/cassandra-2/14:08
gary_posterhatch, http://pastebin.ubuntu.com/5601114/ .  Shouldn't the dependencies be the same because of the shrinkwrap?14:08
hatchyeah they should be14:09
hatchbut outside of the versioning I don't know what else could cause this error14:10
gary_posterI'm looking to see what version of ImageMagick is on a working precise instance14:12
gary_posterI'm also thereby running make distfile there14:13
gary_posterit may be that we don't see this normally because we don't normally start the charm form a branch, and we are just broken on Precise14:13
hatchohh that's possible14:16
hatchlet me know if you want me to look into it more14:17
hatchrick_h_: guichat?14:20
rick_h_hatch: sure14:20
abentleyrick_h_, alejandraobregon, luca_, jovan2: I've fixed staging, so uistage should also be fixed.14:21
rick_h_abentley: thanks! what was up? some charm bomb things out?14:22
abentleyrick_h_: mongodb and elasticsearch had config-changed errors, but I don't know why because the debug logs are jammed up back to April 19.14:23
luca_rick_h_: I sent you an email of what I see in my chrome14:23
luca_rick_h_: the Ubuntu font isn't showing on the headers14:23
luca_rick_h_: but it shows in the body copy14:24
rick_h_luca_: k, otp14:24
hatchgary_poster: I'm going to get back on this service details stuff, but before I do that did you want me to look into this error or do you have it?14:32
rick_h_luca_: after this call can we do a hangout on your machine and try to use screen sharing/etc to walk through checking it out?14:32
gary_posterhatch, problem is in charm making distfile14:33
gary_posterhatch, if you change juju-gui-source=lp:juju-gui14:33
gary_posterthen you can dupe14:33
gary_posterhatch, so simplest solution is to just revert the change14:34
gary_posterhatch I'd like to keep teknico's doc changes14:34
hatchyeah reverting to node 8.x is a quick fix - just need to change to the legacy ppa14:36
hatchalthough that's not going to be available forever14:36
rick_h_hatch: can you look over that branch please?14:36
hatchsure14:36
rick_h_hatch: ty much kind sir14:36
gary_posterhatch we can make our own PPA, as I proposed (on call, sorry)14:40
hatchoh right - yeah that should really help solve a lot of our issues....assuming that making our own PPA isn't a huge PITA (I have no idea how that's done)14:41
gary_postervery easy14:41
hatchohh ok well in that case....maybe I Should do it so I can learn how ;)14:48
MakyoBlack magic going on in this dumb zoom code.15:05
hatchrick_h_: after I reviewed your branch I had an idea about the url stuff re minimized, lemme know when you have a second15:08
rick_h_hatch: yea, I need a few please. Trying to multi task around calls has me in cirlces atm15:09
hatchno rush whenever you have time15:09
hatchbut I would really like some starbucks15:09
hatchhmmm.....15:09
rick_h_hatch: guichat?15:13
hatchsure15:13
gary_posterhatch I'd like to get the charm fixed up ASAP so I can send a juju-core announcement15:26
gary_posterhatch, talk through it when you are available?  you may be on another call now15:27
hatchsorry on call with rick_h_15:27
gary_posterack15:29
gary_posterbcsaller, hey.  I'd like to talk over your current card with you when you get a chance, please15:35
hatchgary_poster: done - did you want to chat?15:43
gary_posterhatch sure15:43
gary_posterguichat15:43
bcsallergary_poster: I'm around to chat after that 15:45
gary_posterthanks bcsaller 15:45
gary_posterwill ping15:45
rick_h_luca_: ping, still around? 15:49
gary_posterjujugui please update kanban15:50
=== matsubara-afk is now known as matsubara
luca_rick_h_: I'm available now, are you?15:57
gary_posterjujugui call in 215:58
luca_rick_h_: https://plus.google.com/hangouts/_/cf491220dfb7736cee2876baf29110eeb52f599715:58
rick_h_luca_: one sec16:02
luca_rick_h_: np16:03
