=== JoshStrobl is now known as JoshStrobl|zzz [01:41] kyrofa snapcraft#1581 should be fine [01:41] PR snapcraft#1581: catkin plugin: support rosdep pip dependencies === chihchun_afk is now known as chihchun === tai271828_ is now known as tai271828 === tai271828 is now known as taihsiang === taihsiang is now known as tai271828 === chihchun is now known as chihchun_afk [04:18] Bug #1577471 changed: Snap applications cannot access the user's normal XDG directories [06:26] PR snapd#3996 opened: snap: refrain from running filepath.Base on random strings [06:34] PR snapd#3997 opened: cmd/snap: completion for alias and unalias [06:43] PR snapd#3998 opened: snap-confine, snap-seccomp: Utilize new seccomp logging features [07:02] PR snapd#3991 closed: wrappers: fail install if exec-line cannot be re-written [08:04] yay, my rpis3 should arrive on friday [08:08] pstolowski: nice :) [08:08] pstolowski: did you get any cases? [08:08] pstolowski: I have a few and I can share experiences [08:10] zyga-ubuntu, yeah, I got 2 different cases - https://www.amazon.de/gp/product/B01F1PSFY6/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1 and https://www.amazon.de/gp/product/B00MQLB1N6/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1 [08:11] pstolowski: I have the first one [08:11] pstolowski: it's easy to assemble and disassemble but doesn't stack [08:12] the second one looks very nice [08:12] right, it's curved [08:58] * zyga-ubuntu -> walk [09:32] re [09:32] ok, now just focus on work :) [10:27] <__chip__> zyga-ubuntu: as opposed to focus on what? [10:28] __chip__: distractions that plague this morning [10:29] heh [10:29] <__chip__> zyga-ubuntu: like https://guacamole.incubator.apache.org ? [10:30] * __chip__ might be eviler than usual today [10:30] hmm? [10:30] what is that? [10:31] interesting software I see === JoshStrobl|zzz is now known as JoshStrobl [12:25] morning all [12:26] Son_Goku: hey [12:26] Morn [12:27] ikey: you're alive! [12:27] yessir - hows it? [12:31] ikey: I got sick literally the moment I got home [12:31] yikes [12:33] my nose is completely stuffed and I have a constant headache :( [12:35] Son_Goku, welcome to the club [12:36] * zyga-ubuntu is sick just because of weather, not any fancy travel smavel [12:36] well, I think I'm sick because of the weather rather than anything else [12:36] I didn't travel very far... [12:38] 7C here atm === JoshStrobl is now known as JoshStrobl|Work [12:41] * ogra_ wonders if xnox has an idea about https://forum.snapcraft.io/t/networkd-fail-to-set-ip-address-between-leases-if-ip-address-changes/2333 ... being fluid in system-networkd [12:42] ikey: 12C here [12:42] but it was 8C a few hours ago [12:44] pstolowski: hey, would you mind looking at https://github.com/snapcore/snapd/pull/3971 [12:44] PR #3971: interfaces/mount: make Change.Perform testable and test it [12:46] zyga-ubuntu, will do [12:49] nacc are you registered on the forum? [12:49] pstolowski: thanks, there's also a small second branch if you want to look [12:51] zyga-ubuntu, so, when are you going to start writing the SELinux backend for snap-confine? :) [12:53] ogra_, i need a dump of /run/systemd/netif/leases/* during all stages (one lease, two leases, first expired, both expired), in a launchpad bug report filed with ubuntu-bug systemd [12:54] xnox, no "ubuntu-bug" on core [12:54] config for the dhcp server / interactive script to reproduce would be helptful too. [12:55] ogra_, if a bug needs to be looked into a bug report is needed [12:55] Son_Goku: not sure, I need to finish layouts still, though I'm not alone anymore so that's a big relief [12:55] xnox, well, the forum kind of became the bugtracker for snap, but i can ask him to file a LLP bug [12:55] -L [12:56] sure but this particular part (systemd-networkd) comes