[00:13] is it ok for me to make a new snapd interface? [00:13] think im gonna need one for steam.. === devil is now known as Guest64572 [01:28] ikey I am just going to mention jdstrand here so he can look in the morning; or maybe it makes itself into a forum post; if its simple enough I'd just create a PR [01:29] ok [01:29] it looks like i have no choice but to make one [01:29] we have weird requirements [01:29] the one im struggling with is the mknod denial.. [01:36] ok nvm fixed that one. lol [01:40] PR snapcraft#1802 closed: metadata: extract metadata from appstream === cwayne_ is now known as cwayne === davidcalle_ is now known as davidcalle === tedg_ is now known as tedg === davdunc_ is now known as davdunc === kalikiana_ is now known as kalikiana === thomi_ is now known as thomi === nsg_ is now known as nsg === victorbjelkholm_ is now known as victorbjelkholm [03:31] PR snapcraft#1823 opened: elf: search for DT_NEEDED in prime and the base [04:43] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEnfdpyaaxi: dasjoe Trevinho Shibe benoitc iatrou fnordahl rob-oi-ma Ama [04:43] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEcgfhe: tanimislam iatrou fnordahl icey nottrobin m [04:43] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEpzzivqqi: kyrofa joedborg Stan [04:43] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODErsoty: iatrou StanleyHsiao paulmey matiasb marcoceppi_ cavea [04:44] ▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEeylvicwdq: Trevinho mpontillo caveat Cust0sLimen sdrobertw StanleyHsiao lfaraone petevg joedborg atriv jacekn [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEfrromdwkuq: JamieBennett AmarokNelg rob- [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEqddcwzw: StanleyHsiao petevg mardy nottrobin dragly cprov la_juyis Kaleo lfaraon [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEevivcj: lool tianon dragly dkessel la_juyis Kaleo petevg rob [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEesoaz: philroche joedborg dragly rob-oi-ma dasjoe Shibe ahayzen ralsin [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEywjvgmzgqn: iatrou icey jamespage AmarokNelg faenil stgraber Shibe loo [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEvxypasdzpn: matiasb JamieBennett Trevinho lool mardy tianon iatrou petevg AndyWo [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODExzsyqeboid: mardy ralsina abne [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEqqdzr: benoitc ahayzen marcoceppi_ blood [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEfcqoxefk: lool jacekn coreycb asac la_juyis matiasb Son_Goku mardy AndyWojo Trevinho paulm [04:44] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEmqivq: la_juyis icey jamespage magicaltrout Shibe tianon abner marcoceppi_ Cust0 [04:44] ▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEuvhdqcp: AmarokNelg ahayzen coreycb magicaltrout philroche la_juyis cprov dkessel StanleyHsiao mpontillo icey [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEdsospbl: thibautr__ mardy kyrofa tanimislam paulmey Kaleo jacekn rob-oi-ma corey [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODErpely: coreycb ralsina paulmey iatrou ahayzen moon127 dkessel petevg bloodearnes [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEnrdbnbutfu: mpontill [04:45] ▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEzfven: ahayzen om26er bloodearnest mpontillo la_juyis ahrs icey iatrou coreycb Cust0sLimen stgraber nottrobin fnordahl K [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEsfcxopsh: ahrs benoitc Son_Goku AndyWojo coreycb tianon la_juyis bloodearnest asac tanimislam rob-oi [04:45] ▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEwylbxiacm: StanleyHsiao joedborg Shibe icey Trevinho bloodearnest cprov caveat sdrobertw paulmey mpontillo la_ [04:45] ▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEbsjxao: stgraber philroche caveat m4sk1n_ kyrofa tanimislam cprov