[00:06] nacc: logout/login again [00:13] sarnold: yeah, that gave some 500s and other errors [00:13] sarnold: it worked again after a bit :) [00:13] nacc: ooh fun! :) [00:15] sarnold: :) [00:28] ginggs: when syncing packages for a lib transition, just a reminder to check that the lib package is built and published on all archs first; uploading no-change rebuilds of gimagereader,openalpr now for arm64+armhf [00:29] nacc: So wait, I can start pestering you soon? :D [00:34] Unit193: well, pending approval eventually :) === athairus is now known as afkathairus === afkathairus is now known as afkthairus [08:18] slangasek: thanks, i did wait and check with rmadison, but must have missed something === fginther` is now known as fginther [15:19] cjwatson, infinity, something is behaving very weird regarding PPA packages and seeds today ... could that be related to archive re-org ? [15:23] ogra_: can you please give more details? [15:24] cjwatson, i uploaded a 0.7.41~ppa1 version to the snappy PPA earlier (deleted it again since that breaks images) ... from that moment on the build didnt pull it in with the normal seed anymore ... [15:24] ogra_: Builds don't know about seeds at all. Do you have the URL to the build? [15:24] i have some hook scripts that apt-get install that pakcage when producing the kernel snaps ... there it is found and actually pulled in from the PPA [15:25] ogra_: I can only investigate this if you point me to a build log or something. [15:26] any of https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/xenial/ubuntu-core-system-image/ ... if you look for initramfs-tools-ubuntu-core in one of teh build logs, you will see it doesnt get installed during the normal seed installation, only later [15:26] (i dont have a working log on that page anymore since i did a lot of re-builds today ...) [15:27] if yu can find the one for 20160409 it should be a good one for comparison [15:30] the first mentioning of the package is in: [15:30] + dpkg -s initramfs-tools-ubuntu-core [15:30] + sed -n /^Version:/{s/^[^: ]*: \([^: ]*\).*/\1/;p;} [15:30] dpkg-query: package 'initramfs-tools-ubuntu-core' is not installed and no information is available [15:30] ogra_: Which PPA in particular do you mean? ppa:snappy-dev/ubuntu/image ? [15:30] (which obviously should error out when it cant find the package) [15:30] yeah [15:30] as i said, i removed the package now [15:31] ogra_: And what was the package name that you uploaded? [15:31] initramfs-tools-ubuntu-core [15:31] You could save me a lot of time by not making me play twenty questions here :P [15:31] sorry [15:31] what i uploaded is the same thing thats in the xenial queue now ... [15:32] (but with a ~ppa1 suffix) [15:32] oh right, you're relying on Task fields to get this package installed? [15:33] (can't be seeds, but could be Task fields derived from seeds) [15:33] yes, no metapackage for snappy [15:33] ah well, there you go [15:33] PPAs have never had Task fields generated for them *shrug* [15:33] nothing to do with the recent reorg [15:33] well [15:34] why does it not pick the archive version then ? [15:34] it simply drops the package [15:34] presumably because it looks at latest versions of things before it considers Task, I don't know [15:34] (and i'm pretty sure it worked before) [15:34] but it's not new [15:35] if it worked before then either apt behaved subtly differently or (more probably) the package was also pulled in by a dependency from something else that still had its Task field [15:35] all this stuff is undefined [15:35] hmm, k [15:35] if you want predictable behaviour then you'll probably have to use a metapackage [15:35] or I suppose you could hardcode the Task field just in the PPA upload [15:36] evil but would work for temporary testing [15:36] in debian/control ? [15:36] yes, in the binary stanza [15:36] might need to be XB-Task, I don't quite recall [15:37] anyway, I have to go again, good luck [15:37] k ... well, i dont think i'll use it ... the thing i wanted to test sits in the queue now ... but i'll note it down === afkthairus is now known as athairus [17:03] darkxst: FYI (from #ubuntu-mirrors) there's an apt-cacher-ng config workaround linked from my latest blog post [17:03] darkxst: https://bugs.debian.org/819852 [17:03] Debian bug 819852 in apt-cacher-ng "apt-cacher-ng: support by-hash index files" [Normal,Open] [22:50] cjwatson, thanks!