hatchoh yay they dropped off our curbside recycling bin.....just what I always wanted...to pay for things I already do now for free16:31
hatchcrap, I pushed to the wrong branch and now lbox thinks that's the one it should push to16:43
hatchanyone run into this before?16:43
benjihatch: I think you can do a bzr push with --remember to the right branch and lbox will use it from then on16:47
hatchahh ok - I just deleted it and started over16:47
hatchbut I'll remember that next time16:47
hatchgary_poster: bcsaller https://codereview.appspot.com/8631048/ plz16:48
hatchrick_h_: is the router approach working?16:50
rick_h_hatch:  no idea yet. Been sidetracked in calls for a bit16:51
hatchgotcha np16:51
hatchoh I gota push this to a trunk branch for it to be able to work17:05
hatchpoop17:05
hatchI forgot about that issue17:05
rick_h_jovan2: looks like luca ran away. Can you verify that the fonts are all ubuntu for you on staging now please? http://uistage.jujucharms.com:8080/bws/sidebar/precise/ceph-7/17:06
jovan2rick_h_ will do17:07
jovan2rick_h_ looking good :)17:08
rick_h_jovan2: yay!17:08
benjigary_poster: did you ever figure out where pyjuju puts the logs?  I'd like to look at the gui charm log output.17:11
gary_posterbenji /var/lib/juju/units/.../charm.log.17:12
rick_h_jovan2: questions posted on https://bugs.launchpad.net/juju-gui/+bug/117279017:12
_mup_Bug #1172790: Lots of empty space below charm details <charmbrowser> <juju-gui:New> <https://launchpad.net/bugs/1172790>17:12
jovan2rick_h_ will send charm icons to you in a mo...17:12
benjithanks17:12
rick_h_jovan2: cool, for the featured charms? Those really go to jcastro and arosales 17:12
rick_h_jovan2: I can't get those in17:13
jovan2rick_h_ yeah but you're being copied in ;)17:13
rick_h_jovan2: ah ok thanks17:13
arosalesrick_h_, I'll work with my team on getting some merge requests for the icons into charms if I don't have time myself today.17:29
rick_h_arosales: thanks! I'm working on getting manage.jujucharms.com up and to the latest release so we can get them marked as featured and add QA to somem of them as well17:29
rick_h_arosales: that'll probably be tomorrow17:29
arosalesrick_h_, sounds good17:31
rick_h_oh yay, the browser toggle branch landed on staging as well jovan...who left17:32
* rick_h_ hits toggle over and over bwuhahaha17:32
hatchlol17:36
hatchthe node downgrade appears to be deploying properly on EC2 once it completes all the tests I'll land that branch and get our CI back on track17:49
hatchbut for now I'm going to grab lunch17:50
rick_h_does JS regex not support /^\/(charms|unit)/ ? for matching a set?17:50
gary_posterrick_h_, yeah, that would match /charms or /unit or /charmskumquat or /charms/kumquat and so on. :-)18:01
rick_h_gary_poster: hmm, regex tester fails only matching charms. I'll mess with it some more18:01
rick_h_I think I'm doomed anyway since I see there's no way to negate that in JS18:02
gary_posterrick_h_, http://pastebin.ubuntu.com/5601849/ wfm <shrug>18:03
rick_h_gary_poster: cool thanks18:03
=== matsubara is now known as matsubara-afk
hatchrick_h_: oh yeah regex tester has some odd issues, it claims to work with everything js regexy but I've found a number of issues, I usually just try it in the console18:41
hatchgary_poster: I read charmskumquat as charmsqat lol18:41
gary_poster:-)18:41
hatchtests all passed and the branch has been submitted18:42
hatchso CI should be back on track now18:43
rick_h_hatch: so I've got a working prototype here but I'm worried if this will always hold true: http://paste.mitechie.com/show/949/18:45
hatchok looking18:46
hatchwhy would it always hold true?18:46
rick_h_I'm worried will :gui: always be in the url?18:46
hatchunless it's on /18:47
hatchat which point that should be false, and then show the browser18:47
rick_h_hatch: not that i'll always be true, but that this regex will always catch the internal views. Does :gui: every go out of the url?18:47
