[02:26] Issue core18#128 opened: Having issues installing core18 [03:03] Issue core18#128 closed: Having issues installing core18 [05:23] Hello [05:23] morning [05:50] hey mborzecki [05:50] * zyga tries to wrap up propagation bug [05:50] zyga: hey [05:50] tests now pass, I'm still adding some more checks though [05:51] and I have a lot of unit tests to adjust for extra calls [05:53] zyga: i poked a bit around /etc/nsswitch.conf on fedora, vanilla 29 image starts with a symlink, but after dnf upgrade /etc/nsswitch.conf becomes a file, probably some scriptlet broke it [05:53] mborzecki: ah, I wonder if it's a design or bug [05:54] mborzecki: perhaps worth reporting a bug to ask the question [05:59] zyga: looks like a bug/issue about authselect which is not installed in cloud images, so never gets to run and update nsswitch.conf to be a symlink [06:00] zyga: found this: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/G564T2VXADHS557BOT5L5K42SSF4QSIE/ and https://bugzilla.redhat.com/show_bug.cgi?id=1622272#c8 [06:00] looks like some monkey business to me [06:00] yeah [06:02] https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/PNKKVG3K6WAU42CCPVIEV6LZY7PWUG4P/ [06:04] meh, installed authselect, ran authselect select --force sssd and it errors out :/ [06:05] nvm, looks like this is fixed in f30, f29 will eol in a couple of months anyway [06:15] yeah, at least that's good [06:15] thanks for chasing this [06:15] * zyga writes docs for test tool helper [06:27] one more tweak for selinux policy for f30 [06:27] guess i'll never stop finding search on dir class confusing :/ [06:33] mborzecki: you should do talks on selinux [06:33] zyga: i'm far from competent in this [06:53] breakfast === pstolowski|afk is now known as pstolowski [06:59] mornings [07:01] hey pawel [07:01] PR snapd#6881 opened: data/selinux: allow snap-confine to do search on snappy_var_t directories [07:02] pstolowski: hey [07:11] mborzecki: last bit in this tool is renumbering for deterministic test [07:11] mborzecki: some heavy lifting from that old tool [08:00] ok, I think the tool is ready now [08:09] mborzecki: mountinfo-tool https://www.irccloud.com/pastebin/N5yQNZPX/ [08:11] PR snapd#6803 closed: daemon, o/snapstate, store: support for installing from cohorts [08:12] PR snapd#6703 closed: tests: add deferred actions <⛔ Blocked> [08:12] bongiorno, principesse e principi! [08:13] * Chipaca wonders if italian has "inclusive language" for plurals [08:16] hey Chipaca [08:16] mvo: 'sup [08:18] mborzecki: #6874 was merged but at least jdstrand should have looked at it [08:18] PR #6874: cmd/snap-confine: do not mount over non files/directories [08:19] mvo: ^ [08:19] pedronis: looks like cachio merged it yesterday, i'll ping jdstrand to have a look still [08:19] pedronis: good morning! I was looking over the spike trello cards this morning [08:19] pedronis: yeah, we need to ask jamie to double check it [08:19] pedronis: and tried to flesh things out a little bit [08:20] mvo: maybe we need to tell cachio not to merge snap-confine/snap-update-ns stuff [08:20] pedronis: while doing so I was stumbling on snap-verify and poked around a bit, it seems like we need something like assertstate.Batch just without the ties to the state, does that sound right? or do you have something else in mind for that? [08:20] pedronis: yeah, I think that is a good idea [08:21] my bad, should have added a blocked label or sth [08:21] it's a judgement call, just don't think he is worth having everybody be able to do that [08:23] mvo: yes, Batch could be useful, it depends what the api is though [08:25] pedronis: yeah, it