[03:46] rick_h__: you around? [03:46] hatch: heh, yea [03:46] hatch: but barely [03:46] heh sucker [03:46] :P [03:46] https://codereview.appspot.com/10019050/diff/1/app/views/databinding.js I don't know what you meant with the top comment [03:47] loading [03:47] hatch: just spacing around the { [03:47] you have { return [03:47] but );} [03:47] the asymetrical-ness of it got my attention :P [03:47] ohh - yeah ben wrote that, that might be his style, I've noticed it a few places :) [03:48] hatch: so yea, let me know if we need to do a hangout or something on my comments tomorrow [03:49] hatch: I wasn't clear on some parts and maybe I'm not getting it. [03:49] sure - this is still under development so it's very likely some will change [03:49] it's gone through a few iterations already [03:49] cool, yea. After going through it maybe I just need to get told to hold on and wait and put my LGTM on there for now [03:50] but figured I'd do my best to bring 'fresh' eyes to it [03:50] since I'm not in all the meetings/discussions and just approaching it as a dude that can read some JS [03:51] and with that it's way past my bedtime after a LUG meeting night. Check ya tomorrow [03:52] :) have a good night thx for the review [12:34] If anyone is interested in leading the charge of switching us from canonistack to ec2, please do. this is not cool, and will be untenable with tarmac [12:36] gary_poster: so I'm looking at hatch's branch he put up and some of the things I comment on are "yea, that's not ideal and will get worked on more in the future". So should reviews kind of be "well, point this out" and let it ride as things shake out then? [12:36] gary_poster: I don't want to hold up things, but wonder which things to say "yea this should be tweaked" vs 'ok, pinky promise this won't hang around forever' [12:38] rick_h__, for this effort, behind a feature flag, yes. For instance, it would be fine to say "please mark this with an XXX and a bug. I'm fine with it landing behind the feature flag, but not with it staying." Other variants/ideas welcome. I want them to feel like they can merge incremental work to trunk for larger efforts like this. [12:38] larger exploratory efforts [12:39] gary_poster: k, cool. [12:39] thanks [12:43] rick_h__, I can't get the serviceInspector flag to work, can you? [12:44] gary_poster: not tried, sec [12:44] k [12:47] gary_poster: yea, works here [12:47] huh [12:47] ok [12:47] http://127.0.0.1:8888/sidebar/:flags:/serviceInspector/ is the final url after I added a service and confirmed it [12:47] rick_h__, using sandbox? [12:48] gary_poster: yea [12:48] k trying again [12:48] still not [12:48] will try make clean? [12:50] nope [12:50] gary_poster: yea, not sure. I just grabbed trunk, updated it, colo-branches, merged jeff's branch and make devel and it worked [12:50] sounds exactly the same [12:51] rick_h__, and you add a service (I added mysql), confirm (ugh I keep forgetting to ask someone to fix the cancel button css), and then click on the service, click view, and...get taken to the old page [12:51] no, you get the right thing at the end :-) [12:53] http://localhost:8888/sidebar/:flags:/serviceInspector/ is my final url.. [12:53] gary_poster: right, I added ceph, top of my sidebar list. Added it, confirmed, clicked on it and hit 'view' and got this floating panel up [12:54] :-/ ok thanks rick_h__ . bztr status shows all the right stuff too. [12:54] I have call soon so will return to that later [13:02] gary_poster: it seems that suddenly ie is not able to click on DOM elements without disconnecting the webdriver :-/ [13:03] frankban, :-( so not canonistack? [13:03] frankban, on call will check in with you after [13:03] gary_poster: I see that behavior in ec2. thanks [13:09] Has anyone else seen this? I'm getting this error on a full test run, but not when the set of tests is run on its own: http://paste.ubuntu.com/5758147/ [13:15] benji: is this updated trunk? [13:15] rick_h__: it is on a branch, which hasn't been updated recently [13:15] benji: I fixed up some failing tests in trunk yesterday in browsers that didn't show in phantomjs. This looks like phantomjs though [13:16] this is Chromium [13:16] benji: ah, check out trunk. With