=== setuid is now known as zZZzzZZetuid [05:09] morning [05:14] mvo: morning :) [05:14] hey mborzecki [05:15] mborzecki: how is life this morning? [05:15] mvo: cold :P 5C outside [05:15] mborzecki: woah, thats cold indeed. we have ~15 here but it will change! [05:17] mvo: how was your train ride? [05:20] mborzecki: all went well, all trains on time [05:20] mborzecki: which is not the default anymore sadly. so it was a pleasant surprise [05:21] mborzecki: also it was the last train so any major delay would have left me stranded in the middle of nowhere :) so I was a little bit nervous [05:22] mvo: nice :) our trip to cdg from the hotel and the flight were smooth too [05:22] mvo: although saturday was apparently rather chaotic in paris [05:24] mborzecki: yeah, I read in the news there was a lot of protesting going on [05:30] mvo: our snapd snap has an 'unset' license in the store https://snapcraft.io/snapd [05:36] PR snapd#7493 opened: snapcraft: set license to GPL-3.0 [05:51] mborzecki: uh, lets me try to fix the license [05:52] mvo: opened a PR with an update in snapcraft.yaml ^^^ [05:52] mborzecki: nice! [05:52] mvo: but maybe the store page needs an edit too [05:52] mborzecki: I have a look now [05:54] mborzecki: I updated the store page now [05:54] mvo: great, thanks! [05:55] mborzecki: not visible in snap info (yet?) though - so maybe the authoritative one is the snapcraft [06:17] hey guys [06:17] eating breakfast [06:17] zyga: hey [06:17] did usual school run [06:17] so need a moment to get to office shape [06:18] mvo: https://paste.ubuntu.com/p/4tjyHvcn2S/ looks like it's still unset [06:18] hey zyga [06:18] mborzecki: yeah, confusing [06:18] everything ok? [06:18] any fires? [06:19] zyga: some confusion about the license of snapd snap, which apaprently is unset, opened #7493 to address that [06:19] PR #7493: snapcraft: set license to GPL-3.0 [06:19] :D [06:20] fwiw, do the review tools complain if license is not specified in the snap? [06:23] mborzecki: I think now [06:23] not [06:52] why is license only as a passthrough? [06:52] zyga: idk [06:53] how do you guys feel today? [06:53] and it's red :/ [06:54] yeah :C [06:54] what was the failure? [06:55] zyga: centos, systemctl daemon-reload [06:58] zyga: feeling a bit sleepy [06:59] mvo: same here [06:59] mvo: I got up at 6 AM [06:59] but cannot stop going to bed before midnigth [06:59] er [06:59] after [07:00] https://github.com/snapcore/snapd/pull/7488 is green [07:00] PR #7488: docs: Update README.md [07:01] mvo: did you guys see https://github.com/snapcore/snapd/pull/7488? [07:02] PR snapd#7409 closed: interfaces/wayland: allow a confined server running in a user session to work with Qt, GTK3 & SDL2 clients [07:03] mvo: policy idea: prioritize contributions from outside of the team above new features from the team [07:03] mvo: prioritize fixes over new features [07:03] zyga: yes, I thnk that is a good idea [07:04] it would improve our interactions with other teams [07:04] and could eventually outpace our own contributions [07:09] https://github.com/snapcore/snapd/pull/7490 this one needs a review [07:09] PR #7490: interfaces/app-launch: support confined snaps launching other snaps [07:16] does anyone know whats up with centos-7 ? I see a lot of errors from systemctl reload [07:17] mvo: we tried to investigate this a bit last week, so far i could not reproduce this running the whole suite [07:21] mborzecki: strange - thanks! [07:24] hmmm https://github.com/poettering/systemd/commit/1f37d342f3a32da5ce52123a5afb3498800097e9 [07:25] mborzecki: this guy again ;) [07:25] mborzecki: I'm doing one now [07:26] mvo: note the super nice github issue integration [07:27] one line in a commit is enough to cross-reference everything [07:27] https://bugzilla.redhat.com/show_bug.cgi?id=1378449 [07:30] hah, got a shell :P though it failed for different reason, but i can see systemctl daemon-reload times out [07:32] https://paste.ubuntu.com/p/MtVJ7j76gc/ [07:32] zyga: mvo: ^^ [07:34] mborzecki: do you know how long it had to run? [07:34] is it a busy vm? slow generator (or deadlocked generator?) [07:34] or something else? [07:34] mborzecki: nice! but does that mean we need the backported fix from systemd? [07:34] mborzecki: also, why does it start to fail now, it used to work, no? [07:34] hmmm systemctl times out all the time [07:37] i see systemd got updated during our test run, there's a yum transaction that pulled in 219-67.el7_7.1.x86_64 [07:38] the command is: yum y install golang >= 1.9 systemd autoconf automake libtool gcc gettext gnupg pkgconfig(glib-2.0) pkgconfig(libcap) pkgconfig(libseccomp) pkgconfig(libselinux) pkgconfig(libudev) pkgconfig(systemd) pkgconfig(udev) xfsprogs-devel glibc-static valgrind /usr/bin/rst2man selinux-policy selinux-policy-devel [07:38] so it's part of our test setup [07:47] mborzecki: reviewed https://github.com/snapcore/snapd/pull/7490#pullrequestreview-291590702 [07:47] PR #7490: interfaces/app-launch: support confined snaps launching other snaps [07:47] brb, shower [07:47] mvo: do you mind if we cancel the 1-2-1 today? [07:48] I'm super sleepy and my plan for the day is to go through reviews and bug triage [07:48] we can do one later during the week if you feel like we need one [07:51] mborzecki: can you review https://github.com/snapcore/snapd/pull/7484 later today? [07:51] mborzecki: I don't want to distract you from centos issue [07:51] PR #7484: osutil: generalize SyncDir with FileState interface [07:51] zyga: ok, will do [07:51] thanks [07:53] PR snapd#7479 closed: tests: skip centos when running interfaces-openvswitch for nigthly suite [07:56] Sep 23 07:22:49 sep230656-524345 systemd[1]: Caught , dumped core as pid 22180. [07:56] Sep 23 07:22:49 sep230656-524345 systemd[1]: Freezing execution. [07:56] Sep 23 07:23:06 sep230656-524345 sshd[19296]: Connection reset by 222.186.175.147 port 7036 [preauth] [07:56] Sep 23 07:23:14 sep230656-524345 dbus[475]: [system] Failed to activate service 'org.freedesktop.systemd1': timed out [07:56] mvo: zyga ^ [07:57] ooo [07:57] abort? [07:57] mhm [07:57] grep systemd [07:57] no coredumps listed by coredumpctl [07:57] bugs are fun [07:58] zyga: the log https://paste.ubuntu.com/p/ynVRcCTBPN/ it goes south after ABRT [07:59] I wonder what "freezing execution" is [07:59] I mean, systemd just aborted, what does it freeze [08:00] duh, wonder where the coredumps are [08:00] normally they'd end up in the journal, but perhaps we cleaned up the journal? [08:05] zyga: I think it freezes because pid-1 canont die or the kernel is very unhappy [08:05] hmmm [08:05] mborzecki: is that really what happened? this is a log from the journal after all [08:07] zyga: that's all there is [08:10] zyga: wonder whether there's some 7.7 related problem, it only started a coulpe of days ago [08:10] zyga: and yum install pulls in a partial update from 7.7 actually [08:21] moin moin [08:24] mborzecki: https://github.com/snapcore/snapd/pull/7475#pullrequestreview-291621672 needs fixing [08:24] hello chipaca [08:24] PR #7475: sandbox/cgroup, usersession/userd: move cgroup related helper to a dedicated package [08:24] now small break for warm tea, I'm cold today :| [08:30] zyga, mvo, if you were me, and wanted to give a golang-related talk, what would you talk about? [08:32] Chipaca: depends a bit on the audience - if its more beginners the export_test.go / package / package_test pattern is a nice one maybe ? I find it super useful but it seems to be not *that* common knowledge [08:32] is that called black-box testing? [08:32] white-box testing? [08:33] * Chipaca