requires refactor because assertstate only exposes a RODatabase so we have to have a custom save function at least. anyway, I can poke around a bit but if you have a firm plan already I can also leave it alone, it seems nice and tracktable though :) [08:25] mvo: please leave it alone, I don't think is a good use of your time, and a script that does exit 0 should unblock you [08:26] pedronis: heh, I can do that [08:26] mvo: yes, Batch is in assertstate, I don't think we need that, a Fetcher is enough [08:27] but anyway [08:27] ok [08:28] mvo: I think we need to understand what we should verify, before writing the tool [08:28] with relates more to the rest of the initramfs [08:28] s/with/which/ [08:29] * mvo nods [08:31] mvo: my biggest worry is that we will need to verify the kernel also on the run path [08:31] I mean worry related to the tool [08:32] pedronis: verify the running kernel? [08:32] maybe [08:32] maybe not, we need to think [08:33] pedronis: hm, interessting. you need to tell me more when we have high bandwidth time in a HO [08:50] * zyga -> quick coffee [08:54] PR snapd#6882 opened: cmd: add snap-verify stub binary (UC20) [09:01] pedronis: where is this meeting? [09:01] Chipaca: added a HO now [09:35] pedronis: https://forum.snapcraft.io/t/downloading-snaps-via-snapd/11449 [09:35] pedronis: maybe point igor at that (dunno his nick) [09:50] mvo: reviewed 6882 [09:51] ta [10:14] ok, *really* time for that coffee [10:45] pstolowski: can you take a quick look at https://github.com/snapcore/snapd/pull/6881 ? [10:45] PR #6881: data/selinux: allow snap-confine to do search on snappy_var_t directories [11:00] PR snapd#6881 closed: data/selinux: allow snap-confine to do search on snappy_var_t directories [11:01] pstolowski: thanks! [11:01] np [11:15] pstolowski, hey, #6859 is ready I think [11:15] PR #6859: tests: new hotplug test executed on ubuntu core [11:16] pstolowski, hold on, I just read the comments you did [11:16] cachio: great! does it pass on both cores? [11:16] * zyga runs spread for propagation test! [11:16] (so so so close now) [11:16] pstolowski, yes [11:17] cachio: awesome! thank you [11:17] pstolowski, for a following PR I'll reuse the tests parts [11:17] pstolowski, but I would keep it working first so then I can work on sharing test code [11:17] cachio: yes, sounds good, let's land it and do a followup [11:17] between tets [11:18] pstolowski, nice [11:18] cachio: i'll take a look today [11:26] hi, is there any way to have snapcraft pull in a deb from a specific URL as stage-packages ? [11:28] (or have it pulled from a repo) [11:29] pedronis: my changes to the 'snap info' will probably conflict with the pr that depends on it (that i'd made to depend to avoid conflicts.... augh) but i think it's a lot nicer now [11:43] morning, question about 2.39, when is it out? [11:53] sergiusens: should go out today , cc mvo [12:02] pedronis, sergiusens correct [12:04] * zyga afk for errand [12:08] Chipaca: #6816 needs master merged, and has some conflicts right now [12:08] PR #6816: daemon, overlord: support for cohort-key in refresh and switch [12:16] pedronis: will do [12:16] * Chipaca finishing lunch [12:16] thx [12:38] re [12:41] not sure, don't know what they scheduled over the standup === ricab is now known as ricab|lunch [12:51] mborzecki: snapd tests now pass :-) [12:51] mborzecki: also in spread [12:51] I will run them on more systems and start to trim crap out of that branch for proposal [12:57] Issue core18#129 opened: Multiarch isn't working [12:59] Chipaca: did a pass over #6564 [12:59] PR #6564: cmd/snap, tests: refactor info to unify handling