[01:44] <mup> PR snapcraft#2437 opened: repo,baseplugin: support trusting repo keys <Created by kyrofa> <https://github.com/snapcore/snapcraft/pull/2437>
[05:35] <zyga> Good morning
[06:07] <mborzecki> morning
[06:34] <zyga> hey mborzecki  :)
[06:34] <mborzecki> zyga: hey hey
[06:35] <zyga> I'd love to land some boring stuff today
[06:35] <zyga> most blocked by https://github.com/snapcore/snapd/pull/6353
[06:35] <mup> PR #6353: cmd/snap-update-ns: move existing code around, renaming some functions <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6353>
[06:35] <mborzecki> zyga: which pr should i start with first?
[06:35] <zyga> I think this one is the only one I really need
[06:35] <zyga> though all the <simple> ones should be a moment to review
[06:36] <mborzecki> ack
[06:40] <mup> PR snapd#6355 closed: release: 2.37~pre1 <Created by mvo5> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6355>
[07:02] <mborzecki> zyga: one suggestion under #6349, other than that it's +1
[07:02] <mup> PR #6349: cmd/snap-update-ns: move XDG code to dedicated file <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6349>
[07:02] <zyga> looking
[07:03] <zyga> nice :)
[07:05] <zyga> pushed
[07:06] <zyga> I prepared another branch but I will need https://github.com/snapcore/snapd/pull/6346 as well
[07:06] <mup> PR #6346: osutil: add helper for loading fstab from string <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6346>
[07:08] <zyga> mborzecki: updated https://github.com/snapcore/snapd/pull/6346 based on your comment
[07:08] <mup> PR #6346: osutil: add helper for loading fstab from string <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6346>
[07:54] <mborzecki> zyga: regarding #6345 none of the patches that start to make use of FromSnapConfine in s-u-n user mounts are up yet, right?
[07:54] <mup> PR #6345: cmd/libsnap: pass --from-snap-confine when calling snap-update-ns as user <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6345>
[07:54] <mborzecki> zyga: same for freezer
[07:55] <zyga> mborzecki: the main branch is
[07:55] <zyga> mborzecki: the only use of --from-snap-confine is to lock/freeze
[07:55] <zyga> but that is arguably one of the last patches in this pile
[07:55] <mborzecki> zyga: mhm
[07:57] <mborzecki> btw. have dropped codecov?
[07:57] <mborzecki> s/have/have we/
[08:01] <zyga> I ... don't remember we did
[08:01] <zyga> but perhaps?
[08:02] <mborzecki> idk, haven't seen any coments being added to the new PRs
[08:02] <pstolowski> morning
[08:04] <mborzecki> pstolowski: heya
[08:04] <zyga> hey Pawel
[08:13] <zyga> pstolowski: hey, could you do a 2nd review for https://github.com/snapcore/snapd/pull/6353
[08:13] <pstolowski> zyga: sure
[08:13] <mup> PR #6353: cmd/snap-update-ns: move existing code around, renaming some functions <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6353>
[08:20] <Chipaca> moin moin
[08:20] <zyga> hey chipaca :)
[08:21] <zyga> Chipaca: can I bug you for one little review in the morning please? https://github.com/snapcore/snapd/pull/6346
[08:21] <mup> PR #6346: osutil: add helper for loading fstab from string <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6346>
[08:21] <Chipaca> zyga: you can bug me all you want
[08:21]  * Chipaca adds zyga to his ignore list
[08:22] <mborzecki> Chipaca: hey
[08:34] <zyga> re
[08:45] <Chipaca> should I have another coffee?
[08:45] <Chipaca> I should have another coffee.
[08:45]  * Chipaca obeys
[08:57] <zyga> Chipaca: I hear you
[08:57] <zyga> coffee is good
[08:57] <mup> PR snapd#6356 opened: overlord/snapstate: during refresh, re-refresh on epoch bump <Created by chipaca> <https://github.com/snapcore/snapd/pull/6356>
[08:58] <Chipaca> pedronis: ^_^
[08:58] <pedronis> Chipaca: thx
[08:58] <Chipaca> pedronis: and good morning
[08:59] <pedronis> we'll see if I get to it today
[08:59] <zyga> pstolowski: https://github.com/snapcore/snapd/pull/6353#issuecomment-453438414
[08:59] <mup> PR #6353: cmd/snap-update-ns: move existing code around, renaming some functions <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6353>
[09:00] <pstolowski> zyga: sure
[09:00] <zyga> thanks!
