[12:32] * benji reboots === benji___ is now known as benji [12:51] This is the second time I couldn't use -bug with lbox because I get this error: error: ERROR: Failed to update bug task: Server returned 400 and body: milestone_link: Constraint not satisfied. [12:57] benji, maybe that would mean something to bac or Curtis? Sounds registry-related. [12:57] It may, I'll check. [12:58] it may be the bug isn't linked to a milestone [12:58] may be worth looking at a bug that lbox creates and see if it is linked vs the pre-exisint one you're trying to use [12:59] on juju-core i don't have permission to manage milestones, priorities, etc on bugs [12:59] === bcsaller_ is now known as bcsaller [13:25] rogpeppe, great that the next two megawatcher branches landed. In https://codereview.appspot.com/7727045/diff/5001/state/megawatcher.go OLD line 13 you deleted a comment that we would need the last revid in the struct. Is that not true (anymore?), or did you just remove an unnecessary note about the future? [13:25] gary_poster: looking [13:26] gary_poster: it's still TODO. in fact, the revno is added in the next branch in the pipeline [13:26] rogpeppe, cool :-) thanks [13:26] gary_poster: BTW even though i've submitted, any comments from yourself would be very welcome. [13:27] gary_poster: i can easily make fixes in upcoming branches [13:27] rogpeppe, understood and thanks. I looked over 241 last night and didn't see anything other than that question. I'm re-reviewing now and then will look at 242 [13:28] +1 on change from Remove to Removed in changes btw. that inconsistency that we introduced bothered me [13:28] gary_poster: cool [13:29] gary_poster: in fact that change was just merging the two Delta implementations which should already have been together, i think. [13:29] yeah, I noticed that [13:29] did we do that? sorry, if so [13:29] I mean, have two copies of the Delta [13:30] gary_poster: i dunno. somewhere between us, i think. [13:30] :-) cool [13:30] gary_poster: it doesn't matter [13:55] morning [13:57] Herro. [13:58] hiya [13:58] after hours of frustration I was able to get my osx > ubuntu vm nfs working [13:59] gary_poster, do we have our call now? [13:59] teknico, we do, and I'm even in the chat room now :-) [14:00] from calendar [14:01] * hatch wonders if gary_poster is always in guichat just hanging out [14:01] heh no hatch, just seems that way :-) [14:02] ohh ok ;) [14:16] teknico, I lost you again === teknico_ is now known as teknico [14:29] oops [14:30] jujugui call now in guichat [14:42] re: "leave your camera home" http://www.nytimes.com/2013/03/04/business/media/photojournalists-in-oakland-face-epidemic-of-camera-robberies.html?pagewanted=all&_r=0 [14:48] bcsaller, I forgot that we needed the VPN running to see it :-) [14:48] the link you gave I mean [14:50] yeah, should have mentioned it [14:50] I haven't turned the VPN off since the sprint though [14:52] gary_poster: at the bottom of that log is Charm panel False Deploy Button False which is the output of is_displayed on those elements. .click doesn't work when thats false [14:53] :-( gotcha [14:54] gary_poster: ping for round 2 [14:55] ack rick_h_ . jujugui, if anyone wants to join discussion with rick_h_ come over to guichat [15:14] so now that I'll be sshing into my vm what shell window manager do you guys prefer? screen? [15:15] tmux, the pragmatic programmers book is really good intro. [15:15] hatch, or byobu if you want a nice wrapper to either of them === teknico_ is now known as teknico [15:18] rick_h_: so why tmux over screen? [15:18] hatch: so honestly a big part was the book got me over the hurdle for scripting/using it [15:18] I can't claim to have ever mastered screen, but my sense is that tmux is more scriptable more easily [15:19] so for instance, this is how I work on jujugui: /bin/tgui http://paste.mitechie.com/show/899/ [15:19] it's scripted startup for a tmux session with 4 windows all setup in the right directory to start work named for function [15:20] oh that's pretty cool [15:20] tgui && ta jujugui (aliased in my .zshrc to tmux attach -t) gets me started in my day [15:20] "j u j u g u i" pings us all :-) [15:21] bah, my apoligies. As someone that doesn't see it I don't realize