=== ackk is now known as ack
=== ack is now known as Guest46801
[05:06] <mborzecki> morning
[05:08] <zyga> Hi
[05:09] <mborzecki> zyga: hey hey
[05:16] <zyga> mborzecki: being so tired somehow I am not looking forward to that 7AM flight tomorrow
[06:32] <jamesh> zyga: so, after refactoring the session agent code as its own process, I've got something that even runs on Ubuntu Core 16
[06:32] <jamesh> fails on Core 18 due to missing unit files.  I guess this is another thing that would need to be symlinked in if we wanted it to work.
[07:02] <mvo> good morning! did I miss anything important yesterday?
[07:03] <mborzecki> mvo: hey
[07:04] <mborzecki> mvo: nothing i'm aware of, somneone found a nice bug in the forums though: https://forum.snapcraft.io/t/snap-help-install-unexpectedly-performs-the-install/11686
[07:07] <mvo> mborzecki: thanks!
=== pstolowski|afk is now known as pstolowski
[07:11] <pstolowski> mornings
[07:12] <mborzecki> pstolowski: hey
[07:12] <mvo> hey pstolowski ! good morning
[07:19] <zyga> jamesh: hey
[07:19] <zyga> jamesh: which unit files are missing on core18?
[07:20] <zyga> jamesh: but that refactor sounds like good progress! thank you
[07:20] <zyga> mvo: hey, good morning
[07:20] <jamesh> zyga: it looks like it is something I could probably fix in wrappers/core18.go, after a bit more investigation
[07:20] <zyga> mvo: I will file yesterday off, personal stuff and I ended up not doing much
[07:20] <zyga> jamesh: I will review your branch soon
[07:21] <jamesh> zyga: I also ran into problems with your "replace ! with not" suggestion.  I was calling a shell function I had defined, which is not available in the context of your "not" shell script
[07:22] <zyga> jamesh: ah indeed
[07:22] <jamesh> but the "not" shell script successfully inverted the "command not found" error into a success
[07:22] <zyga> for shell functions you must refactor the code a little
[07:22] <jamesh> so I didn't notice it until looking at the failures on core18
[07:23] <zyga> jamesh: thank you for reporting that, I will special-case command-not-found to make it less surprising
[07:24] <mvo> zyga: sure, no worries
[07:25] <zyga> mvo: compound iza's school change (meeting the parents) and paperwork for lucy (both parents must be present)
[07:25] <zyga> sorry, it will carry on today for the part of morning, I should be back at noon;
[07:38] <jamesh> zyga: for what it is worth, the following shell function looks like it'd work as a a "not" implementation: not() { ! "$@"; }
[07:38] <jamesh> that can see shell functions, and correctly exits on failure
[07:40] <zyga> jamesh: interesting, perhaps it's worth switching over
[07:40] <zyga> jamesh: thank you
[07:41] <jamesh> it also has 10 consecutive pieces of punctuation, so is almost perl
[08:43] <Paddy_NI> Yaay!
[08:43] <Paddy_NI> First snap https://snapcraft.io/peerflix
[08:44] <Paddy_NI> Some of the functionality is broken for now and there is no "official" icon.
[08:48] <zyga> Paddy_NI: congratulations! How did it feel like?
[08:49] <Paddy_NI> zyga, It's pretty cool, I want to do more. lol
[09:00] <popey_> Paddy_NI: that's rather cool!
[09:01] <Paddy_NI> popey_, I know right!  My wife doesn't get it :-(
[09:02] <popey_> hah :)
[09:02] <Paddy_NI> She's more excited about her new job
[09:02] <Paddy_NI> X-D
[09:03] <Chipaca> popey_: easter egg for you
[09:03] <Chipaca> popey_: snap --help install bofh
[09:04] <popey_> I was just reading that thread :)
[09:04] <Chipaca> popey_: heh
[09:04] <popey_> nice try ;)
[09:04]  * Chipaca would put on a dunce hat but it won't fit over the donkey ears
[09:05] <popey_> reminds me of the time i told someone on the desktop team to hold down the printscreen button. (bug meant it ddosed your laptop with printscreens trying to do gl photo effect). that was fun. I should have warned them :)
[09:05] <Chipaca> popey_: where's the fun in warning people
[09:06] <popey_> exactly
[09:09] <Chipaca> what I'd _rather_ have happen is that 'snap --help foo' and 'snap foo --help' mean the same thing
[09:09] <Chipaca> but, sadly, go-flags does not let me do that
[09:09] <Chipaca> so you get the global help ¯\_(ツ)_/¯
[09:19] <zyga> ok, back in some degree
[09:36] <jamesh> Chipaca: maybe what the world needs is another command line parsing library?
[09:36] <zyga> anyone using qemu for testing: log into your vm (without -snapshot), apt-get build-dep  -d  snapd
[09:36] <zyga> that will save you 0.5GB each time you boot
[09:37] <zyga> obviously setting up apt-cacher-ng is even better but this is low-tech
[09:37] <mvo> mup_: hello
[09:37] <Chipaca> jamesh: I've got one on my list to look into
[09:37] <mvo> niemeyer: hey, could you please restart mup_ ? it seems to be unhappy
[09:37] <Chipaca> jamesh: but it is very tempting
[09:38] <Chipaca> jamesh: 'kingpin' i'm told is alright
[09:41] <zyga> hello mvo
[09:41] <mvo> hello zyga
[09:41] <zyga> mvo: small update, I've created more tests and more feature code for MS_SHARED bug
[09:42] <zyga> it's not fixed yet, I'm struggling with the bit related to propagation and delta algorithm
[09:42] <zyga> on the up side: i've prepared matching unit and integration tests that allow easy iteration
[09:42] <zyga> I'll push it all up today even if it doesn't fix the bug completely
[09:43] <mvo> zyga: thanks for the update!
