[06:10] morning [06:34] mvo: hey [06:38] mborzecki: hey, good morning [06:40] mvo: in the context of https://forum.snapcraft.io/t/apps-not-in-desktop-search-menu/15275 do you recall why we didn't add a user env generator? [06:41] mborzecki: hm, I thought we had a generator [06:41] mvo: we have a system-env generator [06:41] mborzecki: indeed, I think there is no reason then [06:42] mborzecki: maybe it was because older systemds do not support this yet? [06:42] mvo: mhm, i'll add one and we can ship it selectively if needed [06:44] mborzecki: +1 [07:04] Hello :-) [07:05] zyga: hey [07:07] hey zyga [07:15] zyga: do you still have that opensuse vm? https://forum.snapcraft.io/t/error-while-loading-shared-libraries-on-opensuse-tumbleweed/15285 [07:17] PR snapd#8073 closed: daemon: drop support for the DELETE method [07:53] mborzecki: sure [07:53] * zyga is in the office now [07:54] weird [07:56] I'll try [07:57] mborzecki: offtopic, stuff from weekend: http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2018/p1095r0.pdf [07:59] mvo: have you seen https://bugs.launchpad.net/ubuntu/+source/snapd/+bug/1861648 [07:59] Bug #1861648: When booting 20.04 an 'ld-2.23.so' process consumes 100% CPU for minutes [07:59] mvo: could that be fontconfg? [07:59] (or related?) [08:01] morning [08:02] hey pawel, welcome back [08:02] pstolowski: hey, all rested? [08:03] zyga: it could be, slightly strange but possible [08:03] pstolowski: hey, welcome back! we missed you :) [08:03] pstolowski: (in a good way) [08:04] `error: too early for operation, device not yet seeded or device model not acknowledged` wonder if we could prouce something more friendly instead of this error message [08:04] mborzecki: oh yeah [08:04] I saw that a few times [08:04] and it's very precise and very useless [08:05] zyga: annoying when it happens on `snap install` right after you installed the snapd package [08:05] yeah, I wonder if snap from command line could block and hold while stuff like that happens [08:05] but that's a separate concern from just making a readable message [08:06] like "snapd is performing first-boot setup process, please wait (10%)" [08:07] mborzecki: updating suse (882 updates over weekend) [08:07] mborzecki: I'll try those snaps in a moment [08:07] hey guys! yep, rested, i had great time with my family; i only wish there was more snow [08:08] pstolowski: I think you had more snow than the rest of poland :) [08:08] but yeah, I wish this winter was better [08:12] mborzecki wow [08:12] type=AVC msg=audit(1580717508.948:224): apparmor="DENIED" operation="mknod" profile="/usr/lib/snapd/snap-confine" name="/sys/fs/cgroup/devices/snap.poddr.poddr/cgroup.procs" pid=4155 comm="snap-confine" requested_mask="c" denied_mask="c" fsuid=0 ouid=0 [08:12] type=AVC msg=audit(1580717508.948:225): apparmor="DENIED" operation="open" profile="/usr/lib/snapd/snap-confine" name="/sys/fs/cgroup/devices/snap.poddr.poddr/cgroup.procs" pid=4155 comm="snap-confine" requested_mask="wc" denied_mask="wc" fsuid=0 ouid=0 [08:13] zyga: yeah.. but temperatures revolving around 0C (only) were not expected [08:13] mborzecki: we have ... denials? [08:13] ah. wait, maybe I have two profiles [08:13] one left over from work [08:13] sigh [08:16] mborzecki: replied there, I think leap may be broken, TW is okay [08:27] I need a 2nd +1 for https://github.com/snapcore/snapd/pull/8033 [08:27] PR #8033: tests: switch mount-ns test to differential data set [08:27] I'd like to merge it before it drifts [08:45] mvo: how was last week? anything remarkable? [08:45] pedronis: morning! [08:47] pstolowski: hi [09:08] it'd be so nice if systemd manpages indicated whic version of systemd given feature first appeared in === pedronis_ is now known as pedronis [09:19] pstolowski: I will look at the preseed PRs later today, I fixed some conflicts on the first on Friday. [09:20] pstolowski: maybe you could finish reviewing #8036 ? [09:20] PR #8036: snapstate: refactor things to add the re-refresh task last [09:20] pedronis: sure. and thanks for updating preseed PR [09:24] hmm, so we have /lib/environment.d and /etc/profile/snapd.sh but environment for zsh user is still not getting updated [09:35] mborzecki: hi, I can do a follow up about that name, if there's a preference [09:36] pedronis: nah, it's ok, not worth another spread run [09:38] zyga: do you recall whether it's possible to start a user session in the tests somehow? [09:46] re [09:46] mborzecki: I think it is [09:46] mborzecki: I've done something -ish- like that [09:46] mborzecki: well, I didn't start a full user session but I did start a dbus session bus [09:46] mborzecki: what I found out is that some PAM / distro / i-dont-know-what settings do start a user session for the test user [09:46] mborzecki: because we observed that it runs stuff and shuts down asynchronously [09:47] quick errand, back in 30 or so [09:48] lp timeouts [09:48] oh well [09:48] back to work [09:50] pstolowski: thanks [09:50] PR snapd#8036 closed: snapstate: refactor things to add the re-refresh task last [09:59] pstolowski: last week was interesting, 2.43 a wee bit delayed but hopefully just 1 week [10:11] mvo: uhm, just browsed last-week standup notes (they are very useful to catch up btw!) [10:11] pstolowski: yeah, when at the sprint I had the same feeling, very useful [10:13] PR snapd#8080 opened: dirs: manjaro-arm is like manjaro [10:14] mvo: in case 2.43.x will happen ^ [10:15] zyga: thank you [10:28] re [10:49] * zyga goes offline for 2-3 hours [10:58] PR snapd#8081 opened: tests/main/user-session-env: add test verifying environment variables inside the user session [10:58] simple test ^^ [11:17] still, something is off with zsh as indicated in https://bugs.launchpad.net/ubuntu/+source/zsh/+bug/1640514 [11:17] Bug #1640514: /snap/bin is not added to the PATH when using zsh Xenial):Confirmed> [11:39] zyga: little tweak in #8080 [11:39] cachio: hey, welcome back! [11:39] PR #8080: dirs: manjaro-arm is like manjaro [11:41] mborzecki, heu, thanks [11:42] mborzecki, trying to start again :) [11:47] cachio: hi! have you seen my earlier email re nested vm, and comments under #8046 ? [11:47] PR #8046: many, tests: integrate all preseed bits and add spread tests [11:47] pstolowski, no yet [11:47] pstolowski, hello :) [11:48] cachio: ah, you were on vacation as well! [11:48] pstolowski, yes :) [11:49] caching up emails, logs and other stuff yet [11:49] pstolowski, I'll take a look to the PR [11:49] today for sure [11:50] cachio: i'll deconflict it in a moment [11:54] PR snapd#8082 opened: data, packaging: Add sudoers snippet to allow snaps to be run with sudo [12:05] cachio: hey, welcome back! [12:05] mvo [12:05] hello [12:05] PR snapd#8033 closed: tests: switch mount-ns test to differential data set [12:05] mvo, reading logs for 43.2 [12:06] mvo, is .1 going to stable? or we should go with .2? [12:07] cachio: we will need .2 instead, it's already in beta [12:08] cachio: so once beta validation is done it can go into candidate [12:08] cachio: and then we can release to stable in 1 week [12:08] mvo, ok [12:09] mvo, i'll continue with the beta validation in that case [12:10] mborzecki: thank you! [12:10] cachio: welcome back :) [12:10] cachio: thank you! [12:11] zyga, hi [12:19] blergh [12:19] * Eighth_Doctor waves [12:20] it's been a while since I've had to send a patch snapd upstream :P [12:20] mvo: it's been a while :) [12:22] mvo, zyga: by the way, I'm surprised you guys haven't reaped the Ubuntu 14.04 packaging from git [12:22] Eighth_Doctor: it's still supported [12:22] don't tell me you guys are still building for Ubuntu 14.04?! [12:22] Eighth_Doctor: have you heard of ESM? [12:22] Eighth_Doctor: we may have to [12:22] blergh [12:22] https://ubuntu.com/esm [12:23] somebody better giving you a huge bag of money for that [12:23] "Security updates for 14.04 LTS until 2022" [12:23] ugh [12:23] Eighth_Doctor: it's free for personal use [12:24] I can see that, and that deeply concerns me :( [12:24] why? [12:24] disincentive for upgrading [12:25] Eighth_Doctor: it's enterprise stuff [12:25] you can upgrade when you are ready [12:25] Eighth_Doctor: I'm sure that using ESM and paying for it [12:25] yeah, yeah [12:25] Eighth_Doctor: versus using the update for free [12:26] Eighth_Doctor: is a good incentive [12:26] plus, it keeps developers employed :) [12:26] heh, sure [12:26] the only difference between RH/SUSE and this model is that there's a "the first hit is free" thing like Oracle does... [12:27] I just think it probably introduces a stronger disincentive when it's possible to get it for free [12:27] zyga: I don't have a problem with the ESM thing, just the fact it is available for free for personal use [12:27] ah [12:27] well [12:28] you can always pay :) [12:28] you get cool stuff [12:28] haha [12:28] I don't have bags of money [12:28] like kernel live patch [12:28] it's not expensive [12:28] the livepatch stuff is interesting though [12:28] and you get a discount because you met zyga in person ;) [12:28] * Eighth_Doctor still has his legacy KSplice free subscription for Fedora [12:29] haha [12:29] sadly, I know how the sausage is made for kernel livepatches [12:29] shhhh [12:29] I try very hard to avoid them if I can ;) [12:30] * Eighth_Doctor was once asked to make kernel livepatches using kpatch [12:31] zyga: if I could afford to, I'd buy subs from all my favorite Linux distro companies :) [12:31] Eighth_Doctor: if you want to support, $25 gives you desktop ESM license [12:31] nice [12:31] you get fips, livepatch, some other fancy acronyms [12:31] I'll consider it [12:31] probably desirable in corporate world [12:31] for somewhat more you get phone support [12:32] prices for servers and vms are different though [12:32] but yeah [12:32] you can just use it for free because it's pretty damn cool :) [12:32] and it's meant to be adopted by personal users [12:32] because that makes everyone safer [12:32] fair [12:32] zyga: if you guys offered Fedora support... :P [12:33] Eighth_Doctor: you can run it in a container :) [12:33] hah [12:33] I don't think I've run Ubuntu as my desktop OS since 2012 [12:33] Eighth_Doctor: but I think if you want to start supporting fedora IBM will look at you with The Eye [12:33] actually, they're super weird about this [12:33] Eighth_Doctor: I heard 20.04 is looking like a sweet system :) [12:33] they want more people to commercially support Fedora [12:33] pstolowski: answered your question in 7590 [12:34] thx [12:34] but some folks give you the stink-eye when you do it for CentOS [12:34] Eighth_Doctor: are you ubuntu community member? you can get it on up to 50 machines for $0 [12:34] what does that mean? [12:34] Eighth_Doctor: the fedora/centos/rhel split is still making me dizzy [12:34] I have an Ubuntu account on LP [12:35] zyga: it's going to get less confusing in the near future [12:35] Eighth_Doctor: fedora/rhel would be better, if you had a way to get no-cost unsupported rhel [12:35] abyway [12:35] * zyga needs to focus [12:35] fedora -> centos -> rhel [12:35] fedora -> rhel -> centos? [12:35] or more like [12:35] going forward, fedora will branch into centos, and centos will branch into rhel [12:35] fedora <-> rhel <-> centos <- to fedora -> [12:35] that's starting with CentOS 8 [12:35] with the CentOS Stream [12:36] which is essentially what RHers will tell you is rhel-rawhide [12:36] so Fedora is where the cool shit happens, CentOS Stream is where it stabilizes, RHEL/CentOS Linux is the end state [12:37] over the next ~6 months, the RHEL development process is being flipped public as part of the CentOS Stream infrastructure bringup [12:38] zyga: on another note, looking at my LP account, I created it shortly after Ubuntu 5.10 (breezy badger) released [12:38] pedronis: right, ty, i forgot it's for core [12:39] zyga: it actually predates my FAS account and my openSUSE account by two years ;) [12:39] https://launchpad.net/~ngompa13 [12:43] cachio: can you take a look whther