[05:35] <mborzecki> morning
[05:53] <mborzecki> new kernel, brb
[05:56] <mvo> hey mborzecki ! good morning
[05:56] <mborzecki> mvo: hey, morning
[05:56] <mborzecki> mvo: how was your vacation?
[05:57] <mvo> mborzecki: very nice, thank you
[05:58] <mvo> mborzecki: also good to be back :) anything I should look at right away, i.e. anything blocking one of you?
[05:58] <mborzecki> mvo: lots of things were red last week
[05:59] <mvo> mborzecki: yeah, protocol_error is something I saw
[05:59] <mborzecki> mvo: i briefly joined standup on friday and we were still tracking PROTOCOL_ERROR popping up in http client
[05:59] <mvo> mborzecki: and one of the tests failed to remove a dir
[05:59] <mvo> mborzecki: thanks for this, I will check it out - how was flock?
[06:00] <mborzecki> mvo: it was good, really interesting
[06:00] <mborzecki> mvo: got some notes, need to compile a report out of that :)
[06:00] <mvo> mborzecki: cool, looking forward to that :)
[06:31] <mvo> silly question - 7208 looks green and ready but is not merged yet - any reason for this? can I just merge it or is there something I am missing?
[06:52] <zyga> Good morning
[06:52] <zyga> I don’t have my laptop anymore
[06:52] <zyga> Nor am I home
[06:53] <zyga> mvo: would it be terrible if I skipped our 1:1 today? I’m planning on how to get home today
[06:54] <zyga> We can try to sync on telegram audio
[06:55] <mvo> hey zyga ! good mooooorning
[06:55] <mvo> zyga: its fine, we can reschedule. out of curiosity, where are you :) ?
[06:55] <mborzecki> zyga:h hey
[06:58] <zyga> mvo: I’m in Ostróda
[06:58] <zyga> Trains are full of people returning from seaside
[06:58] <zyga> And I was unable to get a ticket for anything yesterday
[06:59] <mvo> zyga: ok
[06:59] <zyga> I’ll get coffee and breakfast, I should be home in the afternoon
[06:59] <mvo> zyga: ok
=== pstolowski|afk is now known as pstolowski
[07:07] <pstolowski> mornings!
[07:07] <pstolowski> mvo: welcome back!
[07:10] <mvo> hey pstolowski ! thank you
[07:11] <mvo> pstolowski: feels good to be back - anything you need from me where I could unblock/help you?
[07:12] <mborzecki> pstolowski: hey
[07:13] <mborzecki> quick errand, back in 30
[07:28] <zyga> mvo: tests are in a sad state
[07:29] <zyga> Store/CDN protocol error
[07:29] <zyga> Unit tests sometimes fail on context cancelled
[07:31] <mvo> zyga: yeah, I noticed :/
[07:31] <mvo> zyga: looking at some of them right now
[07:40] <zyga> Protocol error is a mystery, unclear if anything we changed or if something far side
[07:40] <zyga> Tests are clearly our side, some of the recent changes to daemon code
[07:41] <pstolowski> mvo: Sergio was in touch with store re protocol error but they didn't find anything :(
[07:43] <pstolowski> mvo: sorry, missed your earlier question - you can look at my latest PRs concerning firstboot, and also https://forum.snapcraft.io/t/firstboot-seeding-failure-scenario-possible-fixes-and-boot-process-confusion-question/12698
[07:46] <mvo> pstolowski: great, will do
[07:47] <pstolowski> ty
[07:49] <mborzecki> re
[07:58] <pstolowski> mvo: do you remember the details of be4fc4d117c255cadd697a93f9f94e49c708f2c3 ?
[08:02] <mvo> pstolowski: yes :)
[08:02] <mvo> pstolowski: it was needed because we used a version of mux from 2015
[08:03] <mvo> pstolowski: that pre-dated context.Context
[08:03] <mvo> pstolowski: which caused issues when trying to use context.Context in our http code
[08:03] <mvo> pstolowski: iirc it was breaking passing MuxVars
[08:03] <mvo> pstolowski: why do you ask? is it causing issues?
