[08:34] morning gmb, how are you doing? [08:48] gmb: would you like to pair on some card? there is only one free slot in the lane [11:04] * gmb lunches [11:08] ahoy [12:06] benji, did you ask jml if we should run with the tag leakage work he did, or is he planning to bring that to roost soon? [12:06] gary_poster, I think benji's out at the toothiologist. [12:07] gary_poster: I didn't. [12:07] I stand corrected. [12:07] that was bac, wasn't it? [12:07] * gary_poster goes to look [12:07] :-) [12:08] benji, do you think that's a reasonable thing to do? [12:08] benji gmb frankban call in 2 [12:08] gary_poster: I do [12:09] Meh, people whose name's all start with "B" confuse me. [12:09] excellent, benji. hey, benji, would you please go talk to jml? you could coordinate with him on the... [12:09] gary_poster: sure [12:09] :-P [12:12] frankban, yoo hoo [12:50] gary_poster: Katie came down with what I had last week and I just found out that one of our cats has a vet appointment at 2:30 (our one-on-one time). Do you mind if we reschedule so I can take him to the vet? (It should be a short visit and the vet is just outside the neighborhood.) [12:50] no problem, of course, benji [12:51] benji, 1:15? [12:51] gary_poster: sounds good [12:51] cool, I will adjust calendar [12:57] thanks [13:30] frankban, hey, https://talkgadget.google.com/hangouts/_/extras/canonical.com/goldenhordeoneonone when you are ready. no rush [14:24] i am back from the toothiologist [14:31] gary_poster: look at all of this local cloudiness we're missing out on: http://www.ibm.com/solutions/education/cloudacademy/us/en/cloud_academy_conference_program.html [14:36] gmb, call in 1 or 2 [14:37] bac, interesting. I wish it were easier to keep abreast of all these opportunities in the area [14:38] Cloud "Academy" ? [14:38] gary_poster: yep, i seem to find out about stuff one or two days late [14:38] I guess it is education-y [14:38] and thus "Academy" [14:39] I mean, for education-related organizations [14:39] It looks a bit like an IBM advertisement for their cloud offerings t education people [14:39] the university is particularly bad about advance notice of events [14:39] hm [14:39] gmb you around? if so, https://talkgadget.google.com/hangouts/_/extras/canonical.com/goldenhordeoneonone [14:40] bac, I've done searches for calendars of tech stuff in the area to no avail [14:41] find a need and fool it, as Zippy the Pinhead says: maybe we should make one! [14:41] gary_poster, Sorry, my connection went away and I didn't notice. [14:41] bac, this is where we are with the PPA and dependencies and stuff: https://docs.google.com/a/canonical.com/document/d/1msejam1-AliaJ55lVEYayfWlmHTetjGheF3Tzjm9evA/edit [14:41] I'm back up now, though. [14:41] cool gmb [14:42] https://talkgadget.google.com/hangouts/_/extras/canonical.com/goldenhordeoneonone gmb [14:42] ok [14:44] gary_poster: i can start on those PPA tasks if benji hasn't [14:44] bac, +1 [14:45] he is working on the fix for that MP [14:45] last I checked [14:45] bac also are you done with you slack card? [14:45] gary_poster: i just posted something in #juju about the slack card. i'll move it now. [14:45] and finally you could get a juju 8 core buildbot instance up in (optimistic) preparation for figuring out how to hook all of this up [14:45] bac: I just updated testr in the ppa; I /think/ the others are up to date, but I'm checking against that doc now [14:46] and maybe run a daily test t=while we are at it bac [14:46] gary_poster: ok [14:48] bac: I just kicked off a build of testtools-daily which should take care of the second bullet point on that doc (the first has been done too) [14:49] gary_poster, Still there? [14:49] bac: I have view-only access to the doc for some reason (I'm using my canonical account.) [14:49] I'm guessing no, then. [14:50] gmb, we are into robot lag gmb :-( [14:50] * gary_poster contemplates skype as an alternative [14:50] gary_poster, Skype works, but I don't have anything to add really; I'll be mailing the list with my UDS plan anyway... [14:50] do you have skype gmb? I just got that on my precise recently [14:50] the issue might be on my side, I don't know [14:50] gary_poster, I have skype, and also OSX... [14:51] let's skype anyway, if you don't mind, gmb. lemme log in [14:51] benji: i cannot edit it either. boo