[09:00] <zyga> I didn't notice 2017 :)
[09:00] <mup> PR snapd#6353 closed: cmd/snap-update-ns: move existing code around, renaming some functions <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6353>
[09:01] <zyga> Chipaca: the pic in your response is lovely
[09:01] <zyga> feels like the way to spend time :)
[09:01] <Chipaca> zyga: :-)
[09:01] <mup> PR snapd#6346 closed: osutil: add helper for loading fstab from string <Simple 😃> <Created by zyga> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6346>
[09:03] <pedronis> Chipaca: next on your list, I think, is snapshot user docs?  and maybe picking up/rework #6348? does that sounds right?
[09:03] <mup> PR #6348: snap: split alignment calculation and display for channels <Created by mvo5> <https://github.com/snapcore/snapd/pull/6348>
[09:04] <Chipaca> pedronis: I'd like to know what to do about #6034 also
[09:04] <mup> PR #6034: many: save media info when installing, show it when listing <⛔ Blocked> <Created by chipaca> <https://github.com/snapcore/snapd/pull/6034>
[09:04] <Chipaca> pedronis: but, yes, snapshot user docs, info refactor, then maybe per-architecture info
[09:05] <pedronis> Chipaca: yes, I'm aware of 6034, but indeed needs discussion so it's blocked atm
[09:05] <Chipaca> pedronis: we're not going to have time to have that discussion before the town of capes, right?
[09:05]  * Chipaca hopes everybody packed capes
[09:06] <pedronis> likely not but we'll see
[09:07] <pedronis> Chipaca: also next week sprint is organized a bit differently, with lots of everybody sessions all in the morning, so I might not be busy all the time, just almost all the time
[09:07] <Chipaca> pedronis: also I notice you removed the block on #6280, does that mean it's gtg (pending 2nd review)?
[09:07] <mup> PR #6280: cmd/snap: make 'snap warnings' output yamlish <Created by chipaca> <https://github.com/snapcore/snapd/pull/6280>
[09:07] <Chipaca> pedronis: you're going to a sprint, and think you might not be busy all the time
[09:07] <Chipaca> *giggle*
[09:09] <pedronis> Chipaca: I should probably look at it again (6280) but is not blocked about the --unicode oddities vs non-ttys
[09:09] <Chipaca> pedronis: ah :-) ok
[09:09] <pedronis> Chipaca: mvo asked for gustavo's input but if I understand it is something that was discussed with him, no?
[09:09] <Chipaca> pedronis: yep
[09:09] <Chipaca> at the london sprint (!)
[09:09] <pedronis> also the general principle, if too wide for tabular/naive output, be yaml-like/parsable
[09:13] <niemeyer> Heya
[09:14] <pedronis> niemeyer: hi
[09:48] <zyga> brb
[09:51] <pedronis> Chipaca: can you remind me how to poke at the API with http ?
[09:52] <Chipaca> pedronis: http snapd:///v2/find name==go
[09:52] <Chipaca> pedronis: or, http --pretty=all snapd:///v2/find name==go | less -R
[09:53] <pedronis> Chipaca: thank you
[10:16] <Chipaca> huh, I thought snapshots were in 2.36
[10:17] <Chipaca> just had a user on 2.36.3 tell me it didn't work there
[10:31] <pedronis> Chipaca: no, they are not, not they client bits
[10:31] <pedronis> they are in 2.37
[10:34] <Chipaca> 2.37 is going to be a big'un
[10:39] <pedronis> pstolowski: hi, btw I started yesterday looking again at the hotplug-connect PR but didn't finish, we'll try to do so today
[10:40] <pstolowski> pedronis: ack, thanks
[10:59] <zyga> pedronis: added one suggestion to the doc you just shared, it was something on the back of my mind for several interfaces (e.g. optical drive is the same)
[11:02] <Chipaca> why isn't architecture in `snap version`?
[11:02] <pedronis> zyga: add it "all" to what?
[11:02] <pedronis> the plug, the slot?
