=== axw_ is now known as axw [02:29] wallyworld_: reviewed, let me know what you think [02:29] sure ty [02:36] axw: i had to move providerregistry to avoid circular imports, from memory with LoopProviderType. i think adding a registry package is a good idea [03:47] wallyworld_: thnx for env destroy and suggestions :D [03:47] waigani: ^^ not wallyworld_ [03:47] anastasiamac_: welcome :) [03:47] waigani: i'll put on my todo list [03:48] waigani: but while tests pass with calling RemoveEnvironment [03:48] waigani: probably not so high a priority to change :D [03:48] anastasiamac_: absolutely. Nothing urgent. I just wanted to follow up from last week. [03:49] waigani: it's awesome! thank u :D [03:49] waigani: ur thoughtfulness is highly appreciated! [03:50] :) [06:54] jam: hey, did you discuss with mark the storage CLI? [06:55] axw: ty for review, that branch now merged [06:55] wallyworld_: cool :) [06:58] axw: i'll now convert Pool to storage.Config etc [08:17] dimitern: poke? [08:18] jam, I've got kicked out [08:18] jam, and now I'm the only one in the hangout [08:25] morning [08:28] TheMue, o/ [08:28] dimitern: o/ [10:00] dimitern, TheMue, voidspace: standup? [10:00] dooferlad: sure ;) [10:01] omw === ashipika is now known as ashipika|food [11:21] phew, all those nested references to network parameters :/ [11:40] jam: hey there [11:43] wallyworld: hey [11:44] jam: have you had a chance to talk to mark about the storage CLI changes we are proposing? [11:44] I have not, though I'd still advise you to make progress and I'll chat with him about it. [11:45] ok, will do, thanks [11:50] wallyworld: which changes are those? [11:50] axw: the ones where we use 3 levels [11:50] juju storage pool create [11:50] as oposed to [11:50] juju storage create-pool [11:51] wallyworld: I thought nobody liked it? [11:51] the spec says create-pool [11:51] axw: i sent an email saying we were revisiting it (or at least i thought i did) [11:51] ok, I probably just forgot [11:52] np, i thin storage pool create etc is better [11:52] so want to try and get that approved [11:53] axw: https://docs.google.com/a/canonical.com/document/d/1JMDDOPXHV6SPrTSA3SRb_qyyRUYcz7huo_tTIjD8UOc [11:53] wallyworld: ta [11:56] wallyworld: I think we'd still have "juju storage add", at the top level, allowing you to add storage instances to a unit or service [11:57] wallyworld: just saying since it's missing from the first "--help" output example .. not sure if that's meant to be complete [11:57] wallyworld: axw: jam: thnx :D [11:57] axw: yes, true. i was mainly focusing on what we had implemented so far [11:57] axw: feel free to add to doc to make it more completed [11:58] i just update dsharing [12:53] wallyworld: thanks for the review [12:53] sure, np [12:53] thanks for fix, will be good to get 1.21/1.22 out [13:20] axw: oarsome, thanks for review [13:20] wallyworld: nps [13:42] where can i go to to find out whether juju-core updates are currently blocked? [13:42] rogpeppe: http://goo.gl/4zd1e9 [13:43] jw4: so, any critical bugs == blocked ? [13:43] rogpeppe: yeah, that's basically the same query CI uses [13:43] jw4: ta [13:43] rogpeppe: if you prefer json: https://api.launchpad.net/devel/juju-core?ws.op=searchTasks&status%3Alist=Triaged&status%3Alist=In+Progress&status%3Alist=Fix+Committed&importance%3Alist=Critical&tags%3Alist=regression&tags%3Alist=ci&tags_combinator=All [13:43] * rogpeppe puts the PR on hold for a bit more [13:44] jw4: that json query doesn't give me any results [13:44] rogpeppe: hmm; my JSON query shows zero though... yeag [13:44] yeah [13:45] jw4: (JSON is indeed preferable for me in this case) [13:45] rogpeppe: I personally use the JSON one :) [13:46] jw4: except it doesn't work :) [13:46] rogpeppe: I wonder if the 'Incomplete' status is why it's not showing [13:48] jw4: this one also return empty: http 'https://api.launchpad.net/devel/juju-core?ws.op=searchTasks&status%3Alist=Fix+Committed&importance%3Alist=Critical&tags%3Alist=regression&tags%3Alist=ci&tags_combinator=All' [13:48] jw4: this query is helpful in most cases... however, in this particular case, it's not right [13:48] jw4: TRUNK IS NOT BLOCKED [13:48] jw4: sorry for caps... [13:48] jw4: but removing the Fix+Committed qualification makes it work [13:48] anastasiamac_: that's useful to know! [13:48] anastasiamac_: thanks [13:48] anastasiamac_, rogpeppe yeah - thanks! [13:49] jw4: did not mean to cap :D [13:49] anastasiamac_: lol [13:49] so it seems that it might be nice to have an indepdendent source of truth maintained so that we can actually see whether it's blocked or not [13:49] jw4: but landing has been happening today :D [13:49] rogpeppe: +1 this would be gr* and was the