JamieBennett petevg mardy StanleyHsiao thibautr__ drag [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEianpmabyg: faenil thibautr__ dasjoe StanleyHsiao bloodearnes [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEextdkd: mpontillo ma [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEgfiegfqa: magicaltrout thibautr__ om26er cprov philroche dkessel Trevinho Kaleo paulmey ah [04:45] ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ https://beta.companieshouse.gov.uk/company/10308021/filing-history christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODEqsxsmvtya: dasjoe matiasb Kale === zarcade_droid is now known as Guest45071 [06:17] morning === chihchun_afk is now known as chihchun [07:45] good morning [07:46] kalikiana: morning [07:46] j605: hi, did you manage to start spotify with libpulse logs? [07:57] mvo: morning [07:58] hey mborzecki, good morning [07:58] mborzecki: how are you? [07:58] mvo: good, thank you [07:58] seems like with debian-sid-64 gone, we can actually complete the tests [08:00] mborzecki: yeah, the PRs look much nicer this morning [08:03] PR snapd#4414 closed: tests/main/interfaces-fuse_support: fix confinement, allow unmount, fix spread tests [08:19] mvo: do you have any idea why systemd-remount-fs might be failing like this? https://paste.ubuntu.com/26225909/ [08:20] it's happening in the image that the guy from yocto forum thread uploaded [08:27] mborzecki: hm, we have not touched this at all for core, the message is also strange. something wrong with the permissions maybe in ls -l /lib/systemd/systemd-remount-fs ? [08:27] mborzecki: stracing it might be nice, especially with -e trace=execve -s1024 to see what it passes to mount [08:28] what's funny, in the console, id gives: uid=0(root) gid=0(root) [08:28] mount /dev/hda2 /mnt: mount: only root can do that (effective UID is 1000) [08:28] hmm ok [08:29] but i'm in the maintenance shell started by systemd that could not complete system init :/ [08:29] mborzecki: uhh [08:29] mborzecki: that is strange indeed [08:30] -rwxr-xr-x 1 1000 1000 1148876 Dec 14 10:23 /lib/systemd/systemd [08:31] mborzecki: aha! [08:31] mborzecki: I bet most of the FS is owned by uid 1000 instead 0 [08:31] it is [08:31] nevertheless, the kernel starts init with uid/gid 0 right? [08:32] mborzecki: yes, your /bin/shell is also owned by 1000:1000 I pressume so it seems like this makes the euid also 1000 (which is new to me but I never dealt with a system with wonky permissions like this before) === chihchun is now known as chihchun_afk [08:35] mborzecki: this log is with the debug flags, https://ptpb.pw/VbIw [08:36] Trying to connect to /run/user/1000/snap.spotify/pulse/native... [08:36] connect(): No such file or directory (2) [08:36] this ^^ is interesting [08:37] j605: can you `ls -l /run/user/1000/pulse/` ? [08:37] the directory is empty [08:39] j605: `ps -ef|grep pulse` is there a pulseaudio daemon workign for your user? [08:41] mborzecki: jagan 1960 1954 1 Dec20 ? 00:12:49 /usr/bin/pulseaudio --daemonize=nojagan 1978 1960 0 Dec20 ? 00:00:00 /usr/lib/pulse/gconf-helper [08:42] from systemctl --user status pulseaudio.socket, "Listen: /run/user/1000/pulse/native (Stream)" [08:43] j605: can you try pulseaudio -k && pulseaudio --start and check if a socket appears under /run/user/1000/pulse/native ? [08:43] the socket already exists [08:43] it doesn't exist under snap.spotify [08:44] > srw-rw-rw- 1 jagan jagan 0 Dec 20 20:57 native= === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk [08:48] j605: can you run `snap run --shell spotify` and then `ls -l /run/user/$(id -u)/pulse` inside that shell? [08:54] mborzecki: https://paste.ubuntu.com/26226041/ [08:56] j605: this is using snapd-git right? [08:56] mborzecki: yes [09:01] j605: in your log, spotify is trying /run/user/1000/snap.spotify/pulse/native, can you also `snap run --shell spotify` and ls -l this path? [09:02] mborzecki: it is empty [09:05] mvo: XDG_RUNTIME_DIR