[11:02] <zyga> to the interface, I will clarify my comment
[11:03] <zyga> well, to the slot side
[11:03] <zyga> if you have "all" you are the old legacy camera
[11:03] <zyga> if you don't have that you are a precise camera
[11:03] <zyga> either hotplug or gadget
[11:03] <pedronis> I was basing that on having a path or not
[11:04] <pedronis> but we can do both
[11:04] <pedronis> on the slot side, anyway this is a discussion to have
[11:04] <pedronis> those were just jumping points
[11:07] <zyga> that's good, it's the same idea
[11:07] <zyga> ah, I see it's on the following line
[11:08] <zyga> I'll retract my suggestion
[11:29] <mborzecki> any clue if we'll be merging master to 2.37 again?
[11:31] <zyga> master to 2.37?
[11:31] <zyga> never?
[11:31] <zyga> I think it is cherry picking from now on
[11:31] <mborzecki> hm ok
[11:31] <zyga> proxy leaking from test to test? https://www.irccloud.com/pastebin/cw3P0RWp/
[11:33] <zyga> soliciting 2nd review for https://github.com/snapcore/snapd/pull/6349
[11:33] <mup> PR #6349: cmd/snap-update-ns: move XDG code to dedicated file <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6349>
[11:33] <zyga> pstolowski: do you have a moment?
[11:38] <pstolowski> zyga: ok
[11:42] <zyga> mborzecki: https://github.com/snapcore/snapd/pull/6352#discussion_r
[11:42] <mup> PR #6352: many: remove .user-fstab files from /run/snapd/ns <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6352>
[11:43] <mborzecki> zyga: yup
[11:44] <mborzecki> zyga: wonder if that should use rm -f {} \: instead, so that find does not stop if there's a failure to remove
[11:45] <zyga> those are plain files
[11:45] <zyga> should remove fine
[11:45] <zyga> pushed
[11:45] <zyga> I think we only had -f to "fix" rm -f *.foo expanding to "*.foo"
[11:46] <mborzecki> it'd be good if Chipaca could take a look too
[11:46] <zyga> +1
[11:46] <zyga> I added him to the review now
[11:46] <zyga> +4, -16
[11:46] <zyga> :)
[11:47] <pedronis> pstolowski: +1 with a comment still, also would be good if you got reviews from mborzecki and mvo for it
[11:47] <mborzecki> hehe :)
[11:47] <pstolowski> pedronis: k
[12:20] <zyga> re
[12:20] <zyga> some kitchen emergency, all good now
[12:24]  * pstolowski lunch
[12:31] <zyga> I'm seeing one test fail often today;
[12:32] <zyga> tests/main/xdg-open-compat fails with ERROR 503 while talking to launchpad.nett https://www.irccloud.com/pastebin/74kd9yef/
[12:32] <Chipaca> zyga: .nett?
[12:33] <zyga> no, .net, just a typo
[12:35] <Chipaca> zyga: launchpad might be sad today
[13:01] <mup> PR snapd#6357 opened: cmd/snap-update-ns: make freezer mockable <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6357>
[13:05] <mup> PR snapd#6358 opened: cmd/snap-update-ns: close internal lock on error <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6358>
[13:05] <zyga> mborzecki, pstolowski: I would love a review for the two above, this would make the refactor purely a refactor with 0 new features or fixes
[13:06] <zyga> I broke the two patches out to make that true
[13:06] <zyga> they are both really tiny
[13:07] <mup> PR snapd#6349 closed: cmd/snap-update-ns: move XDG code to dedicated file <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6349>
[13:34] <diddledan> is the store api down (snap find foo isn't working) or just my system having issues?
[13:34] <diddledan> aha. it's gone through now
[13:34] <diddledan> transient
[13:34] <diddledan> ignore me
[13:34] <cmatsuoka> working for me
[13:34] <diddledan> :-)
[13:36] <zyga> re
[13:36] <diddledan> the knights who say "re"?
[13:36] <mup> PR snapd#6359 opened: tests: use double quotes for regex on listing test to match \\* <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/6359>
[13:36] <diddledan> :-p
[13:36] <zyga> store down and wonky errors https://www.irccloud.com/pastebin/pj7DG1v2/
[13:37] <zyga> Chipaca: ^
[14:02] <pedronis> Chipaca: standup?
[14:02] <Chipaca> agh, yes
[14:02] <Chipaca> was distracted by CVEs
[14:02]  * Chipaca goes
[14:21] <ali1234> how do i tell my snapcraft.yaml that i want to stage-packages from 16.04?