[08:04] <Chipaca> moin moin
[08:04] <pstolowski> mvo: the reason i'm asking is I hit a weird situation ~2 weeks ago, where my custom build of snapd & snap would refuse to talk to each other at runtime (i didn't know vendor needed refreshing), the passing of args via api seemed broken (api would not receive snap names with snap ops for example). interestingly, everything built just fine. i thought i messed up my VM and eventually refreshed vendor and everything worked.
[08:04] <pstolowski> but last Firday ijohnson hit the same issue, and this is when I understood it was vendor change.
[08:05] <pstolowski> mvo: so we were scratching our heads last Friday; why evertyhing builds correctly without vendor refresh but fails only at runtime
[08:06] <pstolowski> mvo: and also if we didn't accidentaly break api compatibility with this change?
[08:07] <mvo> pstolowski: oh, interessting - that is worth exploring, let me try to reproduce
[08:19] <pstolowski> mvo: now that i think of it, most likely the semantics of context api changed and without vendor refresh it wouldn't work correctly with current master, but our api is not affected
[08:21] <mvo> pstolowski: yeah, I think it won't break api but confuses our code so I think we need some sort of version check
[08:21] <mvo> pstolowski: i.e. make the build fail if current snapd master is used with an older gorilla
[08:21] <mvo> pstolowski: either via an explicit version check or by using something new from the gorilla code as a test to see that things are current enough
[08:22] <pstolowski> mvo: nice idea, i can take a look at this
[08:22] <mvo> pstolowski: that would be great!
[08:22] <mvo> pstolowski: thank you!
[08:22] <pstolowski> np
[08:22] <mvo> Chipaca: hey, good morning! hope you are well, anything I can help you with to unblock you?
[08:23] <mvo> Chipaca: (if needed :)
[08:26] <Chipaca> mvo: nothing particular! i succumbed to curiosity and caught up with some bits of the forum over the weekend so it's not a terrible backlog today :)
[08:26] <Chipaca> but i am going to need extra coffee
[08:27] <Chipaca> mvo: aren't you also just back?
[08:31] <mvo> Chipaca: great to hear. yes, just returned
[08:32] <mvo> Chipaca: and coffee sounds like a great idea (s/coffee/tea for me but the idea is the same :-D
[08:32] <Chipaca> yeah
[08:32] <Chipaca> mvo: similarly then, holler if i can do anything for you
[08:33] <mvo> Chipaca: thanks!
[08:52] <Chipaca> pedronis: wb!
[08:53] <mborzecki> mvo: added your comments about the http2 problem to the forum topic https://forum.snapcraft.io/t/snap-download-failures-with-protocol-error/12677
[08:54] <mborzecki> looks like the whole team is back :)
[08:54] <mvo> mborzecki: thank you! I'm looking into some sort of unit test right now but it appears tricky
[08:54] <mvo> hey pedronis, welcome back!
[08:57] <mborzecki> mvo: https://github.com/snapcore/snapd/pull/7232
[09:01] <pedronis> Chipaca: mvo: hi
[09:02] <mvo> mborzecki: that looks similar like the one that sergio did some days ago?
[09:02] <mvo> mborzecki: also I suspect this affects real users too, not just tests :/
[09:03] <mborzecki> mvo: yeah, i'm surpised fastly does not log details when error like this appear
[09:05] <mvo> mborzecki: I suspect the http2 spec if not super great at providing extra information
[09:06] <mvo> mborzecki: its also unclear if its their fault or go fault, it might be something that go does wrong and the reply is "protcol_error" (i.e. they tell us "you do something silly")
[09:06] <mvo> mborzecki: but don't take my word for it (yet) :) still trying to get into this issue
[09:06] <mborzecki> mvo: we need log_martians for http2
[09:10] <mvo> mborzecki: haha
[09:14] <pstolowski> pedronis: hi, welcome back!
