rick_hhuwshimi: so thought about it and viewNavigate isn't needed. Home is just the hard coded path of either /sidebar or /fullscreen based on the isFullscreen param00:13
rick_hhuwshimi: so I'd just make the template have those links and not try to fire custom stuff00:13
rick_hthat is what we mean by "Home" right?00:14
huwshimirick_h: Well, the sidebar needs to be able to navigate back from the search results or a category without closing the charm panel...00:18
rick_hbcsaller: still around?00:20
bcsallerrick_h: yes00:20
rick_hbcsaller: looking at your branch with the icon, does that work? We never have an icon value so wondering where it's binding to/from?00:20
rick_hhuwshimi: right, so on search results, home would purely be a link to /sidebar or /fullscreen. 00:21
bcsallerThe ghost inspector sets it from the charm data onto the service so it works in that single case 00:21
rick_hhuwshimi: or what is the 'charm panel'?00:21
bcsallerwe have notes that we need a more general solution00:22
huwshimirick_h: I think it would need to be "/sidebar/precise/ceph-12/"00:22
rick_hbcsaller: right, I'm with you on notes. Known issue and Makyo hit it with his work to include an icon but wasn't sure how this fit00:22
rick_hhuwshimi: ah, gotcha.00:22
rick_hbcsaller: ok, so you're getting it from Makyo's work then which is getting it from the Store if it has it. Cool00:22
bcsallerrick_h: yes, its based on that work00:23
rick_hbcsaller: cool, that helps. Thanks.00:23
rick_hhuwshimi: so it sounds like the only thing we're changing is clearing search. So yea, a viewNavigate with a change object of {search: false; query: {};hash: undefined} would reset it00:28
rick_hor query: undefined I guess is simpler00:29
huwshimirick_h: Ah, that sounds promising.00:29
rick_hthat would leave the viewmode and the charmid alone which are the two parts you want to keep consistent00:30
huwshimirick_h: I was thinking of building this inside charm-search.js. But inside that file, and its corresponding template it doesn't have the isFullscreen variable. Any idea how I might get access there?00:32
rick_hhuwshimi: sec00:33
rick_hhuwshimi: I was thinking it would be part of the sidebar/fullscreen views and that has it. It could go in subapp/views/view.js 00:33
rick_hand be shared between them. 00:33
rick_hNow I guess, when does "home" show up? Is it always there? Or only during a search?00:34
huwshimirick_h: It only apears during a search (including viewing a category), or if you're viewing a charm in full-screen mode00:34
rick_hhuwshimi: who's rendering the html link I guess?00:34
huwshimirick_h: The HTML needs to exist in that header (hence me wanting to put it in charm-search.js)00:35
rick_hhuwshimi: oh, in the charm-search widget?00:35
huwshimirick_h: Yeah00:35
rick_hhuwshimi: ah, so the *right* way to do that then is to create a new event in the widget, like the "search changed" event now that's a "GO_HOME" event. Then anyone that creates a new Search() widget needs to bind to that event and fire the viewNavigate when it happs. 00:36
rick_hhuwshimi: just like the SEARCH_CHANGED event works now00:36
huwshimirick_h: OK.00:36
rick_hhuwshimi: hmmm...so then the trick is that you need to supress that link from the widget when it's not a search/charm details view. :/00:37
huwshimirick_h: I'll still need to know when to hide/show that control, but I guess I can do that in the parent view00:37
rick_hhuwshimi: yea, thinking. We've not really had conditional stuff like that. 00:38
huwshimirick_h: Yeah, there are a few other cases I'm going to need to do similar stuff to change headers etc. depending on the particular view00:38
rick_hhuwshimi: ok, so the subapp/browser/views/view.js has a charmID attribute. So if that's set you're looking at a charm. 00:38
huwshimi(stuff outside the view content)00:39
rick_hhuwshimi: that gets you part way there. 00:39
rick_hhuwshimi: so you'd need to add a flag in the _renderSearchWidget based on if there's a charmID...and then the other one is if you're in search mode or not. 00:39
rick_hhuwshimi: well, you want 1) "isFullscreen and charmID is set...show the link"00:39
rick_h2) "Search results are being shown"00:40
rick_h1) you're set on00:40
rick_h2) I'm looking00:40
rick_hhuwshimi: you can check the filters object. 00:40
rick_hhuwshimi: the same View has a filters ATTR, if those are set to something you should be in a search. 00:40
rick_hhuwshimi: the only problem with that is we still have default filters set. We've not removed them in a branch yet00:41
rick_hhuwshimi: so those have to go as well. 00:41
rick_hthen if any fitler is set, you're doing a search. 00:42
rick_hhuwshimi: because in the filters is the query being search, or the category selected when viewing a category00:42
huwshimirick_h: This is looking like a job that would probably take a day. So might not be feasible once I've gotten through my other list of things...00:42
rick_hhuwshimi: k, yea. It's not as simple as I'd hoped because of the new conditional rules. 00:43
rick_hhuwshimi: the other thing is to always show a home link to start out00:43
rick_hand add the conditions as follow ups00:43
huwshimirick_h: I'm guessing the search/categories bug I filed is probably not trivial either?00:44
rick_hhuwshimi: I don't think it's hard. I think the category links need a more specific viewNavigate change object. 00:46
rick_hhuwshimi: it's not clearing out enough stuff. 00:46
huwshimirick_h: OK, I might take a look at that later if I get a chance.00:47
rick_hhuwshimi: yea, basically in browser/browser.js in the _getStateUrl it's merging the filters where it sounds like it should replace the data. 00:52
rick_hline 81, this._filter.update(change.filter);00:52
rick_hit wasn an update so that as you changed yuor checkboxes in the filter it would add/subtract one item at a time. 00:53
benjibac: I have finished walking through the email quagmire and am soon to reach the caffeine castle.  Once that journey is complete I can help you battle the CSS dragon.12:29
bacbenji: the css quagmire was holding up the branch you reviewed yesterday.  the problem was a setting being inherited from bootstrap.12:30
bacbenji: thanks, though.12:30
benjioh, good!12:31
benjiI'll have to find a different dragon then.12:31
gary_posterbenji, other dragons available.  ping me :-)12:32
benjiwill do12:33
gary_posterplease review and land Huw's most recent branch ASAP: https://codereview.appspot.com/11209043/12:36
baci'm reviewing the above12:40
benjiI'll do a review too.12:41
benjijujugui: it is known that dragging of service inspectors is completely broken?12:58
benjiI thought it was just the branch I am reviewing, but it happens on trunk too12:58
gary_posterluca__, rick_h I am swamped with meetings this morning.  I just sent out an email about the priorities for today.  Please review.  Also, could you please discuss together the home/back button issue and the search interaction issue between categories and text?  For each of these I would like an expedient solution we can implement today in an hour and land, which may require discussion.12:59
gary_posterrick_h, luca, I think these experiences are broken ATM, and my goal is to not be broken for this release.13:00
gary_posterbenji, service inspectors are not draggable in Luca's design.  We may want them to be draggable to demo variation ideas from MS.  "Flexible"13:00
bacbenji: http://uistage.jujucharms.com:8086/:flags:/serviceInspector/ is really borked13:00
gary_posteris goal13:00
luca__gary_poster: sure, I'll get in touch with Rick13:01
benjigary_poster: in that case they are broken on the trunk because they are partially draggable ;)13:01
gary_posterok thanks13:01
bacbenji: do you see the inspectors being completely broken on huw's branch?  no content for most 'tabs'.13:11
rick_hgary_poster: will do, I'm afk a little bit this morning but will try to meet up with luca13:12
benjibac: nope, I didn't click on any of the tabs; let me look real quick13:12
benjibac: yep, very broken13:12
bacbenji: i'm unsure what we're supposed to be looking for.  luca__ do you know?13:12
rick_hsinzui: I'm still getting setup at the dealer here and might not make the standup. My branch for autocomplete is updated ready for final review by you and hatch. 13:20
rick_hsinzui: will try to get into the call as there's some important stuff flying around atm. 13:20
sinzuithank you rick_h13:24
