=== imbrando1 is now known as imbrandon === jovan2_ is now known as jovan2 [13:38] bac: I'm wandering around in testing infrastructure, but we should probably formulate a plan for today soonish [13:38] benji: my plan is to make the changes per review and get my branch landed [13:38] that sounds good; want to sync up after that? [13:38] then i will be an official Go programmer, a juju core contributor, and an all around badass [13:39] morning [13:41] yes, benji, that'll be good [13:48] bac, they teach the secret handshake then ;-) [13:49] excellent. [13:50] bac is so awesome he doesn't call the wrong number, you answer the wrong phone [13:53] Bwa ha ha ha [13:53] Is that taken from a "Clint Eastwood" joke? [13:54] goodspud: Chuck Norris [13:54] benji, er yeah... that's who I meant :) [13:54] :-) [13:54] Clint, Chuck... all the same really (don't hit me Chuck) [13:55] Clint Eastwood doens't hit people, he makes Chuck Norris do it for him. [13:56] Bwa ha ha ha [14:14] bac, care to add a code review link to your card? [14:15] teknico: ok. [14:15] bac, beat you to the punch :-) [14:15] dang [14:16] oh, you got reviews already, ok [14:16] teknico: are you doing juju-core reviews or just curious [14:16] i have an abundance of reviews! :) [14:16] bac, just curious, I don't feel up to the task yet [14:23] bac you available for a call with me and frankban to share your glorious go juju api command knowledge, now that you have a branch in review? [14:24] gary_poster: sure. benji may want to join if he has time. [14:24] let me re-tea and i'll join you there? [14:24] thanks bac, pm'd location [14:24] you too benji [14:24] for you I'll clear my schedule [14:27] gary_poster, ok, where's the party? [14:40] I've noticed that some code uses camelcased and some uses understores fooBar vs foo_bar - is there a prefered method? I typically write camelcased [14:40] hatch, camelcased [14:41] hatch, we transitioned mid dev [14:42] sounds good to me - old habits are hard to break ;) [14:43] Makyo, hey. I have more feedback for your branch--small but nice. Trying to add it now.. [14:50] comment added [14:59] gary_poster, Thanks for the feedback. I'm not sure I understand. Do I need to call _navigate rather than navigate or fire('navigate')? Where do I set the options? Tests for those lines? [15:00] Makyo, about to have call, but you would need your logout to eventually call the app navigate method with that new option [15:01] Makyo, talk to you after this call or after daily? [15:01] gary_poster, Sure. [15:01] thanks [15:09] Will try to coffee in the meantime, so I'm not running so slow [15:17] gary_poster, nvm, I get it now, was misreading Router.combine. Will get a test in for that. [15:27] cool thanks Makyo [15:28] jujugui call in 2 [15:32] sorry, hangout crashed and is not starting anymore, trying again [15:35] gary_poster: branch back in review after making changes. [15:35] gary_poster: proceeding with the next branch, dependent on the first [15:35] cool [15:54] bcsaller_: have time for a chat? [15:55] hatch: sure [15:55] in the boardroom [16:38] if I run --fixes multiple times on a commit will that link to multiple tickets or will it blow up? :) [16:39] heh, dunno hatch. you can do it in lp though [16:39] "link to bug" [16:41] sounds good :) [16:41] I think the arg parser is last write wins [16:42] Have yet to use --fixes, been using -bug=xxxxxx in lbox. That just do the same thing? [16:43] different under the hood in a way you don't care about Makyo. --fixes annotates branch in a way that LP snarfs up, and I suspect lbox uses LP API to directly communicate the connection [16:43] so inetrchangeable effectively [16:59] hatch are you already factoring out the namespace code into a separate module? If have some changes--mostly but not exclusively docs--for that code as well. It would be easier if we did not stomp on one another. [16:59] s/If have/I have/ [17:00] gary_poster: I haven't touched the namespace stuff yet - I'm working on creating the subapp structure right now [17:00] Great hatch. Lemme know when you get there and we can either coordinate or maybe I will already have my changes in. [17:01] right now my changes to app.js are going to be minimal - I'm trying to keep them all in the extension - so update away! [17:02] cool thanks [17:23] hatch: let's talk soon about code reviews. whenever convenient for you. [17:23] can we do it after lunch? I really want to get this subapp stuff working [17:25] Small simple cleanup branch for review. bcsaller needs to look at