=== Makyo is now known as Makyo|out [12:44] * bac is back [13:11] frankban: if you have a couple of minutes, I have a branch up for review that fixes the charmheplers problem with log messages containing "--" (bug 1130793). The MP is at https://code.launchpad.net/~benji/charm-tools/bug-1130793-add-log-option-escaping/+merge/150442 [13:11] <_mup_> Bug #1130793: For safety, Python charm-helpers juju-log command should insert a "--" before the log message. < https://launchpad.net/bugs/1130793 > [13:12] benji, I don't think any of us are reviewers. Was thinking about that last night after the baby woke me up again :-P We may need to ping Mark Mims [13:12] gary_poster: ah; will do [13:21] benji, utlemming *might* also be able to give a review. both he and m_3 are out west though (UT and CO respectively), so probably won't show up till later. [13:22] gary_poster: yeah, after I invoked him I wonderd where he was and looking him up; I'll try again later [13:23] cool benji. [13:40] bac: feel free to submit that CL [13:41] rogpeppe: already done. thanks. [13:42] bac: ah, only just showed up in my inbox. thanks. [13:47] hey frankban and teknico. I was thinking of asking Roger if the watcher branch is far along enough for the three of us to talk with him about what you might do to start exploring the mega watcher. Does that sound good/interesting/workable? [13:47] gary_poster: sounds good [13:47] cool [13:47] gary_poster: do you have a minute? juju-gui? [13:48] gary_poster: the watcher branch is hopefully ready to be submitted - just a prereq to be submitted and a final signoff required. [13:49] sure frankban great rogpeppe. you available for a call with us soonish? Maybe in half an hour or hour, so we have time for teknico to return, and to review the code some more [13:49] gary_poster: yeah. i've got a in 70 mins, otherwise i'm free. [13:49] cool thanks [14:09] about 2" of snow lastnight....of course right after I washed my truck hah [14:12] gary_poster, I'm back [14:13] cool teknico. frankban, teknico, how do you want to do this? I propose that the three of us get on juju gui and look together at https://codereview.appspot.com/7390043/ and talk about it; once we feel good, we ask Roger to join us. I'm open to ther plans too. :-) [14:14] other [14:14] gary_poster: sounds good [14:14] gary_poster, SGTM :-) [14:15] cool frankban teknico let's do it! I'm there [14:18] gary_poster: i'm afraid that CL is showing spurious noise because of the undeclared prereq (oh how i wish we could add prereqs after creating a merge request!) [14:18] rogpeppe, s'ok :-) we're doing our best and will call you in when we give up/understand everything [14:26] is ther any way to adjust the frequency of emails from a group on lp? [14:29] bac: I added some notes to that document [14:31] benji: thanks [14:35] gary_poster: so I have been attempting to write proper tests for this app-extension branch - but I can't really do it properly without the sub-app class and fake subapp which are in other branches. What is the policy regarding this issue? [14:37] hatch will ping [14:44] gary_poster: re-warning: i'm in a call from 15 minutes from now, for probably around 30 mins. [14:45] rogpeppe, ack, we just finished our call, and since you only have 15 minutes now, I was going to suggest we have a call in 1 hour from now. Good? [14:45] gary_poster: how about i ping you when i'm out of my call? it's sometimes pretty quick. [14:45] rogpeppe, we have a 15 minute call in 45 minutes :-) [14:45] lol [14:46] gary_poster: ok, 1600 UTC sounds good [14:46] so one hour works for us [14:46] cool thanks talk to you then [14:46] gary_poster: rather, 1545 UTC, presumably [14:46] rogpeppe, yes, exactly [14:46] hatch, sorry for wait. call in juju gui? [14:46] gary_poster: did you manage to penetrate the code BTW? [14:47] gary_poster: np sure [14:47] rogpeppe, yes, pretty well we think! We have several questions to verify assumptions and so on, so you can evaluate our understanding yourself :-) [14:49] gary_poster: cool === Makyo|out is now known as Makyo [15:04] Makyo: did you seem my message yesterday about the two branches? [15:05] bcsaller_, yeah, looked through