hatchbenji: bac inspectors are not supposed to be dragable13:28
hatch8086 is also working as expected here13:29
bachatch: some content drags and snaps back.  is that expected?13:29
benjibac: did you find out if the empty inspector tab issue was introduced on this branch or pre-existing in trunk?  If not I can test it quickly.13:31
bacbenji: i did not yet13:31
benjibac: I'll take a look.13:31
hatchbenji: it's a known issue - it happens when you open the inspector too quick after deployment13:31
benjiah, ok13:32
benjirace conditions are fun13:32
rick_hrun faster!13:32
luca__gary_poster: me and Ale have just had a meeting with MS and he's very happy with the inspector13:33
luca__gary_poster: there are some content changes but overall he was pretty positive13:33
gary_posterluca__, awesome :-)13:33
luca__gary_poster: said the new Juju looks f'ing amazing13:33
gary_posterlol :-)13:33
hatchbac: the dragging is not expected - that's not a known bug :)13:34
luca__bac: gary_poster do I need to look at anything?13:34
bacluca__: benji and i are looking at huw's branch https://codereview.appspot.com/1120904313:34
bacluca__: the description references some visual changes you requested but doesn't state what they are.  if you know and can verify the fix that would be helpful13:35
luca__bac: I can try to bar it13:35
luca__bac: bzr13:36
luca__bac: checking it now against the list of bugs I sent off13:38
gary_posterluca__, I need to have some help from you with the expedient plans for home/back and search interaction, but I'm swamped and rick_h is afk.  I will ping asap, 'cause these are biggest risks remaining IMO13:39
gary_posterI'm swamped on calls I mean13:39
luca__gary_poster: ok, I'm available most of the afternoon13:39
gary_posterthank you luca__ 13:39
rick_hluca__: gary_poster hopefully I can be free in the next hour13:39
gary_posterthanks rick_h 13:39
luca__bac: gary_poster I can confirm that all of the bugs I submitted yesterday have been fixed in visual-update-5 apart from the 2 that have been stated in gary's email13:41
gary_posterluca__, great, thank you!13:41
gary_posterbac could you land when you are satisfied please?13:41
bacgary_poster: if there are small issues with the branch you want me to just fix them then land, i assume.13:43
gary_posterbac yes please13:43
baccool.  i'll wait for benji's input (if it isn't there yet)13:43
benjibac: I just posted (not LGTM, unfortunately; big problem with the test runner)13:44
bacbenji: rt13:44
benjibac: "rt"?13:44
bacbenji: 'roger that'.  i think i've referenced my helicopter pilot fantasy before?13:45
benjigary_poster: I'm ready for a dragon quest.  Should I take up that NPM cache issue?13:57
gary_posterbac, did you mention to luca about whatever different assets we need for the zoom widget?13:57
gary_posterbenji no one sec13:57
gary_posterbenji, oscon tasks have highest priority today.  I sent email and have new "urgent" kanban section.  If bac could use help in fixing Huw's branch, that would be best use of time.  Otherwise see urgent section of kanban14:01
bacgary_poster: what we have may work, though limited to a fixed size slider.14:01
benjigary_poster: will do14:02
benjibac: I described the problem with the body styling and test runner in the review, but I could easily work up a patch.14:02
bacgary_poster: this page has an example using very similar assets. http://yuilibrary.com/yui/docs/slider/slider-from-markup.html  the issue is the lack of separate images for the rail and end caps14:02
bacbenji: i'll ping you if that'll be more efficient14:03
gary_posterbac, ack. luca, ^^^ is that enough info to ask jaime for what we need?  AIUI, as a convenience it would be great if he could simply slice the images per the example above14:04
gary_postersorry, luca__ ^^^14:04
bacgary_poster: there is also the possible issue of the thumb image being a sprite with embedded shadow.  unsure how much of this YUI skinning should be pushed to jamie14:05
gary_posterbac, when you say "unsure how much" you mean we possibly/probably should not push it to him, because we will be better able to identify what we need?14:05
gary_posterand it is simply a matter of slicing?14:06
bacgary_poster: i'm saying YUI skinning has very specific requirements and it is unclear who should ferret them out.14:06
gary_posterbac ok.  luca__ scratch that scrollbar request for now :-) thank you14:08
jcsackettjujugui: head's up, i have an appliance repair dude showing up at 11, so i'll be absent for the retrospective. i don't see any cards i have a strong feeling about, so you shouldn't feel my loss very keenly. :-P14:09
* benji feels a vague sense of loss.14:09
* jcsackett laughs14:11
benjigary_poster: I would ideally tackle one of the two criticals that have yet to be claimed.  They both list needing to be discussed, have they been?14:15
bacbenji: how do i reproduce the test failure you saw?14:16
benjibac: its not so much a test failure, but a failure of the tests.  Run the tests, wait for them to end, try to scroll up, feel frustrated and listless.14:17
bacbenji: i ran 'make test-server' and i can scroll around a-ok in chrome.  what am i missing?14:18
bacbenji: but i do feel listless14:18
benjibac: let me replicate it again, one second14:18
luca__gary_poster: was dealing with something, is there anything you need?14:20
hatchluca__: I think the 'reviewed' icons look like 'favourite' buttons14:21
hatchjust sayin :)14:21
luca__hatch: your not the only one to say that, we are testing it on monday and tuesday with 10 users from the inductry14:21
hatchrick_h: and I decided a better symbol would be two lions fighting over a steak with a shark jumping overhead14:22
luca__hatch: +1 I'll get jamie to mock something up :D14:22
hatchso huw's branch added overflow: hidden to the body, but something added overflow: auto to the element14:24
hatchcausing the scroll bar to re-appear14:24
gary_posterbenji, they have not yet been discussed.  not ready14:25
gary_posterluca__, nothing I need right now thanks, till Rick is ready14:26
rick_hluca__: ping for when you get time to chat14:26
gary_posterooh ooh!14:26
rick_hgary_poster: too since it makes him so happy :)14:26
gary_posterI will join if I can but I have another call for now.  Very excited about possibility of getting a reasonable quick resolution to those issues though :-)14:27
benjibac: I can't reproduce the error now.  I don't know what happened.  So you can disregard the patch part of that comment, but the part about removing the now-redundant code still applies.14:27
rick_hhatch: would really appreciate if you can re-look at the AC branch this morning so I can get that in/up. 14:27
bacbenji: done14:27
rick_hgary_poster: heh, don't get your hopes way way way way up :P14:28
rick_hsinzui: while I wait to get some time with luca__ did you want to chat for a sec for anything hangout related?14:29
luca__rick_h: I'm free you need anything14:30
rick_hluca__: ok, will send an invite one sec14:30
luca__rick_h: hatch is this better for the reviewed icon? https://docs.google.com/a/canonical.com/file/d/0B7XG_QBXNwY1WWwyd2cyLWtrWWc/edit?usp=sharing14:30
gary_posterluca__, LOL!14:30
rick_hluca__: https://plus.google.com/hangouts/_/0c7a8746abe1e765bf1f06cd256f317702da126d?hl=en14:31
gary_posterantdillon hey https://plus.google.com/hangouts/_/b00dc841a3b70c4f357210452e531a462a28841a14:31
hatchdo it do it!!!!14:31
rick_hluca__: O...M...G on the icon14:31
* gary_poster still giggling14:31
antdillongary_poster, Im there14:32
bacjujugui: what are the steps to use lbox to submit another's branch?  how does it link my local branch to huw's MP?14:32
hatchbac: are you going to land huw's branch that you NLGTM'd?14:33
Makyobac, -adopt, I think14:33
bachatch: catch up man.  I changed my vote.14:33
hatchohh lol ok well one second14:33
bacMakyo: is that really all that is requried14:33
hatchthere is a breaking change in my QA14:34
Makyobac, no clue :D14:34
Makyobac, the docs for lbox are...well...14:34
hatchwhat...there are docs for lbox? ;)14:34
bacMakyo: i branched trunk into a 'review' branch then merged his branch in.  can lbox really link it all together?14:34
bac'lbox submit -h' isn't too clear on what happens.  guess i can try.14:35
bachatch: so i'm waiting on a review from you?14:35
hatchbac: I just qa'd see my comments as one thing needs to be fixed before it lands in trunk14:35
