=== uru_ is now known as urulama === rogpeppe2 is now known as rogpeppe [07:43] urulama: yo! [07:46] rogpeppe: heya! [10:52] urulama: why are we storing machine and unit count in extra-info ? [12:00] morning all [12:01] morning [12:02] hi rick_h_ [12:02] and frankban [12:03] how's europe treating your bac? [12:03] it's great [12:04] bac: welcome back [12:04] ty [12:05] bac: for the charm failure, it seems to only be when you have the cloud-archive enabled that you get this failure per the comment from Nicolas. [12:05] bac: I'm going to try to run the full test suite with the dep removed today and if the tests pass I think I'll just push a new charm version without the dep. [12:06] rick_h_: ok [12:06] I seemed to have added it during the git conversion and maybe it was something I didn't need but ran into during testing/debug [13:54] morning all [14:03] howdy hatch [14:31] phew, ok back in irc. I feel human again [14:32] rick_h_: Can I get you to take a look at something for me? [14:32] https://bugs.launchpad.net/charms/+source/rails/+bug/1366660 [14:32] Bug #1366660: Rack charm does not exist [14:32] lazyPower-sprint: looking [14:32] there's some voodoo going on in the store [14:33] so if it used to exist and the branch was removed it's not gone until it's gone from the charmstore (juju go based store) [14:33] yeah, i figured it was some phantom data left over [14:33] this is pretty old, as rack was renamed to trails before I came on board [14:33] this kinesis keyboard is getting easier to use...but really sucks that it has big issues with OSX [14:33] lazyPower-sprint: looking, sec [14:33] hatch: :( stop using OSX :P [14:33] +1 to rick_h's sentiment ;) [14:34] haha kind of hard being the host OS [14:34] lazyPower-sprint: https://store.juju.ubuntu.com/charm-info?charms=cs:precise/rack [14:34] lazyPower-sprint: so yea, we'd need to work with IS, get the charm removal script run against that charm in teh CS, then we have a delete button in charmworld to remove it from there [14:34] ok, do you need me to retarget the bug? [14:35] i dont think this is mission critical to people actually finding it - as its existed for 8 months before i even became aware of it [14:35] but it might be worth adding it as a low priority item to eject it from the store [14:37] lazyPower-sprint: yea, bac do you know if we ever got the delete docs into the wiki [14:38] rick_h_: i'm unsure. i think that was benji who worked on it ... a long time ago [14:38] bac: it was cmars actually that added it to the core charmstore. There's a script in there. I guess I could just look at the charmstore code now [14:39] rick_h_: yes, it was cmars who had created the script. i think benji worked on doing some deletions. i'm not sure where he would've documented it. [14:39] bac: ah found an email with it in an RT [14:40] lazyPower-sprint: forwarded you an email. If you are interested I think it'd be good for you all to know how this works as well as us. [14:40] lazyPower-sprint: you can file an RT for the removal sometime and once that's complete we'll remove from charmworld [14:42] rick_h_: please forward to me too and i'll put it on the wiki [14:43] bac: note that'll change once we get our charmstore in place of the old one [14:43] bac: but good to have for the next little bit [14:43] ty [14:50] rick_h_: so do you do any key remappings? [14:51] hatch: the kenisis? [14:51] hatch: you can recode any key to anything in the keyboard itself [14:52] https://www.kinesis-ergo.com/support/technical-support/faqs-advantage-keyboard/ [14:52] hatch: ^ [14:52] check out the swap from win/osx and the one on remapping a key [14:53] jujugui call in 8min, kanban please [14:54] rick_h_: right, but have you found any favourable mappings? [14:54] delete is my meta key [14:54] = is too far away and {} are hard to hit ive found [14:54] ls [14:55] = isn't a problem for me I guess. [14:55] the win/meta key was too far so moved that down to delete [14:55] well, swapped those two really [14:55] ahh yeah that's a good idea [14:55] other than that, no I don't think I moved much else around [14:56] i have to plug it in after my laptop boots else it blocks the boot process :) [14:56] oh :/ [14:57] rick_h_: sounds good to me. thanks for the follow up [14:57] lazyPower-sprint: np [14:57] lazyPower-sprint: make sure to poke people about GUI/machine view on the orange box stuff please :) [14:58] jujugui call in 1 go go go [14:59] bac: kadams54 ^ [14:59] rogpeppe: ^ [15:07] rick_h_: oops, was lunching... [15:26] jujugui is there a way to force destroy a service so i don't have to resolve all the failed hooks? [15:29] hatch: you can try with "juju destroy-machine --force {machine where the unit is placed} && juju destroy-service service {service name}", but I am not sure it works [15:32] yeah was hoping for a --force on the destroy-service but it doesn't look like it exists === urulama is now known as urulama-afk [16:18] * rick_h_ goes to get some lunch [16:59] frankban: additionally juju-deployer -T is also gnarly for destroying all machines but your bootstrap node. [17:00] lazyPower-sprint: good to know thanks [17:08] vms..... [17:09] oh back and such [17:21] jujugui I need a review and qa https://github.com/juju/juju-gui/pull/534 [17:22] kadams54_: are you going to review huws brach or do you need to pass it off? [17:23] hatch: FYI while you were away lazyPower-sprint mentioned that juju-deployer -T can be used for destroying all machines but the bootstrap node [17:24] frankban: oh cool thanks [17:33] kadams54_: I did the fina.l review on huws branch [17:33] Makyo: I added your tag to the card [17:33] hatch, oh, thanks [17:33] I forgot :( [17:34] haha np [17:50] rick_h_: https://bugs.launchpad.net/juju-gui/+bug/1364956 I'm not sure what to do here as we don't support this interaction any longer [17:50] Bug #1364956: Removing a unit via the inspector creates a sticky remove unit command in the deployer bar [17:50] close as invalid now? [17:52] hatch: otp, will look in a sec [17:53] sure np [17:59] Makyo: your branch ready for review? [18:00] hatch, should be yep. [18:00] cool I;ll take one [18:00] jujugui need some reviews, QA in real env https://github.com/juju/juju-gui/pull/535 [18:00] now [18:00] u got to do one for me :P [18:00] Sure [18:01] Makyo: you have a merge commit in here? [18:02] hatch, yeah, I was worried there would be conflicts around deployer bar stuff. [18:02] ahh [18:02] Scared after my last branch :P [18:05] haha [18:06] Makyo: one comment while my ec2 instance spins up [18:07] hatch, will look into it (this is the confirm button, not deploy, so I stayed away initially) [18:07] ohh ok that may be it [18:12] hatch: so it seems that'd be a good thing to try to get QA on the orange box stuff. [18:13] hatch: I got a contact for the one in london and I'll try to get that email going [18:13] guihelp: I need two reviews for https://codereview.appspot.com/139350043 (quickstart/python). Thanks! [18:13] sounds good I'll put it back into the queue [18:15] hatch: update the card with a "NEEDS QA" prefix to help others not grab it or such [18:15] ok i blocked it with a note about orange box qa [18:15] cool thanks [18:26] Makyo: so I just qa'd and it looked like it didn't change anything. the commit button was still active long after clicking it and the summary closing [18:27] was a commit missed or something? [18:27] hatch, I just tried on lxc. Let me try again on EC2 [18:28] just cleared cache and got the same result [18:30] * frankban grabs some food [18:35] hatch, check for a config-changed hook failure, I get that a lot when setting juju-gui-source\ [18:37] Makyo: hatch I think you have to be on trusty to get the source change correct [18:38] Makyo: hatch and you can check the version.js to make sure it's setup right [18:41] rick_h_: Makyo the files being sent over the wire have the changes as per the pr [18:41] Huh, okay, digging. [18:44] grabbing lunch [19:14] hatch, I see what's going on, misunderstood. Good news is that toggle deploy button will help :P [19:18] guihelp: looking for QA and reviews on https://github.com/juju/juju-gui/pull/536 [19:20] rick_h_: one of the remaining MV cards is for this bug, which is marked as invalid: https://bugs.launchpad.net/juju-gui/+bug/1365053 [19:20] Bug #1365053: Add units dialogue on pre deployment inspector is confusing [19:21] rick_h_, oddly enough that ddclient icon renders fine in inkscape on my box [19:22] jcastro yea, does in gimp here to [19:22] huh [19:27] jcastro: yea, it's odd. I've not seen it before but I'm not an svg spec expert [19:28] jcastro: I tried to look for an svg validation service or the like w/o any luck [19:28] jcastro: so sorry to punt, but if the browser won't show the icon the gui can't either. [19:28] no worries [19:28] I might just redo the icon entirely [19:28] jcastro: k, let me know if I can help. And by help I mean get luca to send you to someone on the design side that knows image steuff [19:29] kadams54_: looking [19:29] kadams54_: ok, so the question then is the add units collapsed by default? [19:29] kadams54_: I think this might have been left as a place holder to do that work [19:31] kadams54_: so just to verify, it's not yet collapsed like the post-ghost inspector view and needs to be done [19:39] guihelp: need one more review + QA for https://codereview.appspot.com/139350043 . anyone available? thanks [19:40] Makyo: :) cool let me know when you'll be ready for another review [19:40] frankban: I can look at it [19:40] hatch: thanks! [19:41] frankban: I don't have a utopic vm [19:41] do I need one? [19:41] hatch: it's not required [19:41] ok sounds good [19:42] frankban: are you going to need a charm release soon around this? [19:43] frankban: I need to do one for the dep fix and the bug on the cloud-archive and wonder if we should sync up [19:44] rick_h_: no I don't, but I can take care of releasing the charm if you want. what was the fix for libapt stuff? just removing the dep? [19:44] frankban: yea, just removing the dep seems to qa ok and passes tests [19:44] frankban: if you can peek at that it'd be cool thanks [19:45] rick_h_: cool if functional tests pass then sounds good: perhaps that was required at some point (maybe as a support library