the diffs, makes sense. Will try them out before the call. [15:05] Makyo: great, thanks [15:06] gary_poster: I'm prepping two talks for later today... so sorry tomorrow's best for me [15:07] m_3, cool, Are you the only ~charmer who is doing regular reviews now? [15:07] Or is there someone else we can ping [15:08] gary_poster: nope, jamespage, negronjl, marcoceppi, imbrandon, bbcmicrocomputer, SpamapS [15:09] gary_poster: should all be in the rotation [15:09] m_3, oh, cool! thanks. didn't know SpamapS was still active [15:10] gary_poster: np, yup... I'll check out the queue tomorrow for a bit to see if somebody's hit it [15:10] cool [15:10] bbcmicrocomputer is on point this week [15:10] yeah, looks like he is flooded [15:10] ack [15:11] bcsaller_: I'm a little confused as to what you wanted me to do on your last review - were you saying i should change the code? or were you just pointing that out? re the middleware route stuff [15:12] hatch: talking about the problem, if any action was needed it would happen in another branch [15:12] alright thanks [15:13] bcsaller_, I liked both options. I might prefer the second (ordering one) but if the wind blew the other direction I might switch sides. :-) [15:24] bcsaller_, running juju/delta-annotations and merging my work with juju-gui/landscape-links, I now see the landscape controls on the bottom bar as applicable \o/ [15:25] Makyo: sweet [15:28] jujugui call in 2 [15:30] bac not sure if you are here but starting without you. [15:33] i am. have my sound off so i didn't hear the alert. joining now. [15:33] cool [15:39] improv changes lp:~bcsaller/juju/delta-annotations gui-side lp:~bcsaller/juju-gui/landscape-links [15:41] bac: do you call in ondialup? :) [15:42] er, no [15:42] hatch: usually my 4G phone is better, though. i'll start using it for daily. [15:43] hatch: though i had trouble a few weeks ago when in NC using FIOS. go figure. [15:43] heh maybe it's your laptop :) [15:52] benji: i just added the new section. take a look. [15:52] bac: looking now [15:56] gary_poster: are we doing the call? [15:57] rogpeppe, yes. sorry, on call with mramm now, almost done [15:57] gary_poster: np [15:58] bac: looks good, I had a couple of suggestions [15:58] rogpeppe frankban teknico https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AoQnpJ43nBkJdEFIZVg0dnN0SXNNREpTMTd6X1FMS1E#gid=1 [15:58] benji: great [16:00] rogpeppe, frankban teknico oops :-P https://plus.google.com/hangouts/_/7fb7c30f3a232db57dd8549738fb98e723d90d4a [16:01] gary_poster, nice X's though ;-) [16:01] :-) [16:03] does mocha have a way to log to the console? [16:04] their documentation kind of sucks :) [16:08] hatch: console.log [16:08] bcsaller_: yeah odly enough nothing is getting logged to the console [16:08] the test is showing as completing [16:09] using the prod build the console is disabled, maybe thats what you're seeing [16:10] I'm getting loader: Has Skin? cssreset-context : false these logs in there [16:10] you're using make test-server and using the browser? [16:10] `make test-server debug` [16:10] and in the browser [16:10] the browser should have no issue with logs [16:11] mocha-phantomjs can fail on logging objects that can't serialize to JSON (usually objects with cycles) [16:11] but the browser can handle it [16:12] hmm then wth is going on here [16:13] we really need a way to run individual test suites heh [16:14] hatch: its easy, there are a few ways to do this, the simplest is to change describe to describe.only [16:14] or change 'it' to 'it.only' [16:14] to run a suite or a single test [16:14] so there is some type of command which will only run a single js file? [16:15] with .only applied any method of running the tests will do the right thing [16:15] another option is to pass ?grep=somestring at the end of the URL you're accessing [16:15] which also works [16:15] but .only is very simple if you have the test in the editor already [16:15] yeah the grep fails every time with a 404 [16:16] but the only worked nicely [16:17] now to figure out what's going on with this logging [16:17] the current skin around grep puts a trailing slash where there shouldn't be on, remove the slash