calls it orange-box testing to be onbrand [08:33] Chipaca: uh, thats a good question - I would go with black-box [08:33] Chipaca: *but* I'm sure there are people with different opinions [08:33] Chipaca: I would call it "api-surface-based-testing" or something [08:34] :-) [09:02] Chipaca: the language that defied the trend, retrospective from medium sized codebase written in go [09:05] zyga: what trend? :-) [09:06] Go went back to a simpler language [09:06] Language complexity seems to grow almost uniformly otherwise [09:11] hmm, i'll leave that one to you :-) [09:12] zyga: the directory still thinks you're in palamós [09:12] zyga: what city are you in / close to? [09:13] er, maybe i should ask in private :-p [09:13] * mborzecki going through systemd core dumps [09:13] mborzecki: weren't you taking today off? [09:13] Chipaca: nope, i'm swapping on 3.10 [09:14] mborzecki: what happens on 3.10? [09:14] Chipaca: /me gets a day off :P [09:14] mborzecki: :-) [09:18] zyga: mborzecki: pstolowski: there's three GDGs in Poland AFAICT, go talk about golang / spread / snapd / etc [09:18] wait, no, just two [09:19] GDGs? [09:19] silly map confused me :-) (but kaliningrad might be fun to visit) [09:19] zyga: https://devfest.withgoogle.com/ [09:20] iirc they loosened some visa rules if you travel to kaliningrad from PL [09:21] zyga: mvo: so ABRT comes from glibc apparently, let me post the backtrace [09:21] sadly there isn't one in Córdoba for Sergio to go to, otherwise he'd kill it [09:22] zyga: there's a GDG in wroclaw on 15.11 and one in warsaw in dec [09:22] yeah, I was looking at the one in Dec [09:28] zyga: mvo: https://bugs.centos.org/view.php?id=16441 [09:29] it'd be great to try with a 7.7 image or just clean upgrade 7.6 to 7.7 [09:32] mborzecki: memory corruption? [09:32] mhm [09:32] it seems that _int_free is the entry point to glibc [09:32] and ealier it was cg_enumerate_subgroups [09:32] well :/ [09:33] mborzecki: btw, what is that bug tracker [09:33] it's not bugzilla [09:33] is it? [09:33] I like the fields it offers [09:33] mantis? [09:33] oh, is that mantis? [09:33] wow [09:33] that still exists [09:34] it was my first ever bug tracker like 20 years ago [09:39] ok, putting that core dump aside, back to some useful work [09:42] PR snapd#7493 closed: snapcraft: set license to GPL-3.0 [10:00] who to ping for auto-connection requests in the absence of roadmr ? [10:01] context: https://forum.snapcraft.io/t/auto-connection-request-for-deskconnd/13364 -- I am the upstream of both the provider and the peer snap. [10:04] mborzecki: the centos question sounds like something for sergio? [10:05] mvo: yes, we're apparently using the official centos images in gcp, still waiting for 7.7 to be uploaded there [10:05] mvo: discussed taht with sergio in paris, but will ping him when he's back online [10:07] om26er: hmm, good question, I don't know === zZZzzZZetuid is now known as setuid [10:30] * Chipaca takes a break [10:31] om26er: doesn't that fall under the usual process for auto-connections? [10:31] or, actually, isnt' that automatic [10:31] * Chipaca should check [10:31] * Chipaca needs to take a break [10:40] break for tea and more reviews [10:51] @chipaca while I am the upstream for both snaps, they come from different organizations, hence it didnt work automatically === JanC_ is now known as JanC [11:07] zyga: updated #7475 [11:07] PR #7475: sandbox/cgroup, usersession/userd: move cgroup related helper to a dedicated package [11:08] ok, on to the review of 7484 [11:11] thank you, I'll look soon [11:11] mvo: thanks for pushing the spread test to https://github.com/snapcore/spread/pull/85 ! [11:12] PR spread#85: spread/client: workaround bash 4.3 subshell errexit issues [11:15] mborzecki: my pleasure - was looking at bit a spread this