[00:20] hatch, they don't have infrastructure to deal with it, and have to get home somehow :P [00:20] rick_h_, http://www.buzzfeed.com/passantino/potentially-catastrophic-ice-storm-bears-down-on-the-south [00:20] Makyo they should have stayed home [00:21] I remember I was in Toronto once and looked at the news and was like .....nope, staying at the office [00:21] hatch, the city would've shut down in a place where the community would have a hard time with a city being shut down. [00:21] that sounds like a recursive problem [00:22] hatch, you live in a place with infrastructure to support the city shutting down, right? You've got the will and the way to stock up on what you need to deal with a snowstorm, whereas the south is lacking both of those. [00:23] Nevermind the whole problem of local and county level governments fighting with each other and screwing over departments of transportation, as with what happened in Atlanta. [00:23] yeah that's true that's true [00:24] ive been stranded in a car before, it's no fun, definitely not something you want to do if you can avoid it [00:27] now I'm trying to decide if I should buy this humble bundle https://www.humblebundle.com/weekly [00:30] Makyo if you enjoy card games you should check out hearthstone, its actually pretty good [00:30] I've been hearing good things about it, yeah [00:30] Ooo, Psychonauts. [00:30] Ive been playing the beta, I haven't spent any $ yet, but so far it's pretty enjoyable [00:30] haha yeah [00:34] heh it just glitched [00:53] Makyo think you'll get your branch landed and mine reviewed tonight? [00:54] I can never remember when your EOD is :) [00:55] hatch, Was an hour ago, but I'm usually around. QAing your branch now. [00:55] ohh ok so your EOD is an hour after mine, I can remember that I think :) [00:55] thanks === arosales_ is now known as arosales [01:49] Makyo since your branch landed is there a way I can re-trigger the CI in github? [01:50] hatch, yes, hold on. [01:50] I know I can if it fails the :shipit: CI....but not the pre-shipit....afaict [01:50] https://github.com/juju/juju-gui/blob/develop/docs/continuous-integration.rst#helpful-tips [01:51] ahh I gota log in and do it manually [01:51] The value you would use is origin/pr/125/merge [01:51] Uh, and then add ac omment saying it passes [01:52] ok sounds good [01:52] too bad we can't do a "rerun ci" :) comment [01:52] patches welcome :P [01:53] bastard! [01:53] lol [01:53] manually triggereded [01:54] :repeat_one: [01:54] haha yup [01:54] I gota go shovel my parents driveway soon...ugh it's over 2x as big as mine and mine took 45mins [01:55] so....much....snow [01:55] if we could reduce the http requests made in our CI we could probably cut a HUGE amount of time off of it [01:56] yea [01:56] all those 404's and the like [01:56] I wonder why it's so slow [01:56] sauce labs must be throttling it or something [01:58] stackoverflow.... 36M pageviews/day, 334M SQL queries, 23 servers. [01:58] "...strive to maintain 50ms..." [01:58] wowzers [01:59] wow, every pageview averages 10 queries? [01:59] seems like a lot for that many pageviews. You'd think you could cache/etc some to cut that down [02:00] yeah I'm not sure about that [02:00] it's very odd that they could do 10 queries per pageview and still get a 50ms response time [02:00] heh [02:00] http://nickcraver.com/blog/2013/11/22/what-it-takes-to-run-stack-overflow/ [02:01] SQL servers have 384 GB of memory with 1.8TB of SSD storage [02:01] HAHAHA [02:01] 384GB of memory [02:01] how the heck [02:03] 11IIS webservers....they could probably use half if they used nginx :P [02:27] rick_h_ will ci run with a :shipit: if I had to manually trigger the first test? [02:27] I ask because it doesn't look like it is [02:32] hatch___: yea, :shipit: runs any time regardless just by you hitting it [02:32] hmm it didn't, I deleted the accepted message in hopes it'll run [02:32] * rick_h_ goes to look [02:32] the only thing that blocks a shipit is if it has a 'merge request accepted' message already [02:33] so yea, it's running right now [02:33] you're good [02:33] it is? [02:33] yesa [02:33] http://ci.jujugui.org:8080/job/juju-gui-merge/ [02:33] http://ci.jujugui.org:8080/job/juju-gui-merge/117/ [02:34] origin/125/head [02:34] very odd I don't see it [02:34] http://ci.jujugui.org:8080/job/juju-gui/ in the list on the left [02:35] right, there's two jobs [02:35] one for pull request tests [02:35] and one for the merge/shipit tests [02:35] they're different because the merge one triggers the api to merge if it succeeds [02:35] ohh shoot then it was