and it works [16:18] https://gist.github.com/hatched/5638c5dde8fcfeac49b8 [16:19] so those console.logs should output to the browser console? [16:20] I just want to make sure I didn't mess something up which would cause them not to [16:21] if I enable YUI debug logs those end up in the console so I'm going to guess that these console.logs are being trapped somehow [16:22] I should mention that if I use console.log in any other test file the logs aren't outputed to the console [16:28] yeah looks like console.log is being clobbered [16:28] darn [16:30] gary_poster, sorry, lost both land lines at once, connected via mobile now, which is not very much conducive to hangouts === teknico is now known as teknico_mobile [16:31] teknico, :-( sorry. we'll fill you in when we can [16:32] yikes - must be a lot of construction around you guys :) [16:32] or the nom nom fiber eating moles [16:32] hatch, nah, just rotten countryside copper wires :-/ [16:34] ahh that'll happen [16:34] our primary telco is a crown corp so we have pretty unreal communications [16:35] something like 95% of the prov can get high speed internet [16:35] and we only have ~1M people hah [16:36] of course that comes with its own side effects....primarily being that the customer service SUCKS! [16:38] there is only one real competitor for landline/internet but they have bandwidth limits [16:38] and there are a ton of mobile competitors but they don't have many towers === teknico is now known as teknico_mobile [17:05] victory! [17:05] at least debugger; still worked :) [17:16] gary_poster: ok subapp-extension branch now has tests - much to the chagrin of mocha I'm sure [17:16] :P [17:17] :-) will look soon [17:28] gary_poster: I'm looking to start a new card (pausing my slack one) since my charmhelpers branch was reviewed and approved. The "Story 1" lane is the only one not at the WIP limit and it has no high priority cards. Is there one you would prefer I take? [17:28] on call... [17:29] benji, new command? [17:29] gary_poster: sounds good, that's what I was thinking [17:29] cool [17:30] gary_poster: should "get_annotations" be blocked because there aren't "annotations" in go-juju? [17:30] y benji [17:30] * benji blocks that card, and set_annotations [17:31] er... "update_annotations" [17:32] and "remove_annoations" [17:38] bcsaller_, minor on the delta-annotations branch. The annotation is being passed as 'landscape-needs-rebooot'. Can we settle on either two or five 'o's for 'reboot'? [17:38] ha, wonder how that happened, I'll push an update now [17:39] Makyo: pushed [17:39] bcsaller_, thanks. [17:59] bcsaller_, everything's working with one change to the gui branch. Will grab a diff. [17:59] Makyo: excellent [18:00] http://pastebin.ubuntu.com/5568288/ [18:01] bcsaller_, er, the last bit is from merging, never mind that. [18:02] ha, ok [18:06] lunch & [18:54] gary_poster: let me know when you have a second for a post-pre-implementation call [19:02] Makyo: ahh, thanks for the diff, I read it wrong the first time missing _where_ in the file you were doing the update, makes sense and applied [19:02] bcsaller_, thanks. +1 from me, then. [19:03] benji can have quick one now or longer one later [19:03] which do you prefer [19:03] gary_poster: lets try now [19:03] ok [19:03] the regular place? [19:03] yeah [19:05] guihelp - If someone's got time, a quick review of https://codereview.appspot.com/7369054/ would help move the landscape stuff along quickly. [19:06] I'll take one [19:06] give me a few [19:06] hatch, thanks. [19:07] Makyo: does this after update stuff have anything to do with the reason the app call dispatch() so much? [19:08] hatch, possibly on views other than the environment view. [19:09] hatch, that'd be another branch, though. Goal is to try and have branches for the rest of the landscape work today. [19:09] yep that's fine [19:11] so this 'update' event is actually from another view/model? [19:12] benji: do you know how to run 'go test -gocheck.v' for juju-core? it seems to have a different syntax. [19:12] hatch, update happens when a delta is received from the server. [19:12] that is, you can run 'go test ./...' from launchpad.net/juju-core [19:12] but if you add -gocheck.v it says there is no go source there [19:12] annoying [19:13] bac is the jujucore exercise done now that you sent out the email with instructions? (have not yet read and will do so) [19:14] gary_poster: i think yes [19:14] cool thanks [19:14] card moved [19:14] thx [19:16] Makyo: can you point me to the file where it's being fired from? [19:17] hatch: app/models/models.js:Database is the primary source [19:17] hatch, bcsaller_ line 511 [19:18] got it thanks guys - just writing my response right now [19:18] benji: nm, it doesn't look like you can run 'go test -gocheck.v' recursively. [19:23] bcsaller_: Makyo done [19:26] gary_poster: roger is working on 'juju set' config, bug 1130169. not sure how you want to handle the card [19:26] <_mup_> Bug #1130169: Support set_config in juju-core env

< https://launchpad.net/bugs/1130169 > [19:26] i linked the bug to his branch and assigned him the bug [19:26] bac set_config bug 1130169 also right? [19:26] <_mup_> Bug #1130169: Support set_config in juju-core env

< https://launchpad.net/bugs/1130169 > [19:27] gary_poster: that is the same bug [19:31] bac I put it in tracking for now [19:32] gary_poster: because it looks like you aren't doing anything today ;) I have added tests to both of the tasks blocking the kanban [19:32] Makyo: its in if you didn't get that notification [19:32] hatch, heh [19:32] looking [19:32] bcsaller_, thanks, grabbing some food now, will merge after lunch. [19:35] crap sorry I committed the test code with only in it [19:35] hatch, huh? [19:35] describe.only() [19:35] vs describe() [19:35] oh [19:35] re-proposing now [19:36] cool, will be back in a few [19:53] gary_poster: ping me when you're back [19:53] bac, I'm back [19:54] gary_poster: for things like bug 1130159 the work entails only the juju-core side, no? [19:54] <_mup_> Bug #1130159: Support expose in juju-core env

< https://launchpad.net/bugs/1130159 > [19:55] it has that right in the title but the body shows juju-gui env bits [19:55] bac, the bug represents both sides. You are welcome to subdivide it [19:55] gary_poster: is it your expectation that they be done in lockstep? [19:55] or close [19:55] bac, juju-core env can also refer to the JS juju-core env [19:56] bac, close [19:56] bac, we should verify practically each step of the way that JS is working fine with what we implement on the Go side, so we don't get nasty surprises [19:57] I don't expect them, but that's why they are called surprises :-P [19:58] gary_poster: not to nit picks, but those bugs should at least have two bug tasks, one for juju-core and one for juju-gui. i'll update the expose one. [19:58] thanks bac [20:00] rats, i can't change importance on the juju-core bugtask. [20:07] teeny-tiny review up at https://codereview.appspot.com/7377056 [20:24] benji and bcsaller, I philosophically support both of your small branches, though I won't have time for reviewing them till later. hatch, am I right in assuming that your changes to the branch have been ready for me for some time now? [20:24] gary_poster: yeah :) I knew you were busy so that's alright [20:24] gary_poster: my branch has been reviewed and landed. Where is your philosophy now!? [20:24] lol [20:25] benji, lol, happy to see that was unnecessary [20:25] :) [20:27] bcsaller and hatch I have to specify goals for the sprint. I'd like you to focus on getting the continuous integration and IE fixes. The goal would be to have CI working with Jenkins on Canonistack, which will have a few interesting challenges; to have at least some of the integration tests passing on IE, which will also be interesting; to fix at least four IE bugs; and to identify remaining necessary work. [20:27] Is that alright? I can help as possible. We can have some remote help to from Diogo Matsubara [20:27] who has done some Jenkins set up in the past [20:27] gary_poster: I won't have any way to test IE10 there [20:27] hatch, no virtualbox? [20:28] my laptop only has 4GB of ram but I could attempt it :) [20:28] heh [20:28] I hear ya [20:28] if I can get it running then I have no issues with that task [20:30] ok I lied I only have 3GB according