[00:31] Makyo: that error happens once in a blue moon and seems saucelabs related [13:22] guihelp: anyone available for reviewing/QAing https://codereview.appspot.com/83880044 (quickstart)? thanks! [13:32] frankban: i will shortly [13:32] bac: thanks [13:56] morning [13:57] morning hatch [13:57] hey rick_h_ when you did the nr_cpus=1 did you just put that in the list of the commands when editing the startup? [13:57] hatch: have to put it in the right place [13:57] where it says splash and such [13:58] oh ok, I thought that might have been the case - I coudlnt' find any docs online about it [13:58] change those to noslash noxxxx and add nr_cpus=1 [13:58] nosplash that is [13:59] ok thanks - man I just want this darn thing on Ubuntu lol [14:00] lol yea [14:00] thanks for following up and trying [14:00] I've been knocked out dead early this week [14:00] https://bugzilla.kernel.org/show_bug.cgi?id=60635 [14:00] I am using redfind already....not sure how else to do an efi boot [14:01] yea, I think we need to make that clear [14:01] the bug linked might not be valid because he's missed we're using refind already [14:01] I wish I knew -anything- about kernels lol [14:05] Anyone know why MachineList is a LazyModelList and not just a ModelList? [14:06] Doesn't seem like the performance advantage of LazyModelList would be needed here [14:07] And I'm depending on change events firing on the machines in the list in order to keep the MachineView up-to-date, which LazyModelList makes less likely. [14:09] kadams54 there could be thousands of machines [14:09] s/could/will [14:17] hatch: scratching my head over YUI stuff - could I have you take a quick look? [14:18] 2 mins just gota grab my breaky [14:18] np [14:20] ohh kay link? [14:21] http://pastie.org/private/hu1ljam1js0lhdtxozxdbw [14:21] yep [14:21] that's right [14:21] :) [14:21] When I create a Machine and add it to MachineList, the raw object that I get back out of the list has a null displayName [14:22] try foos.create() [14:24] are the returned values actually strings? or is that just in your paste? [14:24] yes [14:24] actually strings [14:25] Paste is taken directly from web inspector console [14:25] does create help? [14:25] No [14:25] hmm ok lemme look at the lml code [14:25] http://pastie.org/private/qgxsrjhx5canncqqdqn7ug [14:26] ok I'm going to fire up a jsfiddle here [14:26] or jsbin for that matter [14:26] kadams54: foos.items(0).get('displayName') [14:26] it a pojo [14:27] * rick_h_ is guessing displayName is an ATTR? [14:27] oh nvm then :) [14:27] bad guess [14:27] so'ok! [14:27] LazyModelList stores the objects as POJOs instead of full Models for performance reasons [14:27] kadams54: the displayName is added to machine objects by MachineList.add [14:27] kadams54: it calls _setDefaultsAndCalculatedValues [14:27] Ah [14:28] And that's probably overwriting my "0" with a calculated value [14:28] oh there we go :) I was just writing a repro with raw YUI [14:28] heh :) [14:28] Which is why we get a "null" string [14:28] Instead of null [14:28] thanks for stepping in frankban [14:29] bleh lifestyle changes taste so bad... [14:29] #dietingisforsuckers [14:29] heh [14:30] Thanks frankban - that was the missing piece [14:31] welcome [14:50] jujugui call in 10 [14:50] I think Makyo has his alarm set for 11min vs 10 :P [14:52] 10 minutes by my phone, just whenever that goes off. [14:52] (but now I'm totally doing that. Thanks rick_h_ :D ) [14:58] rick_h_: Makyo hatch can we have a call once your off your call? :) [14:59] we can [14:59] luca__: rgr [14:59] thanks [15:01] bac: http://bazaar.launchpad.net/~go-bot/juju-core/trunk/view/head:/doc/bazaar-usage.txt [15:04] thanks frankban [15:11] luca__: https://plus.google.com/hangouts/_/canonical.com/daily-standup [15:12] just saving my doc [15:22] * frankban biab [15:54] brb [15:58] rick_h_: so, should we keep joyent support out for now? I know our exception is for fixing quickstart on newer juju-core. I wonder if that's to be intended as exclusive i.e. only the buf fix, no other developments/changes added in