[12:10] seb128, Laney, didrocks: http://people.canonical.com/~ubuntu-archive/component-mismatches-proposed.svg fonts-smc was split. Please could you review the licenses and subscribe desktop-bugs to the split packages? [13:28] sil2100: hi, the trusty version of ubuntu-advantage-tools failed to build: https://bugs.launchpad.net/ubuntu/+source/ubuntu-advantage-tools/+bug/1719671/comments/56 [13:28] Launchpad bug 1719671 in ubuntu-advantage-tools (Ubuntu Artful) "[SRU] include recent version containing fips and livepatch" [Medium,Fix committed] [13:28] is it because of universe? [13:31] andreas: hey, yeah, for a main package an universe package won't really be visible [13:31] I mean, well [13:37] all ppa builds I did before didn't show this problem [13:37] even our daily build ppa [13:37] and how did the other packages build? The ua-tools ones that are in proposed for x, z and a? [13:40] andreas: I guess I need to see first what's happening before I start talking [13:40] eh [13:40] :) [13:41] Anyway, I'll take a look at this in some moments [13:42] thanks [13:49] sil2100, andreas seems like universe hasn't been enable in trusty but xenial yes --> http://pastebin.ubuntu.com/25910954/ [13:51] Anyway, clearly I made a mistake and didn't take notice that u-a-t is actually a main package [13:51] that changed in xenial, but trusty uses the old model [13:52] right [13:52] (sorry, I was a little further back in scrollback without noticing) [13:52] these build deps are only used for tests at buildtime, not at runtime [13:52] does that help? [13:52] irrelevant [13:52] >>> lp.load('/ubuntu/trusty').strict_supported_component_dependencies [13:52] True [13:52] >>> lp.load('/ubuntu/xenial').strict_supported_component_dependencies [13:52] False [13:53] I mean, in terms of policy: a main package having build-deps from universe [13:53] we made the policy more liberal in xenial, but for trusty it does not help [13:54] then let's reject that trusty sru [13:54] you could probably just disable those tests for trusty [13:55] yeah, would that be acceptable from the sru point of view? [13:55] we can show them in the daily build ppa [13:55] andreas: I told you last time it was - if we can't run the tests on a given platform then I'm generally good to skip [13:55] It's just that if a series can run the tests then they shouldn't be disabled [13:56] slashd: could you re-upload u-a-t to trusty with the tests disabled and test deps removed? [13:56] sil2100, sure can take care of it [14:00] slashd: thanks [14:00] sil2100, thanks to you [15:39] does anyone know if https://launchpad.net/~network-manager has a #irc [15:58] !alis | LargePrime [15:58] LargePrime: Alis is an IRC service to help you find channels. For help on using it, see "/msg Alis help list" or ask in #freenode. Example usage: "/msg Alis list http" [16:05] LargePrime: #nm [16:34] thanks nacc cyphermox [16:43] LargePrime: yw === LargePrime is now known as nOOb === nOOb is now known as LargePrime === kurros__ is now known as kurros [19:46] stgraber, infinity, kees, mdeslaur: TB 15 minute warning [19:47] ack [20:31] slangasek: release upgrade failure due to having apt and apt-transport-https:i386 installed. That seems crazy right? [20:31] haha [20:32] bdmurray: yes, that's a crazy state. maybe it's our bug that the system got in that state to begin with, but the failure to upgrade from it is not a bug [20:41] What is the process for moving a package from multiverse to universe if the latest version is fully freely licensed? Asking specifically about the redeclipse package which is licensed under a combination of zlib and CC-BY-SA and which is in the main repository in Debian. [20:41] NewGnuGuy: file a bug and subscribe the ubuntu-archive team [20:42] Which bug tracker should I file under? [20:44] you can run ubuntu-bug redeclipse [20:44] or https://launchpad.net/ubuntu/+source/redeclipse/+filebug [20:49] I'm trying to compile a custom version of python for ubuntu, and I've basically used the .dsc to get all the patches etc, and then dropped in the customizations I need to