[14:22] <diddledan> if you don't tell it you want `base: core18` it will use 16.04 by default (you are using multipass, right?)
[14:22] <ali1234> i don't know
[14:25] <zyga> diddledan: I don't think that's true
[14:25] <zyga> diddledan: if you use base: core18 you will use the new base and the new multipass builder
[14:25] <zyga> diddledan: but in absence of that I think you will build the old way
[14:25] <zyga> diddledan: I may be wrong, but that's what I think was the approach
[14:25] <ali1234> "Issues while validating snapcraft.yaml: The 'version' property does not match the required schema: 0 is not a valid snap version string."
[14:26] <diddledan> zyga: yes, which is why I caveated my comment with the multipass
[14:26] <zyga> right but multipass is not on uless you use base: *
[14:26] <zyga> but there's no published core16 base
[14:26] <ali1234> what is a valid version string then?
[14:26] <zyga> ali1234: perhaps "1"
[14:26] <zyga> not sure why 0 is rejected
[14:26] <ali1234> nope, 1 also does not work
[14:26] <diddledan> if you aren't knowingly using multipass then you'll use the old build which builds against your host. if you specify `base: core18` then you will use multipass
[14:27] <ali1234> i want to use 16.04 as base when building on 18.04
[14:27] <diddledan> the solution to use multipass correctly is `env SNAPCRAFT_BUILD_ENVIRONMENT=multipass snapcraft`
[14:30] <ali1234> yeah what happens is if you don't have "base:" in snapcraft.yaml it builds against the host
[14:30] <diddledan> not if you tell it to use multipass
[14:30] <ali1234> if you have "base" it installs multipass for you, then fails to work because it won't accept any version string
[14:31] <diddledan> the only `base` value that is accepted is `core18`. don't set it but tell snapcraft to use multipass
[14:32] <ali1234> if i set base: core18 then it still fails to accept the version string. but doing it your way works
[14:32] <ali1234> it is now launching a VM
[14:36] <ali1234> it seems to be stuck...
[14:36] <mup> PR snapd#6358 closed: cmd/snap-update-ns: close internal lock on error <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <Closed by zyga> <https://github.com/snapcore/snapd/pull/6358>
[14:36] <zyga> ali1234: quick question, is there a reason you don't want to just say "base: core18"
[14:37] <ali1234> yes, the package isn't available in ubuntu 18.04, if it was i'd just install it normally
[14:38] <ali1234> "launch failed: timed out waiting for instance to respond"
[14:40] <ogra> instance failed instantly :)
[14:41] <zyga> ali1234: I'm not sure I understand
[14:41] <ali1234> which part?
[14:41] <zyga> ali1234: using base: core18 will just build your project against ubuntu 18.04 build system
[14:41] <zyga> ali1234: is that breaking your software?
[14:42] <ali1234> i dont have any software
[14:42] <diddledan> o_O
[14:42] <zyga> ali1234: which package is not available in 18.04?
[14:42] <ali1234> eagle
[14:43] <zyga> eagle?
[14:43] <ali1234> yes, eagle
[14:43] <zyga> and is it advailable in 16.04?
[14:43] <zyga> https://packages.ubuntu.com/search?keywords=eagle&searchon=names&suite=all&section=all
[14:43] <ali1234> yes
[14:43] <ogra> the CAD tool ?
[14:43] <ali1234> yes, the CAD tool
[14:43] <zyga> aaah
[14:43] <ali1234> which is non-free
[14:43] <zyga> well
[14:43] <zyga> that's not from the ubuntu archive :)
[14:43] <zyga> what are you trying to build?
[14:44] <ali1234> eagle
[14:44] <zyga> an eagle snap?
[14:44] <ali1234> yes
[14:44] <ali1234> specifically an eagle 6.6 snap
[14:44] <zyga> I think you can use base: core18 just fine, you need to teach your snapcraft.yaml about where to get eagle from
[14:44] <ali1234> the version that is available for ubuntu 16.04
[14:44] <zyga> unless eagle just plainly doesn't work on any other version of ubuntu
[14:44] <ali1234> it isn't available any more
[14:44] <ali1234> the only place it is available is the ubuntu package
[14:45] <diddledan> eagle isn't available for ubuntu 16.04 specifically - you download it from autodesk as a tarball
[14:45] <zyga> I think the error  you are seeing is unrelated to core18 or eagle and it would be the best to migrate to core18 for your software, integrate fetching the eagle .deb (however legal is that)
[14:46] <zyga> I'm not just saying this because core18 is better
[14:46] <ali1234> diddledan: no, it is available as an ubuntu package for 16.04
[14:46] <diddledan> nope
[14:46] <zyga> but because the old build model (building natively) is going away
[14:46] <zyga> snapcraft switched to build with multipass-managed virtual machines
[14:46] <ali1234> https://launchpad.net/ubuntu/+source/eagle/6.6.0-2
[14:46] <mborzecki> or migrate to kicad :)
[14:47] <diddledan> that's not available in xenial
[14:47] <zyga> i386?