morning, could fully test locally though, so I hope it works ok [11:28] mborzecki: +1 [12:16] hm, travis having a bad hair^Wnetwork day? [12:24] morning folks, perhaps Travis should also get a swap day too :-/ [12:28] haha [12:28] I was thinking that just a second ago :D [12:28] looking at why the branches I review are red [12:29] does anyone know what this is about? [12:29] error: Get https://api.snapcraft.io/api/v1/snaps/download/99T7MUlRhtI3U0QFgl5mXXESAiSwt776_7396.snap: x509: certificate signed by unknown authority (possibly because of "crypto/rsa: verification error" while trying to verify candidate authority certificate "Let's Encrypt Authority X3") [12:29] 1008 [12:29] are we hitting some odd redirect, or something or the sort, that breaks ssl? [12:37] zyga: left some comments under #7484 [12:38] PR #7484: osutil: generalize SyncDir with FileState interface [12:40] mborzecki: I noticed, thank you so much! [12:42] zyga: tbh found the documentation of io.Reader somewhat vague [12:42] yeah, it's an interface after all :) [12:42] and the case of Read() returning , io.EOF is quite fun to handle [12:42] I'll look at the error handling, I agree it's pretty complex [12:54] mborzecki: when does it return n, eof? [12:55] mborzecki: i mean, reading what? [12:55] os.File says it's 0, eof on eof [12:56] Chipaca: right, but we're passing io.Reader around [12:56] fair [12:57] Chipaca: i mean, the func is called StreamsEqual() so it does make sense to use io.Reader in there [12:57] totes [12:58] though that comes with some luggage :/ [13:20] mvo: installed snapd from edge, it's showing GPL-3.0 correctly now https://paste.ubuntu.com/p/65D3S6KqKr/ [13:21] mborzecki: nice, thanks [13:25] mborzecki: so [13:25] mborzecki: why not take an io.Reader and wrap it into a osutil.readerButAlwaysZeroWhenError ? [13:25] mborzecki: wouldn't that keep the code calling Read simpler? [13:39] Chipaca: perhaps, wonder if it's worth it, more code to test [13:39] mborzecki: depends how complex the code without it is [13:39] ¯\_(ツ)_/¯ [13:57] off to pick up the kids [14:03] is roadmr on vacation ? [14:06] om26er: quite likely [14:07] Chipaca in his absence, do you know who to contact regarding store requests ? [14:07] I don't want to bug Jamie for everything ;-) [14:07] om26er: which request is it? [14:07] om26er: the one from this morning? [14:08] about content interface autoconnection between snaps of different publishers? [14:08] Chipaca yes, that and two others (a total of three auto-connection requests) [14:08] om26er: did you read the process for requesting auto-connection? [14:09] I have read, yes. But I have had the process fastened in the past and this is something I wanted sooner because its for a project that I have to show at Ubucon in few weeks [14:10] om26er: I'd be entirely unsurprised if today were a bad day for this [14:10] om26er: a lot of people take the day after a sprint off [14:11] aha, its the "swap day" day ;-) [14:11] yus [14:13] * om26er used to have two swap days after Canonical events due to two weekends consumed as a result of very long flights [14:15] om26er: ssh, don't tell cachio, he'll get ideas [14:16] * diddledan whistles nonchalantly: https://usercontent.irccloud-cdn.com/file/gXAW7XVu/image.png [14:16] the full output: https://usercontent.irccloud-cdn.com/file/TgRSHRNQ/image.png [14:17] start your rippers! [14:17] * Chipaca is already ripped [14:29] PR snapd#7488 closed: docs: Update README.md [14:56] * zyga lunch and walk [15:06] Chipaca: thanks for merging this. are we back to green tests? [15:30] mvo: looked like it! [15:47] brb, need to get two birthday cakes into the oven === genii_ is now known as genii [16:18] PR snapd#7495 opened: [RFC] tests: add gh action for static checks <⛔ Blocked> === lborda is now known as lborda_afk