/srv/irclogs.ubuntu.com/2013/06/05/#juju-gui.txt

=== rogpeppe2 is now known as rogpeppe
=== dimitern_ is now known as dimitern
gary_posterhey bac.  thank you for looking into the GA stuff.  As soon as you know if it will work for us, please let me know: the information (whatever the answer is) will unblock another task.12:35
bacgary_poster: yesterday afternoon i inserted the tracking code on uistage and could see it being served but no data was being collected.  chrome plugins showed that it was installed correctly.  that's all i knew at EOD yesterday12:36
rick_h__luca__: thanks, I see the images in the folder and submitting the category changes now. 12:36
bacgary_poster: i now see that we are collecting data.12:36
bac28 unique visitors for 31 visits12:36
luca__rick_h__: excellent, many thanks Rick12:36
gary_posterbac but that is for uistage only, right?  We now need to discover whether it will still work when deployed elsewhere?12:37
bacgary_poster: my next step is to deploy the charm with the same tracking code and see if data are collected from those deployments.  from further research i'm pretty confident the data will be seen.12:37
bacgary_poster: ack12:37
gary_posteryay, bac!12:37
gary_posterso...well, quick guichat call?12:38
bacgary_poster: i'm not sure yet what the time delay is for reporting.12:38
gary_posterack12:38
gary_posterbac, could I run something past you on guichat?12:39
bacsadly their dashboard still shows the tracking slug is not deployed, which is misleading12:39
bacgary_poster: sue12:39
bacsure12:39
bacdon't sue12:39
gary_postersally12:39
luca__gary_poster: sorry to pester, I should of asked in IRC but would it be possible for you to answer my latest bundles question I sent in email? Really sorry!12:42
gary_posterluca__, yes reply started, will get out in a few12:43
luca__gary_poster: thank you!12:44
gary_posterluca__, could you give bac information on what the privacy policy banner thing needs to be?  If you have standard HTML/CSS that would be even better12:47
gary_posterand JS12:48
luca__gary_poster: the cookie thing?12:50
gary_posterluca__, yeah12:50
luca__bac: just asking the web guys what they have12:52
bacluca__: thanks13:01
bac"Error with ubuntu-doodoo" ... from uistage13:02
=== jcsackett_ is now known as jcsackett
hatchmorning13:26
gary_postermorning13:27
luca__gary_poster: in regards to the feedback stuff, I'll have to complete this early next week as I'm flying out to Norway in a few hours13:31
gary_posterluca__, ok cool, thanks for the heads up13:32
gary_posterenjoy :-)13:32
hatchjealous13:32
BradCrittendengary_poster: good news.  i was able to modify the tracking slug and am seeing reports from me hitting a locally running juju gui.  i deployed 'mysql-bac' and see it in the real-time reports.  the data available in real time is minimal (urls and location) so it'll be interesting to see what shows up later13:33
=== BradCrittenden is now known as bac
gary_postergreat, bac!  thank you for letting me know13:34
bacbut it looks promising13:34
gary_posterexcellent13:34
benjiI have a small branch up for review: https://codereview.appspot.com/998104413:38
* gary_poster stepping out to son's school in about 8 minutes; back at about 20 or 25 past the next hour13:47
gary_posterin time for vitals discussion :-)13:48
gary_posterbiab13:54
abentleyorangesquad: uistage.jujucharms.com is failing to load the browser.  It says "Failed to load editorial content.\nCharm API server did not respond"14:08
rick_h__abentley: k, looking14:08
abentleyThat was at http://uistage.jujucharms.com/:flags:/browser_enabled/14:09
rick_h__abentley: http://uistage.jujucharms.com/:flags:/browser_enabled/ works for me14:09
abentleyrick_h__: Worked this time, failed three times before.14:10
rick_h__abentley: :/ so we're wondering if this is a 504 on m.j.c issue?14:11
sinzuiabentley, I don't see any issue with uistaging or m.jc myself14:11
abentleyrick_h__: That seems plausible.14:11
sinzuiabentley, is it possible there was a 504 from m.jc that broke your uistaging app? Reloading doesn't help?14:12
abentleysinzui: The first three times I loaded the page, I got a failure.  the fourth time, it worked.14:12