from a package in ubuntu, for which bug tracker is launchpad.net/ubuntu/+source/systemd [12:57] and lp:~ubuntu-core-dev/ubuntu/+git/systemd git repository [12:57] sure [12:58] note that there are PPAs for backported packages involved (netplan specifically) [12:59] (also ... i think core is the only flavour that makes use of networkd in xenial) [13:01] zyga-ubuntu, I can try :). nb, I've adressed your review feedback for early-sanitize [13:01] pstolowski: I saw, thank you [13:01] ah, you replied, just noticed === ShalokShalom_ is now known as ShalokShalom [13:30] niemeyer: oh, one more thing [13:30] niemeyer: we need manjaro icon on https://snapcraft.io/ [13:31] zyga-ubuntu: this is coming today/tomorrow [13:35] Mister_Q, should the UBports installer site there for a long time in "Ready to install..." (trying to switch my pro5 to ubports via the ubports-installer snap here) [13:36] *sit there [13:36] (seems to not move at all) [13:37] * zyga-ubuntu -> break for lunch and back to small PRs that improve snap-confine [13:37] PR snapd#3999 opened: cmd/snap-confine: add detection of stale mount namespace [13:46] zyga-ubuntu: And also that other one... [13:52] kyrofa when you are up, can you read through this thread so we can have conversation later in the day on where/how this https://forum.snapcraft.io/t/classic-snaps-failing-on-ubuntu-17-10/2324/37 ? [13:54] niemeyer: indeed :) [13:54] niemeyer: I'll note to do that [13:54] re :) [13:55] zyga-ubuntu: Although I haven't really found good references to snaps in the web page, so maybe let's not add until there's something we can explicitly point to [13:55] niemeyer: I think having install instructions is a must, I'll install them and check in the evening [13:55] I need to find a moment to swap my HDD, I bought a 2TB HDD as update to my laptop [13:56] (I ran out of VM space) [13:56] zyga-ubuntu: I wouldn't worry about it.. layouts are way more important [13:56] i know the snap-update-ns changes bricked solus btw [13:56] i was toying with git master last week [13:56] niemeyer: ok :) [13:56] ikey: oh, can you tell me more? [13:56] ya basically i was getting no such file or directory error from it [13:56] when it was invoked from snapd [13:56] aha... [13:56] (i straced which is how i know it came from snap-update-ns itself) [13:57] ikey: is that on current solus package? I investigate [13:57] ikey: do you have the strace somewhere? [13:57] nah not on the current pkg [13:57] we're at, um [13:57] * ikey looks [13:58] 2.27 [13:58] ill need to rebase it to newer master anyway as some of my PRs are in [13:58] and i gotta figure out what strays i have left from ubuntu rally [13:58] ok [13:58] ikey: Thanks for letting us know. As an aside, please note we're working on major improvements around mounting flexibility so we can support layouts and more. Following git too closely may be a bumpy ride for a while for that reason. [13:58] ikey: let me know if you have some details [13:59] ikey: strace is one thing [13:59] ikey: but golang and strace don't work very well AFAIK [13:59] ikey: you may need some prints here and there [13:59] niemeyer, yeah this was on-the-day-git-master [13:59] zyga-ubuntu, thats what im thinking [13:59] just printf the hell out of it [14:00] but if git is rocky right now then ill leave it a little while longer [14:00] zyga-ubuntu: strace shouldn't care about language [14:00] strace will just follow the underlying syscalls [14:00] zyga-ubuntu: You may be thinking of ltrace? [14:01] niemeyer: hmmm, last time I tried strace and golang didn't work at all, maybe it is because of the threads and I dind't specify -f and I assumed it's something broken somewhere [14:01] ikey: Yeah, I suggest following the releases.. once it reaches candidate we're pretty sure we have a winner [14:01] yeah [14:02] then i can get the last of the needed solus PRs together [14:02] and 2.29 would be full upstream solus support [14:02] \o/ [14:03] https://www.irccloud.com/pastebin/zJXwdBQ8/ [14:03] zyga-ubuntu: ^ [14:04] zyga-ubuntu: It wouldn't really make sense for it not to work.. this traces system calls, not application logic [14:04] zyga-ubuntu: It will likely be much noiser though, due to the runtime [14:06] Good morning! [14:07] elopio: o/ [14:07] I hope everybody is safe back home now. It was nice to meet you all last week. [14:08] zyga-ubuntu: and we missed you. Hopefully next time it won't be in the US [14:08] elopio: or I will be able to come [14:10] A different country will be better 😃 [14:11] costa rica maybe? [14:12] +100 [14:13] zyga-ubuntu: https://github.com/nomad-desktop/nx-software-center [14:14] nx = nixos right? [14:14] er [14:14] nxos [14:14] the kde desktop that looks like budgie but isnt [14:14] nixos [14:14] using the nix package manager [14:14] right its the one that isnt that one [14:14] zyga-ubuntu: Or Puerto Rico.. I've heard the gov doesn't mind as much there.. [14:14] * ikey gets confused on that too [14:15] https://nxos.org/#nomad-desktop [14:15] there, nxos [14:15] the not-budgie-budgie [14:15] https://nixos.org/nix/ [14:15] :P [14:15] not confusing at all [14:15] lol [14:16] ikey: Yeah.. I was surprised to find out about it.. seems like a curious mixture.. claims it's based on Ubuntu, but no debs, pacman, plus snaps [14:16] Wonder if the author is around here [14:16] wonder if they've jumped on the ostree bandwag.. wait.. pacman? [14:16] what the shizz [14:16] Yeah, don't ask me [14:17] pls excuse me while i find a corner to cry in [14:18] ikey, funnily i was looking at nixos before because niemeyer and sergio were discussing their patchelf tool (which is a core part of the nix package manager) on the forum [14:18] yeah the nixos approach honestly confuses me [14:19] seems to be nix and nx day today :) [14:19] (despite bein german re-unification day too :) ) [14:19] instead of actually addressing the root cause of statelessness they worked around it with yocto style sysroots and reinvented subvolumes [14:19] and pattching ninaries [14:19] *binaries [14:19] niemeyer: *nice* [14:20] ninaries [14:20] lol [14:20] heh [14:20] hello jetlag :) [14:20] my old friend ... [14:20] we are at the golden era of OS development [14:20] you sneeze and wham, an OS pops out [14:20] accompanied by a couple dozen symlink farms [14:20] and it has access to all the apps :) [14:21] ikey: ninaries sounds great! [14:21] ikey: it's like a bit, but has nine states [14:21] well i mean thats kinda the "dream" right? package manager is relegated to an implementation detail [14:21] nessita, bitField ^= &derpFace ? [14:21] * zyga-ubuntu thinks "90s" [14:21] er niemeyer * [14:22] somethin somethin jetlag.. [14:22] this is what i get for poking fun. instant karma. [14:22] ikey: Is there any package manager that isn't relegated to an implementation detail? :P [14:22] pacman [14:22] its a badge of honour [14:22] a way of life [14:22] LOL [14:22] I should work on that then [14:23] Because apparently we all work very hard so that our work eventually becomes invisible :) [14:23] yea sorta backfired there on the fame plan.. [14:23] Yeah :P [14:23] niemeyer: if you want to work on something visible [14:23] niemeyer: we should make snap support wallpapers [14:23] ^_^ [14:24] yes. [14:24] Or we just make it crash a lot [14:24] black and white animated tiffs. [14:24] hahaha [14:24] "why is gnome-shell using 48GB RAM" [14:24] ikey: because it can :> [14:24] oh burn [14:24] lol [14:24] it's like vista [14:24] yes [14:25] it sucks [14:25] ram [14:25] literally [14:25] but then you have free ram in laptops [14:25] panic: snapd word