[04:05] mcphail: snapcraft's README.md refers to the Launchpad one [05:06] morning [05:34] mvo: hey, how are things? [05:45] Good morning [05:45] zyga: hey [05:46] zyga: i see we made it to 2.32.6 [05:50] hey zyga and mborzecki ! [05:50] mvo: morning [05:51] mvo: how was the sprint? [05:52] * zyga -> walk [05:52] mborzecki: it was good [05:52] mborzecki: no surprises, focus for the next couple of weeks is making core18 bootable and working on the upgrade path [05:52] mvo: any updates to the roadmap then? [05:52] mborzecki: but we knew that :) [05:53] mborzecki: nothing new in the roadmap I think [05:54] mborzecki: we will need to do a .7 to fix an important usecase (two PRs up for that already) [05:54] mborzecki: and then we are hopefully fully back to 2.33 [05:54] Oh [05:54] What is for .7? [05:54] hm and i thought we were done with .6 :) [05:54] zyga: a systemd target for when the system is fully seeded [05:54] zyga: and a way for snaps to declare an "after: snapd.seeded" [05:55] zyga: it is important for some 18.04 cloud image use-cases [05:55] mborzecki: I thought so too :) but then reality disagreeded [05:55] I think its pretty straightforward, it will need a design ACK from gustavo though [05:56] I See [05:57] mvo: a target taget? as in snapd.seeded.target? [05:57] mborzecki: correct [05:57] mborzecki: so that a unit can say "after=snapd.seeded.service" [05:59] mborzecki: there is a PR up for this (5124) but it was written during the sprint so an extra careful review is needed :) [05:59] the other one is 5132 and we will need 5133 for cosmic [06:09] PR snapd#5136 opened: tests: ubuntu 18.04 or higher does not need linux-image-extra- [06:26] PR snapd#5128 closed: Revert "Skip test lp-1721518 for arch, snapd is failing to start afte… [06:29] PR snapd#5121 closed: interfaces:minor autoconnect cleanup [06:36] PR snapd#5069 closed: configcore: validate experimental.layouts option [06:44] * zyga is back home and working now [06:44] my whole desk was covered with yellow powder [06:45] something must be blooming nearby [06:45] yuck [06:48] * zyga catches up with email [06:49] mvo: please ping me if something urgent is needed, I will start hacking at around noon, I need to get my inbox under control [06:54] zyga: I will focus on .7 this morning and trying to make sure autopkgtests and spread is happy [06:55] zyga: 5131 needs a second review but its pretty trivial and can probably go in with a single review even [06:55] zyga: I got this error in one of my PRs so it might be a prereq for landing the 2.32 ones [06:58] mvo: FYI, I think there's a real error that affects debian [06:58] mvo: but it's rare [06:58] mvo: if you see a network-bind or network test fail there, that's it [06:58] it looks like we become unconfined in some situations [06:59] PR snapd#5136 closed: tests: ubuntu 18.04 or higher does not need linux-image-extra- [07:00] mvo: btw, do you know what is cosmic? [07:00] cosmic ... c-what? [07:00] zyga: I don't [07:00] zyga: canimal? [07:00] zyga: cockroach maybe [07:01] mborzecki: no :) [07:01] or maybe canibal [07:01] haha [07:01] canibal cockroach [07:01] sounds like a great name for a death metal band ;) [07:02] lol [07:02] cosmic cosmonaut would be nice [07:02] humans are animals too [07:02] I like cosmonaut! === pstolowski|off is now known as pstolowski [07:04] mornings! [07:04] hey pstolowski good morning and welcome back [07:04] how was the sprint? [07:05] pstolowski: morning [07:05] good morning pawel [07:05] pstolowski: good, no surprises, roadmap is pretty much what we discussed already [07:12] arch package bumped to 2.32.6 [07:13] and wow, we have quite a lot of open PRs [07:14] PR snapd#5135 closed: Docs moved to forum [07:15] need coffee [07:15] PR snapd#5133 closed: spread.yaml: add cosmic (18.10) to autopkgtest/qemu [07:28] zyga: regarding #1769203, haven't we looked at auto mounting snaps via .automount unit files? [07:28] Bug #1769203: every revision of snaps are mounted