CI not going we had things land that broke tests in FF/chrome but ran in phantomjs so they landed [13:17] thanks, I'll take a look [13:29] rick_h__: no change in test; thanks for the hint though [13:30] benji: :( ok nvm then. Not seen that directly then. [13:30] I really wish we had a decent test runner; doing things like running a subset of the tests is painful. [13:32] benji: +1 [13:36] benji: so looking at your error again I see it's in browser_charm_view and with us removing the browser subapp feature flag I wonder if this is a side effect? I thought I had gotten all that squared away with yesterday's tests updates though. [13:36] benji: so let me know if it doesn't clear up and I can help look at hte branch and peek at it [13:36] maybe, but the tests pass on trunk and that set of tests pass when run independently; I suspect dirty global state [13:36] thanks [13:37] benji: completely possible. There's still fun with that these days. since the browser subapp isn't FF'd it's always in every App() instance and causes side effects sometimes. We've been working on minimizing that. [13:44] frankban, hi. off call. [13:44] so...yuck. [13:45] frankban, so that means IE CI is completely broken until this is fixed, which is completely out of our control? Maybe we simply need to move it to the end of the CI list? That would mean that we would still see the effects [13:45] and tests would still fail :-( [13:46] but FF and chrome would still have a chance to run [13:46] and hopefully pass [13:46] orangesquad: Do you have a few minutes to review https://code.launchpad.net/~sinzui/charmworld/charm-model-updates/+merge/168812 [13:46] gary_poster: I might have a workaround, testing now. Anyway, I was not able to find a specific ie driver bug, only a number of people complaining about weird behaviors in ie click(), more or less similar to this one [13:46] huh [13:47] ok thanks frankban [13:47] sinzui: sure. [13:48] gary_poster: I also saw that the tests output in Jenkins is confusing, maybe stderr is printed before stdout? I am thinking about printing our messages (e.g. restarting API backend, etc..) to stderr [13:48] huh [13:49] probably fine [13:49] ok, I'll try that [13:57] morning [13:58] morning hatch. trying to finish up the rollup review. can't get the demo to work, but rick did, so it's PEBKAC somehow [13:59] hmm that's concerning [13:59] any errors? [13:59] hatch: morning, thanks for some of the updates. I noticed that some of them in https://codereview.appspot.com/10019050/diff/1/app/views/service.js didn't have a reply so not sure if we're set or want to chat more? [13:59] no errors hatch [14:00] rick_h__: oh right sorry - I didn't comment on those other ones because that whole block is a prototype and will change a lot [14:00] hatch: gotcha, ok. [14:01] sinzui: invite on the way [14:01] gary_poster: hmm that's very odd I'll have to step through it to see if there is somewhere where something could happen [14:01] * gary_poster liked that code [14:02] hatch will ping in a few and we can try it out together in hangout? [14:02] sinzui: r=me with some suggestions in the comment. [14:02] gary_poster: sure can you give me 10? morning routine around here is still going on so it's a little hectic :) [14:02] hatch, np, still reviewing :-) [14:16] looking at the comments [14:31] hatch, I have another call in 29. Will try to get your demo working again for a few. if you are up for hangout lemme know, but no pressure. [14:32] sure just replying to the last comment [14:32] cool [14:33] ok done that [14:33] guichat? [14:33] ok [14:33] thx [14:34] rick_h__: I'll add XXX's to the codebase [14:35] hatch: yea cool. I know it's all balancing between moving forward with ideas vs making sure we don't leave things that could be better behind for later people to fix up/complain about :) [14:35] hatch: so take my feedback with a grain of salt and all that. [14:40] this is kind of cool, missed this annoucement. https://plus.google.com/+GoogleChromeDevelopers/posts/644qQuBKZeL [14:40] editing files directly from the chrome dev tools should be kind of neat, especially with css/etc [14:41] guihelp: could you please review https://codereview.appspot.com/10049046 ? This