[00:08] mwhudson: o/ [00:09] mwhudson: is there an index of snaps ? [00:09] lifeless: there is https://uappexplorer.com/apps?type=snappy [00:13] mwhudson: yeah, but its pretty useless AFAICT - this https://uappexplorer.com/apps?type=snappy&q=nginx&sort=relevance [00:13] lifeless: yeah no argument there [00:13] mwhudson: or this - https://uappexplorer.com/app/apache2.mcaps - I can't figure out how to install that [00:14] lifeless: creation date suggests that's from the version 15 era [00:14] lifeless: this timezone is pretty bad for figuring out what's going on in the snappy world sadly === JanC is now known as Guest21246 === JanC_ is now known as JanC [00:39] mwhudson: thats sad; https://bugs.launchpad.net/snappy/+bug/1656976 is the next issue [00:39] Bug #1656976: classic mode cannot start services [00:41] Bug #1656976 opened: classic mode cannot start services [06:10] Could I get someone to peek at my app spacedrive? :) [06:11] Also, the build failed (except for ppc64el) on Launchpad, which is odd because it's a script and some deb packages. [06:15] Not sure if this is public or not... https://code.launchpad.net/~nhaines/+snap/spacedrive [07:49] Bug #1657021 opened: Operator specified filesystem access [07:55] good morning [08:03] zyga: good morning! [08:37] Mirv: hey! did you see my ping yesterday about the new runtime? [08:59] PR snapd#2641 opened: cmd: fix hardcoded paths to rst2man and support rst2man.py [09:03] Mirv: Just did a PR for reference: https://github.com/ubuntu/snapcraft-desktop-helpers/pull/34 [09:03] PR ubuntu/snapcraft-desktop-helpers#34: Runtime refactor [09:20] PR snapd#2642 opened: tests: disable ipv6 before unpacking delta [09:41] Hello, hen building a snap, launchpad throws a GPG error that I suspect is a launchpad problem, not related to my script. Am I correct? It may be related to 1626739? https://paste.ubuntu.com/23815514/ [09:45] r2geo: you mean the F1831DDAFC42E99D warning? [09:46] correct [09:47] when I install on my pi3, it fails on: $ snap install openhab_r2geo_2.0.0.b4-7_armhf.snap [09:47] error: cannot find signatures with metadata for snap "openhab_r2geo_2.0.0.b4-7_armhf.snap" [09:47] r2geo: i understand you get that warning also "outside" of snap, right? [09:48] ? sorry, I don't understand your question [09:48] i you run apt-get update (i.e. nothing snap related) [09:48] you also get that warning, right? [09:49] should be fixed if you run something like [09:50] sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys F1831DDAFC42E99D [09:50] will you make trust the launchpad packages signed by that key [09:50] on the pi I don't use apt-get update. On my pc I do not get that error. The buildlog was from launchpad [09:53] ah [09:53] should that key be added to my pc (from which I request launchpad to build)? Or from the pi? [09:53] then i don't know, sorry [09:53] Just checked, but building on pc does not throw this error [09:54] so ... as the buildlog is created on launchpad, I suspect there is something wrong there. But I may be wrong. I will create a bug report - still if someone knows, I would be very happy to solve it via IRC. Thanks [09:57] PR snapd#2643 opened: many: add stub implementation of snap-alter-ns === chihchun_afk is now known as chihchun [10:50] Bug #1656340 opened: XDG_RUNTIME_DIR is not created on app startup [10:51] jdstrand: hey, can you please review https://github.com/snapcore/snapd/pull/2630 [10:51] PR snapd#2630: many: detect potentially insecure use of snap-confine [10:59] ogra_: about the thing I was pointing out last night... https://bugs.launchpad.net/snappy/+bug/1657080 [10:59] not snap run or unity7 specific [10:59] jdstrand: ^ [11:00] PR snapd#2642 closed: tests: disable ipv6 before unpacking delta [11:09] Bug #1400802 changed: sudo snappy install should autocomplete === chihchun is now known as chihchun_afk [11:12] Bug # changed: 1500755, 1574114, 1574830, 1581596, 1588322 [11:15] Bug # changed: 1593371, 1593956, 1598656, 1611641, 1622782, 1623589, 1627338, 1635264 [11:18] Bug #1641590 changed: snap --help outline text [11:18] Bug #1641885 changed: jailmode doesn't work with snap try or snap install [11:21] Bug # changed: 1643885, 1643888, 1645413, 1646085, 1646479, 1647992, 1653955, 1654585 [11:24] Bug #1537793 changed: Autoupdate should be less frequent and more randomly distributed [11:24] Bug #1654666 changed: snapd-xdg-open doesn't work in strict mode [11:26] Trevinho, ah [11:27] Bug #1595064 changed: 'snap refresh' return code unhelpful

