=== Noskcaj10 is now known as Noskcaj === iulian is now known as Guest40084 === Guest40084 is now known as iulian === FourDoll1rs is now known as FourDollars === vanhoof_ is now known as vanhoof === ev_ is now known as ev === balloons_ is now known as balloons === jpds is now known as Guest46086 [11:47] do i have network connectivity to ports archive, when building arch:all packages on distro builders or ppas? [11:48] Not guaranteed, but the archive in sources.list will be a non-split one so you should use whatever it says [11:49] I remember writing code somewhere to parse sources.list for that ... [11:49] Ah yes [11:49] xnox: see e.g. grub2-signed [11:52] cjwatson: nice. let me see, if I can extend that for my needs. [12:12] could somebody please approve that? ^^ [13:23] chrisccoulson: Do we have new versions for P and Q too? [13:24] infinity, yeah, 1 second [13:24] infinity, ok, those are uploaded as well now === debfx_ is now known as debfx === knome_ is now known as knome === Riddelll is now known as Riddell === henrix_ is now known as henrix === balloons_ is now known as balloons [13:59] cjwatson: is there an easy way to check why a package takes so long in proposed sometimes? just wondering if we have any sort of logs or pending tests and such so I can better track how much time it'd take still to be fully published in the archive [13:59] rsalveti: If it's still in proposed, it's easy to tell why. [13:59] rsthat should be solved since beginning of this week [13:59] just remember that I had to wait quite a few hours for upstart and network-manager over the past few days [13:59] rsalveti: NM was caught in a mini-transition of some sort, I believe. [13:59] i havent had any package that took more than 50min turnaround time since monday [13:59] Not sure what your upstart issue was. [13:59] the same package took between 4 and 6 h last week [13:59] cool then, nm might be some other issue it seems [14:00] will let you guys know if I get such behavior again [14:00] yeah [14:00] it should be gone [14:00] infinity, i thought there was supposed to be an annoucement mail ? [14:00] * ogra_ hasnt seen one [14:01] ogra_: There was, I suppose I should write that this afternoon. Though, I didn't want to write it while there were other kinks being worked out. Things seem smooth now. [14:01] Ish. [14:01] :) [14:02] awesome [14:02] * rsalveti hugs the release team [14:07] rsalveti: We have loads of logs under http://people.canonical.com/~ubuntu-archive/proposed-migration/ [14:07] rsalveti: network-manager got stuck on a bug in our autopkgtest handling, now fixed [14:07] cjwatson: awesome, thanks [14:10] rsalveti: upstart was bitten by a bug in auto-package-testing relating to virtual package handling and had to be manually forced; the underlying bug there has been fixed since Friday [14:11] (Or so I believe, judging from lp:auto-package-testing r212) [14:14] cjwatson: got it, so it's just be being unlucky, but seems it should all be better now [14:14] *me === rbasak_ is now known as rbasak [14:15] rsalveti: The last couple of weeks have had some fairly big (though quiet) archive-infrastructure changes, but the worst of it has settled down [14:16] So yes, I think it should mostly be better, though I'd still appreciate being told about any oddities while they're happening so that I can investigate [14:16] great, sure, will let you know if I get stuck again with any package [14:16] i doubt you will see more than 1h turnaround time with anything now (unless it actually takes hours to build indeed) [14:17] Or unless it's genuinely stuck in a library transition, or with build or test failures [14:19] yeah [14:20] can another AA look at iproute2 in binNEW? I've got a couple of packages that are stuck on the iproute => iproute2 transition so it'd be nice to have iproute2 in the archive === psivaa-afk is now known as psivaa === popey_ is now known as popey [14:31] stgraber: Ew, can we not fix the failure to check return values instead? The security team is usually anal about those in MIR audits. :P [14:31] Also, whoever NEWed that source to universe shouldn't have... [14:33] stgraber: Also a bit suspicious that we had to disable -Werror and Debian didn't? Same glibc version... === shadeslayer_ is now known as shadeslayer [14:35] infinity: well, those warnings also appear in our current iproute but are ignored there, so I thought we wouldn't be any worse [14:35] stgraber: Anyhow. Accepted, but it would be nice to dig deeper into that. [14:36] (And promoted to main, where