is my attempt to fix CI failures, thanks [14:42] http://opensas.wordpress.com/2013/06/12/live-hot-code-reloading-with-chromium-browser-on-linux/ for chromium notes on enabling it [14:42] frankban, taking a look. [14:43] thanks [14:55] orangesquad: could one of you have a look at this MP: https://code.launchpad.net/~adeuring/charmworld/1163544-es-error-monkey-search/+merge/168971 ? [14:55] * sinzui looks [15:01] rick_h__: definitely thanks for the review - normally I would want everything perfect before landing but in this case it's still a WIP [15:01] :) [15:02] rick_h__: there is also a regression when deploying services from the browser - the ghost config doesn't show all the options [15:02] can you get to this this morning? [15:12] hatch: looking [15:12] adeuring, I replied with a question? I don't see the unbalanced quote case tested [15:13] hatch: did that get fixed before? /me doesn't remember what ever happened with that. [15:13] sinzui: it is tested like the other characters: 'foo %s bar' % '"' [15:13] rick_h__: yeah I was sure someone fixed it before [15:13] hatch: k, filing a new bug/card. Will see. [15:14] ah! [15:14] it must have gotten overridden in a merge or something [15:14] sorry adeuring [15:14] sinzui: no problem ;) [15:14] adeuring, r=m with my suggestion to represent the the special chars as all-caps constants. [15:15] sinzui: thanks [15:15] hatch: files as #1190265 [15:15] <_mup_> Bug #1190265: adding a service does not load all config

[15:30] thanks rick_h__ [15:39] I have a spare (rather powerful) server hanging around doing nothing, I wonder if I can instlal openstack on it and run local juju instances [15:39] can I run openstack on a single machine? [15:43] hatch: I think they're working on that but it's not there yet [15:44] hatch: but don't quote me on it [15:44] I don't really want to spend to much time on it, so if I can't `apt-get install openstack` I'm not really interested :) [15:47] benji oops! [15:48] coming back :-) [15:48] control-shift-T [15:53] adeuring, did you manually merge? I am looking at jenkins [15:54] sinzui: no, I did not. [15:54] okay, thank you. We are going to wait and see if tarmac does its job [15:54] jujugui call in 6 [15:54] kanban now [15:58] frankban gave you a second LGTM [15:59] gary_poster: thanks, submitting [15:59] jujugui call in 1 [16:01] benji starting without you (teknico got your message) [16:14] adeuring, I QAed your fix while verifying tarmac was up [16:14] sinzui: thank you [16:16] orangesquad, hatch, Makyo anyone up for some review of a nice small one please? https://codereview.appspot.com/10227045 [16:18] orangesquad: heads up that kanban is acting nuts. I'm taking the bug #1190063 card next but kanban won't shot it atm. [16:18] <_mup_> Bug #1190063: going into a category and then hitting back does not re-render the editorial

[16:19] /shot/show [16:22] rick_h__, looking. [16:22] Makyo: ty [16:25] rick_h__: looking [16:26] * gary_poster feels embarrassed he didn't know about bitnami [16:27] rick_h__: done, just one q [16:28] I didn't know about it either :) [16:29] looks pretty cool on the surface [16:29] and like the primary closed source and pre-existing competitor to juju [16:30] yeah although it appears to be taking a different approach [16:30] to the same problem [16:30] yeah [16:31] once we have bundles done I think our approach is better [16:31] stacks in particular [16:31] er yeah, stacks [16:31] I get the two mixed up :) [16:32] :-) yeah [16:32] * gary_poster runs to lunch for a bit... [16:32] hatch: replied [16:32] hatch: basically hasStateChanged is a question, not mutating [16:33] gotcha thx [16:43] hatch: fixed tests in devel and debug, still the same "not a constructor" error in prod :-/ lp:~teknico/juju-gui/add-cookies-usage-banner [16:43] but it works in devel? [16:43] yes, under "make test-server" and " make debug" [16:44] ok looking [16:44] thanks [16:44] so something isn't making it into the combined js file :/ [16:45] yeah [16:45] teknico: ahh it's missing from the bin/merge-files ln89 [16:46] hatch: ooh, right [16:46] try that out - if that doesn't work then I'll dig further [16:46] too many moving parts :-P [16:46] ""jenkins is back to normal"" yay! [16:46] yeah