google [14:51] boogle [14:51] Sure [14:51] benji: when the ppa finishes i'll kick off a run [14:51] bac: cool [14:52] gary_poster: do you want to do the verification dance you did last night or shall I? [14:52] bac benji I just made doc editable for you [14:52] thx [14:52] benji, if you can do it locally, go for it. If you need an ec2 instance, bac, have you started one yet? [14:53] benji, also, we are not really done until the filter stuff is there [14:53] or is it done now? [14:53] I made the first two bullet points strike-through. [14:59] bac: the ppa builds are done; since you are going to verify that the code in the ppa works I will look at the state of the tag filtering [15:08] bac: darn, the ppa build of testtools didn't work, investigating [15:08] benji: ok, killed my run [15:09] it is deceiving that packages that are not yet published show up in the PPA only to die later [15:15] bac: I need to add a "ppa#" element to the version string of https://code.launchpad.net/~yellow/+recipe/testtools-daily to make this work, any objections? [15:16] I intend to model it after https://code.launchpad.net/~yellow/+recipe/testrepository-latest which is working well. I will attempt to forgo the "++z" bit though. [15:16] benji: there is already a ppa [15:16] bac: right, but I'll be changing it from tracking a revno to being hand-updated. [15:17] The version number not changing is what caused the packaing to fail. [15:18] benji: i think the right thing to do is to merge trunk back into the ~yellow version and use it [15:18] at least until we figure out a better scheme [15:18] bac: ok, will do [15:20] bac: build now pending [15:21] benji: we are screwed now when we want to switch back to trunk as our forked branch will have a higher revno [15:21] "screwed" in that we can't use the current version string and get higher versions from trunk, right? [15:22] I'm starting to think that synthetic, hand-managed version numbers are the only sane approach, that or using the time. [15:24] benji: yeah, but hand-coded ones will fail the first time an automatic build is triggered, so we'd need to change to 'build on request' or just take the failure email as a trigger to manually bump the version [15:25] yep [15:26] makes me think of a hack: have a bot with an email address that gets those failure messages, logs into LP and increments the ppa number and then requests a build [15:26] gary_poster: I am going to create cards for subunit installation in the host if you agree, and I don't remember what you suggested about the lxc-ip branch [15:27] frankban, +1 on subunit card... [15:27] lxc-ip branch we never actually got around to me saying [15:27] I think I outlined the problem and then we got sidetracked [15:28] lemme go check at the door & maybe I'll have an answer when I get back [15:28] :-) [15:28] gary_poster, Hahaha. See Elliot's latest email to the -tech list about bzr vs git. [15:29] cool thanks gary_poster [15:35] gmb, heh, and I'll be interested in seeing Elliot's reply to Chase's follow-on question. It might conceivably also affect your LP presentation re support and the future [15:38] Indeed... [15:39] frankban, we've talked about an answer that I don't love, but I don't see a better one. If you have another suggestion, *please* make it. But if nothing else better comes to mind.... (1) the lpsetup package will have lxc-ip, but named something else, like lp-lxc-ip. (2) we will contribute lxc-ip to lxc for the Q/12.10 release. (3) Q versions of lpsetup would ideally simply have a symlink (or similar) of lp-lx [15:39] c-ip to lxc-ip. Tools using lp-lxc-ip culd gradually migrate to explicitly using lxc-ip at that point, when it makes sense. [15:39] what do you think of that? [15:46] benji: 0.9.14-bzr267~ppa40~precise1 is built and published. i'll kick off a test run now. [15:46] gary_poster: that makes sense IMHO. lpsetup can just ship lp-lxc-ip until lxc-ip will be available, and obtain the ip using subprocess rather than calling internal python function [15:49] gary_poster: once lxc have lxc-ip, we just have to 1. remove lp-lxc-ip from the branch, 2. change packaging, 3. change the script name in lpsetup settings file. [15:49] frankban, cool [15:49] frankban, let's do it then. :-) [16:18] benji: after some fiddling, i finally got a master and slave