[11:27] Bug #1600083 changed: 'snap' tool bash completion does not complete local file names [11:27] Bug #1603481 changed: multiple binaries from the same package [11:28] Issue snapd#2514 closed: seccomp tests fail on Ubuntu 14.04 [11:30] Bug # changed: 1605258, 1606053, 1616657, 1617890 [11:33] Bug # changed: 1635016, 1636847, 1640874, 1645961 [11:36] Bug #1654590 changed: docker interface should account for /run/shm/ in addition to /dev/shm [11:41] hello, I am currently building an embedded system with Yocto/OpenEmbedded. I am considering trying to port Ubuntu Core to my platform to get its update functionality. However I am unsure: does using Ubuntu Core on OEM devices require buying licences from Canonical? [11:41] zyga, bug 1656976 is on purpose, right ? [11:42] Bug #1656976: classic mode cannot start services [11:42] eduardas_m, nope, but we currently only support grub and uboot as bootloaders and you need full access to the source (we require a specific setup for the bootloader) [11:43] beyond this there is nothing required [11:46] ogra_, thank you for the answer... my platform (DART6UL SoM from Variscite) currently has u-boot 2015.04 revision... where can I check if this is supported? [11:46] also, is it possible to do kernel updates via snaps? [11:46] eduardas_m, well, you need the source and need to enable/change some config options [11:46] yes, os and kernel updates [11:47] eduardas_m, https://github.com/snapcore/pi3-gadget/blob/master/prebuilt/uboot.patch has an example patch [11:48] ogra_, thanks a lot... sounds good so far... will give it a shot [11:48] CONFIG_ENV_SIZE (with the right size), CONFIG_SYS_REDUNDAND_ENVIRONMENT, CONFIG_ENV_IS_IN_FAT and CONFIG_SUPPORT_RAW_INITRD are hard requirements [11:50] ogra_: I believe so, yes [11:51] zyga, iirc we support that you install services that you can also (temporary) start ... i wonder if that affects it [11:52] (so that you can experiment with apache until the next reboot where it wont be started anymore) [11:52] at least thats the wording i remember [11:53] PR snapd#2644 opened: snapctl: support running by the snap app itself, not only just from hooks [12:04] PR snapd#2634 closed: tests: improve debug output when reexec is used [12:05] Issue snapd#2541 closed: tests/main/interfaces-network-control-ip-netns test fails on ubuntu 14.04-32 [12:09] Issue snapd#2576 closed: snap-confine cannot be setuid root on openSUSE [12:11] Issue snapd#2569 closed: snap-confine cannot perform namespace capture even with CAP_SYS_ADMIN [12:12] Issue snapd#2568 closed: snapd needs a SELinux profile to run on Fedora [12:13] Issue snapd#2553 closed: snapd is not tested against Debian [12:14] Issue snapd#2552 closed: snapd is not tested against Arch Linux [12:15] Issue snapd#2572 closed: .fstab files generated by snapd for the content interface do not follow the snap. scheme [12:17] Issue snapd#2603 closed: Disk free space left check [12:18] Issue snapd#2625 closed: feature request: ability to talk to sysctl [12:19] Issue snapd#2484 closed: Support removing all unused revisions [12:19] Issue snapd#2594 closed: Please add "install" hook [12:24] PR snapd#2641 closed: cmd: fix hardcoded paths to rst2man and support rst2man.py [12:26] PR snapcraft#1051 opened: misc: remove snapd "submodule" [12:35] PR snapcraft#1039 closed: Check the older ubuntu-core snap name for core dynamic linker [12:38] PR snapcraft#1045 closed: Handle parser errors better [12:45] can i move my scripts and setup dir from snapcraft to just the snap dir in the top level directory of my project now? [12:55] stokachu: I don't know, sorry [12:56] thanks [12:59] PR snapd#2645 opened: spread: exclude .o and .a files [12:59] zyga, where did https://github.com/snapcore/snapd/issue/2625 get moved to? [13:08] stokachu: ohhh [13:08] stokachu: hmm that's terrible,one cannot see the issues anymore [13:08] niemeyer: ^ can we have read only issues so that people can see where they were moved? [13:09] stokachu: all of those moved to launchpad.net/snapd but I cannot even