our build process is too complex [16:47] hatch: now it doesn't find Y.Cookie?!? [16:48] looking [16:48] hatch: does it have to be added to the list on line 74? [16:49] yeah, that did it :-P [16:49] :/ I really don't understand that [16:49] me neither :-/ [16:49] those scripts should not need to be manually added into the reqs [16:50] what if you add 'cookies' into app.js as a requires [16:50] (curious to see if that works) [16:51] hatch: do you mean "cookie", and without adding the line "reqs.push('cookie');" to merge-files? [16:51] yeah [16:51] sorry [16:52] hatch: yep, it works [16:52] which one is better? [16:52] the latter (in app.js) imho [16:53] deal [16:53] that means that our parsing script is not working properly [17:01] I think those problems are because the YUI loader isn't designed to be used this way [17:01] Maybe we could find some nice improvements if we reviewed it though [17:02] frankban, thank you very much for Jenkins fix! yay! [17:02] \o/ [17:04] lint fixes, merge from trunk, test again, propose, yadda yadda :-) [17:06] hatch: proposed again, how about another review? :-) https://codereview.appspot.com/9967044 [17:06] thanks for all the fish, btw :-) [17:06] sure give me a few minutes and I can get on it [17:06] no hurry [17:13] jcastro: is the idea to replace askubuntu with discourse? [17:14] no [17:14] AU is for q+a [17:14] discourse is for watercooling [17:15] "How do I configure Juju to do foo?" goes on AU. "I like Juju let's talk about it." goes on Discourse [17:15] theoretically [17:17] ahh gotcha - I wonder if the two can be integrated at all [17:17] link from one to the other for example [17:17] yeah I have some ideas [17:17] for oneboxing and making it more seamless, as I'm sure there will/is confusion [17:18] discourse confuses me to start with but I'm sure I'll get used to it :) [17:18] it's a fresh start. Brave new world. [17:18] burning of ships. [17:18] I don't like new things... [17:18] * hatch goes back to sit on his porch yelling at the neighbour kids [17:18] lol [17:18] Even I like it, and I hate everything [17:18] jk [17:18] hahahaha [17:18] burning of RAM and CPU, more than anything... [17:19] jcastro: :P [17:19] btw rick_h__ what's with all the underscores? [17:19] hatch: must have been reconnecting at some point === rick_h__ is now known as rick_h [17:22] rick_h: we should finish up your blog post [17:22] jcastro: yea, good call. I'll poke at it tonight at CHC [17:27] teknico: review done with a few comments (LGTM'd) [17:30] Does anyone on orangesquad have time to review https://code.launchpad.net/~sinzui/charmworld/charm-model-fixes-views/+merge/168175 [17:35] I think kanban is on read-only mode [17:36] sinzui: yea, it's borked. Yellow had issues on their call today as well [17:36] I don't even get that far [17:36] yeah, I get loading errors sometimes [17:36] ah, loads but as soon as I *do* something it undoes it and yells at me [17:37] gary_poster: so is your only issue with trello that it doesn't do wip limits? [17:37] hatch: and you've not written lp-trello to replace lp-kanban [17:37] * rick_h does use trello for outside of work stuff though heh [17:38] do we need lp-kanban? It was down for the longest time and I didn't even notice :) [17:38] hatch, no, I also want greater flexibility in arranging the swim lanes than trello provides. [17:38] rick_h: yeah so do I haha [17:38] lp2kanban is less important as we use fewer LP bugs [17:39] gary_poster: ahh ok - I'm a +0 in the matter anyways but just curious [17:40] hatch I looked up WIP limits in trello earlier. whatzizname the big blogger who is part of the company that makes it says that wip should be social anyway. It is social, but don't agree that this means that visually highlighting WIP is unimportant [17:40] somebody has a chrome plugin I think [17:41] https://chrome.google.com/webstore/detail/kanban-wip-for-trello/oekefjibcnongmmmmkdiofgeppfkmdii [17:41] that one? :) [17:41] yeah - that response sounds like an arrogant one [17:41] probably. didn't look closely [17:41] "we aren't going to add a counter.....because I don't believe in it....yeah that's it" [17:42] https://trello.com/card/wip-limits/50e0afbecc2e335d2e000da3/23 [17:42] according