[09:15] <pstolowski> mborzecki: hey, how was flock?
[09:16] <pedronis> pstolowski: hello
[09:17] <mborzecki> pedronis: Chipaca: hey
[09:18] <mborzecki> pstolowski: intersting, i can tell you more about it in the standup
[09:21] <pstolowski> mborzecki: sounds good!
[09:31] <pstolowski> mborzecki: i suppose it's too early for recordings from flock?
[09:31] <mborzecki> pstolowski: yeah
[09:49] <mvo> 7190 needs a second review, should be easy
[09:49] <pstolowski> mvo: re #7234 i'm not sure if there is a better way, but fwtw it did the job in my tests
[10:05] <mborzecki> audio-{playback,record} will replace pulseaudio interface?
[10:10] <pedronis> mborzecki: yes
[10:24] <mvo> pstolowski: thanks, this looks good
[10:39] <mvo> fwiw, I added a tweak to 7231 to disable the interfaces-calendar-service test so that this gets green
[10:39] <mvo> I got two errors in this test in a row
[10:40] <mvo> but no protocol_errors anymore, seems the retry helps short term
[10:45] <pstolowski> +1
[10:52] <Chipaca> api.snapcraft.io struggling, expect tests to flake
[10:56] <mborzecki> quick errand, back in 30
[11:37]  * Chipaca lunches, probably
=== ricab is now known as ricab|lunch
[12:14] <mborzecki> re
[12:45] <jdstrand> mvo: hi! hope you had a nice holiday :)
[12:46] <jdstrand> mvo: so, I committed the audio-playback/audio-record PR on friday, but only after reverting the pulseaudio changes. I squashed and merged, but didn't update the first line, which said I made pulseaudio changes
[12:47] <jdstrand> mvo: so I went to commit --amend, but couldn't commit it since it is protected
[12:48] <jdstrand> mvo: so I created pr 7229 with no code changes (which I see now you approved), and all weekend I've tried to get the thing to pass tests, but it won't ever pass
[12:49] <jdstrand> mvo: (different failures each time. probably tried 10+ 'restart builds' for the individual integration tests
[12:49] <jdstrand> mvo: is there something else I should be doing?
[12:49] <mvo> jdstrand: yeah, its a bit annoying right now, we are working on this currently to fix the test situation - sorry for this
[12:50]  * jdstrand can keep mashing restart build
[12:50] <jdstrand> mvo: ok. I guess from your approval, I did the right thing with an errata commit?
[12:51] <jdstrand> mvo: after doing the wrong thing by not changing the first line of course :)
[12:53] <mvo> jdstrand: just to make sure I understand> the commit that needs an errata is already on master, yes?
[12:57] <abeato> jdstrand, morning! I added a couple of comments in https://github.com/snapcore/snapd/pull/7173 , could you please take another look?
[12:59] <Chipaca> gah, just as i settle in to the hangout, 2fa
[12:59] <Chipaca> and it's at the other end of the house
[12:59]  * Chipaca untangles himself and goes get
[13:01] <zyga> I am not coming today
[13:01] <zyga> Just FYI
[13:03] <pedronis> mvo: standup?
[13:05] <Chipaca> zyga: why tho?
[13:05] <mborzecki> mvo: pulseaudio -k and restart chromium?
[13:06] <pstolowski> ijohnson: wdyt: https://github.com/snapcore/snapd/pull/7234 ?
[13:10] <jdstrand> mvo: yes. I commited to master. the code is correct. all of the commit messages is correct except the first line. I would've done a commit --amend but couldn't due to protected branch, so have the empty errata commit instead
[13:11] <jdstrand> abeato: yes, planning on going through PRs today/tomorrow
[13:12] <abeato> jdstrand, ok, thanks
=== ricab|lunch is now known as ricab
[13:26] <zyga> Chipaca: on a train
[13:26] <zyga> No laptop
[13:27] <Chipaca> zyga: is it a CI train
[13:36] <mvo> Chipaca: lol
[13:59] <tomwardill> any idea why a core refresh will cause my HDD that's in standby to wake up?
