=== kadams54 is now known as kadams54-away === liam_ is now known as Guest65700 [20:00] mwhudson, waigani: email standup again? [20:00] if you like [20:00] menn0: yep, writing mine now [20:03] you guys don't like talking as much as tim, hey? :) [20:04] hehe [20:12] mwhudson: that and, then thumper know what we've been up to while he's away [20:14] fair enough [20:32] menn0: strictLocalID branch not working out? [20:35] waigani: just not seeing the point [20:35] waigani: looking at it now [20:36] waigani: many of the uses of localID still need to be relaxed because they're in generic code that could see old or new style ids [20:36] waigani: so what we have now (localID and strictLocalID) might actually be fine [20:38] menn0: yeah, i thought that is what we concluded on Friday - but use strictLocalID unless you know you have to relax the check [20:38] waigani: yeah we were thinking that might be best [21:55] I keep getting breakage in firewaller_test L139 TestGetMachinePorts ... mismatch at .Results[0].Ports[0].PortRange.FromPort: unequal; obtained 4321; expected 1234 [21:55] anyone else seeing this? [21:57] maybe I just need to go get -u -v, let me see if that fixes it [22:11] bodie_: strange. works for me. [22:15] wallyworld: ping [22:15] hey [22:15] are you still looking at that blocker? [22:16] i fixed it last night, was hoping for a review from tim when woke up, then only just realised he's away this week [22:16] http://reviews.vapour.ws/r/321 [22:16] if you want to look [22:16] looking now [22:16] ty [22:22] there we go. cleared up with a little more careful updating of tools [22:28] wallyworld: why is it important to not write metadata if it hasn't changed? [22:29] menn0: because it otherwise updates the "updated" attribute and caused issues with the scripts used to publish the metadata [22:30] wallyworld: got it. the reason for the change wasn't clear to me. [22:30] np, sorry [22:30] it came up in a conversation over the weekend [22:30] the bug didn't mention it [22:31] wallyworld, ping [22:31] hey [22:36] hey there wallyworld how goes it? [22:36] alright consider it's monday [22:36] yay for monday! [22:36] or sunday evening for sime [22:36] some [22:37] I sent mail regarding an lxc issue, looks like that may be a hot item (or going to get hot) [22:37] I will find out more on monday [22:37] but just giving you an heads up [22:37] sure, ok. if it's the one i'm thinking of, i asked dimiter about it lasy week, and he's doing a temporary fi [22:37] fix [22:37] there's a proper fix as part of the ongoing network stuff coming up [22:38] but right now it's all a bit broken [22:39] yeah there seems to be 3 related bugs (outlined in the email) but I didnt go look at all of them in detail [22:39] it depends on individual machine network setups though as to whether it's broken or not [22:39] wallyworld: review done [22:39] i'll read the email in more detail and see what's up with it [22:39] thanks menn0 [22:40] thanks wallyworld [22:40] np. now go and enjoy the rest of your sunday [22:41] and hopefully we'll have a 1.21 alpha3 elease real soon now [22:53] hm [22:53] I'm trying to add some actions-related testing to charm but it looks like charm master breaks juju master [22:54] so I'm not sure what to base it on -- the v4 branch appears to have the breaking change as well [22:55] charm commit 190b3a5d removes charm/testing/repo/quantal/metered-custom, which is tested against in juju CharmTestHelperSuite.TestTestingCharm [22:56] the charm revision listed in godeps is fine [23:00] waigani: http://reviews.vapour.ws/r/322/ [23:01] dear lazy-irc, is the build blocked ? [23:07] davecheney: yep [23:07] well shit [23:07] davecheney: i think Ian is about to merge the fix though [23:10] menn0: done [23:10] waigani: cheers [23:12] waigani: I like your suggestion. Changing now. [23:22] menn0: ah great [23:28] waigani_: PTAL http://reviews.vapour.ws/r/322/ [23:31] https://github.com/juju/charm/issues/69 [23:31] davecheney: one part of that bug was a map ordering issue, and has been fixed. the other failures are due to compiler issues as per the bug comments [23:33] wallyworld: my mistake [23:33] i missed those in the updates [23:33] davecheney: np, it's a long conversation [23:33] wallyworld: be aware [23:33] gccgo on ppc is completely fucked up [23:33] we don't appear to be able to push the updated revision out properly [23:33] yeah? [23:33] :-( [23:33] this issue is 8 months old now [23:33] it's nearly a toddler [23:34] we need to escalate this then [23:34] ok, not sure who to talk to about this [23:34] alexis [23:35] i haven't been involved in this since vegas [23:35] give her the info and she can prod the right people [23:35] i think you are saying the issues have been fixed but we don't have the latest version in the archives? [23:36] if only it were that simple [23:37] wallyworld: forwarding you mail [23:37] ty [23:37] will send it on [23:40] shall I open a PR against charm v4 so that it is no longer a breaking change but includes the misnamed testing repo, which tests in juju core were written against? [23:40] https://github.com/juju/charm/pull/68 [23:41] or shall we migrate to charm v5, since I also have some changes I would like to make which deprecate some tested bits in charm v4 [23:42] menn0: done, looks good [23:42] wallyworld: ok, done [23:42] importantly, all version of trusty, even 14.04.01 are known to be BROKEN [23:42] davecheney: thanks, i'l follow up [23:42] you must enabled trusty-updates **BEFORE** installing juju [23:42] otherwise -> FAIL [23:43] ouch [23:53] bodie_: what's up? [23:55] bodie_: I'd suggest working with roger and uros on charm updates. The v4 is in the list for core deps. Not sure what 'breaking change' is?