probably running already [02:35] while the test one does not, it's pushing back that the tests passed to the pull request [02:35] yes [02:35] my bad [02:35] all good [02:35] I hope it doesn't merge it in twice now [02:35] it'll handle/blow up itself [02:36] no, it'll be cool [02:36] oh good good [02:36] phew [02:36] :) [02:39] well now I know [02:40] yea, everything has the two jobs. normal and -merge [02:41] kind of makes sense [02:41] I just never thought of it before [02:41] hmm, failing IE tests? [02:41] hatch___: looks like there's an IE issue with your branch [02:41] see https://saucelabs.com/jobs/12ebdad9a20a429a96f161842e9157d0 [02:41] hatch___: mess with it tomorrow [02:42] it's way past EOD, get out of here [02:42] yup thanks [02:42] haha i'm playing hearthstone as well [02:42] k, cool then [02:42] looks like i'm gona lose this match [02:42] you can blame me :) [02:43] lol === _mup__ is now known as _mup_ [10:00] morning dimitern: I see that the charmsHandler parses the revision file in order to fetch the "bundle" revision (falling back to an obsolete meta.revision). How is the case where no revision file is present handled? [10:00] frankban, morning, in a meeting - will reply a bit later [10:01] dimitern: cool thanks np [10:08] ok so I guess the logic is in PrepareLocalCharmUpload. I guess revision is what found in the revision file or 0. Then if revision <= the max revision found in the db for that charm id, revision is set to maxRevision + 1 [11:04] frankban, sorry, i'm back [11:04] frankban, yeah, the revision from the charm is respected, but at the end an unique one is chosen, depending on what charms are in state already [11:05] * dimitern wonders when exactly did the revision file got obsolete [11:38] heh, thanks dimitern [14:26] rick_h_ so IE11 gives totally different failures than IE10, so I'm downloading Win 8 with IE10 now [14:27] hatch: k [14:28] it looks like the failures are because the version of YUI we use doesn't consider IE11 IE [14:28] lovely [14:28] hopefully an upgrade will fix that [14:28] well, that'll be part of updating for IE11 work. [14:28] yea [14:29] although I'm starting to get a little skeptical of YUI's development :/ [14:29] how so? [14:29] they were updating promises in the last sprint. I figured you'd be happy :) [14:31] well it seems to me that they are focusing more on Pure and other Yahoo projects than YUI [14:31] the progress on YUI has been very slow [14:42] modern.ie is pretty cool....just finished downloading the image...will see if it works :) [14:43] I remember the hoops we had to jump through to run various versions of IE [14:44] hatch: bad news: I think we have to scrap the "drop a directory to deploy a charm" feature. [14:45] It turns out that browser's completely ignore symbolic links when dropping files, which many charms depend on (the GUI charm included) [14:45] * hatch ragequits [14:45] ah crap [14:46] benji what if you symlink a zip file and just drop it? What happens then? [14:47] hatch: I don't understand the question. Do you mean create a zip with a symlink in it? === rogpeppe2 is now known as rogpeppe [14:47] benji no I mean I'm wondering if it's an issue with the folders or an issue with symlinks in general [14:47] so take a .zip of a charm and create a symlink to it [14:48] drag and drop that symlink, what does it do [14:48] hatch: I don't know what that does, but I don't see how it applies to this situation. [14:49] well in order to file a bug with chrome we need to know if it's actually an issue with chrome or an issue with the zip lib [14:50] hatch: it is chrome, when a directory containing symlinks is dropped the symlinks are not represented in the datastructures (only real files and directories) [14:50] cool the modernie vm's work [14:50] ohh ok cool, can you file a bug with chrome or would you like me to? [14:50] (and I just tried dropping a symlink to a zip and it is the same as dropping the zip itself) [14:51] ohh ok so it's a combination of a directory and symlinks then [14:51] well, not quite; I suspect if you drop a group of files and symlinks only the files will be recognized [14:52] for the link-to-zip example the browser never sees the symlink, just the contents of the zip [14:52] would you be able to do a bit of qa'ing and create a detailed bug report? [14:53] it's a story I think we would still like so we need them to fix it :) or at least monitor the status of the issue [14:54] is this true of ff as well benji ? [14:54] rick_h_: FF doesn't support enough of these new-ish APIs for this to work anyway [14:54] IOW, I don't