centos-8 is avaialble in gcp already? [12:44] mborzecki, sure, let me check [12:44] ijohnson: hi, already around? :P [12:45] mborzecki, it is available now [12:45] cachio: great, would it be possible for you to add it to the project then? [12:45] do you weant a pr to enable it ?? === ricab is now known as ricab|lunch [12:45] mborzecki, sure [12:45] cachio: i can open a PR once you tell me the image name [12:46] * pstolowski lunch [12:47] mborzecki, try with centos-8 [12:47] cachio: thx [12:47] mborzecki, but I would need to add it to our pool [12:48] mborzecki, but in the meantime you can use that one [12:54] cachio: yay, it works - centos-7-64: [12:54] workers: 4 [12:54] image: centos-7-64 [12:54] PRETTY_NAME="CentOS Linux 8 (Core)" [12:55] mborzecki, centos 7? [12:55] cachio: nah, wrong paste ;) [12:56] cachio: see the PRETTY_PRINT=.. line [12:58] mborzecki, which is this PRETTY_NAME ? [12:58] mborzecki, where do you see it? [12:59] cachio: PRETTY_NAME="CentOS Linux 8 (Core)" <-- [13:00] PR snapd#8083 opened: spread: add CentOS 8 <⛔ Blocked> [13:05] mborzecki, does not work for us that version? [13:05] mborzecki: not quite yet :-) [13:06] mborzecki, well you said that works [13:06] cachio: i was able to boot a node, so a good sign (?), let's how spread tests will fare [13:06] ijohnson: haha, sure will poke you in a bit [13:07] mborzecki, nice, the only change we are doing to centos is to set selinux as permissive [13:49] PR snapcraft#2893 closed: [legacy] meta: include environment in hook wrappers [13:49] morning folks [13:49] PR snapd#8084 opened: many,randutil: centralize and streamline our random value generation [13:55] hey ijohnson ! [13:55] hey pstolowski did you have a good vacation ? [13:57] ijohnson: yep, thanks for asking! [13:57] nice, I saw your picture of the mountains with the snow, very beautiful :-) it actually got really warm here over the weekend and a bunch of our snow melted [13:59] ijohnson: https://drive.google.com/file/d/1g9I2qGgGfpgYCyLWCHbrXlVzRnrRR_tN/view?usp=sharing [14:00] zyga,mborzecki ^ morskie oko :) [14:02] mmm [14:02] I want holidays with snow [14:03] pstolowski: awesome, I'm jealous we don't have mountains nearby me :-) [14:10] mvo, mborzecki: I think this is officially a sign that snapd has too many tests [14:10] when the tests fail because it takes too long to run (!) === ricab|lunch is now known as ricab [14:12] Eighth_Doctor: in a meeting right now, sorry, will get back to you [14:14] Eighth_Doctor: hmm probably snap downloads taking too long [14:15] happens from time to time [14:26] stgraber, so i tried to build a preseeded image with lxd ... which seems to work fine but then console-conf doesnt allow me to add a default user to the core image at all anymore, seems lxd created a user that console-conf now considers the images system user ... https://imgur.com/a/fqu7xrS [14:29] stgraber, how did you test your daemon.preseed change ? i cant really verify it that way ... [14:33] (this is most likely a console-conf bug, to not ignore such users (i bet docker is the same here) i'm just wondering how you verified the poatch without being able to log into the image) [14:34] *patch [14:35] mborzecki, do you need permisive mode on centos-8? [14:36] cachio: yes, we'll probably need it because most tests are not taking care of the labeling properly [14:36] mb [14:36] mborzecki, ok [14:38] pedronis, see above .. if i add lxd to required-snaps in the model (which creates an lxd user in the extrausers db), console-conf doesnt offer me to create a default user ... i assume snapd now considers the image managed, yet it isnt ... would that be a snapd bug, a console-conf one or ... ? [14:39] * ogra isnt sure if console-conf queries for "managed" here [14:39] (or if that is even the right conclusion) [14:40] I figured out why mi microphone was not working [14:40] I was playing with the wireless headphone charging case [14:40] and flipping it open pairs the devices [14:40] so I kept switching the