bachatch: the overflow issue is resolved per benji's review suggestion14:36
Makyobac, I've never had the chance to try, I juist know the option exists.14:36
bacMakyo: ditto.  i'll report back.14:37
hatchbac: ohh ok - I thought that would do the exact opposite of what we wanted14:37
hatchthe bigger issue is the small charm config when not under the flag14:37
bachatch: i don't understand what you mean.  wanna chat?  might be fastest14:38
hatchin guichat14:39
bachatch: rejoining14:44
hatch-mobilebac sorry it all crashed, rebooting14:44
rick_hbac: so when I did his stuff I just created a branch under my name, merged his stuff in, pushed to my account, and re-lbox'd it references his codereview url in a self-LGTM as the dicussion was over there and then submitted it14:46
rick_hsinzui: ok, off that call, shifting up what I'm working on based on emails and such. Let me know if we need to sync up. 14:46
bacrick_h: sweet, thanks for the details14:46
bacbut now i'm super curious about -adopt14:47
sinzuirick_h, I see you have already claimed one of the important issues on the board14:47
rick_hsinzui: so I put back those as they're considered lower priority than fixing the home button according to luca and gary_poster's email14:47
sinzuiyep, I saw you do that14:48
rick_hthe other things lower on the priority list are all dependant on removing the filters UI and the default filter settings. So they're nice to have but will be more work than EOD will allow14:48
sinzuikanban doesn't tell you who is watching14:48
* rick_h feels stalked14:48
hatchbac: ok I'm back now but I have to set my env back up - were you able to reproduce with mysql?14:49
bachatch: no, it looks good to me14:49
hatchok one minute I'll set this back up14:50
Makyojujugui call in 10 kanban now14:50
bcsallerhopefully my lbox submit finishes before then :(14:50
hatchjsbin is down :(14:51
gary_posterrick_h, hey, you still talking?14:52
rick_hgary_poster: done talking atm14:52
antdillongary_poster, Oh forgot to say, I am off on Monday (its my stag do this weekend) I will be starting on Tues14:52
gary_posterantdillon, oh cool, congrats and have fun. :-) see you Tuesday then.  Maybe we'll have more for you to work with by then :-)14:53
gary_posterrick_h, mind joining guichat real quick to summarize plans for me?14:54
rick_hgary_poster: yep one sec14:54
bachatch: you still rebooting?14:54
hatchbac: yup just booting up the env14:54
hatch(i really gota setup a script for this)14:54
rick_hgary_poster: joining14:57
gary_posterjujugui call in 314:57
hatchbac take a look at my screen for the bug15:00
rick_hhatch: sinzui https://codereview.appspot.com/11126043/ is updated, comments replied to, tests updated. let me know if there's anything else needed. 15:03
bacbenji: you still have huw's branch marked bad.  do you want another pass or just land with your dissent?15:04
benjibac: I'll re-mark now15:04
bacbut it'll be a lie!  :)15:04
sinzuirick_h, criss-crossed. You got my LGTM15:04
bacgary_poster: that sounds like a good check for 'charm proof'15:07
adeuringabentley: could you please have another look at https://code.launchpad.net/~adeuring/charmworld/get-file-content-from-bzr/+merge/173987 ?15:21
abentleyadeuring: Sure.15:22
abentleyadeuring: I don't understand "I chose to return the branch's directory instead of the branch itself because the directory is needed as an entry in the charm_data dictionary during test setup".  You can get the directory from the branch, as I explained.15:24
adeuringabentley: I know, but that's not what is needed in tests -- it's the directory15:24
abentleyadeuring: So you're saying it was for convenience?15:26
adeuringabentley: yes15:27
abentleyadeuring: at 103, you are patching BZR_EMAIL into the environment, but bzr_isolation should do that for you.15:28
adeuringabentley: ouch, forgot hat... I'll fix it15:28
abentleyadeuring: r=me.15:44
adeuringabentley: thanks15:44
sinzuiBugger. webops are creating unversioned tarballs.16:01
sinzuigary_poster,  or bac: Webops asked that we add an upgrade hook  to the juju-gui stable charm to handle cases where we want to deploy a new release of the app. This seams easy enough, but now I see the webops scripts create unversioned tarballs. There is never a config change to indicate to the backend to unpack the tarball.16:04
gary_postersinzui, the tarballs are always named the same?16:05
bacyes, :(16:05
sinzuiI see other problems with their script16:05
gary_postersinzui, easiest solution is to add revid to the tarball name16:05
gary_posterif unversioned16:05
sinzuithe source is hardcoded. They need to edit the script each deploy.16:05
sinzuiI'll propose a replacement script that makes it easy for them and us to do a deploy of just the app16:06
gary_postersinzui, cool, thanks.  is this documented on your wiki page?16:07
sinzuiI got a pastebin copy of the script. https://pastebin.canonical.com/94228/16:08
sinzuiThe version was changed yesterday to my request, but the script will need to be edited again for the next request :(16:08
rick_hluca__: around?16:09
gary_posterbcsaller, could you let me know what you think about in terms of difficulty of getting those export issues addressed?16:12
gary_posterwhen you've had time, of course16:13
bcsallergary_poster: I'm reviewing them now, there are some questions, like qualified charm name, precise/mysql-1 or cs:precise/mysql, including rev, etc? Just not sure yet, making notes on some of the bugs16:18
benjiMakyo: what browser were you using when you saw bug 1200412?16:21
_mup_Bug #1200412: drag-and-drop: dropping a service on a non-drop target leaves the token appended to the body <juju-gui:In Progress by benji> <https://launchpad.net/bugs/1200412>16:21
Makyobenji, Chrome v2816:21
benjiMakyo: hmm, I'm running the same and I can't reproduce the symptoms listed in the bug.16:24
rick_hluca__: take two, would love a sec to chat. 16:24
luca__rick_h: sure16:25
luca__rick_h: got a hangout?16:25
rick_hcouple of ? for you luca. Hangout?16:25
Makyobenji, hmm, quick chat so I can screenshare?16:28
benjiMakyo: sure, guichat is open16:28
hatch_sorry was dc'd16:41
hatch_reposting - bcsaller: re your 250ms comment - I'm OK with either approach just curious if you were planning on changing to rate limiting vs debounce or leave it for now?16:42
bcsallerhatch_: I left it for now, I think it will be fine the way it is, but we can keep an eye on it16:42
hatch_alright sounds good :)16:42
hatch_why do I have to add my ssh key to be able to pull from our bzr repo...did it get locked down by accident? :)16:44
bcsallerhazmat: ping16:55
hazmatbcsaller, pong16:56
bcsallerhazmat: can you check the comment in https://bugs.launchpad.net/juju-gui/+bug/1200625 and see if that export matches your needs16:56
_mup_Bug #1200625: gui export is exporting invalid config options / not respecting type when serializng <juju-gui:New> <https://launchpad.net/bugs/1200625>16:56
=== hatch_ is now known as hatch
bcsallerthat includes service output, annotations and config changes16:56
hazmati was going against uistage.jujucharms.com:808616:56
hazmatmaybe that's not on a cron16:56
hazmatbcsaller, might be some noise then.. your comment export looks good to me,16:57
* hazmat checks trunk16:57
bcsallerhazmat: that export is from my current branch, I didn't change the options handling but the other things were changed, I have to propose this16:57
bcsallerif it looks good then it should be ready16:58
bcsallerbut the options handling seemed sane out of the box16:58
hazmatbcsaller, the yes/no thing correspond to booleans in yaml17:00
hazmatbut the type is a string17:00
hazmatie when that gets passed to juju core it will barf on debug setting17:01
bcsallersounds like an import error ;)17:01
hazmatsince it per the charm config schema it should be a string, but comes in as a bool17:01
hazmatnot really17:01
hazmatits two bugs17:01
hazmatsorry phone call17:01
hazmatexporting default 17:01
=== matsubara is now known as matsubara-lunch
bcsallerits going out as a string here, the parse side translates it to a bool in implicit conversion which is wrong 17:02
bcsallerwait, you don't want me to export if its the default value on the charm? more work but I can do that 17:03
sinzuigary_poster, http://comingsoon.jujucharms.com/ is up17:08
gary_posterawesome thanks sinzui!17:08
gary_posterhey luca, you really here? :-)17:09
lucagary_poster: just leaving17:09