[14:01] <popey_> tomwardill: yeah, the whole re-mount the world does that for me too, always has
[14:02] <tomwardill> minorly irritating, as that drive is the noisiest thing in my PC
[14:49] <zyga> tomwardill: probably udev trigger / settle
[14:49] <Chipaca> popey_: quick question about your mysterious numbers-don't-appear thing: do you have the noto font(s) installed? does the issue go away if you remove 'em?
[15:07] <Chipaca> whoa, suddenly very dark here
[15:08] <diddledan> did you unleash the end of days, again?
[15:08] <Chipaca> le ciel va nous tomber sur la tête!
[15:08] <diddledan> tuesday
[15:09] <Chipaca> all I did was e̷̺͚̦̦̬̩̠̽̄͛̀͋̑͌͝m̡̙̳̻̻̩͖̼̒̽̉̀̈͑̄͠e͍͎̫͍̮͉͗̓͛͊̂̓̆́̕͜ͅr̡̥̯̻̣̭̥͉̩̿͌̾̊̾͜g̺̞̜̭̠̈̌̋̀͐̆̌̉̌̕͜e̶̗̭̝͎̞̦̤̺̔͛̈̃͛̃ s̶͔̻̰̰͕̟̣̻̀̑̿̊̃̆͜͝͝͡k̯̜̲͖͉̘̹͌́̐̌̂̔͘̚͞y̛̛̙̤̗̲̖̝̓̔̓͘
[15:10] <diddledan> that looks decidedly like the spell to raise the dead
[15:10] <marcustomlinson> woah
[15:12] <diddledan> related query, you're running Gentoo?! :-o
[15:14] <Chipaca> no, of course not
[15:14] <diddledan> phew
[15:14] <Chipaca> obviously it's sabayon
[15:14] <diddledan> :-p
[15:20] <popey_> Chipaca: lemme try
[15:25] <popey_> Chipaca: you win a cookie
[15:26] <Chipaca> popey_: I'm already 300kcal over my budget, but thanks
[15:29] <cachio> Chipaca, hey, any idea how to skip catalog update?
[15:29] <cachio> https://paste.ubuntu.com/p/bQ7cSX4Dss/
[15:29] <cachio> on our tests to prevent this error?
[15:29] <mvo> Chipaca: can we merge 7208 (green and plenty +1)?
[15:30] <Chipaca> mvo: if that is tab completion, i wanted to add a check that zyga asked about
[15:30] <Chipaca> cachio: we should be able to, gimme a mo'
[15:30] <cachio> Chipaca, tx
[15:32] <mvo> Chipaca: ok
[15:34] <Chipaca> cachio: touch /var/cache/snapd/sections before starting snapd should be enough to not refresh
[15:34] <Chipaca> cachio: OTOH, we should totally catch that 429, now that it's there
[15:35] <cachio> Chipaca, that's true
[15:35] <Chipaca> I'll look into that
[15:36] <Chipaca> cachio: do you have a bit more context around that error output?
[15:36] <cachio> Chipaca, https://travis-ci.org/snapcore/snapd/jobs/570795886#L2314
[15:37] <cachio> this it the error
[15:37] <cachio> I talked to store team
[15:37] <cachio> and they say we should avoid refreshing so much to prevent this error
[15:37] <cachio> the is not other way
[15:46]  * cachio afk
[15:46]  * cachio lunch
[16:00] <Chipaca> ok, i'm off to do gym stuff. I'll sneak in a 429 PR this evening, after closing one of my open ones.
=== pstolowski is now known as pstolowski|afk
[16:15] <ijohnson> hey folks I asked in the PR description for my netplan-apply PR, but is it standard practice to include the source of a complicated snap that needs to be built with snapcraft inside the snapd git tree? I see a number of snapcraft.yaml's there so it seems the answer is yes?
[16:18] <ijohnson> I see some tests also have the snap source inside the ~snappy-dev project, so is that a better place to put snapcraft.yaml sources for test-snapd-* snaps?