undetected by microphones in last 10 days [14:25] the henry hoover of desktops. :p [14:25] not like nowadays when windows 10 makes 4GB laptops common because it's one single time when windows actually improved memory usage so you don't need 8GB in baseline models [14:26] ah damn you ubunturally. [14:26] i got still-face-derp-faced from the chris fisher vlogs [14:26] er stillframe* [14:26] https://plus.google.com/114713706129194876663/posts/PfL3dKtdMXc :( [14:27] "a face you can trust" [14:30] OMG [14:30] can you edit that to prevent mental burn-in ;D [14:32] XD [14:33] I have something appropriate [14:38] sergiusens, alright, I'm up to speed [14:51] PR snapd#3973 closed: cmd/snap-confine: put processes into freezer hierarchy [14:57] elopio, are you in today? [15:03] sergiusens: probably not :/ i'll fix that today [15:05] sergiusens: done, as nacc [15:05] great, in any case, you might be interested in https://forum.snapcraft.io/t/classic-snaps-failing-on-ubuntu-17-10/2324/ [15:05] nacc ^ [15:12] kyrofa: I am here, yes [15:13] sergiusens: thanks will read [15:13] sergiusens: i would highly recommend tests added to fix this [15:14] nacc it might be hard in adt, but certainly doable on travis [15:14] elopio, when you get a few minutes, can I get reviews on snapcraft#1581 and snapcraft#1582? [15:14] PR snapcraft#1581: catkin plugin: support rosdep pip dependencies [15:14] PR snapcraft#1582: plugins: add ros2 boostrapper [15:14] kyrofa (and elopio if you want) hangout? [15:15] sergiusens, sure thing [15:15] the weekly [15:15] omw [15:17] The gardener is cutting the grass just now. You won't be able to hear a thing. [15:17] I can jump in there and listen. [15:21] so... i want to make a snap that comprises a daemon that runs all the time, and a command that talks to the daemon via a unix-domain socket. what's the best way to structure that? === rogpeppe1 is now known as rogpeppe [15:23] do i need to declare a new interface and have two apps in my snapcraft.yaml that use it? [15:25] is network-bind sufficient to create a unix-domain socket? [15:55] rogpeppe, yeah, use network and network-bind [15:55] You should be set [15:56] kyrofa: that will allow me to make a unix-domain socket in $SNAP_COMMON? [15:56] rogpeppe, yep. Although I suggest putting things like that in /tmp [15:56] kyrofa: i saw the listen-stream attribute in https://docs.snapcraft.io/build-snaps/syntax#apps-and-commands but that doesn't appear to be an actual thing [15:57] kyrofa: i suspect the docs might be out of date [15:57] kyrofa: so, do all apps declared in snapcraft.yaml share the same /tmp ? [15:57] rogpeppe, ah, that may be new actually [15:57] rogpeppe, yes indeed [15:58] kyrofa: do they all share the same SNAP_COMMON too? I'm a bit confused as to what an app actually is actually. [15:58] rogpeppe, although it's namespaced when looking at it from outside the snap, so you probably don't want that if you ever want access from outside [15:58] kyrofa: in this case, it's quite nice if only the command can get access to the socket actually [15:58] kyrofa: so if I just listen on /tmp/socket and dial /tmp/socket, then i should be good? [15:59] rogpeppe, a snap is just a big blob of bundled binaries. You define how the user interacts with it using `apps`, which allows you to define either services or commands that are called by the user [16:00] rogpeppe, yep, should work. That's what I do [16:00] kyrofa: ok, cool. it might be nice if the docs mention that "network" means unix sockets too [16:01] rogpeppe, it looks like listen-stream and socket are for socket activation, so those features must be new [16:02] kyrofa: months old iirc [16:02] davidcalle, hahaha [16:03] sergiusens: do you recall? I think you added these ^ [16:03] (in the docs I mean) [16:05] davidcalle not