is something we populate right? [09:09] mborzecki: is that directory a symlink to /run/user/1000/pulse if it works correctly? [09:10] mborzecki: yes, we set this [09:10] mborzecki: via the `snap run` mechanism [09:10] mvo: ^^ i guess it's a bind mount or sth? [09:12] mborzecki: I need to check, there was some discussion in the context of wayland and things changed a bit, let me see [09:13] mborzecki: is that the problem for j605 ? that the pulse socket is in xdg_runtime_dir but because we change it spotify cannot find it? [09:14] mborzecki: https://forum.snapcraft.io/t/wayland-dconf-and-xdg-runtime-dir/186 was the discussion I had in mind [09:15] mvo: something weird, XDG_RUNTIME_DIR=/run/user/1000/snap.spotify and for j605 libpulse tries that location, in my case it's completely ignored and instead /run/user/1000/pulse/native is used [09:17] mborzecki: yeah, so one easy fix would be in the spotify snap to do what is done for the wayland socket, i.e. a tiny snippet that creates a symlink [09:17] for me that is never tried, after snap.spotify it goes to /var/run/pulse/native [09:17] j605: out of curiosity, if you create a symlink from /var/run/pulse/native to /run/user/1000/pulse/native - does that work? [09:18] (not as a permanent fix, more out of curiosity) [09:18] seems messy since I need to be root for this but pulse is started per user [09:19] mvo: it works [09:20] may snapd should look into the user dir before going to /var/run/pulse [09:21] j605: thanks for confirming! yeah, the symlink is not a "solution", just to see if that was the only piece in the puzzle to make things work [09:22] j605: I think we have various ways to fix this, one is to ask spotify to modify the snap (as outlined above) or we could manually migrate the pulse socket into "our" runtime dir [09:22] j605, mborzecki thanks for getting to the bottom of this! will you update the forum topic with the findings? I can quickly create a PR to make the socket appear in the right dir [09:23] I am not sure how it worked for popey in Manjaor. May be it is doing things differently [09:23] may be a log from him would be nice to see [09:23] mvo: sure [09:23] j605: yeah, its all a bit odd, also that it works for mborzecki who is also using arch AIUI. but its fine, we figured it out, now we just need to implement the fix :) [09:24] j605: he should be around soon, we can ask him for logs [09:27] i'm grabbing a cofee :) bb in a while [09:30] mvo: just a thought, but shouldn't pulseaudio 'interface' be doing this? i mean bind mount /run/user/$(id -u)/pulse to $XDG_RUNTIME_DIR/pulse inside the snap? [09:34] so that fixes vlc as well :) [09:37] yay [09:39] mborzecki: tying it to the interface would be ideal, not sure how feasible this is with the current structure we have [09:39] btw. still a mystery why it looks in a different location here [09:39] hmm [09:40] or [10:12] mvo: do you need any help with the PR? [10:14] mborzecki: I have not even started yet, I was starring at command-not-found, feel free to pick it up if you want [10:30] mvo: hm i don't exactly see where we could put the symlink part in current code, maybe snap-confine, there's some xdg setup code but it's currently disabled [10:35] mborzecki: yeah, snap-confine was my thinking, the same way we make the Xauthority file available [10:35] it'd be great to have per-interface exec hooks that are part of snapd [10:47] * kalikiana coffee break === Dmitrii-Sh is now known as Dmitrii-Sh-PTO [12:55] diddledan corebird all of the sudden does not launch for me, I tried a reinstall (refresh data on snap info is no wrong though :-P); I don't see anything wrong, did anyone else mention it not working anymore? [13:11] sergiusens: which version are you using? the version on edge launches for me (r88), despite errors about libopenh264.so.bz2 and libgstalsa.so - I don't use it much, but it's showing tweets fine [13:16] * kalikiana FYI going to have lunch in ~15min [13:27] Hi, I have installed