rick_h_hatch: ok, cool then18:47
hatchif /service /unit or /charms are in the url, it will be under the :gui: namespace18:47
hatchat least unless something else changes of course18:47
hatchanyone have any idea why `this.get('env')` in serviceList will return an array of undefined values equal to the number of services in the environment?19:09
hatchI'm pretty baffled, the envChange event is being fired with an undefined event facade19:12
hatchO K now I'm bonified confused because it's working now lol19:15
rick_h_hatch: hmm, login/logout view?19:15
hatchimho it should be hidden and so should the little toggle button19:16
hatchbut I don't know how much say I have in that decision :D19:16
rick_h_hatch: yea, but there's no url to check on those19:17
rick_h_it's determined mid-request processing it looks like?19:17
hatch{ path: '/logout/', callbacks: 'logout'}19:17
hatch?19:17
rick_h_hatch: oh hmm, didn't see it in the browser when I did it19:17
rick_h_and was checknig for login :)19:17
hatchbut you are right there is no login19:18
hatchsee check_user_credentials19:18
hatchmethod19:18
hatchso we might need manual fn calls in there19:19
hatchdarn!19:19
rick_h_hatch: checking, so logout I do have a url. And logout shows the login at the same time without another request19:19
rick_h_so that just works as is, check for that in the url as well19:20
rick_h_can you get to just login? there's no url for it19:20
gary_posterbenji, would 40 minutes from now work for you?19:20
hatchif there are no credentials then it shows the login page19:20
hatchit should redirect to /login19:20
hatchI've been meaning to bring that up19:20
* hatch creates card19:20
hatchrick_h_: in the meantime you'll need to call the callback manually in the check_user_credentials method19:20
rick_h_hatch: does it auto log you in though? I'm logging myself out and then I see a login screen. When I reload, I get the main site19:21
rick_h_hatch: k, looking for that method19:21
hatchyeah that's in the config.js19:21
hatchconfig-debug.js19:21
hatchhas user: and password: properties19:21
hatchset those to undefined to land on a login page19:21
rick_h_hmm, but if I check user credentials it shows the login. I don't want to check it again19:21
rick_h_ah, but it cancels any other routes so I should be good19:22
rick_h_it'll show the login and kill dispatch so never hit my stuff19:22
hatchyou want to put your callback call on line 682, right after this.show_login();19:23
hatchthat way if there are no credentials and it's showing the login then hide ths browser19:23
hatchat least it makes sense in my head19:24
hatchwhich has been wrong from time to time ;)19:24
rick_h_it seems dirty and another edge case for me to cover. Shouldn't subapps get killed on the login?19:24
rick_h_it seems crazy to need to call another method to force hidden in another corner case. 19:24
hatchwell your routes shouldn't be called if there are no credentials19:25
hatchso....19:25
hatchyeah19:25
hatchignore me19:25
rick_h_right, that's what I was thinking19:25
rick_h_cool, I'm happier now19:26
hatchand I feel dumber19:26
hatchTHANKS!!!19:26
hatchhaha19:26
rick_h_:P19:26
benjigary_poster: my internet connection is back from its brief hiatus19:31
gary_posterhey benji cool.  I asked earlier when I didn't realize that you were gone of we could postpone a half hour from the usual time.  that work?19:32
gary_poster*if19:32
benjigary_poster: yep, that's fine19:32
gary_posterok thanks benji19:32
rick_h_hatch: https://codereview.appspot.com/893704619:47
gary_posterbac I lost you19:47
hatchrick_h_: cool, does it work? ;)19:49
rick_h_hatch: :)19:49
rick_h_hatch: seems to work fine. Do wish multi-dispatch would run away as it hits it 10 times but works19:49
hatchlol.....working on it19:50
rick_h_hatch: if you get a sec for review appreciate it. 19:50
