=== JanC_ is now known as JanC [01:53] so this is my first time using snap... [01:53] trying to install huggle on debian9 [01:53] am getting message "Run configure hook of "core" snap if present" [01:54] with a spinning baton. not sure if I'm supposed to do something? [01:58] Agafnd: that means it's broken :( [01:58] :( [01:58] how do i fix? [01:58] can i fix? [01:59] oh...I ^C'd it, tried again, and now it seems to be installing huggle? [02:00] * Son_Goku shrugs [02:00] it happens occasionally with me in Fedora too [02:00] there's no obvious reason why it happens, either [02:01] I wish everyone used AppImage as the distro-agnostic format [02:01] hmm reminds me i should update snapd in debian now that stretch is out [02:01] never had problems with those [02:02] mwhudson, what is debian on right now? [02:02] in Fedora, we're at 2.26.3 [02:03] looks like 2.24 [02:04] 2.21 [02:04] oh dear [02:04] that's before a number of core transition handling things [02:04] Agafnd: that's the version from the core snap, i presume [02:04] it's from the output of snap --version [02:05] mwhudson, does snapd do re-exec even for "snap --version"? [02:05] "snap 2.24" "snapd 2.24" "series 16" [02:05] Son_Goku: i think so and what Agafnd is saying certainly suggests that [02:05] oh dear [02:05] that's annoying [02:08] well anyway, looks like i got huggle running [02:08] so problem solved for me :D [02:13] hmm maybe i should make zyga do this === chihchun_afk is now known as chihchun [06:23] good morning [06:40] * zyga fetches some coffee [08:48] * zyga breakfast [08:58] PR snapcraft#1412 opened: lxd: Snapcraft update in containers [09:13] pedronis: good morning [09:13] pedronis: how were your holidays? [09:21] zyga: pedronis is off until wednesday, says the calendar [09:34] ah, I though he would be back already [09:34] Chipaca: warsaw looks a lot like impressions of london today [09:36] zyga: ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒? [09:38] PR snapd#3595 opened: debian: update debian/tests/control to use isolation-machine [09:43] Chipaca: exactly [09:43] mvo: approved [09:44] thanks zyga [10:03] PR snapd#3596 opened: tests: disable snapd-notify for the external backend [10:06] mvo: travis seems still to be dead [10:07] zyga: :( [10:24] Son_Goku: o/ [10:46] PR core#49 opened: more detailed build instructions [10:50] ogra_: btw, is travis working for you [10:52] zyga, "pending" ... [11:04] hmm ... i once switched this pi3 to the beta channel, then switched back to edge ... http://paste.ubuntu.com/25111305/ [11:04] shouldnt the refresh to edge have me downgraded to revision 30 again ? [11:05] ogra_: looking [11:05] (at least that would be my user expectation) [11:05] ogra_: yes, I think so, Chipaca ^ ? [11:06] ogra_: you say "switched", how did you switch? [11:06] Chipaca, snap refresh --channelname [11:06] though ... hmm [11:06] ogra_: yes, that would have brought 30 in [11:07] ogra_: revert wouldn't've [11:07] thei SD is very old, could be that i once manually tinkered with some bootloader var [11:07] s/thei/this/ ... tsk [11:07] * Chipaca makes a note to always ask ogra if he tinkered with the bootloader before worrying about weird behaviour [11:07] heh [11:08] well, put it in an envelope and send it to me so i dont ask stupid questions :P [11:08] fgimenez: searching test failed, is that interacting with the store? 