the RPI2 and PRI3 image for ubuntu core. Both on a RPI2 and RPI3. I boot up the PI, follow the instructions. Logon with SSH to the PI with the SSO (ubuntu ont account). After a reboot, I cant logon. The account has disapeard. I have to run the "wizard" again. Any ideas? === Guest45071 is now known as ^arcade_droid [13:36] bbl [13:43] i'm off to pick up the kids [13:54] PR snapcraft#1822 closed: tests: skip classic confined tests on armhf === LyzardKing__ is now known as LyzardKing [14:33] PR snapcraft#1807 closed: tests: run test_cleanbuild in LXD on Travis [14:37] ikey: anyone can propose an interface. if it is easy/straightforward, just submit a PR. if you think there needs to be discussion, I'd suggest a forum topic [14:37] ikey: do note that many of us are either on holidya or going on holiday soon [14:38] re === rharper` is now known as rharper === mbiggers_ is now known as markb1 [15:40] jdstrand, yeah no worries bud :D [15:40] i got the vast majority of it up and running last night [15:40] in the end i couldnt see any other way than the creation of a new interface [15:40] i preliminarily named it "steam" but ill likely rename it "steam-support" [15:41] and its already full of ugly steam specifics: https://hastebin.com/usadagasug.bash [15:44] ikey: interesting. I don't see anything conceptually wrong with a steam-support interface (esp for the shm accesses). we might want to move some rules around/etc, but feel free to send up when you're happy and we can iterate [15:45] sure :D [15:45] it needs some work yet and fleshing out [15:45] *but* i did finally manage to get steam working under full confinement [15:45] cool. thanks for working on it :) [15:45] that's awesome :) [15:45] only real bug i need to work out is why i cant execute things on another partition [15:45] and i think my snapd version is bugged truthfully [15:45] i ran "snap run --shell linux-steam-integration" with fresh installs last night [15:45] hit "ctrl+d" [15:46] then i got invalid syscall errors [15:46] ikey: if you don't know about this already: sudo sysctl -w kernel.printk_ratelimit=0 [15:46] and it was impossible to run "--shell" again [15:46] sometimes denials get rate limited [15:46] aah [15:46] ty ty [15:46] snappy-debug will do that for you, but if you are just taling the log, you'll want to do that [15:46] yeah [15:46] "what went boom" [15:47] ikey: I have an alias for: sudo sysctl -w kernel.printk_ratelimit=0 ; journalctl --follow [15:47] heh [15:47] i think apparmor is gonna need new variables once im done with this.. :p [15:47] heh [15:47] having uid/pgrpid could be useful.. [15:48] yeah [15:48] kernel side variables would be super handy [15:48] it's on the roadmap [15:48] oh fantastic [15:48] i see someone let /usr/share/themes through too [15:48] must be for gnome bases? [15:48] because my bundled themes kept working :p [15:49] was quite happy about that lol [15:49] jhenstridge has been improving fonts, themes, etc [15:49] hehe [15:49] nice! [15:49] ah right - saved me a whole bunch of bother then [15:49] I think if you use the desktop interface with updated desktop snapcraft parts, you get more and more goodness [15:49] oh right [15:50] atm we're not actually using desktop snapcraft at all [15:50] so i might need to merge behaviours [15:50] we have a main entry point, "$SNAP/linux-steam-integration" [15:50] which has lots of magical shim entry work [15:50] I think the biggest improvement lately was the gschemas work [15:50] so you don't have a slow first startup [15:50] AOT compilation? [15:51] tbh i think im even safe in that regard [15:51] all my schemas are inside the runtime [15:51] and the LSI snap is tiny - just the shim and zenity and mesa-demos and such [15:51] * jdstrand nods [15:52] i think once this confinement story is solved i can move onto the last challenges like udev + joystick improvements [15:52] then we could see steam "done" early '18 [15:52] that would be great :) [15:52] didnt mean for