lucagary_poster: if you send me an email ill pick it up17:10
gary_postercool thanks luca ttyl17:10
hatchrick_h: https://bugs.launchpad.net/charmworld/+bug/120070417:23
_mup_Bug #1200704: charm search autocomplete has slow options request times <charmworld:New> <https://launchpad.net/bugs/1200704>17:23
rick_hhatch: cool, will put on the board. Thanks. 17:23
hatchrick_h: I LGTM'd the branch pending the css fix as everything else looked good - the clickoutside if you can't fix we can follow-up with a fix on that17:27
rick_hhatch: ok, will look into it in a little bit. Almost done with this home branch. A 'testing' we shall go...a testing we shall go...17:28
rick_hcurses, luca left17:28
hatchhigh ho the dairy o a testing we shall go17:28
hatchnot sure what a dairy o is17:29
hatchit's probably something horrible17:29
rick_hgary_poster: I've got the home link branch up here. lp:~rharding/juju-gui/home-link I was hoping to get luca to look before he left but :( 17:29
rick_hgary_poster: I also found a decent fix for the search/category issue and have fixed it in there. I'm working on tests to support the changes but if anyone wants to QA in luca's place I'd love to hear it's what's expected.17:30
gary_posterrick_h, home link: thank you!  shoot him a mail.  he said he would be checking.  I'm sending him a couple too.  I'll look soon too, though if you are happy with it then I'm +1 on starting the review process.  search/category: even more awesome! I guess I'm the only one who can at this point unless you get luca's reply.  17:31
gary_posterI will do so asap, though I have two emails I'm trying to plow through before then17:32
rick_hgary_poster: ok, will shoot him an email. Tests will add a little time but trying to get it going and up for review asap. 17:32
gary_postercool thanks you!17:32
rick_hgary_poster: understood, just a heads up on where we're at and such. 17:32
hatchrofl https://twitter.com/horse_yui/status/35574193995554406517:37
gary_posterhey jcsackett , one of our goals that I'd forgotten about was that, when on jujucharms.com, instead of seeing "log out," you would see ...something else that I forgot.  A link to somewhere.  Did you do that as part of your default view work?17:45
rick_htime to acquire food, brb17:45
arosalesgary_poster, could you or bac add additional details to https://bugs.launchpad.net/charm-tools/+bug/1200713 when you have an opportunity17:51
_mup_Bug #1200713: Charm Proof: Ensure the default configs match the defined type <Juju Charm Tools:New> <https://launchpad.net/bugs/1200713>17:51
gary_posterarosales, bac, added18:00
=== matsubara-lunch is now known as matsubara
arosalesgary_poster thank you.18:05
hazmatbcsaller, yes re default18:06
bacarosales: done too18:06
bcsallerhazmat: thats fine but it only masks this issue, doesn't prevent it18:06
hazmatbcsaller, re serialization in general, it would be good to cast to the specified config type as well. arguably its  a bug in core and one in the charm18:06
bcsalleryeah, re core18:07
hatchbcsaller: I see when you added the icon to the inspector you used the databinding for the header information18:12
hatchthat information is not supposed to be in a viewlet18:12
hatchcan you think of a way to bind to the fields without requirng a viewlet?18:12
bcsallerhatch: VC owns databinding, its can either add a slot for header with the viewlet or add the databinding with addBinding directly 18:13
bcsallerI think a slot and a viewlet make sense, its just a slot we always show18:13
hatchyeah slot probably does make the most sense as far as separation of concerns is.....concerned18:16
hatchwill do that, thx18:16
benjijujugui: small branch to fix a drag and drop issue: https://codereview.appspot.com/1113804418:18
bacbenji: i'll do one18:18
hatchbenji: got it18:18
hatchposition fixed on the body eh? heh18:19
hatchI foresee issues in the future18:19
bacbenji: why is that overflow still in service.js?  it should've been removed in trunk18:20
benjibac: it may have been a race with me branching.  I should have merged in trunk before proposing.  I'll do that now.18:20
bacbenji: doesn't lbox produce a diff against trunk?18:21
baci mean, you shouldn't have had to do anything.18:21
benjiI'm re-proposing the merged branch now.18:22
hatchbenji: I'm doing a QA, which revno are you on?18:23
hatchjust to make sure I have the right one18:23
Makyolbox produces a diff against :parent, which points to a revision of trunk.18:23
MakyoThat way new code doesn't show up as deleted.18:23
benjihatch: 824; it's still in the process of proposing18:24
benjihatch: ok, it's finished proposing18:24
hatchgot it thx18:25
bacbenji: cool, your diff is much smaller now18:25
hatchbenji: QA fails spectacularly18:27
benjihatch: how so?18:27
hatchdrag and drop (so fixed gets applied)18:27
hatchthen d click on the inspector18:27
hatchdclick on the service18:27
hatchto view the fullpage inspector18:27
hatchalso the ghost inspector panel is way off position now18:28
Makyohatch, ghost inspector is way off position for me in trunk.18:28
benjihatch: double-click on the ghost inspector?  That doesn't appear to do anything.18:29
MakyoEr.. as in clipped at the bottom; I'm not actually following along, sorry.  I'll keep my yap shut if that's not the case.18:29
hatchsorry I meant dclick on the service after deployed18:30
hatch^ benji18:30
hatchMakyo: the ginspector is probably 100px down from the top nav18:30
hatchsimilar on trunk?18:30
hatchit looks proper to me on trunk18:30
benjihatch: gotcha, yep that's a bug.  I think I have a fix.18:31
Makyo~20px here, but it's clipping over the bottom bar.18:31
hatchMakyo: ahh yeah that happens to me if I shrink my window too18:31
hatchthe overlap18:31
Makyohatch, running full screen full HD here.18:31
hatchbenji: I'm concerned that applying position fixed to the body is going to cause us more issues18:31
hatchas fixed pulls it out of the flow of the page18:31
hatchwhich can cause render performance issues too18:32
hatch(not sure if they apply here)18:32
benjihatch: yep, my fix is going to be to remove the fixed style after the drop, so it will only be applied while dragging18:32
benjiI'm open to other ways to keep the window from scrolling during a drag, but I haven't been able to find a sane one18:32
hatchbenji: so why doesn't the overflow: hidden on the body work? what am I missing? It appears to work as expected here18:34
benjihatch: because if you drag down to the bottom (or right) of the window, the body will scroll18:35
hatchhmm cannot repro - trying in Ubuntu18:35
MakyoI can repro.  No scrollbars, just works the same way as other hidden, where if you select and drag.18:36
hatchI see now18:37
hatchman this DD is suck a mess18:40
hatch(not your fault benji) :)18:40
hatchMakyo: how are you handling dropping off the drop target?18:42
* hatch is curious18:42
MakyoNot sure what you mean.18:43
gary_posternot sure how I left18:43
gary_posterjcsackett, you around?18:43
hatchwell when you drop a service off the drop zone, to remove the thumbnail18:43
hatchI thought you were working on that...18:44
hatchI could have been wrong :)18:44
Makyohatch, no, working on left hand panel for unit view.18:44
hatchgary_poster: that happened to me 3 times today heh18:44
gary_posterhatch benji was doing that18:44
hatchohhh ok18:44
bacbenji: done18:46
bacgary_poster: looking for a task.  i can start the slider skinning but don't think it is landable today.  thoughts?18:46
gary_posterlooking bac18:47
hatchbac: QA'd ok?18:47
hatchshould I re-make my branch?18:47
bachatch: yes.  i reproduced the problem on uistage but not in benji's branch.  did i miss something?18:47
hatchhmm intersting - the full screen inspector views are all broken for me18:48
hatchwill try clearing cache and remaking18:48
bachatch, benji: the problem i saw on uistage was the body scroll up leaving a void.  doesn't happen with fix.18:48
hatchohh so you didn't QA the rest :)18:48
* bac thinks we could all write better merge proposals18:48
gary_posterhatch, bac, Makyo any of following would be good to do:18:49
* bac describing the broken behavior18:49
hatchbac: the merge proposal is fine - but it introduces a regression18:49
gary_poster- QA Rick's branch (I can do that now I think)18:49
benjigary_poster: I wasn't working on the garbage icon; I can take a look.18:49
hatchso that's why we need to do a more thorough QA18:49
gary_poster- sticky header icon18:49