sinzuioh, maybe we got a restart of elasticsearch from the rt I filed?14:13
abentleysinzui: I think the simpler explanation is that we're still getting intermittent failures from m.j.c due to connection time-outs.14:15
sinzuiabentley, I'll talk to webops about what the timeouts are between the services in the stack14:16
sinzuiabentley, rick_h__ has evidence that interesting is not being caches and that there is a 3 second timeout between apache and elasticsearch14:24
rick_h__abentley: yea, I was checking the cache headers and I got a 504 when the req time on the back end was just a hair over 3s14:25
rick_h__abentley: and looking at the x-cache headers I'm not sold it's working right. 14:25
abentleyrick_h__: I see "Cache-control: max-age=0" in the headers, which would defeat caching.14:27
sinzuioops14:28
rick_h__abentley: looking, wonder if YUI auto adds that to the request. 14:28
rick_h__hatch: ^ any inside knowledge there?14:29
abentleyrick_h__: That might be from me hitting reload.  I was doing it manually.14:29
sinzuiyep. I see that for every search I do14:29
hatchreading the backlog14:29
rick_h__abentley: sinzui ok, I'll file a bug and put it on the board to look at the requests going out and look at how to override that 14:29
abentleyrick_h__: Yeah, looks like that was my fault for hitting reload.14:30
sinzuithanks rick.14:30
hatchrick_h__: you wondering if yui auto adds cache-control to the request?14:30
rick_h__hatch: rgr14:30
hatchit doesn't14:31
hatchyou are free to set your own headers though14:31
rick_h__hatch: ok yea, nvm. abentley thinks it was a reload force issue. 14:31
hatchshouldn't that be allowed?14:31
hatchthis sounds like a mjc issue14:31
hatchthe user should be ble to reload and get a site within a second over and over and over14:32
rick_h__hatch: so mjc is trying to cache things but if the client says "give me new stuff" it should skip the cache and never use it14:32
rick_h__hatch: but by default, the stuff in interesting/search results/etc can't change but every 15min anyway14:32
rick_h__hatch: so we're just looking at the caching and if it's in effect/helping/etc. 14:32
hatchahhh ok14:33
hatchyeah that sounds like an issue on mjc side14:33
rick_h__hatch: but nvm, I'm looking now and I don't see the header added. 14:33
rick_h__hatch: well, it's on both ends. the gui has to make the requests with good headers and mjc has to respond approporaitely14:33
rick_h__hatch: and by gui I mean the charmworld store class really14:33
hatchwell yes, but mjc should be smart enough to return the cached version if it's still valid regardless of what the client requests14:34
rick_h__hatch: :P no, give the customer what they ask for 14:34
rick_h__hatch: but anyway, we're all good. 14:34
abentleyrick_h__: One thing we probably should do is support ETags for charm/*/file/*  I see a lot of pointless icon retrievals.14:36
rick_h__abentley: k, sounds like a good plan14:36
hatchMakyo: which card did you just propose? so I can tag it15:16
Makyohatch, oh, sorry, just a sec.15:16
hatchcool thx15:17
Makyohatch, zoom to service if any...15:17
hatchMakyo: do you need a QA?15:20
Makyohatch, yeah, but give me a sec to revert config.15:21
hatchhehe centroid15:21
hatchsounds like a video game15:21
hatchit's metroid's girlfriend15:22
rick_h__hatch: can you link me to the assert node thing you showed me yesterday?15:22
Makyohatch, QA: add services, pan away from them, shift+0 to pan back.15:22
hatchrick_h__: google node.js assert15:22
hatchrequire('assert') it's part of the node core15:22
rick_h__hatch: thanks, was doing javascript and failed15:22
hatchand that's new?15:23
hatchooooo burn15:23
rick_h__lol15:23
* hatch hands some aloe to rick_h__15:23
hatchhaha15:23
hatch^515:23
frankbangary_poster: do you have time for a quick hangout after the call?15:47
gary_postersure frankban 15:48
frankbanthanks15:48
gary_posterjujugui call in 10; kanban now15:50
gary_posterjujugui call in 215:58
gary_posterbenji, call now16:01
rick_h__sinzui: jcsackett can you guys look at https://codereview.appspot.com/10049044 when you get time please? sinzui in particular would appreciate wording and gramatic feedback :)16:02