make, when I compile though the binary behaves differently, the stock python interpretter searches for shared objects with suffixing everything with x86_64-linux-gnu [20:50] while my custom binary does not, and I am unable to figure out how to make my binary search for shared objects the same way [20:51] are the autopkgtest queues actually going down at all? [20:52] mwhudson, yes as fast as fed normalising QE balance sheet. [20:53] mwhudson, it was 14k now it's like 13.9k and things are moving, it's just new things keep being added. [20:53] too. [20:54] ekristen: how are you compilinng? [20:55] nacc: I was using dget -u to pull down everything, then I modified the files necessary, then ran dpkg-buildpackage, then tested the resulting binary [20:56] nacc: it compiles with what I need, it just doesn't search shared objects teh same way. I tried also to compile directly from source from python too, with the same behavior [20:57] I'm sure it's something simple, a small patch or configure option that tells python to look for shared objects in a few different ways, but I cannot seem to figure out where that is setup. [20:57] ekristen: the binary or the .deb? [20:59] nacc: ultimately I'm trying to compile the python interpretter with some custom changes, and install it on ubuntu, ideally I wanted to be able to create a .deb package that contains the binary as well. For my changes to work I've passed the disable-shared, and added my library references to LDFLAGS and CFLAGS, I've tried compiling the binary with and without dpkg-buildpackage. [20:59] It makes the .deb file and the python interpreter with teh changes I need, however when I go to import a module like `gi` for example [21:00] my custom interpreter looks for _gi.so and that's it [21:00] ekristen: i'm asking if you tested the binary (e.g. ./...python) or the binanry package (the .deb) [21:00] however ubuntu's 2.7 python interpetter looks for _gi.so and _gi.x86_64-linux-gnu.so [21:00] nacc: both [21:00] I've tested both [21:02] ekristen: i'm curious what changes you are doing? [21:08] xnox: yeah i know things are being processed, it's just the rate of that vs the rate of things being added that i wasn't sure of [21:09] mwhudson, we are one cloud down. [21:09] still? [21:09] No. [21:09] infinity, it keeps deing taking out the lcy builders for 6 hounts. [21:09] infinity, so after 28 jobs, we are one cloud down for 5 hours. [21:09] xnox: They're being reset more often than that. [21:09] then they come back. [21:09] nacc: hard to explain, some legacy stuff inside my company for a python gtk app we have [21:09] xnox: You're operating on old info. [21:10] ekristen: ok, i'm looking at hte source myself to see [21:10] (Yes, lcy isn't in perfect state, but it's not as dire as that) [21:10] infinity, i'm going by the info, from someone with autopkgtest data, from about two hours ago. [21:10] infinity, there is a problem with lcy when booing bionic guests. [21:10] (all good with xenial guests) [21:10] 11:00 < Laney> nah, I'll make the reset job run hourly so we limp along a bit [21:11] xnox: Trust me, I'm aware of all of this. [21:11] infinity, ok. [21:11] xnox: They're resetting more often than every 6h, as I said. [21:11] slangasek, ^^^ [21:11] infinity, i've been told fake news again then! [21:11] xnox: Not fake, just old. [21:12] xnox: Just irksome when I tell you that it's old news, you keep arguing. :P [21:12] nacc: thanks, any help would be greatly appreciated [21:14] ekristen: i guess my first attempt would be to use a PPA or a good build environment (e.g., sbuild, clean LXD) and see if you can rebuild the existing srcpkg with the same beahvior? [21:14] xnox: it was not fake news when I told you === NCommander is now known as KD2JRT [23:41] cpaelzer: When you're around, would you have a few minutes to talk about the solutions you investigated when fixing https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1620407 ? [23:41] Launchpad bug 1620407 in libvirt (Ubuntu Xenial) "vm startup broken when interface definition has script tag" [Medium,Fix released]