[10:05] <Chipaca> pedronis: you got a minute?
[10:05] <Chipaca> pedronis: about /v2/downloads
[10:05] <Chipaca> pedronis: should we make it a GET method?
[10:05] <Chipaca> current proposal has it as a POST but that feels weird
[10:21] <Chipaca> pedronis: especially because post -> needs auth or user or sth
[10:51]  * zyga goes offline for a while
[10:53] <ogra> You should not run jhbuild as root.
[10:53]  * ogra shakes fist !
[10:54] <ogra> silly thing ... doesnt pick up JHBUILD_RUN_AS_ROOT=1 from the environment :(
[11:15]  * Chipaca ⇝ lunch
[11:30] <Eighth_Doctor> mborzecki: hey, so what kind of talk do we want to propose for Flock?
[12:00] <zyga> back onine
[12:30] <pedronis> Chipaca: I was at my doctor appt, if I remember we had reasons to make it a POST, we can discuss after the standup if it's not too long
=== ricab is now known as ricab|lunch
[12:49] <ackk> mvo, hi, the core18 snap doesn't seem to contain iptables, is that expected?
=== ricab|lunch is now known as ricab
[14:21] <zyga> random failure in device manager:
[14:21] <zyga> https://www.irccloud.com/pastebin/pqgs1A3O/
[14:22]  * zyga dinner
[15:09]  * cachio lunch
[15:38] <zyga> re
=== pstolowski is now known as pstolowski|afk
[16:36]  * zyga runs some more spread tests
[17:14] <ogra> is there any known issue with core18 currently ? seems i'm not alone in hitting:
[17:14] <ogra> Press enter to configure.
[17:14] <ogra> bah ... IRC
[17:14] <ogra> /usr/share/subiquity/console-conf-wrapper: line 13: snap: command not found
[17:14] <ogra> /usr/share/subiquity/console-conf-wrapper: line 13: snap: command not found
[17:14] <ogra> Press enter to configure.
[17:16] <ogra> looking at the SD i see that snapd definitely did its seeding ... so the snapd snap should be there (which makes me think that also the snap command should be there)
[17:17] <ogra> could it be that snap isnt in PATH anymore ?
[17:17] <ogra> (did anything change here ?)
[17:17] <cachio> ogra, I saw that but time ago
[17:17] <cachio> and couldn't reproduce it anymore
[17:17] <ogra> i have reports from customers trying out their own core18 images seeing the same in the stable channel (i see it in edge here)
[17:18] <ogra> and kyleN did hit it too apparently
[17:18] <cachio> ogra, I didn't see the for a while
[17:18] <ogra> strange
[17:18] <cachio> I think I could reproduce it when accidentally caused an ethernet error
[17:19] <cachio> and the board stayed in a weird state
[17:19] <ogra> my ethernet is fine ... i'll re-flash the SD ...
[17:19] <cachio> but then I was trying to reproduce it and I had no luck
[17:27] <ogra> ogra@localhost:~$ fw_printenv
[17:27] <ogra> -bash: fw_printenv: command not found
[17:27] <ogra> sigh ... this is so overly annoying in core18 :(
[17:43] <cachio> zyga, raised the lp:1832024
[17:43] <cachio> I am trying to identify which test executed before is causing the error
[17:43] <cachio> because when I run thtat test alone it works well
[18:06] <zyga> cachio: thank you!
[19:20] <Saviq> hey all, is there somewhere a list of the required / recommended kernel options for snapd to operate?
[19:20] <Saviq> I'm seeing "AppArmor status: apparmor is enabled but required kernel features are missing: file" now, would like to let the ISP know what to enable (I'm in a container)
[19:55] <ogra> Saviq, https://github.com/snapcore/sample-kernels/tree/master (old but still mostly valid)
[20:13] <zyga> Saviq: it's ubuntu sauce patches
[20:13] <zyga> Saviq: ask jjohansen  about those, or anyone from the kernel team
[20:15] <jjohansen> Saviq: can you see /sys/kernel/security/apparmor from in your container?
[20:28]  * zyga EOWs
[20:45] <jjohansen> Saviq: we can't even at this point say the issue is ubuntu's sauce patches
[20:47] <jjohansen> more likely is your ISP is not setting up apparmor for your container. It needs securityfs mapped in, or at least the apparmor subdirectory (/sys/kernel/security/apparmor)
[20:47] <jjohansen> that needs to be done with a bind mount
[20:47] <jjohansen> and it will need an apparmor namespace
[20:48] <jjohansen> currently lxd is the only container that I know of that is setting up the apparmor namespace and bind mounting securityfs into the container
[22:13] <Saviq> jjohansen: sorry for the delay, the dir is there, but yeah I'm getting ENOPERM trying to list its contents
[22:40] <jjohansen> Saviq: yeah apparmor needs access to the contents, that is how it loads policy