to `free -m` [20:30] so it doesn't look like I'll even be able to run win 8 [20:32] or am I going to have to buy a new laptop before Sunday? ;) [20:33] heh [20:33] no, but darn. best laid plans [20:34] maybe I can buy some more ram [20:34] I wonder if my old mac mini ram will fit [20:34] if so then....problem solved [20:38] I only have 4GB [20:38] I think I assigned 2 to the Win8 VM. [20:38] hatch ^^ [20:38] hatch, It'll run IE, at least, but not much more :) [20:38] yeah I only have 3 - I tried to assign less than 2 on my desktop here and it locked up every time [20:40] it appears to be running... [20:40] ubuntu loaded [20:41] free -m gives me..... [20:41] BOOM [20:41] 8GB [20:41] ok problem solved [20:41] pfft and my wife always complains that I keep too much junk around [20:41] :P [21:08] jujugui - any chance I could get final reviews on https://codereview.appspot.com/7384059/ and https://codereview.appspot.com/7381055/ today please ? :) [21:09] hatch: if your original reviewers can't make a return appearance, let me know and I can take a look [21:09] :-) I will do mine starting in 5. [21:09] hatch: I'll review in 5 or 10 as well [21:10] thanks! [22:08] hatch https://codereview.appspot.com/7381055/ is done [22:08] looking at others [22:08] thanks - loosk like I have a few things to add/change I'll be sure to do those tonight [22:13] bcsaller_: thoughts on passing in factories vs strings for the sub app registration? [22:13] gary_poster would prefer factories/constructor, I used strings to match the view syntax [22:13] but I'm impartial either way [22:14] * gary_poster waves [22:15] hatch: I think explicit passing is pretty simple, I went with that for the component system we use as well. [22:15] ok I'll change that syntax then [22:16] bcsaller_: wrt the navigate overwrite - it worked as expected in the prototype but I also anticipate there will be side effects but without a more complex example I have no idea what those might be [22:16] hatch: topo.addModule(views.ServiceModule); It is explicit and can still reveals the namespace. [22:18] hatch: maybe it leaves enough of the code inert, but there are still bound handles to history events I think, guess we'll see [22:19] I'll spend some time tonight to investigate further [22:21] thanks for the reviews! [22:35] ok hatch, did the other one now :-P https://codereview.appspot.com/7384059/ [22:37] thanks :) I want this out of my hair so I'll fix everything in the comments and have it ready in the morning [22:37] hatch, is the namespace you describe just a convention, or part of a YUI feature? https://codereview.appspot.com/7369054/diff/1/app/app.js#newcode259 [22:38] it's a little bit of both actually [22:38] cool hatch. I said LGTM, so for me you don't need me follow-up review unless you want it [22:38] one sec let me find an example [22:38] The "cool" was re the "fixing everything" [22:38] ok thanks [22:40] here is an example of it being part of yui archiecture http://bazaar.launchpad.net/~hatch/juju-gui/1130790-extend-app-subapp/view/head:/test/test_sub_app.js#L30 [22:40] see because I 'published' the event and included a 'facade' it namespaces the events automatically [22:41] ah! [22:41] ok cool [22:41] ty [22:41] http://bazaar.launchpad.net/~hatch/juju-gui/1130790-extend-app-subapp/view/head:/app/assets/javascripts/sub-app.js#L19 [22:41] * gary_poster is exhausted :-P [22:41] I need to run [22:41] have a great evening everyone [22:41] alrighty have a good night :) [22:41] nite [22:42] hatch, yeah I remembered that from review, thx [22:42] bye [22:42] yeah so if you don't publish the event it's usually convention to namespace it [22:42] just to avoid collisions [22:42] :) [22:42] and then anyone reading the code knows where it's from too [22:43] but yeah, my comment was moreso just a `FYI` :) [22:53] Makyo: https://codereview.appspot.com/7381060/ is the menu wip branch, I suspect we both have code to push the landscape object to the topo, but thats a minor merge. [22:54] bcsaller_, Was just about to ping. Running into scope problems. [22:54] Makyo: need a call about it? [22:54] bcsaller_, Yeah, I think I'm basically stuck :P [22:55] Makyo: in hangout