hatchyup doing it right now19:52
rick_h_hatch: ty19:52
rick_h_gary_poster: did you email luca over the landscape stuff from https://bugs.launchpad.net/juju-gui/+bug/1171531 ?19:55
_mup_Bug #1171531: Zoom control <charmbrowser> <juju-gui:New> <https://launchpad.net/bugs/1171531>19:55
gary_posterrick_h_, I talked to jovan about it.  He said to put the zoom control in the center.  sorry for not updating bug.  would you mind? :-P19:56
rick_h_not at all, will add note about moving zoom control to the center then. Thanks19:57
hatchrick_h_: review done20:00
rick_h_hatch: thanks, looking. 20:01
=== matsubara-afk is now known as matsubara
Makyojujugui: can I get one more review on https://codereview.appspot.com/8961044/ to help clear a card?20:12
* hatch will do it20:12
Makyo\o/20:13
hatchMakyo: done like dinnah20:16
gary_posterok benji, sorry, joining20:20
bachi hatch can i chat with you for a bit?20:20
* benji looks up the URL20:20
hatchbac: sure, couple mins, just need to get the dogs in20:22
bacok20:22
hatchok guichat20:24
hatch^ bac20:25
hatchgary_poster: bac: here was the tiny RTE I was talking about http://mindmup.github.io/bootstrap-wysiwyg/21:07
hatchI/we could likely convert that to YUI pretty easily21:08
hatchgary_poster: bac: There is also https://github.com/smugmug/yui-gallery/tree/master/src/sm-editor which is based on Y.Editor21:10
hatchdemo if it http://jsbin.com/abaqiq/221:11
rick_h_hatch: some replies sent your way. Will work on adding a functional test. Was having a hard time finding a good home for it/how to get at it but working on it21:22
hatchthanks - yeah functional tests are hard but I wanted to avoid the case where we release the app and noone noticed that breaking :)21:22
rick_h_hatch: no, you're right. I did look around but test_app seemed bare and tried to sneak by :P21:23
hatchhaha - nobody lets me sneak by so I'm just passing on the love rofl21:23
rick_h_hatch: ok, actually found an easy way to test yay. So code going through lbox now :)21:34
hatchawesome21:35
hatch!21:35
rick_h_hatch: if you can get a second to look over before EOD so I can get it through that'd be great please21:35
hatchyep as soon as it updates I'll do it21:35
rick_h_doh, linter fail21:35
rick_h_thanks, clears the way for me to work on making it the default after this :)21:36
hatchhaha - I have started to lint and make all tests pass before final commit so I don't get the last commit being "f'n linter" ;)21:37
hatchrick_h_: you're gona be irritated but you forgot to destroy app in your new test :)21:42
rick_h_hatch: lol, well crap. I was trying to use makeApp at first which auto destroyed21:42
rick_h_but it's nto in this closure so swapped out but missed that my bad21:42
rick_h_oh wtf, the tests before it app.destroy first?21:43
hatchafterEach has an app.destroy()21:44
rick_h_oh hatch there's an ap.destroy in afterEach21:44
hatchbut you created it with var21:44
hatch:)21:44
rick_h_ok, I can remove that :P21:44
hatchhaha - it's been LGTM'd21:44
hatchcharmstore is sure gona look good once it has some icons with some color21:45
hatchlike in the mockups21:45
rick_h_hatch: yea, they got some 10 icons in today but not through injest yet21:45
rick_h_hopefully tomorrow they'll start to show up21:45
rick_h_thanks for helping me out with this one today hatch 21:45
hatchanytime21:46
hatchthanks for the implementation :)21:46
rick_h_ah crap, test-prod failure now? wtf21:58
hatchcan anyone tell me why charm provides and requires would be undefined in a charm = db.charms.add()22:31
hatchahh nm got it....newly introduced byg22:45
hatchbug22:45
hatcharg so close to finishing this22:59
gary_posterhey hatch.  we explicitly don't want a RTE AFAIK, but I imagine bac already said that to you.  23:53
gary_posterIs Rick's recent branch supposed to fix the CI failure?23:53

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