[14:47] <ali1234> yes it is: https://launchpad.net/ubuntu/+source/eagle/
[14:47] <zyga> Eagle is a powerful IM client designed to fit in perfectly in an enterprise environment.
[14:47] <ali1234> yes, it's i386 only
[14:47] <zyga> weird
[14:48] <zyga> anyway, I'm back to coding
[14:48] <mup> PR snapd#6357 closed: cmd/snap-update-ns: make freezer mockable <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6357>
[14:49] <zyga> pstolowski: could you do a 2nd review for https://github.com/snapcore/snapd/pull/6345
[14:49] <mup> PR #6345: cmd/libsnap: pass --from-snap-confine when calling snap-update-ns as user <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6345>
[14:52] <ogra> ali1234, regarding your version issue, i can build just fine here using "version: 0" in a snap
[14:52] <ali1234> yes, here it won't accept any version string at all
[14:52] <ogra> did you put any quotes areound the version string or some such ?
[14:52] <ali1234> no
[14:52] <ogra> *around
[14:52] <ogra> very weird
[14:53] <ali1234> what is the most simple possible snapcraft.yaml that should always be guaranteed to build?
[14:53] <ogra> what snapcraft version isthat ?
[14:53] <ogra> $ snap info snapcraft|grep installed
[14:53] <ogra> installed:   3.0.1                (2374) 28MB classic
[14:53] <ali1234> whatever you get with 18.04
[14:53] <ogra> you dont use the snap ?
[14:53] <mup> PR snapd#6345 closed: cmd/libsnap: pass --from-snap-confine when calling snap-update-ns as user <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <Merged by stolowski> <https://github.com/snapcore/snapd/pull/6345>
[14:53] <zyga> \o/
[14:53] <zyga> thanks!
[14:54]  * ogra has long stopped using snapcraft from the archive ... never had issues 
[14:54] <ali1234> snapcraft, version 3.0.1
[14:54] <diddledan> I try to use a snap if one is available unless I hit issues with it that I can't fudge
[14:54] <ogra> well, then i dont get why it doesnt work the same way as mine
[14:55] <ali1234> actually i do have it installed through snap: installed:   3.0.1                (2374) 28MB classic
[14:55] <ogra> probably some other typo in the snapcraft.yaml ? (and the version thingie just being fallout)
[14:58] <ali1234> this is my entire snapcraft.yaml: http://paste.ubuntu.com/p/cQZFmSnm49/
[14:58] <ogra> i doubt thats valid
[14:59] <ogra> better use snapcraft init and modify the created file then
[14:59] <mup> PR snapd#6360 opened: cmd/snap-update-ns: refactor of profile application <Per-user mount ns  🐎> <Created by zyga> <https://github.com/snapcore/snapd/pull/6360>
[14:59] <ogra> you are surely missing a lot of required fields
[15:00] <ogra> (description, confinement ... not sure but probably also grade)
[15:00] <diddledan> `version: 0` is actually invalid because snapcraft (or rather the python yaml parser) reads it as a number when it's expecting a string
[15:00] <ogra> diddledan, it doesnt for me
[15:00] <diddledan> replacing it with `version: '0'` works
[15:01] <ogra> i just took one of my snaps changed version to 0 and called snapcraft ... no errors for me
[15:02] <ali1234> okay i did "snapcraft init" and then "snapcraft" - it's hanging at launching the VM again
[15:04] <diddledan> ogra: with ali1234's yaml: https://www.irccloud.com/pastebin/OeZVLYOb/
[15:05] <ali1234> "snapcraft init" also generates a quoted version field
[15:05] <zyga> niemeyer: the wikipedia reference on that bug is fantastic
[15:05] <zyga> "That looks great. Just one thing: get rid of the duck."