it. [15:21] np :-) === teknico_ is now known as teknico [15:32] For future projects we should add an explicit "ping" to the ping-the-world keyword. [15:33] should we though? [15:33] :) [15:45] benji: i just gave you a review. there's one issue we need to solve though. (https://codereview.appspot.com/7554046/diff/1/state/apiserver/api_test.go#newcode404) [15:45] rogpeppe: cool, I'll take a look right now [15:47] * gary_poster heads to agile meetup lunch. biab [16:01] rick_h_ you're flying into SFO you said? [16:01] hatch: rgr [16:02] I can't fly out of OAK without flying a milk run so I was hoping SFO wasn't too far away [16:02] hatch: yea, 55min http://goo.gl/maps/kVzHQ [16:03] oh that should be alright then [16:04] hatch: yep, just watch out for time of day. Transit shuts down after 1am and starts around 4am [16:08] ahh] [16:27] so in sshing into my vm and trying to bzr pull I'm given a permission denied error because the keyring is locked (I'm assuming) [16:27] any way I can get that to request a pw? [16:29] hmm, it auto does that for me. is python-gnomekeyring installed? `dpkg -l | grep keyring` [16:30] that it is [16:30] it of course works while in the vm via a gui window [16:30] yea, I know what you're talking about. I've hit it before but don't recall 'fixing' it. Just it asking for a keyring password the first time and setting it up [16:31] abently might know for sure but he's out today [16:31] alright well I'll make a note to research it tonight [16:31] thx [16:36] goodspud: ping, any requirements we know about for the charm icons we need to make sure to put to paper in the docs? [16:37] goodspud: we're going to require a icon.svg in the charm, so does size/etc matter at that point? [16:41] rick_h_, svg files should be scalable. We have specific sizes for different presentations of the icons ( 32, 48, 64, 96, 160 ) but if it's a well constructed SVG it should scale to all those sizes [16:42] goodspud: ok, if you come across any 'guidelines' we should share with charm authors please let me know so we can make sure we get those into the docs for them. [16:42] hatch: speaking of abentley :) [16:42] rick_h_: Hi. [16:43] haha [16:43] abentley: hatch was having bzr + keyring issues on a vm he's ssh'd into. Do you have super secret knowledge of how that works? [16:43] abentley: I've always just had it prompt me for a keyring pass and start working [16:43] rick_h_, our designer (Greg... who hates IRC so you can't ping him) is going to produce a series of style guides and guidelines "at some stage" [16:44] goodspud: ok, well we're adding the support and starting to work with charm authors to make sure 13.04 isn't a giant page of 'no icon available' default images. So the sooner the better. [16:44] goodspud: we're working with jcastro to help us work with upstream authors [16:44] rick_h_: I'm not sure what would cause that. Was this with GPG or SSH? [16:44] I'm ssh'd into the machine [16:44] abentley: assuming it's bzr + ssh? hatch, more details? [16:44] yeah I'm ssh'd into the vm [16:45] hatch: and what command were you running exacly? [16:45] when I tried bzr pull I got 'permission denied public key' [16:45] `bzr pull` [16:45] rick_h_, the juju design team has a catch-up meeting tomorrow morning so I'll raise it then [16:45] goodspud: thanks! appreciate it. So many moving parts to sync. [16:46] hatch: Could be a username issue. Is your username on that machine the same as you normally use? [16:46] nope totally different [16:46] although I ssh'd using my typical username on the vm [16:46] rick_h_, absolutely. We can't leave our charm authors and Jorge without help [16:46] Are you pulling from Launchpad? [16:47] yes [16:47] I had to go into the vm and bzr pull - it poped up the gui keyring unlock window [16:47] then pulled [16:47] even now I can't bzr pull from the ssh [16:48] hatch: Your LP userid is hatch? [16:48] yes [16:49] hatch: Could you try "bzr info bzr+ssh://hatch@bazaar.launchpad.net/~abentley/bzrtools/bzrtools.dev" please? [16:50] permission denied from the ssh; works from the vm [16:51] hatch: I don't understand. [16:51] hatch: what is "the ssh"? [16:51] my ssh'd terminal window into the vm [16:51] hatch: Okay, did you enable agent forwarding when