give you the URL now :( [13:18] zyga, ah ok i figured you consolidated issue tracking [13:37] ogra_, my current u-boot has CONFIG_ENV_IS_IN_MMC since the board boots from SD card... is this option compatible with CONFIG_ENV_IS_IN_FAT? [13:37] nope [13:37] you need to use a uboot.env blob file [13:38] the userspace needs to be able to find it where it expects it to change variables on the fly [13:38] (there is interaction going on betweern snapd and the bootloader that needs to work) === JanC_ is now known as JanC [13:39] ogra_, this blob is available for download? [13:39] look in the branch that has the patch above ... there is a uboot.env.in and there is the command you need to turn it into a uboot.env blob in the README.md [13:40] (the uboot.env.in should contain your default env like you get it on a serial console with printenv) === ben_r_ is now known as ben_r [13:50] ogra_, thank you... that makes things clearer === chihchun_afk is now known as chihchun [13:55] Hi folks; I do have a question: snapd reads environment variables from /etc/environment file, how could I modify this file since fs is mounted as 'read only'? ;) [13:56] this is the first question. Second one is about to modifying the console command that connects to ubuntu-one so it obtains public key associated with a given account. [13:56] If I would be able to modify the above mentioned file (/etc/environment) so I would put there http_proxy settings, would then that initial script connect via the proxy? [13:57] I'm asking, because I'm not able to install ubuntucore in my lab infrastructure (http proxy is required) [13:58] So, maybe it would be a good idea if that script would take also the http proxy parameter? [14:06] Hi all! I'm experiencing with Ubuntu Core on RPi3 and have a few question, first I can't install webdm, it throws: (installation not allowed by "snapd-control" plug rule of interface "snapd-control") [14:06] so I've used snapweb instead [14:06] but is it webdm replacement? [14:06] is the error seen above normal? [14:15] Hi gibsson. Unfortunately, it appears that noone is here ;) [14:16] (i am not counting - newbie) [14:17] env_ro: thanks for your reply though, we're at least 2 listening ;) [14:17] I'll wait for a little bit, it can be early for some timezones [14:21] PR snapd#2368 closed: tests: parameterize remote store [14:23] gbisson, it was renamed to snapweb a while ago [14:32] PR snapcraft#1031 closed: store: fix sso_host for dev sso [14:32] jdstrand: good morning, thank you for triaging that bug [14:33] np [14:33] it will be nice to have it fixed, but that work is behind upstreaming (at least) atm afaik [14:34] jdstrand: that's a good plan IMHO [14:35] jdstrand: btw, do you know how far are we from upstreaming everything? [14:39] has anyone run into https://bugs.launchpad.net/ubuntu/+source/snapd/+bug/1656580? [14:39] Bug #1656580: snapd on trusty fails to install snap with error loading shared library [14:40] references some apparmor issue [14:40] zyga: I know a fair bit made the merge window. I don't have the details. perhaps tyhicks does? [14:43] jdstrand: thanks, I'll ask jjohansen and tyhicks later today [14:43] stokachu: fyi, I asked for more information in the bug [14:44] jdstrand, thanks redeploying 14.04 in my maas lab will get you that info [14:46] zyga (cc jdstrand): a ton of patches will land in 4.11 (https://marc.info/?l=linux-kernel&m=148455878701987&w=2) but we're still months away from upstreaming everything [14:47] zyga: now that the ball is rolling on the upstreaming effort, I think that the kernel development cycle cadence is going to limit how fast we can upstream things [14:48] zyga: while that was a large number of patches, that's all we'll be able to get into the 4.11 kernel so we'll be queueing the next set of patches up for the 4.12 kernel security subsystem merge window [14:51] jdstrand, hmm 2.21~14.04.2 does't seem to show this problem [14:51] think that was released this morning [14:51] still testing though [14:51] PR snapd#2645 closed: spread: exclude .o and .a files [14:51] I know there has been a lot of 14.04 work lately. it doesn't surprise me it might be fixed [14:52] having systemd issues now but i dont think that apparmor is an issue any longer [14:53] where does