[15:05] <zyga> :"D
[15:05] <ogra> diddledan, well, but that snapcraft.yaml is generalyl broken
[15:06] <diddledan> the rest of the brokenness is besides the point though. snapcraft _expects_ a string.
[15:06] <diddledan> if you managed to get a working build with a number then you've hit a bug
[15:06] <ali1234> i just tried removing the quotes from the generated snapcraft.yaml and it produces the same error
[15:07] <ali1234> the error message could be more helpful.
[15:09] <ali1234> is multipass using docker?
[15:09] <diddledan> no, kvm
[15:11] <ali1234> "multipass launch" works, it is actually doing something
[15:13] <ali1234> i can see qemu using CPU, it's still spinning though
[15:17] <ali1234> nope, failed
[15:20] <ogra> how ?
[15:20] <ogra> (surely different than before ?)
[15:20] <ali1234> nope, identical. it spins for about 8 minutes then says "launch failed: timed out waiting for instance to respond"
[15:22] <ali1234> now running it with verbose
[15:22] <ogra> have you tried a reboot ... or logout/login cycle, after installing multipass ?
[15:22] <ali1234> no
[15:22] <ogra> (did you wiggle the cable ?) :P
[15:23] <ali1234> https://paste.ubuntu.com/p/26DQgRrWKQ/
[15:24] <ogra> wow
[15:25] <ogra> classic snaps ... the pain ...
[15:25] <ogra> looks like it is somehow clashing with your deb-installed qemu
[15:26] <ali1234> i have qemu arm installed
[15:26] <ogra> there is a #multipass channel
[16:12] <ali1234> okay so multipass works now. how do i tell it to stage i386 packages?
[16:12] <ali1234> i tried "--target-arch=i386" but it had no effect. dpkg still only supports amd64 inside the instance
[16:13]  * cachio lunch
[16:15] <ali1234> ah there we go. run snapcraft with --debug, then run "sudo dpkg --add-architecture i386" inside the instance, then exit and run snapcraft again
[16:30] <zyga> I found a bug in go :)
[16:30] <zyga> but that's for next week
[16:30] <zyga> I'm happy I also found a workaround
[16:30] <Crocodillian> I'm trying to install spotify from snap, when it tries to download a core it says x509: certificate has expired or is not yet valid
[16:31] <Crocodillian> is ssl just broken on my system
[16:32] <zyga> Crocodillian: is the time set correctly?
[16:32] <zyga> perhaps some certs *are* invalid though
[16:32] <Crocodillian> no, it looks like my timezone is wrong
[16:32] <Crocodillian> thanks!
[16:39]  * cmatsuoka creates his first wine-based snap package
[16:39] <cmatsuoka> lots of room for improvements tho
[16:42]  * zyga takes a small break
[16:42] <mup> PR snapd#6361 opened: kvm: load required kernel modules if necessary <Created by gerboland> <https://github.com/snapcore/snapd/pull/6361>
[17:22] <zyga> mwhudson: hey
[17:22] <zyga> not sure if you are the person to go to
[17:22] <zyga> but I'm pretty much hopeless now :)
[17:22] <zyga> https://github.com/golang/go/issues/29689
[17:35] <mup> PR snapd#6362 opened: cmd/snap-update-ns: explicitly check for return value from parse_arg_u <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6362>
[17:35] <mup> PR snapd#6363 opened: cmd/snap-update-ns: save errno from strtoul <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6363>
[17:36] <mup> PR snapd#6364 opened: cmd/snap-update-ns: let the go parser know we are parsing -u <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6364>
[17:41] <mup> PR snapd#6365 opened: cmd/snap-update-ns: manually implement isspace <Per-user mount ns  🐎> <Created by zyga> <https://github.com/snapcore/snapd/pull/6365>
[17:41] <zyga> Chipaca: ^ (that last one is cool0
[17:42]  * Chipaca is suspicious of C code being called  "cool0"
[17:42] <Chipaca> zyga: is the C string being passed zero-terminated?
[17:43] <zyga> Chipaca: yes
[17:43] <Chipaca> zyga: does it still crash if you -O0?