it to take so long :/ [16:02] kalikiana, sergiusens I'll be right there [16:02] firefox isn't working fwiw [16:04] Oh darn, I saw a tweet about it and wanted to try [16:10] kyrofa: bug #1739260 fix just got released. [16:10] Bug #1739260: Add macaroon expiration date to acl/verify endpoint [16:13] sergiusens: I had a report a week or so ago, but they rebooted and it started working again (ref: Corebird not working) [16:15] sounds scarily like windows behaviour to me :-p [16:17] ikey: did you get zenity to work in snap? the ubuntu variant has hardcoded it's location so you can't get ubuntu snaps to use zenity [16:18] yeah i have some patches [16:18] aha [16:18] here lemme find linky [16:18] popey: ping [16:18] https://github.com/solus-project/runtime-snaps/tree/master/support_packages/zenity/files [16:18] the most important patch is the 0001 patch [16:18] I wonder if we could get those upstreamed... *scratchy head* [16:18] it allows us to build a relocatable zenity binary with gresource compiled in [16:19] popey: thoughts on promoting sublime-text and android-studio to 'stable' channel ? [16:19] id like to for zenity but i couldnt get distcheck unbroken lol [16:21] om26er: how robust are they both? [16:21] Any feedback from users? [16:21] popey: there was an issue with sublime that was fixed quite a few days ago. [16:22] popey: Android Studio seems to be running solid. I gave it a try on a clean VM today, just to be sure. [16:24] Nice one. I will do the same [16:34] Out of interest, why does the "joystick" interface not autoconnect? it would seem like a much smaller security risk than "network"... [16:56] not sure as for the why right now, but i imagine joystick is going to grow in future [16:56] i.e. to support wheels and steam controller [16:56] though some of that is likely up in hidraw too.. [17:03] mcphail, can't speak for the security team (I'll let jdstrand speak up), but it's my understanding that all interfaces are considered risky. We don't want to speak for the users regarding the risk they're willing to accept as much as possible [17:04] mcphail, there are some interfaces we auto-connect as snaps would be useless without them [17:05] But the rest, we leave up to them [17:10] kyrofa: yes, i think paranoia by default is sensible. but, as a user, i'd be more worried by the network interface than the joystick one. and the syntax to connect these things is pretty horrible. are there plans for a gui so users could drag and drop connections? [17:12] mcphail, yeah I actually dislike some of the ones we autoconnect as well: https://forum.snapcraft.io/t/autoconnection-override/2465/4 [17:12] Regarding the GUI, I think so, but you'll need to talk to the desktop folks [17:13] (or maybe jdstrand has more insight there) [17:13] aah. it'd be nice to have a panel in the gnome settings thingy where users could set their own defaults, then separate tabs for individual snaps [17:15] i'd probably have joystick as an autoconnect and network as manual on my default panel :) [17:16] kyrofa: wrt snapcraft#1817 ironically, the raw properties are not preserved in pluginhandler, but preprocessed with default values... so I added another property with the original properties and a check if it was specified... lemme know what you think [17:16] PR snapcraft#1817: grammar: support strings [17:19] the gui will make that better. it is best to not autoconnect hardware by default. we can use snap declarations for autoconnection, but very much looking forward to the gui [17:19] jdstrand: cool. thanks! [17:20] jdstrand once apps don't get sigkilled autoconnection wouldn't be such a big problem [17:20] not having autoconnection [17:21] well, people will always want this or that (not everything is sigkilled-- mostly just process-control) [17:22] I'm looking forward to that PR too :) [17:22] sergiusens: kyrofa wrapping for the year now. FYI confirmed lxd/core from stable work fine with persistent containers, whatever is