microphone back and forth [14:40] and that apparently confused hangouts [14:54] hangouts gets confused quite easily [15:04] it's so confused it isn't even called hangouts anymore 😆 [15:05] * zyga goes to check on his son [15:05] Janek has fever and stayed at home today [15:06] zyga: my sympathies and a bowl of warm soup - fever sucks [15:21] yeah, we just gave him some meds to reduce the fever as it's been too long (since 6AM) [15:40] PR snapd#8085 opened: [RFC] netutil: add default gateway monitor [15:40] mvo: looking [15:45] ijohnson: 8082 hit the timeout again? [15:45] mborzecki: uh yes [15:46] did you already restart it before ? [15:46] ijohnson: yeah :/ [15:46] I've got a bad feeling about this :-( [15:46] ijohnson: let's see, maybe 3rd time's the charm :P [15:47] mvo: reviewed [15:47] mborzecki: fingers crossed [15:48] mborzecki, centos-8 is failing to install yum install system-lsb-core [15:49] mborzecki, perhaps we could install redhat-lsb-core instead [15:50] mvo: do you have an LTE modem? [15:50] mvo: I think it would be great to test it on one [15:50] mvo: if you need I can test it locally with mine [15:54] * cachio lunch === heather is now known as hellsworth [16:13] zyga: I have one, haven't used it in a while, can do the test in a wee bit [16:30] back from lunch [16:33] mvo: cool, let me know if you want assistance, I can help as well [16:56] PR snapd#8072 closed: daemon, store: better expose single action errors [17:02] mmm I've got a really bad feeling about these tests that depend on the store [17:03] ijohnson: [17:04] :-) === pedronis_ is now known as pedronis [17:39] * zyga takes a coffee break [17:53] anybody have an idea why all of a sudden LP jobs can't git clone snapd ? https://pastebin.ubuntu.com/p/CDJG8DVyPV/ [17:54] hmm I guess I can't clone the master branch in LP either [17:59] yes git.launchpad.net is down, see #is [17:59] ijohnson: ^- [17:59] cjwatson: ah thanks [17:59] (BTW I only see stuff here by chance - if you want LP staff to see stuff it's more reliable to ask in LP channels) [18:00] cjwatson: I checked is-outage and didn't see anything about it in the topic, perhaps #is is a better place to check [18:00] #is-outage isn't a place where customers of IS are supposed to report things, normally [18:00] so I don't :) [18:01] cjwatson: fair enough :-) also did you mean the internal launchpad channel or freenode one? [18:01] Either [18:01] #launchpad-ops on irc.c.c, #launchpad on freenode [18:02] Anyway, should be fixed soonish [18:02] great, thanks [18:43] PR snapd#8082 closed: data, packaging: Add sudoers snippet to allow snaps to be run with sudo [19:51] ijohnson: I did another pass on 8077 [19:53] pedronis: thanks responding now [20:00] ijohnson: I'm not sure whether revisions should error on incosistent trySnap vs status [20:00] in general we should try not to get stuck [20:01] I guess it would be a question of where we would get stuck [20:01] ijohnson: if we are trying or try but there's no try kernel [20:02] I suppose we behave like status is just = "" [20:02] we should clean it up [20:02] we would get stuck in InUse, and GetCurrentBoot it seems [20:02] yea [20:02] so no upgrades [20:02] we should probably not error [20:02] I think cleaning up in GetCurrentBoot makes sense, but not so sure about InUse, seems odd if InUse auto-cleans things [20:02] but make it a task for mark successful [20:02] to clean it up [20:02] ijohnson: well we always do a mark sucessful pass [20:02] before doing other things [20:03] it's run early in snapd [20:03] hmm [20:03] that would make sense [20:03] just to clarify, when you say "make it a task for mark successful" you don't mean an actual manager task right? just that mark successful should clean up ? [20:03] yes [20:03] ok [20:04] I think it has code already in that direction [20:04] but in uc20 [20:04] we have links and status [20:04] that are not quite in the same spot [20:04] maybe the code is already there [20:04] just a matter of having a test [20:05]