it should have been) [14:39] infinity: No human NEWed it into universe - it was auto-synced in bulk [14:42] cjwatson: Ahh. That makes more sense. [14:43] infinity: I'm not seeing those warnings in the Debian sbuild log at all (and it's clearly building with -Wall -Werror there too) ... [14:47] stgraber: So, this could be because it's not been built with gcc-4.8 in Debian. Maybe. [14:48] Oh, hrm. No. The armhf buildd, at least, was 4.8 [14:48] infinity: yeah and 4.7 also had the warning, our old iproute from raring has the warnings in its buildlog [14:48] Is Debian gcc on armhf Linaro or FSF? [14:49] ScottK: I think doko just went Linaro across the board for Debian to save his sanity. [14:49] OK. [14:49] yeah [14:50] stgraber: Weird. [14:51] stgraber: Maybe those functions only grow a __wur when built with FORTIFY_SOURCE or some such. [14:51] * infinity is too lazy to look right now. [14:53] Oh, or maybe __wur itself only triggers when building with FORTIFY_SOURCE. :P [14:53] It's amazing the things one forgets. [14:54] The latter, I think. [14:54] And how quickly one forgets them at my advanced age. [14:54] Using dpkg-buildflags in Debian would probably show it up. [14:54] cjwatson: Yeah, that "or maybe" was more of a "in fact, it's like this". === shadeslayer is now known as Guest33135 === charles_ is now known as Guest68669 === yofel_ is now known as Guest48668 === Termana is now known as Guest35591 === jbicha is now known as Guest68084 === Guest33135 is now known as shadeslayer === Guest68084 is now known as jbicha === Guest48668 is now known as yofel === medberry is now known as med_ === kees_ is now known as kees [19:05] infinity, do you know if livefs builders can see jenkins ? [19:05] or cjwatson ^^^ [19:05] better said lillypilly or nusakan [19:06] sergiusens needs to get the click packages into the build somehow ... preferably without having to post-process them on nusakan indeed [19:06] it's for adding click packages to the touch builds, I'm creating them on jenkins and can copy them over to nusakan scriptiseally [19:10] They almost certainly can't see jenkins. [19:11] They *can* see archive-team.internal (which is ~ubuntu-archive/public_html on lillypilly), but we could poke a hole for something more appropriate, perhaps. [19:11] I assume these click packages will have some slightly more public and sane archive than jenkins at some point? :P [19:12] infinity: they certainly should (I'm not on the hook for that one though) [19:20] when you say poke a hole, it means all the way to jenkins or some other location? [19:25] or I suppose we could have something pulling from Jenkins and pushing to ~ubuntu-archive/public_html on lillypilly (seems vaguely cleaner than direct access from the buildds to jenkins, though adds an extra delay) [19:25] it'd also be slightly better for non-Canonical people as they'd actually be able to see what's being pulled in [19:31] Yeah, that would be preferable as an interim solution. But I think we probably need to discuss how these packages are being built and published. [19:31] I'm not a huge fan of all of this happening outside the usual developer playground. [19:49] infinity: cjwatson would have the long term plan for that I think [19:49] click is just too new to have a nice workflow === med_ is now known as Guest36906 === blair is now known as Guest80769 === yofel is now known as Guest395 === hyperair is now known as Guest96566 === jbicha is now known as Guest14814 === Guest395 is now known as yofel [20:29] apparently, none of the flavors still show up at status.ubuntu.com === barry` is now known as barry === jokerdino_ is now known as jokerdino === stgraber_ is now known as stgraber [22:56] gnome-online-accounts seems stuck in saucy-proposed, the libgdata and e-d-s tests seem to have passed [23:00] update-excuses says libgadata is still running. [23:00] See what happens after the next publisher run, I'd say. [23:04] It's been allegedly RUNNING for a good ten runs now ... [23:08] ./data/adt/saucy-proposed/amd64/archive/2013/07/10/saucy_amd64_libgdata_20130710-205401.result:saucy amd64 libgdata 0.13.3-2 PASS gobject-introspection 1.36.0-2 liboauth 1.0.1-1 libgdata 0.13.3-2 libxml2 2.9.0+dfsg1-4ubuntu5 libsoup2.4 2.42.2-6 glib2.0 2.37.3-1ubuntu2 eglibc 2.17-0ubuntu5 gcr 3.8.2-4 gnome-online-accounts 3.8.2-1ubuntu1 [23:08] so, huh, what's it doing [23:08] Oh, it thinks it's still running on i386 [23:10] I've forced it and will check with jibel in the morning === elmo_ is now known as elmo [23:36] thanks