broken with edge will need to be investigated in the new year [17:22] Alright kalikiana, thank you! [17:23] kalikiana: enjoy your holiday :) [17:26] Cheers [17:27] Same to you (soon) [17:27] o/ [17:48] kyrofa #1823 is good for a review [17:48] PR #1823: overlord/devicestate: try to fetch/refresh the signing key of serial (also in case is not there yet) [17:49] kyrofa a final one, the snapcraft one, not that [18:02] looks like another of my snaps made it to the 10 featured in the monthly summary: https://insights.ubuntu.com/2017/12/20/top-snaps-in-november [18:02] specifically mars the ridiculous shooter [18:18] diddledan, awesome [18:33] ogra_: do you know what work is needed to make your wine snap strict? [18:49] I vote for whips and chains [18:49] kenvandine, I am running in a linode image a gsettings app and it is not creating the dconf dir in ~/snap/test-snapd-gsettings/27/.config/ [18:50] kenvandine, any idea why it could be happening? [18:51] cachio, do you have a user session running? [18:53] cachio, can i see your yaml? [18:53] kenvandine, https://paste.ubuntu.com/26228811/ [18:54] in the test I create a user [18:54] and execute the gesettings command from that user session [18:54] cachio, so dbus is running? [18:55] https://paste.ubuntu.com/26228820/ [18:55] kenvandine, this is what I see [18:57] cachio, iirc you need to do something like this [18:57] eval `dbus-launch --sh-syntax` [18:57] cachio, https://dbus.freedesktop.org/doc/dbus-launch.1.html [18:57] cachio, oh... this is interesting [18:57] To start a D-Bus session within a text-mode session, do not use dbus-launch. Instead, see dbus-run-session(1). [18:58] ok, trying it [18:59] cachio, read the man page on that [19:00] dbus-run-session -- bash [19:00] might be what you want [19:00] or... in your snap [19:00] kenvandine, shoiuld I run it as part of the snap command? [19:01] command: dbus-run-session -- desktop-launch desktop-launch $SNAP/gsettings "$@" [19:01] maybe :) [19:01] if not, you probably need to use dbus-run-session to setup the environment before executing the snap [19:02] cachio, if you add it to the command in the snap, it won't behave well for tests trying to run on the desktop [19:04] kenvandine, i'll try this dbus-run-session -- desktop-launch $SNAP/gsettings "$@" [19:16] kenvandine, I am getting dbus-run-session: Permission denied [19:16] dunno [19:16] kenvandine, should I plug another interface? [19:17] i think you need to run that before running your snap [19:17] so pull it out of the yaml [19:17] dbus-run-session -- bash [19:17] before running the snap [19:17] i think that might work [19:18] ok [19:31] kenvandine, nothing [19:31] kenvandine, it is not saving the values [19:32] I save a value for a key and it is not saved [19:32] but I can access to the default value [19:33] kenvandine, is there any log to see? [19:35] cachio, oh.. it can read the the values? [19:35] kenvandine, yes, but the default ones [19:35] kenvandine, it is not able to save [19:35] must be using the memory backend [19:38] cachio, is this file getting created? [19:38] ~/snap/test-snapd-gsettings/current/.config/dconf/user [19:38] kenvandine, no [19:39] kenvandine, this dir does not exist ~/snap/test-snapd-gsettings/current/.config/dconf [19:39] does ~/snap/test-snapd-gsettings/current/.config ? [19:39] yes [19:40] https://paste.ubuntu.com/26229018/ [19:40] cachio, what do you have in /run/user/$(id -u)/ [19:41] kenvandine, I manually created the dir snap.test-snapd-gsettings on there [19:43] kenvandine, https://paste.ubuntu.com/26229029/ [19:45] cachio, and in your snap, is gsettings a script or binary? [19:46] kenvandine, it is a script [19:46] can you try running it without the snap? [19:46] rule out any of the desktop helpers or snap [19:47] maybe it's related to the session [19:47] I'll try it [19:56] kenvandine, not working [19:57] ok, so that rules out the snap itself [19:57] I am executing the script and it is not saving the valus [19:57] yeah, it's probably falling back to the memory backend [19:57] and if I run /usr/bin/gsettings it works [19:57] not sure how to set that [19:57] cachio, can i see your script? [19:59] kenvandine, http://bazaar.launchpad.net/~snappy-dev/snappy-hub/test-snapd-gsettings/files [20:01] PR snapcraft#1824 opened: Update test_rpm.py [20:04] Issue snapcraft#1664 closed: Search prime for DT_NEEDED and add rpath [20:04] PR snapcraft#1823 closed: elf: search for DT_NEEDED in prime and the base [20:05] cachio, i think you need new_with_backend there [20:05] https://people.gnome.org/~gcampagna/docs/Gio-2.0/Gio.Settings.new_with_backend.html [20:06] kenvandine, ok [20:07] cachio, better docs https://lazka.github.io/pgi-docs/Gio-2.0/classes/Settings.html#Gio.Settings.new_with_backend [20:09] cachio, oh... that might not be the issue at all [20:09] you don't have a main loop [20:09] your script is going to exit before the change can be made [20:09] i think [20:10] for gsettings, you really need a main loop and wait until you get the changed notification from gsettings before exiting [20:12] kenvandine, ok, I'll work on that [20:12] kenvandine, I'll ping you tomorrow [20:12] cachio, i'm off tomorrow :) [20:13] kenvandine, ok :( [20:13] cachio, hopefully you won't need me tomorrow :) [20:13] kenvandine, hopefully [20:13] thanks for helping me [20:14] happy to help [20:15] * cachio afk [20:37] PR snapcraft#1825 opened: metadata: add infrastructure for extracting metadata from parts [20:40] Issue snapcraft#1686 closed: Support strings in grammar [21:37] I'm working on getting snap/snapd working under NixOS. When a snap is mounted, it tries to create a symlink to /etc/system/d/system representing the mount... On NixOS, most of /etc (including this path) is read-only. Can I set a flag or something to get the systemd unit written somewhere else? I could maybe link a directory in and get it working that way. [21:44] cmiles74: no [21:44] you'll need to patch it to put it somewhere else [21:44] snapd assumes it has control of the system [21:49] Son_Goku: Thank you. This seems like something I could make work, like it would work. Does that seem reasonable? I'm hoping it won't lead to another issue, etc. [21:49] cmiles74: it's possible, at least you're not trying to rip out systemd requirements like the Slackware guys... [21:49] snapd heavily leverages systemd functionality [21:50] but if you've got a different writable path, you'll need to patch systemd to use that path instead [21:50] also note that snapd has to operate privileged [21:50] as it manipulates seccomp filters, among other things [21:52] Son_Goku: I don't want to remove systemd, but NixOS has some pretty particular ideas about how things in /etc, /lib, and so on should be managed. I can write things in there at install time, but at run time writing things to /etc isn't allowed. My idea is to get snap to write the system files to another location and then I will symlink them into the proper location in /etc afterwards. [21:53] if snapd is running as root, it should have write access to /etc [22:40] kyrofa: hey, would you mind voting on https://forum.snapcraft.io/t/alias-request-for-snap-john-the-ripper/3169/2? [22:40] jdstrand, sure, let me take a look [22:40] kyrofa: it should be all of 10 seconds. pinging directly since so many or on holiday [22:40] s/or/are/ [22:41] jdstrand, you never need to apologize for pinging me :) [22:41] thanks :) [22:44] jdstrand, how high-traffic is that topic? I'm not subscribed, but perhaps I should be [22:44] Er, not topic, but category [22:45] not very [22:48] Alright, subscribed [22:49] cool. thanks for the vote [22:49] and with that, I am eoy [22:49] happy holidays everyone :) === devil is now known as Guest51955 [23:17] jdstrand: Happy christmas and a guid new year! [23:18] but not a setguid new year [23:32] Seeya in the new year, jdstrand. Merry Christmas! [23:41] Issue snapcraft#1687 closed: Make source grammar-powered