know :) [14:55] rick_h_ I am able to reproduce the test failure in Win 8 IE10 so will get on fixing it now [14:55] benji: ok, then yea. Let's hit the stop button if we cna't support it. [14:55] benji: not sense spinning our wheels [14:55] benji: but thanks for taking it through and getting it as far as we can. It would be something cool to have. We should note where things stopped for future reference [14:56] wait, I get a totally different test failure [14:56] ugh [14:56] do we know what OS sauce labs IE is running on? [14:56] er [14:56] 7 or 8 [14:56] hatch: win7 and ie10 [14:56] damn [14:56] hatch: I think, check the browser.py file in tests [14:57] frankban: how does this effect your current/recent work? ^ [14:57] hatch: your hunch was a good one; dopping a set of files, one of which is a symlink works (i.e., the symlink is represented in the browser as a file with the contents of the pointed-to file), but when the directory containing those files is dropped, the symlink is missing from the set of files [14:57] I'll look to see if this bug has been reported in Chrome yet. [14:57] https://code.google.com/p/chromium/issues/list [14:58] benji thanks.....this is unfortunate :( [14:59] rick_h_: reading [15:00] hatch: benji frankban time for a hangout? I want to chat if there's another way to cover the use case [15:00] rick_h_: sure [15:00] hop in the staundup hangout please [15:00] rick_h_: sure, just a minute to read [15:02] rick_h_: ready when you want [15:02] frankban: cool, https://plus.google.com/hangouts/_/calendar/Z2FyeS5wb3N0ZXJAY2Fub25pY2FsLmNvbQ.j0rk5d371ph8331ijtf48t2uj0?authuser=1 [15:03] benji: ^ [15:06] rick_h_: my machine froze, one second [15:18] *sigh* IE testing [15:19] test broken....put a debugger.....test fixed [15:19] if it wasn't for IE you wouldn't be able to be a Web Development Engineer, just "Web Dude" [15:20] lol...actually I'm a Web Master, I've had enough years experience copying files via FTP and doing folder based version control [15:20] haha [15:22] hatch: if you're getting different errors in win7 and win8 we should try to see if we can make both pass ok [15:22] maybe add both to CI if they're acting that different for IE10 :/ [15:23] rick_h_ yep I'm fixing the test in Win 8 right now....it's a stupid failure, the widget positioning is off, so I am changing the assertion to a range [15:23] ugh, how far off? e.g. does it look right in real life to the user? [15:23] we need to get in touch with someone to find out which version if IE we are supporting [15:23] 2px-ish, it changes every run [15:23] k, well right now we're doing IE10 [15:24] that should be regardless of win7 or 8 [15:24] yeah, I'm not even sure you can get IE10 in windows 8 though [15:24] I think it auto updates [15:24] at least it did for me... [15:25] I'm fixing the test regardless....but curious [15:25] ugh, ok. Yea, I mean the goal is to get IE11 support in, but it's after wrapping up local charms/relation lines, and starting inspector on left === lazyPower_ is now known as lazyPower [15:50] jujugui call in 10 === makyo_ is now known as Makyo [15:50] Much better. [15:51] :) [15:51] Makyo: can you run the standup? I've got the IS meeting and will try to hop over before the end [15:53] Yep, sure thing. [15:58] jujugui call in 2 [15:59] benji ^ [16:08] benji when you get around to filing that chrome bug can you link it plz so I can star it [16:08] that's two chrome bugs in a week....jeesh! [16:08] we are bleeding edge yo! [16:09] Makyo I ended up picking this up https://www.humblebundle.com/weekly last night [16:09] standup over? [16:09] rick_h_ yup, redo? [16:10] hatch: naw, I'll add notes for tomorrow. [16:10] ok [16:10] rick_h_ did you create the juju-core card in Project 1? [16:11] hatch: yes, it's realted and we're not tracking any more. It's something we can work on/do [16:11] so moved it out of tracking and into todo [16:11] ok it's not supported by juju-core right now [16:11] right, the card is to add support [16:11] to core [16:11] ok I'll put frankban's head on it [16:12] well if benji or someone else wants to hack on it they can :) [16:13] oh I didn't know anyone else had enough juju-core experience [16:13] meh, we've got to learn and get more [16:13] we should have a lesson to get more of us up to speed [16:14] a 1 week sprint in Hawaii to get us up to speed on it :P [16:14] heh, feel free to submit it...on a day while I'm away on holiday [16:14] lol [16:17] rick_h_, https://bugs.launchpad.net/juju-gui/+bug/1279873 [16:17] <_mup_> Bug #1279873: Bundle deploy instructions still refer to beta PPA [16:18] we