me, it was someone working on lxd that proposed a new implementation for sockets [16:06] sanpcraft 2.0 never had socket support, snap.yaml had it hidden though [16:08] davidcalle here's the long thread for it https://forum.snapcraft.io/t/socket-activation-support/2050/31 [16:08] * sergiusens takes a lunch break [16:08] sergiusens: thank you [16:14] elopio, this PR is failing the CLA check with an email I don't recognize: https://github.com/snapcore/snapcraft/pull/1580 [16:14] PR snapcraft#1580: cli: setup gitignore when working from git directory at init [16:14] None of the commits seem to have it [16:18] kyrofa: that is Jeff, from heroku. [16:18] not sure where is it coming from. [16:18] What the heck. Master? [16:19] if you tell him to squash and rebase, don't they go away? [16:20] I'll try updating the branch first [16:37] sergiusens, elopio two OSRF +1s on snapcraft#1582 [16:37] PR snapcraft#1582: plugins: add ros2 boostrapper [16:38] Hey zyga-ubuntu, I seem to remember you poking at snapd on gentoo a while back. Did that seem to work? [16:55] jdstrand: hello [16:57] jdstrand: would you have the time to look at https://github.com/snapcore/snapd/pull/3999 today, I'd like to get initial feedback and iterate daily, the branch is rather small and instead of a full security review I'd like to just ack the basic principle [16:57] PR #3999: cmd/snap-confine: add detection of stale mount namespace [17:00] sergiusens: so i've made progress, but the client command gets a "permission denied" error trying to connect to the socket [17:00] sergiusens: do I need to manually chmod it after listening, do you think? [17:01] oops, kyrofa ^ [17:18] kyrofa: the socket mode is Srwxr-xr-x so I think it *should* be open. I can only suspect app armor [17:19] rogpeppe, if you suspect apparmor, check the syslog, or use snappy-debug [17:19] (are you familiar with that handy tool?) [17:26] kyrofa: no, i didn't know about it. will take a look tomorrow === JanC is now known as Guest6853 === JanC_ is now known as JanC [18:03] zyga-ubuntu: done (I can't commit to being able to do this daily, but I'lltry) [18:50] jdstrand: thank you :) [19:13] Issue snapcraft#1444 closed: catkin pip support [19:13] PR snapcraft#1581 closed: catkin plugin: support rosdep pip dependencies [19:20] kyrofa mind taking another look at snapcraft#1565 ? [19:20] PR snapcraft#1565: cli: add the pack command [19:20] sergiusens, on it [19:21] elopio mind following up on snapcraft#1567 ? [19:21] PR snapcraft#1567: recording: record the snaps installed on the machine [19:24] sergiusens: I'm on that after I finish with the dotnet tests. [19:31] elopio, can you take another pass at snapcraft#1582? Or are you good with it? [19:31] PR snapcraft#1582: plugins: add ros2 boostrapper [19:33] kyrofa: I'm good with it. I'll leave my +1. [19:33] sergiusens, okay if that one goes in next? [19:35] kyrofa do you have the follow-up ready already? [19:35] sergiusens, mostly [19:36] sergiusens, turning the demo into an integration test [19:36] \o/ [19:37] But yeah, I had it done for ROSCon [19:37] Then the stupid lighting talks... [19:37] Who raffles those [19:37] PR snapcraft#1582 closed: plugins: add ros2 boostrapper [19:37] \o/ thanks :) === JoshStrobl|Work is now known as JoshStrobl [20:05] kyrofa mind having a second look on snapcraft#1578 ? [20:05] PR snapcraft#1578: project_loader: quote more environment variable values [20:06] Will do, just a minute === ikey is now known as ikey|zzz [23:20] PR core-build#11 closed: remove cruft from the writable-paths [23:20] PR core-build#22 closed: unit testing for sync_dir() [23:21] PR core-build#11 opened: remove cruft from the writable-paths [23:21] PR core-build#22 opened: unit testing for sync_dir() [23:35] PR snapcraft#1583 opened: ament plugin: new plugin