the meanwhile [15:58] I was just getting ready to submit a pull request, but after rebasing to latest on develop, I can't get to the machine view anymore [15:59] kadams54: hit the machine view tab in the nav [15:59] Yeah, I did. [15:59] Nothing [15:59] kadams54: oh hmm, that works in comingsoon [15:59] http://comingsoon.jujucharms.com/:flags:/il/mv/ [15:59] hit machine view button? [16:00] kadams54: if it doesn't work locally make clean-all maybe? Or check for errors in console/etc [16:00] Will do [16:01] *sigh* [16:01] Error in my code [16:01] frankban: can you write up an email about the joyent provider now landing and our wish to have that in quickstart to james page and robie and copy me? [16:01] frankban: we'll ask if it'll be possible, if not. Let's hold on landing that branch until the bugs land first [16:01] so we can release just what we promised them, a bugfix release [16:02] frankban: and if so, then we can follow up with the joyent under whatever rules we can get that in under [16:02] frankban: and if not, then landing joyent in quickstart will be the start of the new ppa releases, but not in trusty [16:02] rick_h_: sounds good [16:10] guihelp https://github.com/juju/juju-gui/pull/222 is ready for review and QA (list machines in machine view) [16:11] kadams54, on it [16:14] thanks [16:21] jujugui going to go afk for some lunchable. biab [16:21] enjoy [16:22] kadams54: so, do we display both top level machines and containers in that part of the machine view? [16:23] Not entirely sure. There's a machine column and a container column, so I'd guess so? [16:23] My card was for listing the machines in the machine column, so the container column is still non-functional [16:26] kadams54: if I read your diff correctly, you are listing all the machines (db.machines.toArray()). That would include also the containers. It is possible to only list top level machines by using machines.filterByParent(null) [16:27] Ah, did not realize the collection contained both machines and containers… [16:27] Is there a reason containers aren't in a separate collection? [16:28] kadams54: it's not a separate entity in juju-core itself, and also they share almost all the properties. it's a tree, where a phisical machine can host containers and containers can host other containers etc. [16:38] rick_h_ “Arguing with an engineer is like wrestling with a pig in mud – you realise far too late it is enjoying it” rofl [16:45] frankban: got it. Will modify the code accordingly. [16:45] kadams54: cool thanks [16:46] hatch: Nice. I'll have to remember that one. [16:46] I want a tattoo of that haha [16:46] Your wife might argue with you about that. [16:47] what if it's in Japanese? Chinese? Then I'd be part of the cool kids [16:47] I'd just tell everyone it means "Love" [16:47] lol [16:49] Makyo I just booted up vagrant and it's creating a new image..... [16:49] any idea how I would point it to the old one again? [16:49] Um. [16:50] I think it's in a .vagrant or something folder in ~? [16:51] hatch: :) [16:52] rick_h_: had a chat with robie in #juju, joyent is out, the other parts of the plan sound great to him. I think landing joyent later next week is fine. [16:52] frankban: rgr thanks [17:00] rick_h_ want to have a quick discussion about this branch [17:10] hatch: sure link [17:10] https://plus.google.com/hangouts/_/7ecpj9soau6943l97kuv7qnknc?authuser=1&hl=en [17:23] hi frankban, when bootstrapping joyent, after creating an environment with quickstart, i get http://paste.ubuntu.com/7199671/ [17:24] bac: uhm, is this the 1.18 branch? [17:24] i think so. let me double check [17:25] frankban: yes, it is using the version i built using the 1.18 branch. however, 'juju version' shows 1.19.0-trusty-amd64 [17:26] frankban: i merged 1.18 into my lw checkout of juju-core trunk. is that correct? [17:28] bac: I don't think so, you should just branch the 1.18 branch and switch to that. I guess trunk includes other changes [17:28] bac: my 1.18 branch is 1.17.8-saucy-amd64, bzr revno 2257 [17:29] frankban: i'm unsure how to do that with