need this one fixed before EOM too [16:18] jcastro_: yea, I'm waiting to see if it's in universe or gets to main [16:18] jcastro_: because in trusty the stable ppa isnt' needed any more either right? [16:19] guess we cna just roll with it [16:19] correct [16:20] it won't matter if it's in main/universe for the end user as both are enabled by defaulty [16:20] just not installed by default [16:20] k [16:28] wow time warner is buying comcast? wow yet another huge hit to net neutrality [16:34] http://gizmodo.com/the-40-best-tweets-about-the-comcast-time-warner-cable-1522034013 lol [16:41] git help: I tried "git push juju archived-drag-and-drop-directory-to-deploy-charm" to push my branch to the main project but it complained about the branch name (error: src refspec archived-drag-and-drop-directory-to-deploy-charm does not match any.) Any ideas? [16:42] benji did you create the branch first? [16:43] git push origin current_branch:refs/heads/branch_to_create [16:43] that's what grb runs to create the remote branch [16:43] hatch: nope; something like "git branch juju NAME-OF-BRANCH" [16:43] but for you origin will be your fork (likely) [16:44] origin == juju [16:44] yeah that won't create the remote branch [16:44] "git push juju current_branch:refs/heads/archived-drag-and-drop-directory-to-deploy-charm" results in the same error [16:45] git push [16:45] git push

: [16:45] git push : [16:45] ah, push; trying that [16:45] benji: ^ [16:45] :D [16:45] https://help.github.com/articles/pushing-to-a-remote [16:45] apparently the new Windows 7 vm I just started thinks that I have an HP printer [16:46] I wonder if I send pictures to it where they will get printed out too [16:46] lol [16:46] that worked; thanks guys [16:46] :) np, git is an odd duck [16:46] but you can pretty much do whatever you want if you know the commands hah [16:48] I continue to belive that git has more power than I want. [16:48] oh, and the UI sucks [16:48] true....and true [16:48] although...git is just a db....so it really needs an ORM....like GRB (which I use) [16:48] er....ORM-like-tool [16:48] benji: it gets to be like vim, you find the 11% you use [16:49] 11%? That's pretty specific :) [16:49] except that I like power in my editor [16:49] hatch: he thinks internally in fractions and converts to percentages for presentation; he uses 1/9th of vim [16:50] haha [16:51] rick_h_ u so smart [16:52] son-of-a, I get no failures in WIn 7 IE10 [16:52] well I'll push my changes and see what happens [17:14] jujugui machine view run through with UX 8:30am est tomorrow if anyone can/wants to make it [17:15] oh balls that's early [17:15] what's the current time in EST? [17:15] 12:14 [17:15] (as a international company we should really deal in UTC) [17:15] yea, we should but I fail [17:15] ok 7:30 for me, not so bad, I thought it was 2H [17:15] I've got to learn [17:15] haha so'k I'll let it pass.....this time [17:15] heh [17:15] rick_h_ send me an invite i'll be there [17:16] hatch: rgr [17:31] wth I think sauce labs is broken or something [17:31] incredibly frustrating [17:33] hatch: port forward and run make test-server locally and use the 'start a sauce instance' to debug from sauce? [17:33] rick_h_ do you know how I may run our unit tests using sauce labs on IE 10? [17:33] heh [17:33] hatch: :) [17:33] are there docs somewhere? [17:33] no, they're kind of custom [17:33] hatch: I use https://chrome.google.com/webstore/detail/sauce-launcher/mmcebionhdleomnkegjcoadpghnmcebl [17:34] and launch a sauce instance of the right machine type [17:34] then in that remote window go to http://yourhomeip:8888 (you setup port forward from your router to your dev machine) [17:34] and you can control the sauce machine and run the tests/debug your local running instance [17:35] ohh yeah that's not going to work unfortunately....my router is a pos and won't let internal connections no matter what I do [17:35] hatch: then run it from an ec2 instance? [17:35] or something [17:35] maybe I can spin up a ec2 instance [17:35] :) [17:35] ugh I hate it when tooling gets in my way [17:35] just remember in ec2 you have to expose the port in the security group [17:35] time to router upgrade [17:37] it's also the modem [17:37] no can do [17:37] hmm it's not accepting the juju-gui creds from the wiki [17:37] hatch: right, what I do is set it up to send all traffic thorugh and put my own router right inside the one given to me [17:37] ohh I see do a DMZ to your own router [17:37] yeah I should do that [17:38] hatch: yea, I don't trust their routers they make me have [17:38] well mine is