bachatch: by 'better mp' i mean explicitly state the problem for people doing qa18:49
gary_posterbenji, bug 1200412?18:49
_mup_Bug #1200412: drag-and-drop: dropping a service on a non-drop target leaves the token appended to the body <juju-gui:In Progress by benji> <https://launchpad.net/bugs/1200412>18:49
gary_postermaybe I misunderstood the conversation, sorry18:50
gary_poster- charm title ellipsis18:50
gary_poster- CSS transitions18:50
bachatch: but you're right, i tried to exercise his problem but didn't do full qa of the app18:50
gary_posterbcsaller, do you have the feedback you need from Kapil about the export issues?18:50
MakyoFor this next week, gary_poster ?  Don't want to switch now.18:50
bcsallergary_poster: I think so yes, it looks like we want some special processing around service options which is in progress, the rest already appears correct in my branch18:51
hatchbac: this brings up a good question - should the reviewers be required to do that qa? or should the author of the proposal?18:51
gary_posterMakyo, ack.  sorry, I was not clear that we could/should put aside inspector tasks if you've only just started.  My fault.  as you were--keep doing what you are doing18:51
bcsallergary_poster: even with the special processing for options though the real issue will still be in core18:52
rick_habentley: ignore my comment on that bug, I mis-read it at first. 18:52
benjigary_poster: nope that bug is about scrolling; well that's the way I interpreted it, the scrolling breaks the app and is ugly, the left over icon isn't worth worrying about in my opinion.  People could drag hundreds of icons into non-target parts of the app and not be an issue18:52
bachatch: author should but he can't be trusted so the reviewers should.18:52
abentleyrick_h: sure thing.18:52
hatchso after 3 qa's we better not be introducing any regressions lol18:52
gary_posterbcsaller, ok cool thank you18:52
gary_posterhatch, bac, not sure each review should be full exploratory review18:53
gary_posterexploratory testing I mea18:53
bacgary_poster: the ellipsis card says you have some secret sauce18:53
gary_posterbac, barely. was hoping luca could provide.  However...18:53
bacgary_poster: i just meant to say if anyone is going to do it, it should be one of the reviewers.  as a developer i am a very tender QA person so i am not to be trusted, and i suspect others may have the same affliction.18:54
gary_posterbac ack18:54
hatch*note to self - get bac to review my horrible code*18:54
bachatch: no, i'm very thorough with other's code!18:55
hatchohhhh I see, you just let yourself off easy18:55
gary_posterhatch, could you review https://codereview.appspot.com/11138044/ pls?  want to make sure you don't have secret sauce about deleting elements there 18:55
hatchgary_poster: I did, it failed QA , benji is working on the fix now18:56
gary_posterhatch ok thaks18:56
gary_posterrietveld is weird18:56
hatchweird, or broken? ;)18:56
hatchI swore that it used to tell me which areas I had commented on without refreshing18:57
hatchthen one day it just stopped18:57
hatch*use tests to write code - who needs docs anyways!*18:59
benjihatch: the branch has been updated19:02
sinzuithank you gary_poster, I did my Mp in the wrong order...I create the reitveld review request afterwards. I will remove it since the Lp MP was sufficient19:06
bacgary_poster: so is the goal to have charm title in charm browser be a single line with ellipsis or two lines with ellipsis? if the former then 'postgresql-psql' would be one affected.19:06
gary_postercool sinzui19:07
gary_posterbac single I think19:07
bacgary_poster: k19:07
hatchbenji: nogo (sorry :( ) see review for the comment19:09
benjihatch: thanks19:09
benjiI'm going to eat a late lunch now.19:09
hatchI swear I'm not trying to be a pain19:09
hatchbenji: alright - I am meeting some friends for lunch in about 30mins so just push up wheever you get to and I'll get to it as soon as I get back19:10
hatchpending of course you find a solution19:10
* hatch contemplates making Benji a drag and drop dart board19:13
rick_hhatch: bcsaller jcsackett anyone up for a review please? https://codereview.appspot.com/1122404419:24
rick_hgary_poster: ^^ that's the home stuff19:24
hatchOh Kay19:25
gary_posterthanks rick_h 19:25
rick_hsinzui: I'm not 100% sure I can get autocomplete landed. We've hit yui3-skin-sam issues and a bug in clicking off AC suggestions. Will work on that here, but not sure how long I've got. 19:25
rick_hyay long running branches that pick up more issues the longer it lives19:25
sinzuiindeed they do.19:25
gary_posterrick_h, sinzui I've seen a few core charm browser issues too; and have been trying to ask jcsackett about a change I had forgotten about19:25
sinzuirick_h, I think you are EoD19:25
gary_posteryeah I was wondering about that :-)19:26
rick_hsinzui: I've got some extra time today since I was afk this morning. 19:26
gary_posteroh cool19:26
rick_hgary_poster: what's that?19:26
jcsackettgary_poster: what change is that?19:26
gary_posteryay jcsackett!  ok first you19:26
gary_posterwe were supposed to make it so that, on jujucharms.com, logout button was replaced with [something I forget, maybe a link].  Was that part of your changes?19:27
gary_posterI mean, the changes to set default view and stuff19:27
sinzuirick_h, we let's emphasise quality and stability. I would rather postpone autocomplete if the effort to get it production ready would delay of issues we want in production early net week19:27
rick_hsinzui: rgr, just wanted to give you a heads up. If the css issues are lighter than they look, and a workaround for the d3/svg/click bubbling is available in the next hour or so I'll try to get it landed. It's qa'd ok aside from those 19:28
rick_hand I understand home/etc was higher priority so will get that landed before EOD if possible19:28
gary_posterrick_h, ok you now :-)19:28
gary_posterI will qa in a sec but these are core charm browser issues I found.19:28
rick_hgary_poster: woot19:28
gary_poster(1) go to http://uistage.jujucharms.com:8086/fullscreen/search/precise/cinder-8/19:29
gary_poster(2) click on Browse19:29
gary_posterI get "Failed to load search results.19:29
gary_posterCharm API server did not respond:19:29
gary_posterAnother one...19:29
rick_hgary_poster: hold up, I can't dupe19:29
jcsackettgary_poster: no, that wasn't communicated in the card i had. all i've ended up doing was setup the option for fullscreen to be default and expose that as a config for the charm.19:29
gary_posterrick_h, cool, will clear cache19:30
rick_hgary_poster: clicking on browser does nothing for me. 19:30
rick_hgary_poster: did you mean search? or something else?19:30
gary_posterrick_h, oh sorry I meant build19:30
jcsackettgary_poster: i can possibly take care of that change on monday, if i get more specifics.19:30
hatchgary_poster: rick_h: I can repro19:30
rick_hgary_poster: ok, known issue with bad charm and the work for that is in progress19:30
gary_posterjcsackett, ack I'll try to dig it up19:30
gary_posterthanks jcsackett 19:31
hatchrick_h: it's an access control origin issue19:31
rick_hgary_poster: though to be fair, how did you get to that url to start with?19:31
rick_hhatch: no, it's an upstream error in the api because of bad data. See bug #119978019:31
_mup_Bug #1199780: search requests errors when a charm has no last_change <charmworld:Triaged> <https://launchpad.net/bugs/1199780>19:31
hatchthe browser is looking awesome btw :)19:31
rick_hhatch: any search for "all charms" will error right now19:31
hatchohhh ok19:31
rick_hdue to that issue19:31
rick_hgary_poster: ok, so 2) ?19:32
gary_posterrick_h, my other story is how I got to that story actually.  so...19:32
gary_poster(1) go to http://uistage.jujucharms.com:8086/precise/cinder-8/19:32
gary_posterrick_h, scratch that, it is http://uistage.jujucharms.com:8086/fullscreen/precise/cinder-8/19:33
rick_hgary_poster: rgr19:33
gary_poster(2) search for cinder19:33
gary_poster(3) click on cinder in results19:34
gary_posternothing happens19:34
rick_hgary_poster: yea, that's a bug. 19:34
gary_posterrick_h: Variant of that19:34
gary_poster(1) go to http://uistage.jujucharms.com:8086/fullscreen/search/precise/cinder-8/?series=precise&text=cinder&type=approved19:34
gary_poster(2) click on cinder19:34
gary_posternothing happens19:34
gary_posterwell, the url changes19:34