for python-apt). who knows [19:46] frankban: yes, git blame says I last edited the line but can't recall/find any reason for it to be there [19:46] err bzr blame [19:46] rick_h_: ok, I'll start a charm release, which includes running the ftests on precise, and if that works the it's all good [19:46] frankban: ty much [19:46] frankban: there's a card in the top of the board for it atm [19:47] rick_h_: cool [19:47] hmm I also haven't got bzr set up either hah [19:49] hatch: can't get rid of that yet :P [19:49] frankban: can you remind me the best steps to take to get your code down? [19:50] hatch: bzr init-repo juju-quickstart [19:50] cd juju-quickstart [19:51] bzr branch lp:juju-quickstart trunk [19:51] bzr checkout --lightweight trunk sandbox [19:51] * rick_h_ runs away for the day. Have a good night all [19:52] bzr branch lp:~frankban/juju-quickstart/utopic-update-dependencies [19:52] cd sandbox [19:52] bzr switch ../utopic-update-dependencies [19:52] hatch: ^^^ [19:52] oh wow all the steps :) thanks [19:53] it's been a while since i've bzr'd [19:53] hatch: and then, in the future, you just have to start from branching the code you need to review [19:55] excellent.....can we switch to git yet? :P [20:02] frankban: I don;t have hp creds, did you test hp? [20:02] hatch: in theory we can... the only missing thing is a way to automatically build deb packages in the PPA. anyway, I still like the bzr+reitveld workflow (especially the reitveld part) ;-) [20:02] hatch: I did, you can skip that step if you want [20:04] ok I'll test the rest thx [20:04] you LIKE that workflow? ugh I want to rip my eyes out lol [20:05] hatch: heh, github code review is less than optimal IMHO [20:06] true but the bzr workflow leaves quite a bit to be desired [20:07] hatch: uhm... you convinced me, I'll move quickstart to bitbucket so that we can all enjoy mercurial ;-) [20:07] lol!! [20:07] victory is mine [20:08] :-) [20:11] frankban: lgtm'd [20:11] hatch: thanks! [20:14] rick_h_: are you ok with us no longer testing non mv stuff? [20:14] oh he's gone [20:14] heh [20:23] kadams54_: review done - I don't see any tests for anything other than the changes made to state.js [20:23] Makyo: is yours ready for another qa? [20:58] jujugui i still need reviews on my last branch [21:07] Makyo: hey u kickin around? Can you remind me of the details for the card "Clear config-changed ECS entries" [21:07] hatch, we need to be able to back out config changed entries in the ECS. It's the only thing we can't reverse. [21:08] oh right because we don't know the original values [21:08] Correct. [21:08] ok I'll work on creating the cloned config card [21:29] guihelp: Need reviews and QA on: https://github.com/juju/juju-gui/pull/537 [21:52] Makyo: ok so for this cloned local config store... [21:52] I've come up with 2 rules [21:52] local config changes do NOT modify it [21:53] delta changes do....even if said changes aren't represented in the inspector because of a conflict [21:53] sound good? [21:54] that way we always have a record of the 'true' config values in juju [21:54] so reverting any changes will revert to what juju has, not what was local [21:54] Yeah, I think that's what we need; we already have a conflict resolution story in place. [21:54] YEp [21:54] RFC? [21:55] rfc? [22:59] Morning [23:03] morning huwshimi [23:05] hatch: How's things? [23:05] they are ok....still getting used to my new kinesis advantage keyboard...geting better though [23:06] I cam type pretty well but special chars are still hard heh [23:07] woah, that thing looks crazy [23:09] yeah it is, but besides the cognitive load being exhausting now my fingers don't really feel like they have done anything all day compared to my old ergo keyboard [23:11] I'll likely have to remap some of the keys for coding but other than that it's pretty awesome....well except that my latptop won't boot with is plugged in [23:12] ...lol [23:12] oh, hah [23:13] yeah...not sure waht to do about that [23:13] they say it's apples fault....but I don't know of any other keybaord that does that [23:14] nice customer service [23:15] yeah...especially at [23:15] this price...i [23:15] ouch [23:22] hatch: Do you happen to know how to delete individual units using the ecs? If I do a db.removeUnits(...) it doesn't seem to go through the ecs. [23:22] you have to make the call of the env [23:22] the env then hits the ecs [23:23] huwshimi: https://github.com/juju/juju-gui/blob/develop/app/store/env/go.js#L1310 [23:23] hatch: Ah great! [23:29] hatch: Strange, if you remove a deployed unit it removes the machine as well! [23:30] (deployer bar just says it will remove the unit) [23:32] hmm that shouldn't be correct [23:32] is this in a real env or just sandbox? [23:32] because I did that exact interaction a few times today and it left the machine running [23:32] hatch: sandbox [23:34] hmm lemme take a look [23:35] you are correct [23:35] that is odd heh [23:37] :( [23:37] I'm guessing you need that machine to stay working? [23:37] :) [23:40] hatch: Not for this branch, but it seems like a bug to me :) [23:43] yup, but a sandbox bug at least