definitely a pos [17:40] great the sauce launcher thing won't accept the creds [17:40] * hatch flips desk [17:40] hmm, sec [17:41] it's ok i'm doing it through the actual sauce labs website [17:41] now [17:42] k [17:42] yea, seems it doesn't like the password from the extension now [17:58] wow ec2 really likes to hide what the username is on these instances hey? [17:58] ubuntu [17:58] always ubuntu? [17:58] or you mean the dns names? [18:00] rick_h_ well their documentation says ec2-user [18:00] apparently it's different for every distro, but there is no documentation to that effect [18:13] *sigh* cannot connect to the ec2 instance [18:15] hatch: you opened the port? [18:15] in the security group? [18:15] yeah it seems the entire instance crashed [18:15] it kicked me out of ssh [18:16] and now I can't get back in lol [18:16] today is not my day haha [18:17] ok trying with a new image [18:17] 3rd time is the charm! [18:19] hatch: you know there's official ubuntu images you should use right? [18:19] http://cloud-images.ubuntu.com/locator/ec2/ [18:19] I just picked 12.04 from the list in aws [18:20] but very cool [18:20] I didn't know about that list [18:20] yea, we put out official ones that always have the same setup (thus always ubuntu user) and usch [18:20] such [18:21] ok finally got it [18:21] yay [18:21] now to sauce lab [18:21] Makyo: you're on the databinding card? Can I stick your head on there? [18:21] rick_h_, oh, whoops. Yeah [18:21] cool [18:22] Makyo: is this the last card? [18:22] rick_h_ the tests pass.... [18:22] hatch: they don't run that fast :P [18:22] it's well past the ones which failed [18:23] hmm ok well I guess I can trigger another jenkins run and see if it works [18:23] hatch: ugh ok. so you've got a {'platform': 'Windows 7', 'version': '10'}, selenium instance passing on your branch? [18:23] is it somtehing to do with prod vs debug then? [18:23] rick_h_, maybe. There might be a clean-up branch with a few small things, but if needed, we can release the feature flag before then. [18:24] Makyo: k, I'm going to move it to maint then so I can start adding cards in here today/tomorrow for the machine view/inspector stuff if that's cool? [18:24] hmm maybe it's a prod vs debug thing [18:24] rick_h_, Yep! [18:24] Makyo: thanks! [18:25] trying prod [18:28] ok prod passes as well [18:29] rick_h_ https://saucelabs.com/tests/340a9b1951a649a1957f624555cc4254 see 100% IE10 WIn 7 [18:29] I'm out of ideas now [18:29] :) [18:29] hatch: ok, so final thing that'll rely on timing :) [18:30] hatch: trigger build, watch when it starts the IE tests and outputs the url to the console Starting unit tests against IE [18:30] hatch: and go to the url on the CI server [18:30] hatch: http://ci.jujugui.com:8888/test/index.html [18:31] well ok...but why? We already know it fails because you can see in the video that there is the error [18:31] * rick_h_ gets the popcorn to watch it [18:31] hatch: meh, yea I was thinking your's is working but it's not different [18:32] I will though, maybe it will give SOME indication [18:32] heh [18:32] maybe it's loading a different version of IE or something [18:33] * rick_h_ goes to look at your diff to look for setHTML bits [18:37] I 'think' the error is coming from the Y.one().append() [18:37] but that should be stubbed out [18:37] and is everywhere else... [18:38] 'should be stubbed out'? [18:38] yeah, like changed for a stub method [18:40] append shouldn't call setHTML though. Looks like it uses addHTML [18:40] yeah that makes more sense doesn't it? hah...that's all I had as far as ideas where setHTML comes from [18:40] if (node.nextSibling) { // IE errors if refNode is null [18:41] try to go about that a diff way [18:41] add it to a diff node, not body [18:41] isn't there a container or anything we can use? [18:41] well it shouldn't be doing any of that [18:41] I stubbed out Y.one with a new object with an append() function [18:42] so it shouldn't be actually 'appending' anything [18:42] or one()ing for that matter [18:42] hmm, worth a debug then I guess. [18:42] to trace the call in the test [18:42] why would you stub out Y.one? I don't follow [18:43] shhh watching movie [18:43] ;0 [18:43] :) [18:43] heh [18:46] apparently it was a schrodinbug the act of watching it caused it to change....and pass [18:47] ugh [18:47] rick_h_ quick hangout I'll explain the Y.one stubbing? [18:47] hatch: k, yea I've got a couple of ? [18:47] https://plus.google.com/hangouts/_/76cpi0qie5fr86lftidmcjfgcg?hl=en [19:12] * hatch remembers to go shut his ec2 instance down [19:12] good call [19:13]