[07:29] mborzecki: I mentioned this and I think gustavo was sceptical about it [07:33] zyga: do you recall what was his reasoning? [07:37] mborzecki: we did use automount a long time ago, it may even be in the git history (a looooong time ago, very early days of the project) [07:37] mborzecki: that it is not fully transparent [07:37] mborzecki: you can observe automount points [07:37] mborzecki: you can unmount them to break them [07:38] zyga: hm, unmount them to break them is also true for .mount services, no? [07:40] PR snapd#5131 closed: tests: fix interfaces-network test for systems with partial confinement [08:06] mvo: are we pushing .6 to stable [08:25] zyga: is there any concern still about https://github.com/snapcore/snapd/pull/5075 ? [08:25] PR #5075: snap/env: fix env duplication logic [08:30] zyga: yes, today when cachio is around (probably after the standup [08:30] mvo: brilliant [08:32] zyga: is there any work to make 2.32 tested again that I can build upon? or just go over the master PRs and cherry-pick the right stuff? [08:35] btw, a review for 5132 would be great. it need a final +1 from niemeyer to ensure he is happy with the language but beside that it should be ready for review [08:42] doing [08:43] ta [08:52] mvo: done [08:56] ta [09:06] PR snapd#5137 opened: tests: cherry-pick commits to move spread to google backend [09:14] hmmmmm [09:14] mvo: I have a debian 10 vm with two snaps but no core [09:14] * zyga wonders how that happened [09:15] snap changes from debian 10 https://www.irccloud.com/pastebin/oU28h1jS/ [09:15] * mvo hugs mborzecki and zyga for the excellent review in 5131 [09:16] zyga: strange, what version of snapd is running? [09:22] 2.30 [09:23] 3.30-5+b1 [09:38] mwhudson: hey [09:38] mwhudson: not super urgent but it would be good to update snapd in debian to 2.32.6 or .7 [09:39] mwhudson: also (as before) not sure how to maintain this package [09:40] mwhudson: what is the process for doing a new release? merge upstream and massage? [09:43] we should also check if re-exec is enabled [09:44] yay, cosmic autopkgtest for ppc64/s390x is green [09:55] mborzecki: are you running arch or ubuntu now [09:55] mborzecki: I may need your help [09:55] arch [09:56] zyga: what do you need help with? [09:56] Bug #1593141 changed: Each run of a snap app leak directories in /tmp [09:57] mborzecki: there's a bug, I cannot find it now [09:57] maybe on the forum, maybe on laucnhpad [09:57] it relates to this call [09:57] PR snapd#5137 closed: tests: cherry-pick commits to move spread to google backend [09:58] https://www.irccloud.com/pastebin/Wygky688/ [09:58] someone got a denial on that [09:58] I wonder if the "none" is significant [09:58] can you see if you got that denial? [09:58] * zyga boots 18.04 live usb to test something else [10:00] wow, I didn't know there are so many snaps on the live ISO [10:00] zyga: ok, let me sync some stuff first [10:01] mborzecki: thanks [10:01] snap list truncates "tracking" too much [10:03] PR snapd#5129 closed: cmd/snap-confine: allow any base snap to provide /etc/alternatives [10:03] mvo: thank you [10:03] mvo: I was working on a new base snap last weekend [10:03] mvo: I ran into a serious bug, I didn't manage to debug it yet though [10:04] zyga: oh? ok - anything you can share? [10:04] zyga: what was the error etc? [10:04] mvo: snaps based on that snap didn't start at all, something wrong in the deep guts of snap-confine [10:04] I didn't get to the bottom of it, I can share later today if you want to play [10:07] some issues from the live ISO [10:08] failed refresh on live iso https://www.irccloud.com/pastebin/bLhbMBWf/ [10:08] failed mount on live iso (curious) https://www.irccloud.com/pastebin/h2R6g2mI/ [10:09] afaik refreshes are not working on the live image, that's why they asked us how to turn them off [10:09] ureadahead is creazy on the live iso, spamming with errors [10:09] pedronis: they work on 2nd try [10:10] what I mean, afaik they are not expected to work [10:10] I see [10:11] I don't know the details, but they didn't ask to help fix them, but how to turn off auto-refreshes [10:11] I think there's a basic space problem as well [10:15] Bug #1443612 changed: please set /run/shm/snaps/@{APP_PKGNAME}/@{APP_VERSION}/ for apps [10:18] Bug #1650389 changed: Installing snapd on 14.04.5 desktop downgrades xorg et al. <14.04> [10:21] Bug #1659924 changed: Snap failures in 16.04 [10:30] mvo: damn, that template is tricky [10:30] Bug #1458866 changed: hangs in uboot boot prompt if dtbs dir is missing [10:33] cjwatson: thanks. that's the one i've used [10:33] Bug #1464396 changed: "sudo: unable to resolve host ..." when custom hostname is used. [10:36] Bug #1471868 changed: automatic reboot fails with non executable empty systemd [10:36] Bug #1473465 changed: kvm/generic-amd64: system got in a broken state in the second boot [10:36] Bug #1650671 changed: Content sharing from snap common is broken [10:39] pedronis: can you take a look at https://github.com/snapcore/snapd/pull/4844#issuecomment-384969859 ? [10:39] PR #4844: overlord/snapstate: allow core defaults configuration via 'system' key [10:39] Bug #1641132 changed: no way to include devmode snaps in snap prepare-image? [10:39] Bug #1666074 changed: Can't update "snapweb' snap [10:39] Bug #1752031 changed: core `service.ssh.disable` key not taken into account [10:41] zyga: any particular revision you'd want to me check? [10:42] Bug #1467553 changed: automatic reboot fails with zero size kernel, no watchdog in grub [10:42] Bug #1588192 changed: GL interfaces seem wedged for Krita on nvidia

[10:42]

[10:45] Bug #1484898 changed: device tarball needs to allow setting sysctl defaults [10:46] mvo: https://play.golang.org/p/QERhAshfyMC [10:46] best i could come up with [10:46] mborzecki-ubuntu: nope [10:46] mborzecki-ubuntu: just anything [10:46] mborzecki-ubuntu: maybe try latest stable + krita [10:48] zyga: stable is 2.32.5+18.04 (distro package) [10:48] Bug #1507693 changed: please add gdb, strace, ltrace, etc to snappy-debug [10:48] Bug #1668738 changed: core snap with configure hook fails for some people [10:51] ogra_: do you know if this is still an issue in practice? https://bugs.launchpad.net/snappy/+bug/1496141 [10:51] Bug #1496141: /etc/adduser.conf should default to use extra_users [10:51] Bug #1474463 changed: a networking service snap like xkcd doesn't wait for the server to be listening [10:53] zyga, well, the tools we use all use the --extrausers option ... switching that default would allow to avoid this [10:53] zyga: yes it would, wouldn't it [10:53] so nothing has changed but if you want to close it ... [10:53] zyga: yes, merge + massage i think, we need to move it salsa too... [10:53] mwhudson: I'm sooo out of date on this process [10:53] zyga: i think there are some new deps that need to be packaged? [10:53] mwhudson: I can merge and push my tree somewhere after some basic massaging [10:53] mwhudson: I bet [10:54] zyga: go for it! [10:55] Bug #1543764 changed: snappy classic must use officially supported lxd images from simplestream; not unsupported ones from linux-containers.org [10:55] mborzecki thank you! [10:55] zyga: it's probably easier for me to do the new packages as a DD i guess [10:56] mwhudson: yeah, for sure [10:56] mwhudson: I'll merge and see what comes out, I'm very out of touch with that aspect, I'll send you a mail with the updates (or forum thread) [10:56] zyga: +1 [10:56] zyga: i don't see any denials [10:56] * mwhudson makes a thing in a trello [10:57] need to zzz now [10:57] o/ [10:57] mvo: who maintains the classic snap [10:57] mvo: and what should it do since we now have 18 looming [10:57] zyga, mvo and me ... kind of [10:57] should it be magic smart and follow the booted base? [10:57] or ... what? [10:57] we should have a classic-18 snap i guess [10:58] Bug #1551747 changed: ubuntu-fan causes issues during network configuration

Released by apw>