snapd install the systemd files for 14.04? [14:53] stokachu: cool. can you followup in the bug? note that when I was playing with 14.04, cgmanager would get in the way of systemd [14:56] jdstrand, http://paste.ubuntu.com/23816658/ thats what i see with systemd and my snap [14:56] journalctl -xn didn't give me much info though [14:59] can't seem to find my systemd service file [14:59] not in /lib/systemd [15:00] it looks like the systemd file shouldve been linked into /lib/systemd/upstart on 14.04 [15:07] tyhicks: that's some great news, thank you [15:09] ogra_: thanks! [15:10] in that case maybe webdm should be removed from the list or have a more explicit error message like "not supported any more, please use snapweb" [15:10] jdstrand, tyhicks: FYI, I have extra data for the apparmor bug at https://github.com/snapcore/snapd/pull/2624 but I need some help to figure out what to collect [15:10] PR snapd#2624: Re-associate with pid-1 mount namespace if required [15:10] I made it easy to re-run the spread test that uses the test kernel from jjohansen [15:10] and to collect various things [15:11] after running it the erorr I see is different, not sure if that is because of the older base kernel [15:12] this log file has useful references: https://s3.amazonaws.com/archive.travis-ci.org/jobs/192346094/log.txt [15:12] I am trying out Ubuntu Core image on Virtual Box... Ubuntu store account is mandatory to even launch Ubuntu Core? [15:13] scroll to the bottom please then look for: linode:ubuntu-16.04-64:tests/regression/lp-1644439 [15:13] (hmm, I guess I need to stil tweak that to restore correctly) [15:13] the results are much more comprehensive if the single regression test is executed [15:14] eduardas_m: to log in, yes, otherwise anyone on the same network could "own" your device and that's not great for security [15:14] zyga, stokachu: No, no way around that unfortunately [15:14] jdstrand, tyhicks: let me know when would be a good time to try some interactive triage/debugging on this [15:15] niemeyer: ah, too bad; on the up side all the reporters got the email that included the new URLs so I hope they can find that [15:15] Yeah [15:15] and we also only had two issues which were not filed by you or mvo [15:15] zyga, but if I want to use Ubuntu Core to be ported for an OEM device that is really bad [15:15] eduardas_m, it will launch without it ... but to log in you need one [15:17] ogra_, so lets assume that Ubuntu Core ships with hundreds of OEM devices..each one of these has to be tied to an Ubuntu Store account? [15:17] eduardas_m: if you have something specific in mind I can get you in touch with our commercial people; this scheme is designed to be secure for end users; [15:17] thre is some way to set a local acocunt using a configuration on a usb stick on first boot [15:18] but i dont think thats documented or promoted yet [15:18] ogra_: only the owner of the device brand can issue this assertion [15:18] ogra_: and it only (sensibly) applies to unowned devices [15:19] ogra_: and canonical does not issue this assertion yet [15:19] right [15:19] eduardas_m: being tied an account also lets the user manage the device better, depending on the form factor and interaction methods it would be impossible to do otherwise [15:20] ogra_: I just ping you in the github PR but there's a test kernel for you to test: https://bugs.launchpad.net/apparmor/+bug/1656121/comments/2 [15:20] Bug #1656121: unexpected errno=13 and disconnected path when trying to open /proc/1/ns/mnt from a unshared mount namespace [15:20] ogra_: bah, my bad... that message wasn't meant for you [15:20] zyga: I just ping you in the github PR but there's a test kernel for you to test: https://bugs.launchpad.net/apparmor/+bug/1656121/comments/2 [15:21] tyhicks: I tested this kernel, all the stuff I said above was based on that [15:21] oh [15:21] * tyhicks rereads [15:21] see ... undocumented, so i know nothing about it obviously ;) [15:21] tyhicks: one sec: [15:21] tyhicks, heh, i was starting to wonder already :) [15:22] tyhicks: https://github.com/snapcore/snapd/pull/2624/files#diff-f3378ae971fb046662010c5acfd2c35cR21 [15:22] PR snapd#2624: Re-associate with pid-1 mount namespace if required [15:22] PR snapd#2347 closed: overlord: flag