you sshed? [16:53] hatch: (agent forwarding can be enabled with -A ) [16:53] I didn't, just checking the config [16:53] looks like no - let me give that a go [16:54] same error [16:54] hatch: but it works in the VM's terminal window? Odd. [16:54] yeah that's what I thought haha [16:57] hatch: can you check whether agent forwarding is disabled in the VM's ssh server? It's "ForwardAgent" in /etc/ssh/sshd_config [16:57] it's commented out [16:58] hatch: uncomment it and set it to yes. [16:59] hatch: Then "sudo service reload ssh" [16:59] hatch: Then log in again with -A [17:00] s/log in/ssh in/ [17:00] heh connection refused [17:01] sec ssh might not be running [17:01] hatch: sudo service ssh start [17:02] yeah when I do that it tells me the id but then when I grep ps all I get is ssh-agent [17:03] hatch: You should be able to see what's going on in /var/log/daemon.log [17:05] when I remove that line it starts working again [17:05] oy [17:05] that file is empty heh [17:05] hatch: Odd. [17:06] yeah this is pretty standard for me - my sysadmin luck is negative [17:06] :) [17:06] hatch: Is the private key also on the vm? [17:06] oh sure - I used to use the VM to code in [17:07] hatch: Could you try sshing into the vm, then using ssh-add to add the key? [17:10] that did it [17:10] :) [17:10] hatch: Okay. Now we should fix your LP username permanently. [17:11] hatch: I think the simplest way is "bzr lp-login hatch" [17:11] hatch: After that, try ""bzr info bzr+ssh://bazaar.launchpad.net/~abentley/bzrtools/bzrtools.dev" (i.e. without the @hatch) [17:12] that works [17:12] spews out a bunch of auth-agent stuff [17:12] Okay, try your original pull. [17:12] great and we are now working! [17:12] thanks :) [17:13] Good stuff. [17:13] You're welcome. [17:13] can you explain what the issue was? [17:13] as in why the key wasn't being applied by default? [17:13] hatch: The issue was that your SSH username was not "hatch". [17:14] ohh [17:14] hatch: So your key didn't match your username. [17:15] ahh I gotcha that makes perfect sense [17:15] * hatch wishes for helpful error messages [17:15] thanks again :D [17:15] hatch: got a sec to chat subapps? [17:15] hatch: np. [17:16] rick_h_ see u in guichat [18:02] bcsaller: should I plan on working on the CI tomorrow? [18:07] hatch: I am about to propose what I have, I think should go with it, I have the one failing IE test as conditional now [18:07] hatch: last two runs were successful [18:08] oh excellent! [18:16] great bcsaller. were you able to isolate the problem to file a bug with selenium? [18:17] or come to some other longer term plan? [18:17] jujugui FYI Node 0.10 was released keep an eye out for odd errors in our build processes please :) [18:18] uh oh :-) [18:18] gary_poster: at this point I only commented it in the code. I think creating an isolated test might be a good next step, not sure though. [18:18] k [18:19] gary is our call in 12 minutes or 60*n+12 minutes? [18:19] bac, 11, unless that's a problem [18:19] nope, just didn't know if it shifted [18:19] cool, yeah [18:38] lunching [19:10] back [19:12] bcsaller: I can assume that this code works as advertised right? You're not going to make me test it are you? :P [19:12] lol [19:12] funny [19:13] haha [19:13] I'll try it :-) at leats in an ec2 environment [19:13] I think its possible there are intermittent failures, but the last two runs went fine. [19:14] bcsaller, is it trivial for us to have jenkins email us on duccess/fail? [19:14] success, even [19:22] hatch, where is your doc branch? [19:23] https://code.launchpad.net/~hatch/juju-gui/ci-documentation [19:23] ty [19:27] curses! mifi hangs when I'm in the middlsle of a lbox propose [19:31] gary_poster: didn't see that message but yes, there are outgoing options, email and IRC both maybe configured it looks like [19:31] bcsaller, ow that would be great. trying out the test now on ec2 [19:32] hatch: jcsackett if you guys get a few min please? https://codereview.appspot.com/7554047 [19:33] I'm on it [19:33] hatch: ty much kind sir [19:34] hatch: if you know how I can make the test more Mock friendly then cool, but failed to see a good way so went full bore with app/subapp objects. [19:34] sure I'll take a look [19:34] I