to this it's done? [17:45] hatch stared at that but couldn't make much sense of it. Looked for wip before. https://trello.com/board/trello-development/4d5ea62fd76aa1136000000c is trello dev board [17:45] yeah the comments are....out there... but it was in the 'done' category heh [17:46] hatch, I agree with gary_poster that WIP limits are fundamental to the intent of kanban. I am glad to see a rival to leankanban though [17:46] but of what board, hatch? does that actually have anything to do with trello dev? I don't see any wip-style config in "list actions" [17:49] on the trello-development board under 'Live' there is (5/31) possible WIP limit? [17:49] I agree, I don't see it in the config either [17:51] oh that's probably a date [17:51] haha [17:53] gary_poster: I think here is the feed you were taking about https://twitter.com/spolsky/status/306719908245417984 [17:54] hatch, yup, that's it. [17:54] hatch, where's that "possible WIP limit" thing? don't see it in https://trello.com/board/trello-development/4d5ea62fd76aa1136000000c [17:54] yeah I don't either [17:54] I'm not sure what that other thing was from then [17:55] ah ok [18:03] I'm going to take off for lunch, bbl [18:07] sinzui: got a moment to chat? [18:09] I do [18:12] jcsackett, ^ [18:12] cool. [18:18] ccccccbljtrvgrbjhhhbvihigrddbeglevvfgnevhttk [18:35] OTOH, if leankit doesn't get their act together, I will be tempted to find other alternatives :-/ [18:50] debug prints (that raise exceptions) left in tests make Benji cry [18:55] gary_poster: I am in the process of rating a bunch of charms [18:55] yay! thanks jcastro :-) [18:56] jujugui, orangesquad, ding dong the witch is dead, or something. leankit appears to be back. [18:57] \o/ [18:57] Undead..? [18:57] :-) [18:57] thank you for your diligence gary_poster [18:57] jcastro: comming on your MP, I think there's a wire crossed [18:57] lol, I can reload web pages with the best of 'em [18:58] rick_h: what do you mean? [18:58] sorry, jcsackett ^^ not jcastro [18:58] jcastro: you carry on, rate away kthx :)( [18:58] heh [18:58] I like how you guys made you guys not landing ratings now blocking on us instead of you guys [18:58] well done sir. [18:59] * rick_h goes to look for anything else we can shove off on someone else [18:59] rick_h: my understanding was we used icon internally, and had the link used by clients of the API. [18:59] jcastro, :-P :-) [18:59] jcsackett: I thought icon was gone. it's no longer an attribute [18:59] or was going away, abentley can you verify? [18:59] * benji read that as "shave off of" [18:59] jcsackett: I thought sinzui was removing it from the model because it had no use [19:00] rick_h: could be; i missed that, if that's the case. [19:00] benji: I've got my merkur safety razor ready and willing [19:00] icon is gone from staging charm data and will be gone from production data in the next deploy [19:00] :) [19:00] jcsackett: ^^ [19:01] sinzui: i'm running off a fresh ingest, so icon is still being brought in somewhere. [19:01] jcsackett, migration 7 cleanup devs and staging, migration 8 cleans up production [19:01] and by fresh, i mean: i had no data before running this ingest this morning. [19:01] * gary_poster crosses fingers and toes on Makyo's behalf [19:02] Crossing everything, over here. Fingers, toes, the dogs' tails.. [19:02] rick_h: hey something to think about wrt. bundles [19:02] so I want to rate "logstash" [19:03] but it's a series of like 4 charms [19:03] maybe we should think about rating the entire thing as a whole [19:03] I think people will want a general view of quality of the entire service, not just the indexer, the webfront end, etc. [19:03] jcastro: hmm, not sure what's up with bundles at this point. I know we talked about forming a 'meta charm' format and getting it into the store like others and you'd just rat the 'meta charm' [19:04] oh ok [19:04] jcastro: so it would work just like it does now for you if that goes through, but I don't think we're doing that any more so not sure what the story is [19:04] well whatever we call them [19:04] jcastro: maybe gary_poster or sinzui can better fill you in what checklist to toss that onto [19:04] jcsackett, charmworld/charmworld/ingest only knows two things about icons. 