of 'direct' snaps [13:00] pedronis: thx [13:31] is the standup over? [13:31] zyga: yes [13:31] ok [13:35] Issue # closed: core18#56, core18#86, core18#89, core18#117, core18#129 [13:35] PR # closed: core18#43, core18#72, core18#90, core18#98, core18#122, core18#126, core18#127 [13:36] Issue # opened: core18#56, core18#86, core18#89, core18#117, core18#129 [13:36] PR # opened: core18#43, core18#72, core18#90, core18#98, core18#122, core18#126, core18#127 [13:37] PR snapd#6867 closed: gadget: offset-write: fix validation, calculate absolute position [13:38] * zyga lunch break [13:38] starving! [13:39] mborzecki: lol, the branch has fixed one known test that just shows how layouts are broken in one case :) [13:45] mborzecki, m [13:45] https://travis-ci.org/snapcore/snapd/jobs/535240901 [13:45] thanks for the help on that one [13:46] cachio: yay! [13:46] cachio: land it while it's green :) [14:02] mborzecki, done [14:02] PR snapd#6860 closed: tests: running tests on fedora 30 === harrisj_ is now known as harrisj === ^arcade_droid is now known as zarcade_droid === ricab_ is now known as ricab [14:36] kenvandine: I updated bug 1825883 - its in 2.39 [14:36] Bug #1825883: stale copy of plug and slot attributes is kept in connection state [14:37] mvo: great, thanks! [14:49] Chipaca: oops, I was almost forgetting, this is my PR that changes the signatures of the users methods: https://github.com/snapcore/snapd/pull/6834 [14:49] PR #6834: daemon: pass the model to the create known user helpers (instead of full Overlord) [14:49] Chipaca: as I said maybe it's easier if you just incorporate that change in the refactor [14:49] your are doing [14:50] looks like it :-| [14:50] need to wrap up the info one :) [14:51] PR snapcraft#2568 opened: docs: consolidate on a simple HACKING.md [14:52] Chipaca: no hurry, it's orthogonal to my other open PRs, I'm not blocked on it (I think) [14:54] snap-confine has elevated permissions and is not confined but should be. Refusing to continue to avoid permission escalation attacks [14:54] hmmm === alan_g is now known as alan_g_ [14:59] mvo: found a small bug in core18 [14:59] mvo: there's no profile for /usr/lib/snapd/snap-confine [14:59] look at this plese [14:59] *please [14:59] https://www.irccloud.com/pastebin/MnHLkuvk/ [15:00] I think we don't notice because spread tests run as root [15:00] zyga: ok, after the meeting (we have now) [15:00] oh [15:00] joining [15:01] that should be written by snapd, no? [15:01] it shouldn't be in the core18 itself === jdstrand_ is now known as jdstrand [15:20] PR snapd#6859 closed: tests: new hotplug test executed on ubuntu core [15:22] * cachio lunch [15:24] cachio: i've merged your nested tests PR [15:25] PR # closed: snapd#5644, snapd#5822, snapd#5915, snapd#6108, snapd#6258, snapd#6325, snapd#6327, snapd#6341, snapd#6347, snapd#6360, snapd#6367, snapd#6404, snapd#6436, snapd#6541, snapd#6564, snapd#6588, snapd#6648, snapd#6666, snapd#6680, snapd#6681, snapd#6691, snapd#6695, snapd#6697, [15:25] snapd#6705, snapd#6708, snapd#6714, snapd#6721, snapd#6734, snapd#6750, snapd#6759, snapd#6760, snapd#6767, snapd#6804, snapd#6805, snapd#6816, snapd#6825, snapd#6834, snapd#6835, [15:25] snapd#6836, snapd#6838, snapd#6839, snapd#6841, snapd#6848, snapd#6855, snapd#6870, snapd#6871, snapd#6875, snapd#6876, snapd#6878, snapd#6879, snapd#6880, snapd#6882 [15:26] PR # opened: snapd#5644, snapd#5822, snapd#5915, snapd#6108, snapd#6258, snapd#6325, snapd#6327, snapd#6341, snapd#6347, snapd#6360, snapd#6367, snapd#6404, snapd#6436, snapd#6541, snapd#6564, snapd#6588, snapd#6648, snapd#6666, snapd#6680, snapd#6681, snapd#6691, snapd#6695, snapd#6697, [15:26] snapd#6705, snapd#6708, snapd#6714, snapd#6721, snapd#6734, snapd#6750, snapd#6759, snapd#6760, snapd#6767, snapd#6804, snapd#6805, snapd#6816, snapd#6825, snapd#6834, snapd#6835, [15:26] snapd#6836, snapd#6838, snapd#6839, snapd#6841, snapd#6848, snapd#6855, snapd#6870, snapd#6871, snapd#6875, snapd#6876, snapd#6878, snapd#6879, snapd#6880, snapd#6882 === cjwatson_ is now known as cjwatson [15:33] pedronis_: your suggestion re RemoveFlags & flags.go on https://github.com/snapcore/snapd/pull/6870 was to simply move RemoveFlags there (or to extend Flags and not introduce RemoveFlags)? [15:33] PR #6870: cmd/snap, api, snapstate: implement "snap remove --purge" [15:36] degville: hey, any thoughts on https://github.com/snapcore/snapd/pull/6870#discussion_r285659028 ? [15:36] PR #6870: cmd/snap, api, snapstate: implement "snap remove --purge" [15:38] pstolowski: to move it there, but as I said I would like Chipaca's input on some of my remarks [15:43] ack [15:54] Chipaca: I asked your 2nd opinion on a couple of things in #6870 from pstolowski [15:54] PR #6870: cmd/snap, api, snapstate: implement "snap remove --purge" [15:54] * Chipaca looks [15:55] pstolowski: looking... [15:58] Chipaca: made a small comment about the building of infoWriter discussion in 6564 [16:02] hah, degville just snuck in as i was writing it :-D [16:02] degville: wdyt of the 'save' comment (now sitting below yours)? [16:03] Chipaca: It's a really good point. I think you're right. [16:04] (about saving. I like the idea of using consistent verbs too). [16:04] Chipaca, degville thanks for comments; Chipaca, are you suggesting to simply reuse Flags (rather than just move RemoveFlags?) [16:06] pstolowski: yes [16:07] k [16:07] pstolowski: dunno if pedronis agrees tho :-) [16:08] Chipaca: pstolowski: I don't [16:08] what other flags do they share? [16:10] if it needed to go into SnapSetup there would be a point, but that's not the case afaiu [16:10] none. Flags seems to install related [16:10] instlal/refresh/etc [16:11] I don't mind if it's separate :) [16:11] (there is ForSnapSetup that zeros out unwanted flags for snap setup fwiw) [16:11] I know [16:11] though is often forgotten [16:11] yes :-| [16:11] pstolowski: my feeling right now is to keep it but move close to the function [16:11] right now is miles away [16:12] afair [16:13] pedronis: done [16:15] pstolowski: thx === ErichEickmeyer is now known as Eickmeyer [16:19] hah, finding bugs by writing tests [16:19] a novel endeavour [16:19] Chipaca: where? :) [16:19] pedronis: maybePrintStandaloneVersion [16:19] when no version in the snap file [16:20] would produce bad yaml [16:20] ah [16:20] not an easy path to hit as a versionless snap is invalid [16:20] but, hey :) [16:20] :) [16:20] robustness in depth [16:20] or something [16:31] pedronis, cmatsuoka, zyga I updated the gadget PR with most^Wsome^Wmost(?) things we discussed [16:31] mvo: ack, thank you! === pstolowski is now known as pstolowski|afk [17:15] pedronis: so many unit tests it's almost embarrassing [17:15] * Chipaca EODs [17:16] mvo: silly and low priority reminder about sprint approvals (just so they don't pile up on you :-p) [17:16] * Chipaca again pretends to EOD [17:27] Issue core18#129 closed: Multiarch isn't working [17:46] PR snapd#6883 opened: tests: fix how strings are matched on auto-refresh-retry test [18:36] * zyga is back home :-) [18:39] tomorrow morning will be busy :) [19:11] PR snapd#6870 closed: cmd/snap, api, snapstate: implement "snap remove --purge" [21:11] * cachio afk [21:46] PR snapcraft#2569 opened: lifecycle: warn about bases