hatchMakyo: QA looks good16:09
Makyohatch, have a q on one of your comment,w ill ask after call16:10
hatchdeal16:10
hatchjcsackett: I think you need a new machine :)16:11
jcsacketthatch: google needs a new hangout app.16:11
rick_h__says the man that needs a new machine because tests are too slow :P16:11
jcsackettbut! i get a new machine in 2 months. and it will have a webcam.16:11
hatchlol noone else drops16:11
hatchrick_h__: lol!16:11
jcsacketthatch: bcsaller dropped at the start of this call.16:12
hatchrumor has it the new MBP will be coming out next week so I should have a new one within a couple months16:12
hatchme hopes16:12
* gary_poster has similar plans16:13
jcsacketthatch: my problem is just that my laptop doesn't have a webcam, so i'm using the mobile hangouts app.16:13
jcsackettoddly, it only really seems to have problems with y'all's standup. :-P16:13
benjiI so want the banner to say "This site uses cookies, but only to determine if you have dismissed this banner or not."16:13
gary_posterlol16:14
=== benji is now known as Guest54884
hatchbenji hahaha16:14
Makyohatch, is your ATTRS comment with regards to this.get('component')?16:14
=== Guest54884 is now known as benji
frankbanbenji: where is your branch?16:15
hatchMakyo: well we are 'getting' attributes but none are defined16:15
hatchso someone coming in later will be like 'wth are these coming from' ?16:15
gary_posterhey rick_h__ thank you for quality message change, awesome.  did luca direct/approve the presentation of that?16:15
benjifrankban: https://codereview.appspot.com/9981044/16:16
hatchjcsackett: ahh yeah I found out the other day (when the power kept goign out) that the mobile app requires some serious mobile cpu power16:16
Makyohatch, just component.  The other attributes are on topology16:16
rick_h__gary_poster: not really. It's a stab at "look at this, does this make people happy?"16:16
benjihmm, I guess lp2kanban hasn't updated it; that's odd16:16
Makyohatch, component is set by the topology framework.16:16
rick_h__gary_poster: e.g. it's not any worse and might just work so check it out.16:16
hatchMakyo: hmm I thought the ATTRS was empty16:16
jcsacketthatch: yeah. i plug it in for the standups, but that hasn't really helped.16:16
Makyohatch, app/views/topology/topology.js16:16
gary_posterrick_h__, lol ok, cool.  unfortunately luca is out rest of week.  maybe ask alejandra for a once over in his absence?  she will be back tomorrow16:17
rick_h__gary_poster: rgr, can do16:17
gary_posterty16:17
hatchMakyo: oh right these things are used as extensions not individual modules16:17
Makyohatch, I mean, sure, I can add component to all the files and document it as being set by the d3ns.Module stuff?16:18
Makyohatch, but that's the only attribute.  Everything else is on topo, not this.16:18
Makyoall the files ATTRS, that is.16:18
hatchyeah - maybe just add component to the ATTRS16:19
Makyohatch, alright.16:19
benjiThe power is flickering here.  I may drop off.16:22
hatchheh ours never flickers, it just goes down without warning16:22
bcsallerhatch: in chat again, branch pushed16:23
hatchkewl16:23
frankbangary_poster: call?16:27
gary_postersorry frankban 16:29
gary_posteryes16:29
=== arosales_ is now known as arosales
frankbangary_poster: https://plus.google.com/hangouts/_/02bb45411739e441fe107c9f66e2a8cc36ba4ba7?authuser=0&hl=en16:31
rick_h__abentley: have sec?16:35
abentleyrick_h__: sure.16:35
jcsackettrick_h__, hatch: can you take a look at https://codereview.appspot.com/9964044/16:38
rick_h__jcsackett: will do16:40
abentleyrick_h__: I tried a hyphenated category and it worked fine: http://staging.jujucharms.com/api/1/charms?categories=cache-proxy16:46
jcsackettabentley, rick_h__: staging looks like it handles it, manage doesn't.16:48
rick_h__abentley: hmm, so something is up wih that request then. http://staging.jujucharms.com/api/1/charms?categories=app-servers&series=precise&text=&type=approved is returning results not too16:48
rick_h__jcsackett: ah!16:48