required-snaps from model as required and prevent removing them [15:22] tyhicks: I've automated the testing process for that issue [15:23] zyga: since it is a kernel bug, I think you're better off working directly with jjohansen on this when he starts his day [15:24] tyhicks: yes, I agree [15:25] zyga, hey, could you check if you are happy with this MR: https://github.com/snapcore/snapd/pull/2226 [15:25] PR snapd#2226: interfaces/builtin: add evolution interfaces [15:26] renatu: yes [15:28] renatu: done [15:28] zyga, thanks [15:48] I am having trouble logging in to Ubuntu Core on VirtualBox... I am trying my Ubuntu One username and password [15:49] I used a public ssh key I generated on my Ubuntu 14.04 host to put into my Ubuntu One account [15:49] Chipaca, unfortunately the bug you referenced was just discussing how unhelpful the error is, not that it was impossible to cleanbuild [15:50] Chipaca, but it's a known issue, there was a ML thread about it and we're working on it [15:59] PR snapcraft#1050 closed: repo: add multiarch support for stage packages [15:59] PR snapcraft#1051 closed: misc: remove snapd "submodule" [16:03] zyga, jdstrand: Do you want to have a quick call to discuss the new interface system? [16:08] niemeyer: I guess a call now is fine [16:08] tyhicks: fyi ^. I can handle it but if you want to be there ^ [16:10] yes [16:10] niemeyer: where shall we meet? [16:10] https://hangouts.google.com/hangouts/_/canonical.com/new-interfaces?authuser=1 === chihchun is now known as chihchun_afk [16:12] jdstrand: ^ [16:13] what is supposed to be the username/password combination for the ubuntu-core-16-amd64.img? [16:13] is it the username and password of my Ubuntu One account? [16:14] it is your username ... no password ... it uses the ssh key that you provided in your U1 account to provide ssh login [16:15] eduardas_m, there's no local login by default [16:16] right, only ssh [16:16] eduardas_m, the idea is, you go through the first boot wizard, provide your Ubuntu One login info, and it'll pull down the SSH keys you have uploaded. Then you can use those to SSH in [16:17] kyrofa, so I can not log in through tty1 in VirtualBox? [16:17] ogra_, how is SSH configured by default? If a password is set, will one be able to SSH in with that password? [16:17] because it prompts me for a password there [16:17] eduardas_m, not by default, but you can always set a password once you SSH in [16:18] kyrofa, is this kind of interesting behaviour documented somewhere? [16:18] because I believe a lot of people would benefit from instructions on how to test Ubuntu Core on VirtualBox [16:20] eduardas_m, yes indeed. I assume you started with the KVM option? https://developer.ubuntu.com/core/get-started/kvm [16:21] PR snapd#2646 opened: tests: fix failing snapd-reexec test [16:21] eduardas_m, note that this behavior is a little out of the ordinary, but it's done to ship a secure image by default [16:22] eduardas_m, default passwords etc. are why we have botnets of CCD cameras [16:24] eduardas_m, to be honest I am quite new to linux in general (haven't even spent a year with it) and have very limited experience with VirtualBox and VMware as it is...since KVM is totally unfamiliar to me, I did not even consider it [16:24] eduardas_m, oh I don't blame you. I'm NOT new to Linux and I prefer vbox [16:25] eduardas_m, but what image did you start with, then? [16:27] kyrofa, I did a VBoxManage convertdd ubuntu-core-16-amd64.img ubuntu-core-16-amd64.vdi --format VDI [16:27] eduardas_m, or did you do something directly like this https://kyrofa.com/posts/ubuntu-core-on-virtualbox ? [16:27] I found these instructions online and it was the first thing I tried [16:27] eduardas_m, ah [16:27] eduardas_m, well, had you gone through the developer.ubuntu.com path, you would have seen docs for this behavior. But no matter! [16:31] kyrofa, thank you for the help so far... will continue exploring Ubuntu Core tomorrow [17:00] kyrofa, hmm, no idea, since the key always preceeds password i have never tested that [17:01] PR snapd#2647 opened: cmd: add autogen.sh [17:05] PR snapcraft#1052 opened: Make git pulls less error prone due to history changes [17:23] jdstrand: trivial: https://github.com/snapcore/snapd/pull/2648 [17:23]