1 ICON_FILENAME which can be removed (it is unused) and a test verifying icon is not in the charm data [19:05] jcastro I'll be sending an email out to juju-dev about bundles in the next few days [19:06] jcastro, I'll include that use case [19:06] the beginings of Go sandbox branch is up for review: https://codereview.appspot.com/10235045 [19:06] and you can follow on as you think makes sense. Does that sound good? [19:06] gary_poster: excellent, I will chime in as necessary [19:06] thanks jcastro [19:06] I hadn't thought about it until I started rating [19:07] then it's like, who cares about the logstash-indexer charm outside of it's stack [19:07] yeah [19:07] jcastro: yea, I know we talked about it from a charmworld POV in oakland [19:07] I should have taken a vacation day tomorrow. I'm getting a new mixing board and I can't wait to play with it. [19:07] benji: woot, what did you get? [19:07] heh [19:09] Behringer x32 (http://www.youtube.com/watch?v=A5v-JnAvipw) [19:09] whoa, nice [19:11] I have a Behringer Xenyx 1222. Love it :) [19:11] Boy howdy is that one nice, though, benji [19:11] yea, I was thinking like a little 8input board but that's crazy [19:13] jcsackett: by "icon is still being brought in", you mean the base64-encoded contents of the icon? That should be impossible. We don't even import base64 anymore. [19:14] I have 22 inputs but only a 16 channel board now, so the 40 inputs of that guy will be very nice to have; plus I get compression and a nice parametric EQ on each channel and 8 FX slots, so I can stop using over half of the stuff in my rack [19:14] abentley: i mean right now, i can do img src="data+xml/svg,{$charm ... }" to get an image in a template. [19:14] abentley: but it doesn't matter. i'll take it on faith that i've got something unusual going on and i'm going to change the branch. [19:19] jcsackett: You can look at the contents of the charm by accessing /charms/precise/wordpress/json . If there's a member named "icon" in there, I don't believe that was fresh ingest. [19:20] at least not with trunk r245 or a descendant. [19:28] benji: cool mixer :) I just got back from lunch....bought a new kiteboard lol [19:28] went out for lunch, forgot about lunch, bought a kiteboard....how does that work? [19:28] distracting hobbies are distracting [19:28] heh [19:29] haha yup [19:29] r u in a band? [19:30] you've heard of zz top? [19:30] he has a beard wig [19:30] lol [19:30] it's called 'benji top' and it's a cover band? [19:30] heh [19:31] hatch: I am the tech director (and primary mixer) for my church: 2 guitarists, keys, piano, drummer, and five vocalists [19:31] coolio! [19:32] (the base player will be back in the fall; mission trips, the bane of church music programs ;P) [19:32] haha [19:33] [wow, misspelling "bass" is a new low, even for me] [19:33] ba-dum-dum [19:33] it's alright I figured he wasn't the guy everyone stood on :P [19:33] "bass," "new low," get it? get it? [19:34] lol! [19:34] *rimshot* [19:34] :-) [19:34] heh [19:35] interesting email on ubuntu tech about git [19:35] and I mean't canonical tech [19:35] benji lgtm with questions [19:35] :) [19:35] * benji looks at questions. [19:37] oo I think I have finished this panel removal [19:38] annnnnd I spoke to soon [19:38] who needs tests anyways! [19:47] these tests are driving me bonkers [20:32] jcsackett: ping, just a double check. Did you make sure that hasIcon isn't dumped in the API and might cause front end issues? [20:32] does anyone know off the top of their head what causes the ghost services to be removed on cancel? [20:32] oh hmm [20:33] with my new iteration on canceling a deployment the ghost sticks around until the next delta [20:33] hatch, I think I knew at one point. Makyo should know. [20:33] One sec. [20:33] thanks [20:34] hatch, When the charm configuration panel is removed. [20:35] ohhh I forgot to fire 'update' [20:35] Okay. [20:35] thanks :) [20:36] I just needed that 'click' I guess heh [20:36] Makyo: the ghost placement is a little odd though [20:36] once deployed it bounces to the proper place though [20:36] known issue? [20:36] hatch, That was a content-free statement :) What's 'odd' mean? [20:37]