intention of jw4's query... [13:49] and so that the criteria can be changed without everyone needing to update their search query [13:50] perhaps someone can grab are-juju-core-commits-blocked.com :) [13:51] rogpeppe: and make it branch specific too, so if u were curious about 1.21, u'd b able to "know" at a glance.. [13:51] it would be even better if the 'bot automatically retried previously blocked commits after blocking is removed [13:51] anastasiamac_: yeah [13:52] rogpeppe: not sure about automatically re-try but at least notifying the person that tried to merge :D [13:53] rogpeppe: it's nice to dream... especially since it's midnight here :D m eod-ing [13:53] anastasiamac_: i think an automatic retry would be best, particularly if no further commits have been pushed or comments made since the block message [13:53] anastasiamac_: good night [13:53] rogpeppe: nite [13:53] anastasiamac_: i had no idea you were that far east! [13:56] rogpeppe: Surprise :D [13:58] anastasiamac_: eastern australia? [14:01] yummy, api/apiserver tests are passing again [14:12] dimitern: I'll keep the NetworkInfo. we've several other types and fields named based on it. so we (a) would get inconsistent or (b) change the wire protocol [14:15] TheMue, what are these cases? [14:16] dimitern: e.g. MachineNetworkInfoResult with the Info field and MachineNetworkInfoResults containing the former [14:18] TheMue, that's fine - the field can be called Info, but the internal type should be params.NetworkConfig [14:19] dimitern: and so don't change the name of the types too? [14:19] TheMue, I'd add a TODO comment there to change the field from Info to Config [14:19] TheMue, or even name the field Config but leave the json:"Info" tag on it [14:20] TheMue, which types? [14:20] dimitern: MachineNetworkInfoResult in MachineNetworkConfigResult, and ...Results too [14:21] TheMue, ah, ok - let's call them MachineNetworkConfigResult etc. as the type names don't matter [14:21] dimitern: also in the code we reference it as info :D [14:21] dimitern: but that's easy to change [14:21] TheMue, cheers :) [14:24] so can someone tell me a definitive query that lets me know whether juju-core commits are blocked, please? [14:24] dimitern: networks counterpart in one place is InterfaceInfo *lol* the trail of the info [14:29] TheMue, indeed, we'll likely change that at some point as well [14:29] rogpeppe: `bzr branch lp:juju-ci-tools; cd juju-ci-tools; ./check-blockers.py master ${YOUR_PR}` [14:29] TheMue, but the params.NetworkConfig should correspond to network.InterfaceInfo for now [14:29] mgz: thank you [14:29] dimitern: I'll add TODOs [14:30] TheMue, +1 [14:32] mgz: one tiny change - the "-" should actually be "_" there [14:32] mgz: thanks, i've been looking this morning for that repo again, but I couldn't remember where it was [14:32] ups yeah, underscores [14:34] mgz: it would probably be better overall if it was an online status, because if someone changes the criteria, i'll have to re-pull that repo [14:39] we've talked about putting it on reports [14:41] you could still have an out of date page, running the script is just what the lander does in practice [14:41] I struggle to believe pull/run script is that complex for a bunch of elite programmers [14:42] mgz: tsk tsk.. .snarkiness ill becomes you ;) [14:43] hey, I struggle to get underscores vs hyphens right :) [14:43] haha === kadams54 is now known as kadams54-away === kadams54 is now known as kadams54-away === kadams54-away is now known as kadams54 === kadams54 is now known as kadams54-away [16:41] * dimitern eod-s === kadams54-away is now known as kadams54 === kadams54 is now known as kadams54-away [17:41] anyone know how to enable feature flags? === kadams54 is now known as kadams54-away [17:49] rogpeppe: export JUJU_DEV_FEATURE_FLAGS="" [17:50] jw4: i just managed to work that out... [17:50] heh [17:50] :) [17:50] jw4: if i set it when bootstrapping an environment, it'll be inherited by all that environment's instances, right? [17:50] jw4: (i *think* i saw that in the code, but just making sure) [17:50] rogpeppe: that is beyond my ken [17:50] rogpeppe: but that sounds plausible :) [17:51] jw4: well... we'll see what happens if/when this bootstrap works [17:56] jw4: it worked! [17:56] rogpeppe: suhweet [18:43] ok, EOD [18:43] g'night [20:31] wallyworld: thanks for the review === kadams54 is now known as kadams54-away === kadams54 is now known as kadams54-away [22:29] wallyworld_: I've addressed your review, thanks: http://reviews.vapour.ws/r/928/ [22:30] ok [22:40] waigani: reviewed, have to head to dentist in about 20 mins, will check as soon as i'm back if not before I go [23:02] wallyworld_: thanks