rick_hgary_poster: right, that url change is a bug as well in the routing 19:34
rick_hgary_poster: see #119939219:35
_mup_Bug #1199392: can't view fullscreen charm details from url <charmbrowser> <juju-gui:Triaged> <https://launchpad.net/bugs/1199392>19:35
rick_hgary_poster: but that first variant is legit. I'll file a bug on it now. Should be some corner case around the state deciding what view to show. 19:35
rick_hgary_poster: e.g. easy fix once found19:35
gary_posterrick_h, cool thanks.  ok, going to look at your home branch...19:35
rick_hgary_poster: #1200743 for tracking purposes. Thanks for the nice repo instructions19:37
_mup_Bug #1200743: complex search/view interactions in fullscreen fail to work <charmbrowser> <juju-gui:Triaged> <https://launchpad.net/bugs/1200743>19:37
gary_postercool thank you!19:37
rick_hsinzui: any word on the misbehaving charm data bug? I thought I heard something about abel's work might fix and saw he got a branch into review today?19:37
rick_hsinzui: sorry to bug you but I don't see any card for him on the board19:38
hatchrick_h: I'm about half way through review (then to QA) but I have to take off now for lunch, so you might be past EOD by the time I get it done, is that ok?19:38
sinzuirick_h, he didn't get his branch landed. He was concerned about a corner case19:38
rick_hhatch: yea, don't worry. I'll make sure I get it through before I hit bed tonight19:38
sinzuirick_h, I will ask webops to delete the branch that is not a charm19:38
hatchok great19:39
rick_hsinzui: ok, no problem. Just wanted to let gary_poster at least one of his three bugs was close to working :)19:39
rick_hone easy fix, and one's a pita wheeee19:39
rick_hI feel the need to get a 'bug hunting' fun to drop on the desk for times like this 19:39
gary_posterrick_h, I merged your branch with trunk and got conflict.  http://pastebin.ubuntu.com/5869067/ (tree is trunk,merge-source is you).  should I really delete that padding?19:41
rick_hgary_poster: sec, looking19:41
rick_hgary_poster: so I think that neesd to stay? I didn't add that code so guessing we both added at the same spot?19:42
gary_posterrick_h, cool agree19:42
gary_posterrick_h, with a similar level of confidence ;-)19:43
rick_hI'll pull trunk here locally and merge in and make sure things are resolved19:43
rick_hgary_poster: ok, wtf. I edited the file 30 lines below...no idea why bzr hates me19:44
sinzuirick_h, we will now wait for ingest to notice the empty branch is gone19:45
rick_hbcsaller: thanks for the catch. I had a container.one vs a container.all on the event for that. Updating and adding another test now19:45
rick_hgary_poster: so the Home link doesn't work :) but will in a sec19:45
bcsallerthats an easy one to make, ha ha, get it ;)19:46
rick_hho ho ho ho19:46
gary_postercool rick_h 19:46
bcsallerenough punishment 19:47
jcsackettrick_h: sounds like you're addressing the one thing i found in my review. :-P19:48
rick_hjcsackett: yep, pushing now, will have to wait for lbox to go through again before it shows up in reitveld19:48
bacgary_poster: css-only ellipsis works for that field.  \o/  i now see that no one appears to be using gallery-ellipsis any more.  i'd like to rip it out.19:51
gary_posterbac +119:51
bacgary_poster: will do in two branches19:52
gary_posterbac +119:52
jcsackettrick_h: yay, lbox.19:52
jcsackettit has been very slow for me today.19:53
rick_hcrap, now test failures. looks like something is cascading down19:54
gary_posterrick_h, I see a number of issues--did you address them in the merge/update?  example: (1) http://localhost:8888/fullscreen/ (2) search for glance (3) click on glance. Problem: competing home links in header.  Example 2: (1) go to http://localhost:8888/ (we really can't hide the home link just on the home page, trivially? :-/ ) (2) search for mysql. Problem: a line covers up "search results"19:57
rick_hgary_poster: looking, fighting test failures atm now. Will look at those in a sec. 19:57
gary_posterrick_h, cool, sorry19:57
* benji dives back into the drag and drop swamp.20:04
rick_hgary_poster: looking at issues. Thanks 1) is from breadcrumbs work. Looking for where huw put that since he didn't put it in the widget here. 2) there might be a hacky way, but nothing that I'd like to see live in trunk. 3) Looking into that. I'm double checking to see if that border rule can come out all together.20:06
sinzuirick_h, the query is fixed in production and staging.20:07
rick_hoh come on, he freaking absolute positioned the breadcrumbs way out of the View itself?! 20:07
rick_hsinzui: awesome, thanks. 20:07
rick_hgary_poster: I'm going to pull breadcrumbs right now. They're only on the charm details view and in a wrong place. 20:07
rick_hgary_poster: any complains on trading Home link for breadcrumb on charm details?20:08
rick_hcomplaints that is...ugh20:08
sinzuiabentley, rick_h, I took this opportunity to measure the time of deletion to ingestion, and cache clearing. Ingest takes 17 minutes in production. The search/cache was fixed within a minute of ingestion20:08
rick_hsinzui: very cool. 20:09
abentleysinzui: That is scary, because we ingest on 15 minute cycles.20:10
bacjujugui: get the easiest review of the day here: https://codereview.appspot.com/1096404520:11
sinzuiabentley, We changed production to be just-in-time cycling I thought20:11
benjibac: I'll take a look.20:11
* sinzui looks for rt20:11
abentleysinzui: I assume production still runs enqueue every 15 minutes.  If it's doing it more often, then we're in really big trouble.20:13
benjiI learned a new term that I really like over lunch, the "grep test": http://jamie-wong.com/2013/07/12/grep-test/20:16
sinzuiabentley, worker-interval=1 in m.jc.com20:17
abentleysinzui: So if I understand you correctly, we're filling the queue every 1 minute, but it takes 17 minutes to remove that from the queue.  Is that right?20:18
sinzuiabentley, I thought we had changed the process to get a lock  so that the queue would not be appended to while the lock was kept20:19
gary_posterrick_h, I don't think so?20:19
gary_posterrick_h, that's a step up, right?20:19
rick_hgary_poster: I think so. That's updated, #3 is corrected as well. Just fighting test-debug and delegate for a final update. 20:20
abentleysinzui: I don't know anything about the locking.  If so, we should be fine.20:20
sinzuiDo I dare go into #webops one more and ask for the count of what is in the queue?20:20
bacjujugui: one more for https://codereview.appspot.com/10964045/ ?20:20
sinzuimaybe I should get a job as a webop20:21
gary_posterbac, will look.  code looks fine; will try qa :-)20:21
sinzuioh, I could put together the diagnostic page to tell us that all process are operating and what their size/capacity is20:22
rick_hoh I'm a moron...ugh20:23
rick_hgary_poster: updated branch pushing now. 20:24
gary_posterbac LGTM20:24
gary_postercool rick_h .  so it has fixes for #1 and # 3?20:24
rick_hgary_poster: correct20:24
rick_hgary_poster: looking close for #2, maybe a temp hack with an XXX20:24
gary_posterrick_h, cool.  +1 on temp hack for #2.  pretty important to visuals to hide I think20:25
bacsinzui: why does searching for 'postgres' not return 'postgresql-psql' ?20:27
gary_posterbac postgresql does.  apparently not partial text20:27
gary_posterdiscovered as well20:28
rick_hyea, search doesn't do paritals atm20:28
sinzuibac, elasticsearch is not configured to search substring20:28
sinzuibac, but it is configured to do that on autocomplete cases20:28
* gary_poster is happy with new alias:20:28
gary_posteralias bzrevert='bzt revert && bzt revert --forget-merges'20:28
gary_posterwhere bzt == bzr20:28
* bac bzr help revert20:29
bacer, bzr revert help20:29
abentleygary_poster: Since revert automatically forgets merges, I don't see the advantage of this.20:31
gary_posterabentley, oh, it doesn't/didn't for me20:31
sinzuiabentley, ingest could has a lot of steps that are surely collecting redundant data. If the branch has not changed and proof has not changed, don't run proof. The same is also true for jenkins.20:31
gary_posterabentley, perhaps this is merely remembered or even misremembered pain.  willinvestigate later20:32
abentleygary_poster: If you revert a particular file or directory, it won't forget merges.  Otherwise, it will.20:32
sinzuiabentley, quick look, these have ingest times: http://manage.jujucharms.com/tools/store-missing20:32