lightweight checkouts. usually i do a 'bzr switch -b', which gives me a copy of trunk. how would i get a checkout of 1.18 in place? [17:30] bac: also, before compiling "godeps -u dependencies.tsv" can be a good sanity check (go get launchpad.net/godeps) [17:32] poor bws-icon [17:32] delete! [17:32] bac: I followed the bzr docs in trunk, and so I created a shared repo outside the $GOPATH, in ~/devel/juju-core. There I can just bzr branch some-branch. at that point you can "bzr switch {target}" from the juju-core in $GOPATH/src/... [17:33] frankban: ah, gotcha. i'm used to just working in the other workspace [17:35] bac: yeah the configuration in that doc is cool because you have both the structure go requires ($GOPATH/src/launchpad.net/juju-core) and a real shared repo to work on, given that the former is bound to the real repo as a lightweight checkout. So, basically, $GOPATH/src/launchpad.net/juju-core is just what we usually call "sandbox" [17:36] frankban: right. i just didn't know you could manually insert a different branch in ~/src/juju-core [17:36] time to go, have a nice evening [17:37] frankban: wait! [17:37] * frankban waits [17:37] frankban: for joyent are the sdc user/keys the same ones for manta? [17:37] bac: yes [17:37] ok. perhaps we can make that clearer in the field descriptions [17:38] or maybe make the manta ones optional and force them to be the same if blank [17:39] bac: yes, that can be an improvement. FWIW my envs.yamls looks like http://pastebin.ubuntu.com/7199750/ [17:46] frankban: if you're still around have you seen this before http://paste.ubuntu.com/7199773/ [17:47] Makyo, frankban: FYI, updated https://github.com/juju/juju-gui/pull/222 to only pull top-level machines. Also realized I had one commit that didn't get pushed. [17:47] bac: no, but I guess your previous attempt left something there, and juju-core got confused by the binary change. [17:47] kadams54, cool, thanks. Was having trouble on this comp, switched to Air to find vagrant out of date. Will switch back. [17:48] frankban: huh. ok, i'll poke around. thanks. won't bug you again. good night. [17:48] bac: you can try "juju destroy-environment joyent -y --force" and check nothing is in ~/.juju/environments/ [17:48] kadams54: tossed a few comments in for fun [17:48] kadams54: which might be out of date now it sounds like === lazyPower is now known as lazyPower-travel [17:52] bac: cool, there is no rush in landing this branch, so if you are blocked we can take a look at that together tomorrow [17:53] rick_h_: yeah, saw those. I caught the constructor one early enough, but will update the tests as requested on the next pass through the code (after another other reviews). [17:53] kadams54: cool [17:53] kadams54: remind me in the call I'll mention how the initializer and some other 'magic' methods work [17:54] think C classes [17:54] that's how initializers work :P\\ [17:56] hatch: I haven't programmed in C since college, over a decade ago :-) [17:56] hatch: right but other things like render and such can be cascaded (at least for widgets/etc) [17:56] base derrived classes yeah [17:56] kadams54 haha yeah same here [17:56] Yeah, I was thinking that constructors needed to cascade when I wrote that [17:57] Then I remembered this is YUI, not plain JS. [17:57] there's a good yuiconf talk on video for how base works [17:57] And this was an initialize method, not a constructor [17:57] Hit me with the link [17:59] rick_h_ kadams54 https://docs.google.com/file/d/0By84yWo_5i6LYXNsdnhHV2Z2TUU aww yeah C++ [18:00] see that one has a 3.5" floppy on it! [18:00] lol [18:18] kadams54: https://www.youtube.com/watch?v=_zhQIfT7g58&index=6&list=PLF49A40DA7A11F08A ftw [18:19] Woot [18:19] ohh yeah I remember that talk [18:19] one of my fav yui talks [18:19] yeah luke is great [18:19] at smugmug now [18:54] rick_h_: just a heads up, manage.jujucharms.com/heartbeat currently shows an API2 failure. it is a false positive caused by us currently not featuring any charms, just bundles. [18:57] No plumber yet, but we did get our new doors