[10:58] ogra_: but you cannot install arbitrary classic on arbitrary system, can you? [10:58] if it is suppsed do that magically we'd have to ship both chroots [10:58] (and have it pik the right one at runtime) [10:59] s/ship both/ship support for both/ [10:59] i suspect 18 looks a litte different to 16 :) [11:04] zyga: tried with "none" and NULL, no denials in either case [11:11] Thanks [11:14] hm, snapcraft still doesn't support license? [11:14] kalikiana: is this fixed or am I using it wrongly? [11:16] popey: offtopic [11:16] popey: I have https://github.com/snaphub [11:16] feels like something you should have instead [11:22] PR core#87 closed: snapcraft.yaml: update stage-packages during build [11:24] I updated 5132 based on the excellent feedback I got [11:27] looking [11:29] zyga: looks empty? [11:30] What is it (intended to be)? [11:35] I think it is very old [11:35] popey: it was supposed to be a place for snaps under shared maintenance [11:35] like snapcrafters today [11:38] mvo, I'm looking forward to seeing a dnf plugin for snapd from you :P [11:38] mborzecki: I'm not sure merging them would be useful, I would suspect people putting a copy or more stuff under system [11:40] pedronis: hm sounds fair, very well, i'll push a change replacing error with the log [11:46] Zyga feel free to delete if it's not used [11:48] popey: ack [11:57] mvo: zyga: any PR I should prioritize vs starting to look at the full queue? [11:58] pedronis: maybe pass through all simple first [11:58] to allow them to land [11:58] mborzecki, mvo: I will fix master as arch moved to a more restrictive compiler now [12:01] PR snapd#5138 opened: cmd/libsnap: fix compile error on more restrictive gcc [12:01] https://github.com/snapcore/snapd/pull/5138 [12:01] PR #5138: cmd/libsnap: fix compile error on more restrictive gcc [12:01] * zyga -> walk dog [12:08] mborzecki, hey, build fail on arch [12:08] failing [12:08] here https://api.travis-ci.org/v3/job/375775038/log.txt [12:08] cachio_: yes, zyga is working on it [12:08] mborzecki, great, thanks [12:08] i've updated to gcc 8.1 just now :) [12:09] Fixed above :-) [12:09] off to pick up my daughter [12:09] pedronis: pushed an update to #4844 [12:09] PR #4844: overlord/snapstate: allow core defaults configuration via 'system' key [12:12] PR snapd#5130 closed: interfaces/apparmor: allow bash and dash to be in /usr/bin/ === pstolowski is now known as pstolowski|lunch [12:20] zyga, you waited until Arch had it too until you'd fix the problem I've had in Fedora 28 since February? [12:20] I'm hurt :( [12:20] That is the problem? [12:21] It broke master :/) [12:21] yes [12:21] Hey jdstrand [12:21] you promised me you'd fix it back then, and you didn't :( [12:22] and you guys keep disabling the Fedora CI instead of fixing it, so I have nothing left to make you guys notice these things [12:23] Hmmm? We only disabled it when golang was broken [12:23] zyga, I've been forced to modify the autofoo to drop -Werror since February so that snapd releases would even build [12:24] I lost track of that [12:24] :'( [12:24] We should have f28 in CI [12:24] you should have Rawhide in CI too [12:24] I distinctly remember help setting that up at the last sprint [12:31] Son_Goku: heh, will look into that once the fires are out [12:32] pedronis: the 2.32 ones would be great [12:32] yay, cosmic autopkgtests are all green now [12:32] Son_Goku: so to recap (I was afk) [12:32] Son_Goku: it would be great if we had the error in CI so that it gets fixed instantly and not forgotten [12:33] Son_Goku: I'll ask if we can move to F28 testing [12:33] zyga: aha, the arch failure(s) I see right now are known and you take care of them [12:33] ? [12:33] mvo: correct [12:33] Son_Goku: I encourage you to nag us more if there's a known issue [12:33] I gave up after three weeks [12:33] zyga: ta, approved [12:34] that's why I keep doing this: https://src.fedoraproject.org/rpms/snapd/blob/master/f/snapd-2.32.4-Drop-Werror.patch [12:34] that started in 2.31.1 [12:34] Son_Goku: also the patch is pretty trivial, we could have merged it long ago [12:34] sorry for dropping it [12:34] yes, but you didn't want to [12:34] you said you'd rather fix the bugs [12:35] I mean the patch for fixing this, not dropping -Werror [12:35] cachio_: do we have a f28 image? [12:35] and of course, I keep this little bugger alive: https://src.fedoraproject.org/rpms/snapd/blob/master/f/0001-cmd-use-libtool-for-the-internal-library.patch :) [12:35] I would swap f26 for f28 (26 is about to EOL) [12:37]