2017-07-17 10:04:03 Error executing autopkgtest:ubuntu-16.04-i386:tests/main/searching : [11:08] “snap refresh --channelname snapname” should leave you with what “snap info” says is current for that channel [11:08] Chipaca, though doesnt snapd just care for the state file ? [11:09] manual tinklering with the bootloader shouldnt chnage the metadata, shoould it ? [11:09] ogra_: that might be true if your view is narrow enough, but it isn't true in general [11:09] I mean, I don't know of a bootloader var that could influence this behaviour [11:09] other than telling it "use this one" [11:09] _however_ [11:09] you're assuming the refresh worked :-) [11:10] that is [11:10] heh, indeed [11:10] if you refresh core on a device, reboot, and the reboot fails, the core will be rolled back [11:10] but I'm not 100% sure the "track this channel" change is rolled back; I'd have to look [11:11] (i think it is, unless there's a use case for it being special and is thus special-cased) [11:11] ah, yeah, that install had inssues often enough (bacause of me breaking it) so there will have been plenty of rollbacks [11:11] zyga: yes it interacts with the store, i just got an error for that test on i386 too while validating 2.27~rc3 http://paste.ubuntu.com/25111252/ [11:11] ogra_: aren't 30 and 34 both very very old? [11:11] no featured snaps for that arch it seems [11:12] Chipaca, no, 34 is current [11:12] ah this is pi2-kernel, not core [11:12] 30 is pretty behind ... i need to sync again [11:12] not sure why i thought it was core [11:13] changing subjects completely, everything is terrible [11:13] :-) [11:13] i'm going to go for a run and think about this stuff (what i'm working on, not ogra_'s thing :-) ) [11:13] Chipaca: enjoy! [11:13] fgimenez: aha [11:13] fgimenez: anyway, feels like a store-side issue [11:14] trying to sort out the client/snap/systemd imports is soul-sucking [11:14] Chipaca: hmm? [11:14] Chipaca: would it help if I was the garden troll and you'd explain the problem to me? [11:14] Chipaca: you can even record audio while running for extra panting drama [11:15] zyga: maybe. When I get back :-) [11:15] zyga: basically, I had client importing systemd, and need to change that around, but client imports snap and snap imports systemd [11:16] zyga: cleaning that up leads to less code, less duplication, woo [11:16] right? [11:16] but then... all the little changes... [11:17] aw [11:17] nothing serious, just soul-sucking (and needing a firm hand to not do more while i'm in there, or to do more in self-contained reviewable chunks) [11:17] would it help if snap imports were copied into client? [11:17] no, that would actually make it worse :-) [11:17] mhm [11:17] we already have two json serialization structs for apps, for example [11:17] and they're _different_ [11:17] I'll gladly review the patches [11:17] enjoy your run [11:18] * zyga looks outside at the rain [11:18] there's client.AppInfo that expects name and daemon, and there's daemon's appJSON that gives you name, daemon, and desktop-file [11:18] anyhow! [11:18] * Chipaca afk [11:27] zyga: hey [11:28] Son_Goku: hey, interesting thread about flatpak for graphical apps [11:29] Son_Goku: I'm working on voicing my opinion === svij is now known as internethulk_svi === internethulk_svi is now known as internethulksvij [11:36] * zyga needs a drink [11:41] * ogra_ hands zyga some schnaps :P [11:55] ogra_: nah, I'm just drinking water today [11:55] :) [12:04] zyga, btw, doesnt look like travis will ever pick up on my branch ... still in "Waiting to be queued" state [12:10] ogra_: I think travis was massively down [12:10] ogra_: and they are replaying the whole requests that happened since [12:10] ogra_: so they have a huge backlog [12:10] ogra_: and will eventually get to each test [12:10] ogra_: but I don't know how soon or not soon that is [12:11] zyga, ah ... [12:26] PR snapd#3575 closed: overlord: populate interface label with summary [12:28] zyga: Are you working on #3399? [12:32] PR core#50 opened: fix snapd.core-fixup.service symlink on core devices [12:32] zyga, have you had a chance to look at the 32-bit failures? [12:33] for snapd 2.27 and fedora [12:33] PR snapd#3399 opened: many: add the interface