it, and then anyone else (hi, jujugui!). https://codereview.appspot.com/7372044/ [17:26] hatch: ok [17:26] ping me [17:27] gary_poster: i'll look at it. there is no card, right? [17:27] right bac. was a review follow-on for bcsaller's branch of some changes I wanted but didn't make it [17:32] *cough* sorry about that === bcsaller_ is now known as bcsaller [17:33] bac: is it required to ping someone to get a juju-core review? [17:34] frankban: i didn't and i got four! [17:34] bac: cool! [17:34] may not hurt to ping fwreade, though [17:35] bac: I have no rush, approaching EOD [17:35] bcsaller, :-) np. I read your reveiew, thanks. In the current structure how does one register a namespaced route? I assumed that you would register something for the path ":your_namespace_here:/your/path" or for something similar. Is that it? [17:35] /:your_namespace_here:/your/path I guess [17:35] gary_poster: that will work, but I think specifying it as an attr of the route definition will be better and will work the subapp stuff cleanly [17:36] The change to match is a few minutes of work, I can propose it after you land your changes [17:36] bcsaller, sounds good, thanks. [17:36] gary_poster: done [17:37] Cool, thank you bac. [18:35] Makyo, approved with some small changes [18:35] hatch/bac, please take a look at https://codereview.appspot.com/7395043/ very soon so Makyo can land [18:38] alright [18:38] hi benji [18:38] hey bac [18:38] i've got a branch that implements the server and client functions for the ws/rpc get command [18:39] benji: would you like to try poking at it with your ws client and see if it behaves? [18:39] benji: or show me how to? [18:39] i.e., gimme your client [18:40] bac: sure, shall we relocate to the hangout to discuss? [18:40] i'm sure others are on the edges of their seats... [18:40] but, ok [18:40] I don't even have a seat! [18:41] bac: here is the code: http://paste.ubuntu.com/1700814/ [18:42] benji: invited you [18:44] the code looks good - I'll do a functionality check now [18:56] I really hate how harddrives advertise 3TB but are really only 2.73TB ... 270GB is a lot of missing space ;) [19:04] benji: bzr push lp:~bac/juju-core/add-get-2 [19:04] thanks bac [19:15] gary_poster, hatch: when you have time can you check that the rules for ns route matching make sense to you in https://codereview.appspot.com/7396050/ [19:19] I must be missing something [19:19] because it looks like it just returns true all the time :) [19:21] that was a poorly formed joke [19:21] and poorly executed [19:24] bcsaller: just to verify when you said [19:24] So /:inspector/service/mysql/ would match while [19:24] you meant :inspector: right? [19:24] yeah [19:25] unfortunate typo [19:26] revieweddd [19:27] benji: which websocket package does your script use? the one from pip doesn't have 'create_connection' [19:27] bac: hmm, let me look [19:28] gary_poster: are we chatting in 1 min? [19:30] bac, no. I hoped we would be, but house person is here. will ping? [19:30] should be within 15 min [19:31] ok [19:31] bac: "websocket-client" [19:48] bac, ready now. joining. you ready? [19:57] gary_poster: joining now [20:15] bac https://codereview.appspot.com/7390043/ [20:21] hi benji, i'm getting {"RequestId":1,"Error":"no such request \"ServiceGet\" on Client","Response":{}} [20:22] benji: is there some other registration step i might be forgetting? [20:22] bac: I didn't have to do anything for mine, let me look at your code a second [20:24] bac: the code looks good to me; did you use --upload-tools when you bootstrapped? [20:24] bcsaller, that namespace-restricted route matching looks good to me. When you prepare for landing, please do include some documentation [20:24] even just comments in the routes section [20:24] gary_poster: thanks, on it [20:24] benji: yes [20:25] benji: i just tried 'Status' and it works [20:25] hmm, well, that's something [20:28] bac: it might be your return signature, you should be returning a struct, not a pointer to a struct (I am far from certain that this is it, but that is the only candidate I have for what the problem might be) [20:29] hmm [20:29] ok [20:32] benji: that is why rog was returning Status{} [20:32] he had to return a struct so he couldn't return nil [20:33] that would explain it (although, I wonder why not just make the standard return signature a pointer to a struct, then returning nil is fine) [20:34] benji: well there is that [20:34] go is weird in that the caller looks the same whether it is a struct or pointer returned [20:36] yeah, that's something that I'm not sure if I like or not; not having to do the "." vs. "->" dance like in C++ is nice, but it might be confusing having the DWIM [20:41] benji: that was it! fixed the return signature and it works. thanks! [20:41] cool [20:54] benji: are you okay running with my add-get-2 branch? [20:55] bac: I'm ok enough to try :) please write down my goal(s) in an email to me and I'll tackle them [20:56] benji: ok, the main thing is to figure out api_test and add relevant test for the new command [20:59] benji: writing email now [20:59] thanks [20:59] tests I think I can handle [21:00] benji: did my msgs just show up in #juju-dev? [21:00] bac: I'm not in juju-dev at the moment [21:00] my client is reporting Your message couldn't be sent to the channel [21:00] ok [21:00] I guess I should add that to my auto-join (it's just so noisy over there) [21:15] O K lunching [21:15] bbiab [21:29] gary_poster: should we be added to ~gophers? if not we cannot submit to juju-core [21:29] BradCrittenden, yes we should. I will ping mramm about it unless you tell me you are doing so [21:29] gary_poster: he is currently on #juju-dev. === BradCrittenden is now known as bac [21:30] hmm, lbox submit is failing for me, trying to figure out why, sorry about any noise its generating [21:33] bcsaller: how is it failing? [21:33] i got a read-only transport failure, which i attributed to not having write permission to lp:juju-core [21:33] bac: http://paste.ubuntu.com/1701833/ [21:34] bac: sh ./.lbox.check works fine though [21:34] bcsaller: ok, unrelated [21:35] no clue as to what is happening with yours [21:35] bac: thanks for looking :) [21:48] gary_poster: well, thanks for trying. :) [21:49] bac :-( [22:03] ok!!! [22:03] sooo [22:04] how do I go about merging trunk into my branch? [22:04] well the proper way of doing so [22:05] I need to make sure I have all of the most recent routing code [22:06] hatch, If you're in your branch, you can do bzr merge ../../trunk or wherever your copy of trunk is [22:06] hatch, Make sure trunk is up to date by doing `bzr pull` in there, first. [22:06] ahh ok that's simple enough thx [22:07] bcsaller, new version of jshint changed the command name, just running into that now. [22:12] guihelp, Newest version of jshint dies spectacularly on some of our files - namely JSON, but a few JS files it was fine with before. Can bcsaller or I set the version in our current branch to an older one for the time being? [22:12] Makyo, +1 [22:12] go for it [22:12] although what's it barfing on? [22:12] just curious [22:15] hatch, **.json, a few redefinitions across files (YUI, etc.), one instance of 'char' as a var name. I'd like to get both bcsaller and I's branches landed (or proposed) today, so I'm proposing we drop the version for now, make a card to investigate moving to current when we've got time. [22:15] oh yeah for sure - drop it down [22:16] jshint isn't one that you wouldn think would all of a sudden start breaking heh [22:16] They just hit 1.0.0, so I'm assuming something big happened :_ [22:16] :) [22:17] bcsaller, when you get back, in package.json, set jshint's version to "0.9.1" rather than ">=0.9.1" [22:19] haha [22:19] aww man this routes stuff is kicking my a$$ [22:35] bcsaller: is the _dispatch method in app.js a monkeypatch for the YUI one in Router? [22:35] I now finally have all of the route parsing business done (and only 2 lines lol) [22:35] but now it's barfing on dispatch [22:39] bcsaller: maybe we can pair tomorrow to solve this? [22:52] bcsaller, I made a review of your Landscape aggregation branch. Please rustle up another one. [22:55] hatch: yes, its overridden [22:55] yeah I attempted to make my way through it but I think it's best if we work together on this integration part [22:55] would that be alright? [22:56] sure [22:56] great - I figure we should be able to get it done in <1h tomorrow [22:58] Makyo, appcache info: thanks! LGTM with trivials [22:58] (and actually fly-bys) [23:03] if one reviewer has comments and the next has nothing to add but agrees with reviewer 1 [23:03] what should they put? land as is? or.... [23:03] land with other changes? [23:08] I chose "Land with Gary's trivials" :) [23:14] has anyone ever evaluated using https://github.com/airportyh/testem ? A bunch of my local dev friends use it and really like it [23:30] never hatch. If you look at it and think it is worth incorporating into our tools you could do so and propose it some Friday or other. [23:30] sure I can look into it