[16:30] <mborzecki> ijohnson: i'd say that keeping the sources with snapcraft.yaml in tests/lib/snaps makes most sense
[16:30] <ijohnson> mborzecki: ack I'll add my snapcraft.yaml for the test-snapd-netplan-apply snap to the PR then
[16:30] <ijohnson> thanks
[16:31] <ijohnson> also who should own the test-snapd-* snaps?
[16:31] <mborzecki> ijohnson: hah, great question :P
[16:31] <mborzecki> ijohnson: the owner should be canonical, but you have to transfer the snap
[16:32] <mborzecki> ijohnson: let me look it up, there was some email you need to send to the store ml
[16:33] <ijohnson> got it, thanks mborzecki
[16:44]  * ijohnson lunches
[17:35] <Chipaca> popey_: ok, went to the gym, now where's my cookies at
[17:46] <popey_> Chipaca: i would give you an unicode cookie, but....
[17:46] <Chipaca> popey_: too late, now i'm all fancy and only accept macaroons
[17:47] <Chipaca> also it's tea time so pulled pork and veg is already in prep
[17:47]  * Chipaca might be salivating in preparation
[18:45] <mvo> I just merged 7231 so with a bit of luck master is green again
[18:49] <ek926m> hey, I have a question, why my snapd service failed?
[18:49] <ek926m> snapd.service: Triggering OnFailure= dependencies.
[18:49] <ek926m> Failed to start Snappy daemon.
[18:50] <ek926m> ● snapd.service - Snappy daemon
[18:50] <ek926m>  systemd[1]: snapd.service: Service RestartSec=100ms expired, scheduling restart.
[18:50] <ek926m> pd.service: Failed with result 'signal'.
[18:51] <ek926m> is there a way to find out, what caused this failure?
[19:20]  * ijohnson relocates to a coffee shop
[19:47] <diddledan> who maintains the docker snap? it's listed as canonical with an obsolete email address (snappy-devel@lists.ubuntu.com)
[20:00] <roadmr> diddledan: talk to ijohnson
[20:01]  * ijohnson hides
[20:01] <ijohnson> what's up?
[20:01] <roadmr> 😎
[20:01] <diddledan> ello. was wondering if we're gonna get any version bumps any time in the next year? ;-p
[20:01] <ijohnson> for what?
[20:01] <diddledan> </sarcasm>
[20:01] <diddledan> docker
[20:02] <ijohnson> ah okay, so actually the maintenance of the docker snap now falls on tianon
[20:02] <diddledan> the contact info is out of date on it, too
[20:02] <ijohnson> I know that there was some updates to the docker-support interface needed for 18.09 I think
[20:02] <diddledan> aha
[20:02] <ijohnson> those should have landed today in the stable channel though
[20:02] <ijohnson> so hopefully tianon will be able to release
[20:02] <diddledan> we're on 19.03 upstream now IIRC
[20:03] <ijohnson> I will make sure the contact info is updated though
[20:03] <diddledan> \o/ danke
[20:03]  * diddledan hopes for new shiny features :-po
[20:04] <diddledan> all the features!
[20:26] <diddledan> something odd here:
[20:26] <diddledan> https://www.irccloud.com/pastebin/QNEGGvGW/odd.txt
[20:27] <ijohnson> diddledan: where did that come from?
[20:27] <diddledan> I was running `snap install mycroft_19.2.14_amd64.snap --dangerous1
[20:27] <diddledan> --1
[20:27] <diddledan> bad typo :-p
[20:27] <diddledan> `snap install mycroft_19.2.14_amd64.snap --dangerous`
[20:28] <diddledan> my layouts are:
[20:28] <ijohnson> diddledan: what's `snap tasks --last=install` show?
[20:28] <diddledan> https://www.irccloud.com/pastebin/qUSq5vIT/layouts.yaml
[20:28] <ijohnson> ah you have layouts
[21:37]  * cachio afk