command [12:34] Son_Goku: no, not yet, but I already spoke with morphis about looking at them (today) [12:34] Son_Goku: I'm wrapping up my last PR [12:35] niemeyer: yes [12:36] zyga: I just reopened it and merged it into the review sprint [12:37] niemeyer: thank oyu [12:37] niemeyer: I have a few things to tidy and re-check all the points [12:37] niemeyer: but it is looking good [12:37] zyga: Super, thanks! [12:48] niemeyer: are we skipping standup today? === chihchun is now known as chihchun_afk [12:49] niemeyer, linode machines not working today, https://travis-ci.org/snapcore/snapd/builds/253590022 [12:50] zyga: I'll be on a plenary with mvo, and pawel and pedronis are out, so we can probably skip it [12:51] niemeyer: ack [12:51] niemeyer: I'm going out with my family, I could drop by in 2-3 hours [12:52] zyga: That looks like it's working, but too much concurrency maybe.. [12:52] Please don't cancel builds, folks! [12:58] niemeyer: i thought spread dealt alright with canceling builds now? (to be clear, i haven't canceled builds) [12:59] Chipaca: It usually does, and some cancels won't really do any harm.. I'm just reviewing the systems now and so far they are all clean [12:59] Chipaca: This might become an issue if somebody starts cancelling jobs pathologically [12:59] Chipaca: Because that removes the chance of spread itself even cleaning the system [13:00] good ol' pathos [13:00] ! [13:00] :-) [13:01] I can tell for instance that 4 hours ago several builds were interrupted.. [13:02] hey cachio, what's the status of the package install centralization that niemeyer mentions here https://github.com/snapcore/snapd/pull/3484/files#diff-5cb500f3c949d5be548c6dc2556eb6c1R12 is it already merged? [13:02] PR snapd#3484: tests: add autopilot-introspection interface test [13:03] * zyga_ codes from his car [13:03] fgimenez, no yet [13:04] cachio: ok thanks! ping me when ready [13:05] fgimenez, it needs approvals :) === zyga_ is now known as zyag === zyag is now known as zyga [13:18] zyga-ubuntu: I just went through and cleaned all systems.. also deallocated one of the systems which seemed to have potential hardware issues based on the logs, and reallocated it elsewhere [13:18] Please let me know if you see any failures that feel like Linode problems [13:18] fgimenez, cachio ^ [13:18] niemeyer, sure [13:18] cachio: Is the PR that needs approval already in the review board? [13:19] niemeyer, it is the dependencies one, it is under review [13:19] niemeyer, https://github.com/snapcore/snapd/pull/3483 [13:19] PR snapd#3483: tests: dependency packages installed during prepare-project [13:22] cachio: Thanks [13:23] niemeyer: ack [13:23] fgimenez, comment on review addressed, thanks for reviewing [13:23] cachio: np thank you [13:32] Is it necessary that snaps be installed on snappy core only via the Ubuntu snap store [13:32] ? [13:33] Is it not possible to have a snap store that has got nothing to do with Ubuntu or Canonical? [13:36] Vamsi: hey, you can install snaps locally without using the snap store [13:36] Vamsi: there's ongoing work to add support for stores other than the ubuntu one [13:36] Vamsi: but that's aimed at enterprise stores that sit on LAN and serve local devices [13:38] zyga-ubuntu: Thank you for the info. I was indeed looking at the perspective of an enterprise. We are looking seriously into Ubuntu core for our IoT usecases and wanted to know about this piece of information. [13:39] zyga-ubuntu: Could you please direct me to any documentation about this ongoing work that you were talkking about? [13:49] niemeyer: hi, if you have a moment would you be able to review this? https://forum.snapcraft.io/t/track-request-for-openstack-core-snaps/1282 [13:54] * zyga-ubuntu lunches [13:55] Vamsi: yep, please wait 15 min [14:02] zyga-ubuntu: No problem. Thanks mate! [14:02] coreycb: Sure, sent a question there [14:05] mvo: We've got some space for hacking, right outside of room 9 [14:06] niemeyer: I could join you if you have any need of my hlep [14:06] * zyga-ubuntu hacks on interface PR [14:06] Vamsi: ok, let me try to help you [14:07] zyga-ubuntu: We can *always* use your help :P [14:07] Vamsi: there are some PRs in flight https://github.com/snapcore/snapd/pulls?utf8=%E2%9C%93&q=is%3Apr%20is%3Aopen%20custom%20store%20 but I think for a business use-case you can already use an enterprise store today [14:08] Vamsi: if you want I can get you in touch with some business people at Canonical [14:08] * zyga-ubuntu hugs niemeyer :-) [14:09] zyga-ubuntu: We'll have the snapd session tomorrow.. maybe plan to be here the whole day tomorrow? [14:09] niemeyer: sure, sounds good [14:09] niemeyer: 8:30? [14:09] zyga-ubuntu: Yeah [14:10] zyga-ubuntu: That would be awesome! [14:10] niemeyer: perfect, I'll see you there [14:17] niemeyer: thanks, responded [14:22] cachio: #3483 reviewed again [14:22] snapd#3483 [14:22] PR snapd#3483: tests: dependency packages installed during prepare-project [14:22] niemeyer, I'll take a look tx [14:23] coreycb: Responded to the response response :) [14:23] niemeyer: thanks :) [14:28] niemeyer: store doens't return featured snaps for i386 anymore [14:28] niemeyer: any store people there you can poke? [14:28] https://travis-ci.org/snapcore/snapd/builds/254377475?utm_source=github_status&utm_medium=notification [14:29] Huh [14:29] Yeah, let me find somebody [14:30] zyga-ubuntu: niemeyer they are also missing for dragonboard http://paste.ubuntu.com/25111989/ [14:30] fgimenez: thank you! [14:33] zyga-ubuntu: niemeyer according to wgrant it is expected, there are no featured snaps supporting i386 [14:34] fgimenez: has that changed? [14:34] zyga-ubuntu: indeed, the test was passing on friday [14:35] AFAIK Evan's team manages that set, but I'm not 100% sure on that. [14:36] ev is at the sprint so niemeyer should be able to reach him [14:37] fgimenez, zyga-ubuntu perhaps we should check in the test if there are featured apps first [14:37] But yes, given the current set of snaps in the featured section, those failures are expected. The store is operating correctly here; the curated data just has some unusual properties. [14:37] fgimenez, zyga-ubuntu make a call to the api or something similat [14:37] that could fail also for other distros [14:41] cachio: iirc we considered that a while ago when we hit issues like this, but at the end we prefrerred to left the test as is, there should always be featured snaps for all the architectures, maybe we could rethink that approach now [14:43] ogra_: https://github.com/snapcore/core/pull/50 can you look please? [14:43] PR core#50: fix snapd.core-fixup.service symlink on core devices [14:47] fgimenez, agree, do you know if someone is going to add a featured app to fix it? or we should put it as manual [14:48] I'd wait till we have a response from store people what changed since last week [14:48] cachio: i think Michael is trying to ask Ecosystem people about this right now [14:52] fgimenez, ok [14:58] ogra_: ? [14:58] zyga-ubuntu, lool king [14:59] lol [14:59] looking [14:59] how did my finget hit the tab key there ? [14:59] ogra_: just use your finger, not your finget :D [15:00] * zyga-ubuntu hugs ogra_ [15:00] lol [15:00] zyga-ubuntu, approved [15:00] :-) [15:00] PR core#50 closed: fix snapd.core-fixup.service symlink on core devices [15:21] Hi! I want my snap to only build on armhf, is there a config to force that. [15:22] architectures: [15:22] right now all my uploads are built for amd64 and armhf, my app is only relevant to RaspberryPi so armhf is the only arch I am interested in. [15:22] - armhf [15:22] ah, that's