[17:43] <zyga> look at https://github.com/zyga/go-bug-29689
[17:43] <zyga> isspace doesn't care
[17:43] <zyga> calling issspace(0) is enough
[17:43] <zyga> or whatever
[17:43] <zyga> ideas welcome :)
[17:44] <zyga> you can also shrink the reproducer even more if you want
[17:52] <Chipaca> zyga: yeah that's not a "minimal" reproducer
[17:52] <mborzecki> zyga: that go bug, doesn't crash here
[17:52] <Chipaca> still, strange
[17:52] <Chipaca> mborzecki: 1.10?
[17:52] <zyga> wait, you mean you can run my reproducer
[17:52] <zyga> and not crash>
[17:52] <mborzecki> 1.11.4
[17:52] <zyga> I triied on 1.10 and 1.11.4
[17:53] <zyga> and my keyboard seems to have some repeat issues sometimes, weird
[17:53] <zyga> mborzecki: go build && ./go-bug 1000
[17:53] <zyga> that doesn't crash for you?
[17:53] <zyga> maybe depends on locale/
[17:53] <mup> Bug #1000: There are too many bug reports in Malone <lp-foundations> <NULL Project:Invalid> <https://launchpad.net/bugs/1000>
[17:53] <mborzecki> uh, ok, that crashes
[17:54] <mborzecki> haven't passed any args before :P
[17:55] <zyga> :D
[17:55] <zyga> README
[17:55] <zyga> but that's good
[17:55] <zyga> it's just broken
[17:56] <mborzecki> hm, idk why but it doesn't crash under dlv, or maybe i'm not passing the command line args again
[17:57] <Chipaca> hm
[17:57] <Chipaca> zyga: if I remove the __attribute__ bit it no longer crashes
[17:58] <zyga> because that code doesn't run anymore
[17:58] <zyga> add printf to check
[17:58] <mborzecki> zyga: must be some weird libc interaction, when built statically it doesn't crash
[17:58] <zyga> perhaps the problem is related to how isspace is implemented
[17:58] <zyga> it uses a lookup table
[17:59] <zyga> perhaps that code runs before various relocations are applied?
[17:59] <zyga> but meh
[17:59] <zyga> not fun
[17:59] <zyga> and makes me think we should split out the C code to snap-run-in-namespace --user --snap-name
[18:00] <mborzecki> zyga: that's from s-u-n right?
[18:00] <zyga> yes
[18:01] <mborzecki> zyga: i recall having some issues with code crashing in isalpha in s-u-n back when i was working on parallel instances
[18:01] <mborzecki> would crash unless i build the binary statically
[18:02] <Chipaca> building it statically has it not working again
[18:02] <zyga> ha
[18:02] <zyga> interesting
[18:02] <Chipaca> zyga: what is it that calls bootstrap()?
[18:02] <zyga> not working?
[18:03] <zyga> Chipaca: the dynamic linker
[18:03] <Chipaca> zyga: as in it isn't called
[18:03] <zyga> look at main.go
[18:03] <zyga> er
[18:03] <zyga> bootstrap.go
[18:03] <zyga> it's explained there
[18:03] <zyga> *MAGIC*
[18:06] <mborzecki> zyga: looking at glibc implementation, macro black magic f****ery, but it looks liek there's an array of points to actual functions that implement each op
[18:06] <zyga> you mean issspace/
[18:06] <zyga> isspace?
[18:06] <mborzecki> zyga: yeah
[18:07] <zyga> it looks like a single function call + array lookup on the result
[18:16] <mborzecki> zyga: right, so it's a LUT indexed by char, the table is in thread specific data (?), and that's set up by some early locale init
[18:17] <mborzecki> zyga: wonder what would hapend if i  add setlocale(LC_ALL, "") somewhere early
[18:18] <mborzecki> zyga: and it works now :P
[18:19] <mborzecki> zyga: https://paste.ubuntu.com/p/X2vgcX5Fr8/
[18:21] <Chipaca> zyga: so
[18:21] <Chipaca> zyga: you got a reply :-)
[18:22] <mborzecki> hah, makes sense
[18:25] <mup> PR snapcraft#2435 closed: Appstream desktop <Created by sergiusens> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/2435>
[18:34] <mborzecki> zyga: something to keep you busy on the flight, dump .preinit_array and .init_array and see where bootstrap ends up and where locale tables are set up
[18:37] <mup> PR snapcraft#2438 opened: rust plugin: new link for rustup <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/2438>
[18:51] <zyga> Chipaca, mborzecki: so we were just lucky so far
[18:53] <mborzecki> zyga: and built statically, cause we had to
[18:53] <zyga> I think I will write a snap-lowlevel-exec
[18:53] <mborzecki> haha
[18:54] <zyga> I'm serious
[18:54] <zyga> this sucks
[18:54] <zyga> we can drop the c code from snap-update-ns this way
[18:54] <mborzecki> zyga: srsly, maybe we should take a look at what podman/docker/crio
[18:54] <zyga> we did
[18:54] <zyga> they do setns the same way
[18:55] <mborzecki> same preinit_array trick?