delivered. Bonus. [18:59] Makyo might be time to fire this guy [18:59] :) [18:59] Makyo: so you've had this leak at the meter for weeks now? [18:59] Yep. [18:59] yowza [18:59] I just got a water bill, actually. [19:00] $286.04 [19:01] https://www.dropbox.com/s/jregju31mdzuuhc/IMG_20140403_130049.jpg [19:02] ouch! [19:06] rick_h_ https://github.com/juju/juju-gui/pull/223 WIP [19:06] this doesn't have the tests fixed but I'd like some third party input [19:18] hatch, ping when you have a second. [19:19] Want to get a good ECS card. [19:19] hmm I'm just cooking lunch atm [19:20] can we do it in 45mins or are you not doing anything now? [19:20] Will do lunch too, finish this review. [19:20] 45 mins is fine [19:20] sounds good [19:35] rick_h_: i'm stuck working with the juju guys trying to get the joyent provider to work. that's kind of like QA, isn't it? :) [19:47] bac: hah, sorry catching up [19:49] bac: oh lol for your qa day, yea we'll call it [20:02] Makyo whenever you're ready, rick_h_ you may want to be part of this discussion as well [20:03] hatch, sure [20:03] hatch: rgr [20:03] creating room [20:03] https://plus.google.com/hangouts/_/7acpjdgv54v5pnjt0fc9v0snjs?authuser=1&hl=en [20:17] rick_h_ btw there have been some developments wrt the kernel bits - tonight I'm going to give them a try [20:17] hatch: thanks, let me know how it goes [20:17] apparently both of us have been installing Ubuntu incorrectly [20:18] we need to use EFI mode instead of legacy [20:18] yea, not sure how that happened [20:18] yeah I'll definitely keep notes [20:18] didn't really see an 'efi' mode during install from the usb cd [20:18] yeah neither did I... [20:40] w00t down to 9 failures [20:45] hatch: that sounds darn right reasonable :) [20:45] :shipit: amiright? [20:45] and lies [20:47] oo the tests found a real bug [20:47] lookatthat [20:48] #so-thats-how-it-works :) [20:49] haha [20:49] as much I tests are a pita, I so love them when they catch a bug before it lands [20:57] somhow I deleted three tests and fixed one and still have 8 failures.... === tvansteenburgh is now known as tvan-afk [21:36] rick_h_ in test_browser_app.js browser subapp display tree there is a `hits` object [21:37] why do they default to true? [21:37] That seems backwards? [21:37] I feel like I'm missing something [21:37] hatch: k [21:38] hatch: they default to true? [21:38] hatch: there's a 'resetHits() method [21:38] ohh ok I broke something [21:38] right [21:38] that gets called in setup and manually for multi dispatch tests [21:38] right [21:38] sorry [21:38] hmm [21:41] test failures when things work in the real world...not fun [21:41] hatch: let me know if you want another set of eyes [21:48] thanks but I just skipped the final two because they would need state [21:55] rick_h_ ok pushed....would be awesome if you could find some time to qa/review [21:55] keeping in mind the follow-up will be to properly implement the state [21:55] hatch: rgr [21:55] thanks [21:55] yea, realize the whole 'test and build state' card here didn't get there because of the viewmode removal [21:55] but smaller branches == better [21:56] yeah agreed [21:57] I did remove over 1000 lines in total [21:57] that's awesome lol [21:57] yea, I mean this is good, just time consuming [21:57] between this and inspector we're paying a price for the better code hopefully [21:58] but let's try not to find anything else to rewrite :) [21:58] lol is there anythign else? [21:58] shush, you'll send back luck our way [21:59] haha [21:59] hehe...except I'm not kidding [22:01] Morning [22:01] morning huwshimi [22:03] rick_h_: Has our ci broken? [22:03] huwshimi: not that I'm aware of. What's up? [22:03] huwshimi: you had a legit failing test. You replied that it could be removed, but never updated and removed the test [22:03] rick_h_: Well, all the current MPs have builds that have started but not completed. [22:03] huwshimi: I was going to ask what happened yesterday and if there was something that got stuck [22:04] huwshimi: oh heh, yea that's broken [22:04] sec, have to stop [22:05] huwshimi: should be good to go hopefully [22:05] Ah great [22:15] hatch: replies in bound, will start qa [22:15] thanks I'll take a look [22:15] rick_h_: What should I tackle today? I see kyle is already working on the machine panel tokens... [22:16] huwshimi: he's working on the data, but not the tokens [22:16] bah, he's gone. I thought he would get that landed [22:16] huwshimi: what were you working on yesterday? Is that done then? [22:16] rick_h_: Ah, I thought there were two cards about tokens and now there's only one so I assumed he must have done it. I must have mis-remembered. [22:17] huwshimi: ah well his card says 'using deployed unit token' but it's not ready yet so he's just dumping out text [22:18] huwshimi: so I'd say to start looking at the token I suppose. [22:19] huwshimi: https://drive.google.com/a/canonical.com/?urp=https://docs.google.com/&pli=1#folders/0B_l975nRB6BuM2p4eTZMSXpWRTQ [22:19] we can go over them during hte call [22:20] rick_h_: That'd be great, thanks! [22:28] hatch: qa notes added [22:29] hatch: I ack that not all are in/for this branch. Just noting things that are broken. [22:29] ok cool [22:29] looking [22:30] line reflow? [22:31] otp [22:31] I actually prefer when there isn't one long line and then another word on the next line :) [22:31] hatch: oh yea, well the limit is 79, not 20 and 15 [22:31] limit.....not required :P [22:32] it looks odd imo and breads readability and I bet one or two of those will fit on one line now [22:32] fighting on everything :P [22:34] hah no it's just going to take me a while to reflow all of these [22:34] open in vim, run gq [22:34] it's one command :) [22:35] that doesn't require a plugin? [22:35] no, I don't believe it does [22:35] hmm ok I'll have to download vim [22:35] http://www.methods.co.nz/asciidoc/chunked/ch36.html [22:35] w...t...f doesn't it exist in your virtualmachine? [22:35] or did you just want me to do those two lines [22:35] ssh, vim xxxx.js [22:35] not the entire file? [22:35] the cases of those comments [22:35] oh ok, then done [22:35] :) [22:36] some of them jump out as looking stupid so I noted them in review [22:37] that's an interesting functionality of vim [22:37] to reflow a line [22:37] very handy when editing blocks of text/comments and emails [22:38] yeah if you don't use a real client :P [22:47] rick_h_ so changes are made - I'm not clear on your qa notes, did you want me to fix those things? [22:51] hatch: https://plus.google.com/hangouts/_/calendar/cmljay5oYXJkaW5nQGNhbm9uaWNhbC5jb20.dd77sn7kjl6unba21lutdr0p70?authuser=1 [23:07] is there an aus call? [23:07] or did I miss it? [23:38] achievement: Ubuntu 14.04 on MBP [23:46] hatch: missed it, it before you and I talked [23:50] hatch: Did you get all the touch packages installed as well? No idea why I have them... [23:51] huwshimi I don't know what that is [23:52] :) [23:52] every few minutes it pops up with a system error report [23:52] so it's not really usable [23:52] hatch: You don't know what Ubuntu Touch is? [23:52] and I can't actually shut it down outside of the terminal either [23:52] oh the Ubuntu Touch packages [23:52] I don't think I got those [23:53] So strange... [23:53] I have to find where I file bugs [23:53] the inability to shut down is kind of a big one [23:54] hah [23:55] switching to wifi [23:56] sudo shutdown -h now [23:56] or per zsh alias 'shut' [23:56] kadams54: hey, around? [23:56] I am now :-)_ [23:56] kadams54: how's your branch? Is it ready to land? huwshimi is doing some work with the token part that will build on it [23:57] kadams54: I had thought your branch might get in before he got running [23:57] Ah, yeah… let me get that pushed out [23:57] kadams54: ty much [23:57] huwshimi: ^ [23:57] np [23:58] huwshimi: what did you need to do? Just install the packages mentioned in the bug report? [23:58] kadams54, rick_h_: Brilliant, thankyou! [23:59] rick_h_: Somehow when I upgraded to Trusty it installed all the Ubuntu Touch packages as well. [23:59] huwshimi: hah [23:59] huwshimi does it do anything different with those packages?