gary_posterabentley, cool thanks20:32
sinzuilooks like this phase is exactly 15 minutes20:32
abentleysinzui: what about cs:~benji/precise/juju-gui ?  Looks more like 12 minutes to me.20:33
sinzuiwe are network bound for Lp and jenkins so we can expect variations20:34
abentleysinzui: Anyhow, I guess the locking acts as a de-facto ingest delay.20:34
abentleysinzui: There is *lots* we are doing redundantly.  Like ingesting the charm files.  Every. Single. Time.20:35
sinzuiabentley, we clear the queue with every deploy, so al is we is we deploy often.20:35
rick_hgary_poster: can't. Even an XXX hack will not work since the height is hard coded, the widget isn't redrawn in a search content vs a home context to catch the update. 20:35
gary_posterrick_h, :-( I dunno if it is ok to land if it is on home. :-(  other issues are fine btw, though I think breadcrumbs will be better in future when we replace.20:36
gary_posterI mean, I ad t20:36
rick_hgary_poster: definitely agree on breadcumbs. 20:36
rick_hgary_poster: just not enough time to get it in there right atm. 20:37
rick_hgary_poster: luca signed off on permanent home. I made sure to express to him that was the result of getting it in.20:37
gary_posterrick_h, oh, you mean getting it at the top?20:37
gary_postersorry that made no sense20:37
gary_posterother than <surprise> ;-)20:38
* rick_h isn't following still20:38
bacjujugui: and another exciting review.  almost 500 lines of deletions.  https://codereview.appspot.com/11216044/20:38
rick_hgary_poster: guichat?20:39
gary_posterbcsaller, btw you should probably tell us how to qa service-options-viewlet branch20:39
gary_posterrick_h, +1 thanks20:39
bcsallergary_poster: It was something I saw when looking at the home-link branch, but there may have been another issue there, the charm I thought would trigger it didn't on trunk just now20:47
rick_hbcsaller: jcsackett updates up there20:48
jcsackettrick_h: ack.20:49
jcsackettrick_h: i'm seeing you completely removing the home button now, but not adding it back anywhere.20:52
rick_hjcsackett: huh?20:56
gary_posterrick_h, I ended up by saying everything was looking great :-)20:56
rick_hgary_poster: ah, cool. Thanks, trying :)20:56
rick_hjcsackett: you mean the event catch? It's changed to a delegate that catches both cases20:56
rick_hjcsackett: and both of them are tested in the tests now20:57
rick_hjcsackett: if you mean something else let me know 20:57
gary_posterbcsaller, ok so I should just to code lgtm?  bac you need review too?20:57
jcsackettrick_h: i am looking at the template.20:57
bacgary_poster: i do, should be cake20:57
jcsackettah, rick_h, ignore me.20:57
jcsackettwrong template.20:57
rick_hjcsackett: ah yea, that' the breadcrumbs huw added I removed since they conflict20:57
jcsackettrick_h: i was looking at where you ripped out breadcrumns.20:57
bcsallergary_poster: yeah, I think its a win, somehow that triggered before and you'd get an empty inspector.20:58
gary_postercool, on it20:58
gary_posterbac, how was it actually turned on, anyway?20:59
bacgary_poster: how was that ellipsis code used?21:00
rick_hbcsaller: can you peek at the review now that the one/all issue is adjusted? https://codereview.appspot.com/11224044/21:00
rick_hbcsaller: changed to .delegate and skip all that one/all business :)21:00
bcsallergood call21:00
gary_posterbac, I mean, what was the mechanism to turn the ellipsis code on?21:00
bacgary_poster: like this:21:01
bac container.all('.charm-summary').ellipsis({'lines': 2});21:01
bacgary_poster: you can see it in action with 'bzr diff -c 245'21:01
gary_posterbac cool thx21:01
gary_posterbac LGTM get somebody else :-)21:02
gary_posterbenji how did drag thing turn out?  did you get that landed?21:02
benjigary_poster: nope, stilly trying to fix very subtile breakage21:02
gary_posterbenji, :-/ k.  when do you give up? :-)21:03
gary_posterlemme know21:03
bacbenji: subtile, like in subflooring?21:04
rick_hcement board21:04
rick_hvery nasty buggy stuff21:04
bacis that the same as mud board?21:04
bacjujugui: one more for https://codereview.appspot.com/11216044/ -- it's another fun one21:06
gary_posterbcsaller, LGTM with a bug fix and maybe a test21:08
* bac has to step out. bbiab to land my branch.21:11
gary_posterbcsaller, ugh, missed your other one.  I will review it.  jujugui, could we have a second review of https://codereview.appspot.com/11227043/ ?  critical21:11
bacbac: ok21:11
rick_hgary_poster: looking21:11
gary_posterbac, Makyo gave you an LGTM fwiw21:12
gary_posterthanks rick_h 21:12
bacgary_poster: just saw.  thanks Makyo21:12
bacfwiw, i liked gallery-ellipsis... (no pun)21:13
jcsackettjujugui: can i get two reviews on https://codereview.appspot.com/11230043 ? it's mostly deletions and icon changes.21:13
gary_posterjcsackett, if no one has soon I will.  one or two ahead of you21:14
jcsackettgary_poster: dig, thanks.21:14
jcsackettsinzui: can you pick mine up, perchance? https://codereview.appspot.com/1123004321:15
sinzuiI can21:15
jcsackettawesome, thanks.21:15
rick_hgary_poster: can I get a second LGTM on here then? I've got to run get the boy from day care soon. https://codereview.appspot.com/11224044/21:16
gary_posterrick_h, on it21:16
jcastroman just our luck21:17
jcastrofinally get all those categories21:17
jcastroand the dang icon guy goes on holiday for 2 weeks21:17
rick_hjcastro: heh21:17
benjihatch and gary_poster: as best I can tell, https://codereview.appspot.com/11138044 is fixed; still needs a good QAing though21:18
gary_posterjcastro, heh :-)21:18
gary_posterhatch, can you do that?  I've got a queue of three atm21:18
gary_posterand I want rick and benji to be able to EoD21:19
rick_hgary_poster: I know hatch said he was afk for a little bit and would get back. 21:19
gary_posteroh right21:19
gary_posterok benji, I'll try to qa it soon; if you need to run lemme know21:19
gary_posterwe can land it for you if need be21:19
gary_posterrick_h, hack idea for later: listen for EV_SEARCH_...and friends and add and remove class there?21:20
rick_hgary_poster: yea, code might be able to go into the code that fires the viewNavigate based on that. 21:21
jcsackettsinzui: i have to run afk for a moment. if you have questions on the branch just drop 'em on rietveld.21:22
gary_posterrick_h, LGTM with one old observation and one new observation21:24
gary_posterthank you21:24
rick_hgary_poster: the search clearning should happen now. It was part of the last minute squeeze in21:25
rick_hgary_poster: and ty for the LGTM21:25
gary_posterrick_h, oh sweet :-)21:25
rick_hok, I'm out. See you all on the other side...of the appalacians next week...21:26
gary_posterbcsaller, LGTM with trivials21:32
gary_posterrick_h, thanks!  see you sson21:32
gary_posteron to jon's21:32
hatchgary_poster: I am still not sure about this position fixed thing21:34
hatchI'm pretty sure it'll cause issues on mobile heh21:34
sinzuigary_poster, what is the lowest version of Firefox that is supported?21:36
hatch sinzui I want to say 2121:39
hatchbut I can't actually confirm that21:39
sinzui21 is the magic number for me21:39
sinzui20 is not magic at all21:39
hatchhaha well there ya go21:40
hatch21 there were some large changes with how FF works, and 22 is the most recent21:40
hatchso you would have to be pretty far behind to be <2121:40
gary_postersinzui, 21 is good21:40
hatchyusss i got it right21:40
gary_posterhatch, position fixed for what, the headers?21:40
hatchthe body21:40
hatchon benji's branch21:40
hatchI can't suggest an alternative21:41
hatchbut I think that that's going to cause us issues in the future21:41
hatchso it may just be a bandaid21:41
sinzuijcsackett, okay, I am will to trade a LGTM if you revise the use of window.location.origin to a cover IE1021:41
gary_posterhatch, :-/ dunno.  if you don't have another suggestion though, maybe we should wait to fight that battle21:42
hatchthat's kind of what I was thinking21:43
hatchdo we have a tablet to test on?21:43
hatchI only have my N7 with Android21:43
gary_posterMakyo has an ubuntu touch.  I have an ipad.  one of those work for what you have in mind?21:43
hatchwell since the support matrix only includes the touch right now I was hoping we could QA in there to make sure there isn't any wakyness21:45
hatchor do you think that's not necessary?21:46
MakyoShould I bring the tablet, gary_poster ?21:48