simple. [15:22] another: what's the recommended way to build armhf snaps on amd64 system ? [15:22] note though that there is a bug in build.snapcraft.io that will make it run the amd64 build regardless ... [15:22] just ignore that one [15:23] hmm [15:23] can I use lxd for that ? [15:23] (it will eventually respect the architectures filed in snapcraft.yaml) [15:38] does build.snapcraft.io support delta uploads ? [15:45] ogra_: can i set a variable in the prepare scriptlet and use it later in snapcraft.yaml? [15:45] ogra_: something like - https://git.launchpad.net/~ubuntu-kernel/ubuntu/+source/linux-snap/+git/xenial/tree/snapcraft.yaml?h=pc-test [15:45] ogra_: look PKGNAME [15:49] ppisati, dunno, try it ... (i'd prefix it with "export" though ... to make sure ) [15:53] ogra_,om26er: architectures isn't correct for this case. See https://github.com/canonical-websites/build.snapcraft.io/issues/556 [15:53] ogra_: unfortunately it doesn't work as it is... :( /me does some more tests [15:53] om26er: Why do you care whether there's an optimisation of the internal upload of snaps from BSI (actually Launchpad) to the store? [15:54] cjwatson: I expect the download size of the update to be smaller as a result. [15:54] om26er: That's delta downloads, not delta uploads [15:54] om26er: Download deltas are automatically computed by the store as needed - BSI doesn't need to support them [15:55] cjwatson: ok, so is that enabled today ? I have a snap on edge channel that I need to quickly test and in many cases its just a single line change so I would enjoy faster downloads. [15:55] om26er: I haven't explicitly checked recently but I don't see why it wouldn't be; that work was done months ago [15:56] PR snapd#3597 opened: arch,release: map armv6 correctly [15:56] obviously a saner thing to do would be to build the snap locally instead of waiting for buid.snapcraft [15:56] om26er: (though it's also not something I've worked on myself) [15:58] ppisati, well, ask the snapcraft guys in rocketchat ... [16:07] om26er: if it's not working then that would be something to bring up as a store issue I think [16:08] deltas are disabled on core [16:09] (downloads that is) [16:09] xdiff3 can add minutes to the install process when they are (depending on the hardware) [16:10] err [16:10] s/xdiff3/xdelta3/ [16:17] ogra_: isn't that just on pi? [16:17] all of armhf maybe? [16:17] or is it all of core? [16:19] all of core [16:19] well ... not sure about amd64 (i rarely use that) [16:43] * Chipaca kicks off a spread run and starts to wonder about dinner [16:45] ogra_: re your forum.snapcraft reply about RPi GPIO interface. I updated my RPi candidate channel but I still don't see the GPIO interface [16:46] * zyga-ubuntu needs a coffee [16:46] did `snap refresh core --candidate` === chihchun_afk is now known as chihchun [16:46] same for pi2 and pi2-kernel [16:46] (and rebooted) [16:46] om26er, as i said, gadgets do nnot get updated [16:46] you need to build an image from one of the mentioned channels [16:47] (well, gadgts get technically updated ... but not their contents) [16:48] ogra_: so once the candidate becomes stable it will get fixed then ? [16:48] for now I'll export GPIO pins manually. [16:49] `echo pin_number > /sys/class/gpio/export` to the rescue. [16:50] om26er, no [16:51] once gadget updating has been implemented in snapd it will work [16:51] only then [16:52] hmm, building my own image might not be feasible because I have to distribute my app and I don't expect everyone to rebuild the image. [16:52] also ... you wont have access to /sys/class/gpio/export without a connected gpio interface (unless you use devmode) [16:52] well, they can use my daily edge images or one of the candidate builds that cdimage provides [16:52] * ogra_ goes afk [16:52] ogra_: doesn't the gpio interface get declared in the gadget snap.yaml which gets updated? [16:53] * zyga-ubuntu is busy on interfaces and hungry/thirsty [16:53] (and on his way home) [16:53] zyga-ubuntu, if the gadget gets updated, for safety reasons we didnt do that yet [16:53] even if I may export pins on OS startup using a udev rule, the problem will still remain i.e. I won't be able to actually write to GPIO config unless in devmode. [16:53] * ogra_ really needs to go now, i'm 330min late already [16:53] *0 [16:53] bah [16:53] 30 [16:53] -> off [16:53] ogra_: sure, will catch you later. [17:11] Hi! Does the network-observe interface support signalling or is that only based on request-response based ? [17:12] -based [17:13] om26er: can you be more specific? [17:13] what do you intend to do? [17:14] I have a snap that runs as a daemon, I want that snap to keep waiting for internet before trying to connect [17:14] also when internet is lost I don't want it to keep trying crazily, rather get "notified" that connection is back. I may then ping to check if internet is really working. [17:14] zyga-ubuntu: ^ [17:16] re [17:17] om26er: I see, let me check [17:18] om26er: you want network-status interface [17:18] om26er: not network-observe [17:18] om26er: but I don't see any signals there [17:18] zyga-ubuntu: is that an unreleased interface ? Its not mentioned here https://snapcraft.io/docs/reference/interfaces [17:18] om26er: which is really unfortunate because it means people have to poll [17:18] yeah [17:19] I don't know what that list contains or who maintains it [17:19] om26er: on the up side I'm working on a way to ask snapd (via snap) about known interface [17:19] interfaces* [17:19] and any documentation they contain [17:33] * zyga-ubuntu supper [17:35] zyga-ubuntu: I'm surprised you haven't commented on the thread yet... [17:37] Pharaoh_Atem: I wrote a few sentences but it's I'm not done yet [17:37] Pharaoh_Atem: I still plan to but I want to write a good post, not something "yeah, mee too" [17:52] anyone around that can look at https://forum.snapcraft.io/t/snapd-2-26-9-and-conjure-up-no-longer-work/1348 [17:57] * zyga-ubuntu [17:58] stokachu: replied on the forum [18:00] zyga-ubuntu:thanks updated [18:02] who's calling me from spain [18:17] zyga-ubuntu:updated the forum post [18:18] Can not open /var/lib/snapd/seccomp/profiles//snap.conjure-up.lxd (No such file or directory) aborting: No such file or directory [18:25] stokachu: aha [18:26] stokachu: interesting, it should have used the new seccomp binary profiles? [18:30] zyga-ubuntu:thanks for the quick reply! [18:32] stokachu: I'm investigating [18:33] zyga-ubu1tu: thanks, just lemme know if you need me to do anything [18:57] stokachu: I updated my analysis, this is a (new) bug [18:57] stokachu: and a pretty obscure one at that [18:57] Ok [18:57] stokachu: I'm looking at fixing it ASAP [18:58] zyga-ubu1tu: ok thank you! I sent out a heads up email to juju lists so people are aware we're working on it [19:02] stokachu: thank you [19:07] stokachu: can you please report it and give me the launchpad bug number [19:08] zyga-ubu1tu:to the snappy project> [19:12] stokachu: on snapd please [19:17] I wrote a test case, trying it now [19:17] Ok [19:48] PR snapd#3598 opened: tests: add test case for classic-schizofrenia-bug [20:07] zyga-ubu1tu:https://bugs.launchpad.net/snapd/+bug/1704860 [20:07] Bug #1704860: classic confinement reexec [20:09] stokachu: thank you [20:26] zyga-ubu1tu:im at the sprint if you want to sync tom [20:27] i thought i saw you earlier [20:31] stokachu: I'm not at the sprint, I'll be there tomorrow though (all day) [20:32] zyga-ubu1tu:ok cool [21:27] * zyga-ubu1tu EODs [21:55] Bug #1555569 changed: [snaps] Show human-readable names for store apps

[21:55]