[05:58] doko, I think I have fixed them, but a double check would be awesome [05:58] kazoo python-tenacity and psycopg2 [07:05] It seems one of the hosts hosting archive.ubuntu.com is outdated, in 4 runs with xenial-proposed, I only got apt to upgrade itself to the proposed version once. [07:36] juliank: btw curl's --resolve is the easiest way to track down which servers have which contents [07:37] sarnold: nice, letme check [07:39] hmm [07:39] I see it everywhere with curl [07:40] hrm, I wonder if they got in sync inthe meantime or .. [07:40] oh well, that's for someone more awake than me :) [07:52] for IP in `dig -t A archive.ubuntu.com +short` `dig -t AAAA archive.ubuntu.com +short`; do echo "$IP: $(curl --silent --resolve archive.ubuntu.com:80:$IP http://archive.ubuntu.com/ubuntu/dists/xenial-proposed/Release | grep ^Date: | cut -d ' ' -f2-)" ; done [07:52] looks identical to me [07:53] I still get some runs where apt install apt does not find it [07:53] maybe I'm hitting a bug in apt somewhere [07:54] Yup, it seems to be a bug in apt [07:56] 'grep ^Package.*apt /var/lib/apt/lists/archive.ubuntu.com*Packages' [07:56] Package: apt [07:56] succeeds [07:56] but apt-cache shows no apt version in a repository [07:58] In fact, the cache only contains the status file [08:00] might be a race with some cloud stuff when initializing the container [08:02] simple test case: https://paste.ubuntu.com/p/4Zzw6mnhKg/ [08:02] just launch a new container, wait 5 seconds [08:02] replaces sources.list, run update [08:03] => you only see dpkg status file [08:20] sarnold: tomreyn so the problem I was seeing is that cloud-init rewrites the sources.list file after I rewrote it, and thus my newly added proposed repo goes missing [08:20] I guess I have to modify my scripts to wait for cloud-init to settle first [11:12] juliank: why dont you place yours in /etc/apt/sources.list.d/ instead? [11:25] tomreyn: mostly don't want to download all the other stuff in the default sources.list [11:25] performance! [11:25] could -o dir::etc::sourcelist=/dev/null, though [13:22] xnox: do you mind merging https://code.launchpad.net/~didrocks/apport/whoopsie-auto-ui/+merge/348479? We would like to upload to cosmic today and seb128 approved it but can't merge (nor can I) [13:27] or maybe juliank, I saw you touched apport recently :) [13:28] juliank: so, if I take the bzr branch (lp:apport), all the snap part isn't in it, and the bzr-ubuntu branch is outdated versus the repo [13:28] I'm happy to apply a debdiff, but it seems we don't have any source of truth anymore? [13:28] didrocks: apport is complicated [13:30] didrocks: the ubuntu branch seems up to date? https://code.launchpad.net/~ubuntu-core-dev/ubuntu/cosmic/apport/ubuntu [13:30] not sure what you were looking at [13:30] juliank: the cosmic one is, not binoic [13:30] bionic* [13:30] juliank: but what do you generally do, merge lp:apport into this one? [13:31] and so, you have the snap stuff into /ubuntu but not upstream apport? [13:31] didrocks: I leave it to bdmurray to handle that [13:31] but yes, the snap stuff is in ubuntu only [13:32] didrocks: sometimes people don't push bzr branches for stable stuff [13:32] juliank: yeah, it's kind of annoying as we have inline deltas with upstream [13:32] Yeah [13:32] ok, I'll upload a debdiff then for now (at least cosmic) [13:33] the branch is up for upstream to merge [13:33] and I'm happy to have a bionic/ branch ready if up to date [13:33] maybe cyphermox and bdmurray have some branches locally for their uploads [13:34] (apport 2.20.9-0ubuntu{7,7.1,7.2} for bionic) [13:34] Otherwise, I'd say import the dscs [13:34] :D [13:34] yeah, also those are native packages, which isn't expected from the version, as told the other day [13:34] (and no release :p) [13:34] yes [13:34] but yeah, let me create a debdiff, that way, I'm sure to not revert your work ;) [13:34] they were not initially [13:35] yeah, and we didn't have distro-patches [13:35] and .pot was updated ;) [13:35] (not only when building the package) [13:35] anyway, not that important for now, but I'm happy to help fixing anything that needs to be fixed in case [15:04] did didrocks quit? [15:15] yes, he called it a week [15:16] he had the apport changes uploaded which was a good point to stop working [15:16] :-) [18:18] juliank: ugh :/ thanks for the feedback === imcleod_ is now known as imcleod [21:12] cjwatson: BTW, thanks for putting some time in on new SSH key types. [21:13] incremental progress :) [21:16] Yes, I've been following upstream get stuck on it, out of your court and you're still prepping for when they do add support. [21:34] ECDSA is doable now, though I'm not especially wild about that format. [21:37] Unfortunately I'm waiting for Ed25519. [21:37] Yeah, I use that most other places now. [21:37] (At least merely offering it to Launchpad no longer causes an SSH hang.) [21:37] Huh, cool. [22:35] infinity: would you be interested in reviewing the xubuntu core/base patches? https://code.launchpad.net/~xubuntu-dev/debian-cd/xubuntu-base/+merge/347414 - https://code.launchpad.net/~xubuntu-dev/livecd-rootfs/xubuntu-base/+merge/347319 - https://code.launchpad.net/~xubuntu-dev/ubuntu-cdimage/xubuntu-base/+merge/347320 (or do you know who might be? :)) [22:37] bluesabre: TBH, I've been dragging my feet a bit because there's been some discussion that we really should do stacked livefses in ubiquity, which would allow you to have both "flavours" in one ISO. [22:38] bluesabre: Of course, if the only real goal here is a smaller download, then you really do want two different ISOs. [22:42] infinity: I think the team is interested in the dual-iso solution. We've been spinning the Core iso for each release since the 2015 announcement and getting testing on it. One of the reasons it's promoted is the smaller size, for those with limited bandwidth or CD-drive requirements [22:43] bluesabre: Kay, fair enough. Icky, but understandable. [22:43] :)