abentleyrick_h__: It's possible that the ES mappings didn't get updated properly on manage.  I guess we'll need some help from the ops of the web.  I'm going on lunch.16:49
rick_h__abentley: ok, updating the bug to note the diff in staging/manage16:50
* gary_poster has lunch17:11
sinzuiorangesquad, gary_poster. Our tarmac is down. So is red's It appears all landing bots cannot ssh to launchpad. I will be in #is to learn how this will be resolved17:15
rick_h__thanks sinzui 17:17
jcsacketthatch: can i get a review on https://codereview.appspot.com/9964044/ ?17:21
hatchyup doing now17:21
hatchdone17:21
jcsacketthatch: awesome, thanks! :-)17:24
rick_h__hatch: why have you not tried http://facebook.github.io/react/blog/2013/06/05/why-react.html yet :P 17:44
rick_h__"React really shines when your data changes over time."17:44
gary_posterthanks sinzui 17:45
hatchrick_h__: haven't but I'll look into it17:56
hatchI have so many libs/frameworks to review17:57
hatchI need to take a month off to get caught up on everything17:57
hatchhah17:57
* gary_poster restarts18:09
=== gary_poster is now known as gary_poster|away
* jcsackett assumes rick was trolling with the react thing18:13
=== gary_poster|away is now known as gary_poster
hatchprobably was but I always try to keep my finger on the pulse18:15
hatchthe pulse just moves damn fast18:15
hatchhaha18:15
hatchoh jeeze they are confusing their lib with JSX18:17
rick_h__hatch: yea, just a little bit. it's a new thing FB released in the last couple of weeks18:17
rick_h__jcsackett: ^^18:17
hatchit also re-renders the whole 'nontemplate' when data changes18:18
rick_h__I had to laugh when I grabbed the quote out18:18
rick_h__right, but it's got 'smarts' :P18:18
hatchmaybe, but on face value it misses the mark18:18
rick_h__and they don't confuse it, but they support it in the html syntax18:18
hatchsure but their examples use JSX....why the **** would you use another technology while trying to demo your new technology18:19
hatchthat's like someone using coffeescript to demo their new js framework18:19
rick_h__because it solves the "ugh, html in JS strings sucks"18:19
rick_h__and people run away18:19
hatchyeah it's almost like we solved that.....with templates18:19
rick_h__lol, but this is especially *not* templates :)18:20
rick_h__templates are too static (in this lib's eyes)18:20
hatchsure - why update values in a template when you can rerender the whole thing...18:20
rick_h__lol, you're mixing templates in again18:20
rick_h__their fiddle demo notes that it doesn't re-render since it didn't change :p18:21
hatch<textarea onKeyUp={this.handleKeyUp} ref="textarea">18:21
hatchO K I'm done18:21
rick_h__lmao18:21
rick_h__anyway, just joking that you didn't grab the absolute latest lib from a big company (FB) for your demo. 18:22
hatchhaha yeah I know :)18:23
hatchangular has been around for a while though18:23
rick_h__it's been out for weeks! 18:23
hatcha couple years I think18:23
rick_h__it's not been 'released' for multiple years has it?18:23
hatchyep18:24
hatchit was pretty basic before though18:24
rick_h__ok, I mean it's only come across it more and more the last year. I figured it was internal and then open sourced just lately18:24
rick_h__so v1 was a year ago, maybe that's what I'm thinking18:25
hatchoh well yeah I think it only hit 1.0 last year - but 1.0 in js world is really like v100 because they all start at 0.0.1 lol18:25
rick_h__I really need to use test-debug more. 13s for tests vs 30 (though I think it's skipping some setup time in that time)18:34
abentleysinzui: Could you please review https://code.launchpad.net/~abentley/charmworld/stop-ingesting-icon/+merge/167613 ?18:36
* sinzui does18:36
gary_posterrick_h__, I suggested other changes than what you had.  Fine with what Curtis gave you.  Was just giving ideas.18:41
rick_h__gary_poster: See that. Not a problem. Thanks. 18:42
gary_posterrick_h__, thank you very much for the changes.  great to have.  I am hopeful that Luca and/or Alejandra will bless with minimum/no changes18:42
rick_h__gary_poster: yea, hopefully seeing will help make it click. I'm not a big fan of moving the UI around and would hate to remove a tab *sometimes*18:42