gary_posterMakyo, hatch, only for fun :-)  If you don't bring it no biggie Makyo, but +1 if it is easy21:48
hatchgary_poster: Makyo: I can tell you that the UI is all sorts of broken with the position fixed on the body in Chrome for Android21:48
jcsackettsinzui: by revise you mean replace it with using either origin or protocol + host?21:49
hatchbut on trunk I can use it :)21:49
hatchit's actually pretty cool heh21:49
gary_posterhatch, that's pretty bad if this one change suddenly breaks Android :-/21:49
hatchwell it's because it pulls the whole body element out of the flow of the page which I'm going to assume cascades into some weirdness because that's never done21:50
hatchI've never even heard of people going position fixed on the body before21:50
bcsallerwhy does the token need to append to the body anyway rather than be its own floating div?21:52
hatchI was thinking if we could position it where the mouse is...but then we might as well use the YUI DD module21:52
sinzuijcsackett, yes, I think we should follow the example in the nod.js code21:52
jcsackettsinzui: ok, per your comment?21:53
jcsackettpushing that change now.21:53
hatchbcsaller: it has to be visible in the DOM to be the dragable proxy21:53
hatchit's pretty odd...21:53
bcsallertop: -1000px left -100px with position absolute?21:54
rick_hbcsaller: replies inbound21:54
hatchbcsaller: I thought I tried that but maybe it was another bug21:55
hatchone sec21:55
gary_posterjcsackett, code LGTM but qa bad: "X" to close is in wrong place.  I also can't find an image from luca of where the icons should go but it doesn't look right there in the build mode--it looks like it should really be above the tabs21:56
gary_posterjcsackett, want to guichat to see what I'm seeing?  I tried clearing cache21:56
jcsackettgary_poster: per luca, below the tabs.21:56
jcsackettgary_poster: sure, one moment21:57
gary_posterjcsackett, does that seem as odd to you as it does to me?  is there an image?21:57
jcsackettgary_poster: there is not. is guichat open?21:58
gary_posterjcsackett, it is, see you there21:58
hatchbcsaller: yep as I thought, it doesn't work - when it's positioned off the screen the drag proxy is invisible21:59
* bac gotta run. see y'all in raleigh22:00
hatchcya bac22:00
hatchsafe travels22:00
hatchdoes anyone remember why we didn't use YUI DD for this drag stuff?22:01
bcsallerwas it just a preference for the HTML5 stuff? Not sure. really does seem like that native stuff should work but itsn't baked yet. 22:03
hatchyeah I think the issues we are having are caused by workarounds around chrome being stupid22:04
hatchwhereas the YUI DD code absolutely positions an element under the cursor22:04
hatchI gota say - with the exception of there being too much UI chrome on this new design, it works pretty darn well on a tablet22:06
hatchside effect of us just being so awesome....right?? right????22:07
hatchawwww yeahhhhh22:07
bcsallerI'd like to cut over to the pointer events polyfill or something similar so we can pull in better touch support more easily 22:07
hatchI'd like developing on something that isn't a moving target lol22:08
hatchahh who am I kidding....I'd get bored22:08
hatch"what do you mean this just works?"22:08
jcsackettgary_poster, sinzui: changes you've requested are up https://codereview.appspot.com/1123004322:18
* sinzui looks22:18
gary_posterjcsackett, on it thanks22:18
gary_posterhatch, what's the story for benji's branch?22:19
hatchif we use it - it breaks mobile...bigtime - if we don't, we don't have a better sollution besides switching to YUI DD (which I'm not sure why we didn't go that route in the beginning)22:19
hatchso maybe there was a reason YUI DD didn't work22:20
hatchI can think of alternatives, but none that would be trivial (positioning the drag proxy under the cursor)22:20
benjirick and I couldn't get YUI DND to work so we went with HTML5 instead22:21
hatchohh ok, do you remember what the issue was?22:21
hatchand as a follow-up - can we use the actual icon in other browsers and just 'deal with it' until chrome fixes the cutoff icons issue?22:22
benjithe last thing I remember was trying to get the tokens to drag out of the sidebar , they wouldn't leave their container22:22
benjiactually we are dealing with the cutoff issue, the cloning of the icon is to deal with firefox; it uses the default drag icon if we don't22:23
hatchohh odd, I've never seen the cutoff22:24
hatchbut hmm22:24
hatchbasically if we want to support mobile we can't use your fix22:24
hatchbenji: so in FF we can't point it to the icon in the DOM already?22:26
benjithe cutoff only happens with background images (not the per-charm custom icons)22:27
hatchok I'll hack on this a bit22:28
benjihatch: right, if I recall correctly, FF didn't like that22:28
hatchit's past your EOD22:28
benjimay the schwartz be with you22:28
hatchlol well you are right, it doesn't like that22:28
hatchok that was an OSX bug22:30
hatchgot it to work22:30
hatchwithout an extra cloned node22:30
hatchI'll try these changes on trunk and then propose the branch if it works in all 3 browsers22:32
gary_posterthanks benji, hatch.22:33
gary_posterjcsackett, LGTM with one additional style hack that I give you specific numbers for.  Thank you very much!22:33
gary_posterhatch, will be back soon if you need LGTM22:34
hatchsure np - this is taking a little longer than expected because I have too many branches active heh22:42
hatchjcsackett: except use a single line to apply those padding styles :P22:49
gary_posterwell, yeah, that would be fine :-P22:49
hatchfor some reason WIndows 8 is telling me what the weather is in New York22:50
hatchit's raining if anyone cares22:50
hatchoh IE you suck so much22:53
hatchwe had a comment in our ghost-inspector.js file which caused a js parse error....22:53
hatcha comment....22:53
hatchthis is a disaster lol22:55
hatchgary_poster: would you like me to file IE10 bugs?22:55
gary_posterhatch, yes22:55
* hatch schedules off next week to file bugs22:56
hatchtee hee22:56
gary_posterhatch, that bad, eh? :-/22:56
hatchyeah it's pretty bad - good thing it's so standards compliant22:56
hatchok so I have fixed the primary issue but caused another one23:06
hatchthe icon in the middle of the ghost in FF is too large23:06
hatchit's svg <image> tag has the proper dimensions23:07
hatchbut the image isn't scaling properly inside it23:07
hatchhas this happened to anyone else?23:08
bcsallerhatch: I just took a peek at this myself, what about http://jsfiddle.net/p33KJ/1/ doesn't do what we need?23:09
gary_posterMaybe Makyo has svg insight, hatch, dunno?23:09
hatchbcsaller: I've actually already solved that issue23:10
hatchheh sorry23:10
bcsallerwith the HTML5 codepath?23:10
bcsallerahh, ok23:10
hatchbcsaller: ^23:11
hatchlook at line 8.....that caused IE to explode23:12
gary_posterpoor IE.  such a fragile flower23:12
hatchso yeah this diff works....with the exception of the svg image resize issue in FF23:13
Makyogary_poster, hatch - Oh, foo.  Yeah.  We have to append 'px' to width and height attributes for FF to parse them.23:14
gary_posterah!  cool, thanks Makyo!  /me senses happy branch landing approaching...23:15
MakyoI thiiink that's it.23:16
gary_posterbcsaller, when are you landing your branches?  you have two sets of LGTMs waiting for you. :-)23:19
hatchMakyo: ahh cool I'll try that23:19
hatchthanks :D23:19
bcsallergary_poster: I thought I had for some reason, must have failed on a merge, checking23:21
hatchMakyo: so I don't see the x and y being set on the old version23:21
MakyoLines 1186-1187 in app/views/topology/service.js23:24
bcsallergary_poster: if the second LGTM is the options fix that needs one more review unless you think doesn't23:26
gary_posterbcsaller, oh, right.  I personally think it verges on trivial.  I say try to get one more and if you don't land it anyway :-)23:29
* gary_poster steps away23:30
bcsallerahh, I see what happened, the trunk merge needed an npm update to run the tests successfully and submit died in one of my many open terminal tabs :(23:30
hatchMakyo: ok thanks23:41
hatchhave a good weekend all - safe travels!23:54
gary_posteryou too hatch.  you closing up on the dnd branch till Monday?23:58
hatchgary_poster: nope, I plan to propose it but I have supper plans which start in 2mins23:58
gary_posterok enjoy23:58
hatch:) thanks23:59

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