deployed. setuplxc is now running for the next hour or so. [17:01] sitrep: in presuing tag support for libcppunit-dev I managed to get the tests to run on Jonathan's branch, with 10 failures/errors. I'll looking at them now. [17:14] libcppunit-dev? [17:14] benji, preparing, then call in a minute [17:15] gary_poster: perfect [17:16] benji, no rush, ready when you are https://talkgadget.google.com/hangouts/_/extras/canonical.com/goldenhordeoneonone [17:25] gmb, if at all possible please try to have your card out of review by your EoD. If that's not reasonable, np, but everything else is blocked by out-of-squad issues [17:30] gary_poster, benji: my slave had an install error b/c launchpad-developer-dependencies was not found. i'm trying to manually locate it and cannot. shouldn't it be in the launchpad PPA at https://launchpad.net/~launchpad/+archive/ppa ? [17:31] yes bac [17:31] weird [17:31] no idea on the face of it [17:31] we are on call but if you need us s'ok [17:31] well that looks quite bad. i'll investigate [18:09] turns out the package 'launchpad-dependencies' we see in the PPA actually builds all of the various 'launchpad-dependencies-*' packages [18:22] bac, right, sorry, I knew that. But so why did the installation fail? [18:24] gary_poster: unclear. http://paste.ubuntu.com/937232/ [18:25] "it is not going to be installed" [18:25] thank you, error message [18:25] bac, have you tried installing manually? maybe a fluke? [18:27] gary_poster: i didn't and the instance is gone now. if this repeats i'll go into the container and try manual [18:47] bac, btw, feel free to tell me you want the call at any time. (I do need to leave about 4:10 or 4:15, but that shouldn't affect our regular call time if you just want to have it then.) [18:48] gary_poster: [18:48] ok [18:48] :-) [18:48] hey my slave just died again [18:48] with install error [18:48] :-( [18:48] so how does one get into lptests? uname/passwd ? [18:51] gary_poster: ^^ [18:52] bac, (1) on the host as root change your password (passwd) [18:52] (2) get hash from /etc/shadow [18:53] (3) copy hash over to /var/lib/lxc/lptests/rootfs/etc/shadow for root [18:53] gary_poster: gah, right. [18:53] Profit! [18:53] pita [18:53] y, but quick [18:56] gary_poster: The following packages have unmet dependencies: [18:57] launchpad-dependencies: Depends: python-apt (>= 0.7.94.2ubuntu6.4) but 0.7.94.2ubuntu6 is to be installed [18:57] so it looks like a problem with our ppa, as i think we maintain python-apt [19:00] gary_poster: you want to talk now? [19:07] I'm back from the vet. [19:10] benji: did you get dipped? [19:10] gary_poster: i can have our talk whenever you're ready [19:14] bac, sorry, was indisposed. yeah, let's go to oneonone [19:19] :) [19:23] bac, is this your network, or should we try skype? google hangouts have been sadly sucking today [19:24] google hangouts says the hangout room is too full for me to join [19:24] haven't see that before [19:25] * benji looks at the Go tutorial. [19:25] skype? [19:31] gary_poster: skype me then [19:33] bac bzr+ssh://bazaar.launchpad.net/+branch/ubuntu/lxc/ but also... [19:34] https://code.launchpad.net/~ubuntu-branches/ubuntu/precise/lxc/precise [20:10] bac, TEMP=$(pwd)/temp testr run --parallel --subunit --full-results -- -t stories/gpg [20:51] gary_poster: there should be no manual tweaks to the slave required now, correct? [21:20] gary_poster: i'm seeing zope:layer tag [21:20] time: 2012-04-19 20:56:35.592186Z [21:20] time: 2012-04-19 20:57:09.998124Z [21:20] test: lp.testing.layers.BaseLayer:setUp [21:20] time: 2012-04-19 20:57:10.072074Z [21:20] tags: worker-4 zope:layer [21:20] successful: lp.testing.layers.BaseLayer:setUp [ multipart [21:20] ] [21:24] at least we're getting the worker tags [21:43] yay worker tags [22:48] bac, that means we should verify that we have jono's fix. if we do, it apparently is not a fix [22:48] to state the obvious [22:48] but my point was, let's first verify that we have his fix [22:52] oh, bac, actually, that's fine [22:52] this is the important part, and it is correct [22:52] test: lib/lp/registry/stories/gpg-coc/xx-ubuntu-codeofconduct-signer.txt [22:52] time: 2012-04-19 22:11:16.022175Z [22:52] tags: worker-0 [22:52] successful: lib/lp/registry/stories/gpg-coc/xx-ubuntu-codeofconduct-signer.txt [ multipart [22:52] ]