gary_postercool18:43
sinzuiabentley, I am not sure I understand the need for --prefix18:45
sinzuiabentley, well, the default rule looks like a filter that is not useful to this case18:46
abentleysinzui: It's a convenience for testing, so that you can do enqueue --prefix '~charmers' and get charms with icons more quickly.18:46
rick_h__sinzui: can I get a LGTM when you get a sec per your changes. Then I'd like to move this card nito acceptance vs landing so I can work on meeting with UX to get pre-landing acceptance per gary_poster if that's cool18:46
sinzuiabentley, okay, works for me18:47
abentleysinzui: it was already using CHARM_IMPORT_FILTER, I just made it possible to override that.18:47
sinzuiabentley, r=me18:48
abentleysinzui: But if you want, we can scrap CHARM_IMPORT_FILTER entirely.18:48
sinzuiabentley, I got confused for a moment. I was thinking about how the filter relates to the icon problem. It doesn't, and that is good because there are non~charmers charms with icons now18:49
* sinzui returns to rick_h__'s branch18:49
abentleysinzui: Cool.18:49
* abentley emulates Tarmac18:49
sinzuiorangesquad. Looks like matsubara is a single point of failure. IS doesn't have access tot the machine.  I have pinged rfowler, my only lead. The loss of ssh though makes me ponder a firewall change18:54
sinzuiabentley, wasn't a firewall change the cause a issues in Feb/Mar?18:55
abentleysinzui: Sorry, I don't remember.18:55
sinzuioh, that change was between tarmac and staging.18:56
sinzuiabentley, I will add an options property to the Charm model in my next branch. *Every* use of config is to get to a dict names options in the code.19:08
abentleysinzui: +119:08
=== swebb_ is now known as swebb
bacgary_poster: i've added comments to https://docs.google.com/a/canonical.com/document/d/15o7YX4HFuuVuzYzKmhMRBfvKkca6Srwq8WdyR6YDL44/edit19:33
gary_posterthank you bac19:34
benjiI have another small branch that adds yuidoc linting up for review: https://codereview.appspot.com/1004504619:39
hatchbenji: on it19:43
benjithanks19:43
hatchbenji: is ther any way to make it lint the prefixed *'s on each line?19:48
benjihatch: I can easily do that, but we would have to finish the argument about whether or not we want the *s first... or did we finish that?  I can't remember.19:49
* benji looks at the blog.19:49
hatchI thought we did but I have seen new code with it hah19:49
benjihatch: the blog doesn't mention it.  I vaguely recall us settling on /** to start, two space indents for the body, no star on prose lines, and /* (indented at the same level as the opening /**) to end.  Does that sound right?19:51
benjiI could make that happen pretty easily, but fixing all the existing comments might be a tad tricky.  Hmm, well maybe not; I might be able to do it relatively quickly.19:52
hatchbenji: I thought that's what we had agreed uppon as well19:53
hatchthere was some pushback because vim's commenting system19:54
benjiright, but I remember providing a setting tweak that fixed that objection19:54
hatchholy schmoly blueprint emails batman19:55
benjihatch: thanks for the review, I had a question for you about the read-only attribute.20:01
hatchsure20:07
hatchshoot20:07
hatchoh I see in the email20:09
hatchnope that just means that the initial value is provided by a function20:09
hatchthere is actually a setting for readonly attributes20:09
hatch^ benji20:09
hatchhttp://yuilibrary.com/yui/docs/api/classes/Attribute.html#method_addAttr see readOnly20:10
benjihatch: oh! ok; thanks for the explanation20:12
hatchnp - fixing the parser about those quoted functions would be great20:13
hatchbtw :)20:13
hatchhttp://analytics.blogspot.ca/2013/06/google-analytics-becomes-robust-testing.html20:27
hatchgary_poster benji: ^20:28
hatchmaybe something in these new changes we can take advantage of20:28
gary_postervery cool hatch, thx.  bac, you also ^^20:29
hatchso generators are coming to ES6, I figured you perlites would enjoy that20:59
hatchpythonites I mean20:59
hatch:)20:59
hatchI actually had a use for them yesterday - then I worked around it20:59

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