would also like to see these changes reflected in the subapp scaffold branch [19:34] any chance after this lands you can copy/paste the changes there? [19:34] ah, ok and this doesn't do anything to the charm yet obviously [19:35] hatch: yea, diff isn't too big there at all. [19:35] rick_h_: looking now as well. [19:35] hatch: though can't you just merge trunk into the scaffold branch once it lands? [19:36] I guess you'd get the browser subapp in there as well, but meh [19:36] yeah I was hoping to keep it 'clean' [19:36] could we merge this commit in? [19:36] I know how to do it in git but not bzr [19:36] :) [19:37] h/me wants git cherry-pick :/ [19:38] gary_poster: next branch *nearly* ready - one test still to be written (most recent test usefully bailed out some lurking bugs, making me quite happy) [19:38] gary_poster: so heads up. I was going to create cards on your board for this config change and for the follow up tweak to the juju-gui charm but I don't have perms. I'm not sure if it matters to you or not? [19:38] rick_h_ darn I saw your merge code and was hoping Y.mix had a blacklist....sadly only whitelist :( [19:38] gary_poster: should be ready by the time you come online tomorrow :-) [19:38] rick_h_, will give you perms [19:38] gary_poster: 'cos that's my eod [19:39] hatch: yea :( [19:39] rogpeppe, awesome! Will be out tomorrow but will check [19:39] hatch: what are you trying to do? [19:39] bac: merge two objects witch a blacklist of property names [19:40] i don't understand [19:40] this is a bzr problem? [19:42] nope [19:42] bac: no, just I've got an update he wants in a side branch he doesn't want to pull all of trunk into [19:43] hatch: you could just merge the specific revision from rick_h_'s branch... but i'm sure you know that. nm. [19:44] that's what I meant, I know how to do that in git, I don't know how in bzr....YET! [19:44] bac yea, I figure I'll do a patch and apply it to his branch [19:44] rick_h_, you and orange have perms [19:44] bac: I didn't think bzr supported a 'cherr-pick-like' single commit merge [19:44] gary_poster: thanks, will add cards on your board as well then. [19:44] cool [19:45] from 'bzr help merge': To merge the changes introduced by 82, without previous changes: [19:45] gary_poster: this is all story1 stuff? /me isn't caught up on the board terminology [19:45] bzr merge -r 81..82 ../bzr.dev [19:45] bac: cool, will try it [19:46] rick_h_, sorry this is for orange? you would be maintenance for us (top lane). Maybe we should have separate lane for orange stuff. I'm on call, so let's talk about after [19:46] gary_poster: k, going to EOD but yea it's adding features to subapp for our use [19:47] rick_h_, ok top lane [19:47] gary_poster: rgr, thanks [19:48] thanks jcsackett. [19:48] rick_h_: yw. [19:48] hatch: packing up for EOD but will look at any comments later tonight. Thanks for looking things over. [19:48] np thanks for writing that [19:49] bac: excellent [19:49] bah, camel case. retraining is hard [19:49] rick_h_ lol the funny part is half of your stuff IS camelCased haha [19:57] I'm trying! [19:59] hatch: it's still unatural to camelcase variables. that's for method names. :-P [19:59] jcsackett: yeah when I was writing python at the sprint I was always going back and adding in _ haha [19:59] * jcsackett laughs [21:05] hatch, I reviewed https://code.launchpad.net/~hatch/juju-gui/ci-documentation/+merge/153165 . [21:05] make some changes, get one more reviewer, and you have a landing. :-) [21:05] thank yous [21:05] :) [21:06] * gary_poster is tired [21:06] * hatch hands gary a coffee [21:06] so long folks! Have a great Friday and weekend! See you Monday [21:06] :-) [21:06] you too! [21:06] thanks [22:15] gary_poster: hey, do y'all have a team review process before changes land in lp:~juju-gui/charms/precise/juju-gui/trunk? [22:15] gary_poster: or is the charm review-queue the first review pass on anything [22:16] doh, just read backscroll [22:16] gary_poster: nm... catch you on the flipside [22:24] m_3: yes, we do reviews of the charm; the "Checklist for Preparing for a Review" section of https://bazaar.launchpad.net/~juju-gui/juju-gui/trunk/view/head:/docs/process.rst largely applies [22:24] benji: thanks! [22:26] np