[18:55] <zyga> at least AFAIK
[18:55] <zyga> I can be wrong
[18:55] <mborzecki> zyga: otoh, there's no pressing need to fix it right away, so we can take time and research
[18:56] <zyga> yeah
[18:56] <zyga> as long as we don't need to call more functions :)
[18:56] <zyga> can you review https://github.com/snapcore/snapd/pull/6365 please
[18:56] <mup> PR #6365: cmd/snap-update-ns: manually implement isspace <Per-user mount ns  🐎> <Created by zyga> <https://github.com/snapcore/snapd/pull/6365>
[18:56] <zyga> I'm still pushing out the user-mount patches
[18:56] <mup> PR snapd#6352 closed: many: remove .user-fstab files from /run/snapd/ns <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6352>
[18:59] <mborzecki> zyga: even if you drop isspace, there's still islower & isdigit in the code
[18:59] <zyga> but it doesn't crash
[18:59] <zyga> perhaps doesn't need a table the same way?
[18:59] <zyga> I'm after getting fixes in place :)
[18:59] <mborzecki> zyga: i suppose you need to use a snap name with instance key
[18:59] <zyga> oh
[18:59] <zyga> I see
[19:02] <mborzecki> wrapping it up
[19:02] <mborzecki> zyga: pedronis: save travel
[19:02] <zyga> thank you!
[19:02] <mborzecki> s/save/safe/ :)
[19:02] <mborzecki> enjoy the summer
[19:02] <zyga> HAHA
[19:02] <zyga> I think I will :)
[19:50] <mup> PR # closed: core-build#11, core-build#22, core-build#26, core-build#37
[19:51] <mup> PR # opened: core-build#11, core-build#22, core-build#26, core-build#37
[19:58] <mup> PR snapd#6366 opened: cmd/snap-discard-ns: fix name of user fstab files <Per-user mount ns  🐎> <Simple 😃> <Created by zyga> <https://github.com/snapcore/snapd/pull/6366>
[19:59] <mup> PR snapd#6367 opened: cmd/snap-update-ns: allow switching to user mount namespace <Created by zyga> <https://github.com/snapcore/snapd/pull/6367>
[20:37] <mup> PR core#38 closed: Add another pi-config option <Created by sergey-borovkov> <https://github.com/snapcore/core/pull/38>
[20:37] <mup> PR core#83 closed: move most of the ubuntu-core config deb into the snap snap build <Created by mvo5> <https://github.com/snapcore/core/pull/83>
[20:37] <mup> PR core#98 closed: Add force_turbo rpi option <Created by sergey-borovkov> <https://github.com/snapcore/core/pull/98>
[20:38] <mup> PR core#38 opened: Add another pi-config option <Created by sergey-borovkov> <https://github.com/snapcore/core/pull/38>
[20:38] <mup> PR core#83 opened: move most of the ubuntu-core config deb into the snap snap build <Created by mvo5> <https://github.com/snapcore/core/pull/83>
[20:38] <mup> PR core#98 opened: Add force_turbo rpi option <Created by sergey-borovkov> <https://github.com/snapcore/core/pull/98>
[20:55] <diddledan> is mup ok?
[20:55] <diddledan> 4 successive "this was closed.. oh no, psych! it's open really :-p" messages
[21:00] <mup> PR snapcraft#2439 closed: snap: add xslt dependencies for lxml <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/2439>
[21:03] <mup> PR snapcraft#2439 opened: snap: add xslt dependencies for lxml <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/2439>
[21:19] <mup> PR snapd#6368 opened: tests: check denials considering all the log lines instead of last 100 lines <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/6368>
[22:22] <om26er> popey, ping
[22:48] <mup> PR snapd#6369 opened: Add check for snap binaries dir not being in path <Created by liamg> <https://github.com/snapcore/snapd/pull/6369>