[00:01] Anyone know how to force install snap? Snap still thinks snappy-debug is still installed, but it's not. So I can't reinstall. [00:23] PR snapcraft#648 closed: Implemented `snapcraft release` [00:26] PR snapcraft#649 closed: Use '/usr/bin/env python3' [00:57] sergiusens: thank you for taking the time to look over my PR, I really appreciate it :) [01:20] PR snapcraft#644 closed: Feature/ant plugin test [03:08] PR snapcraft#646 closed: Feature/gradle plugin === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === PatrizioQON is now known as pbek === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === JanC_ is now known as JanC [06:00] PR snapd#1536 opened: tests: stop using hello-world.echo in the tests === chihchun_afk is now known as chihchun [06:28] morning [06:28] is there a way to lock source for each part to specific commit? [06:28] something like github.com/prokject/repo.git#v1.24? [06:38] you mean a specific tag? [06:39] if so, we have source-tag [06:42] yep [06:42] master is not stable :) [06:42] well in most cases or whatever repo defaults === chihchun is now known as chihchun_afk [06:48] I'm trying to create my first snap, but I have this weird problem where all the parts of the snap appear to work, but using the snap itself gives a bunch of java errors [06:48] http://pastebin.com/Tcy8KvyX [06:50] trijntje: Anything in "dmesg"? [06:51] trijntje: put "set" in your wrapper to see if the environment you have is what you expect. [06:54] also, I get a java crash error when I install it using devmode http://pastebin.com/AniwGkBp [06:55] qengho: as far as I can tell it only shows apparmor allowing the snap: http://pastebin.com/YJSmPVzf [06:57] trijntje: I see fonts in the name in that crash. Do you have fonts installed in your snap? [06:59] "fc" is probably fontconfig, too. [06:59] qengho: I don't. I tried asking about the errors in #java since I'm not a java programmer, but they said my java install was broken [07:00] It is broken. You don't have fonts installed. If you don't put it in the snap, it doesn't exist. === chihchun_afk is now known as chihchun [07:02] trijntje: You should run your program normally under "strace -f -e trace=file -o ~/trace ....", and see what font files it touches, and then "stage" the packages that provide those fonts. That will probably fix it. [07:03] they thought that ubuntu messed up java to the point that it didn't contain any GUI code at all [07:03] they explicitly said it wasn't a font problem, but ill give strace a go to be sure. I don't think they wanted to help to be honest [07:05] That "strace" will tell you a lot. Use "dpkg -S $filenamementionedinstrace" to map back to package names that could be in your snapcraft.yaml . [07:06] trijntje: Also add to your wrapper, [07:06] # Font Config [07:06] export FONTCONFIG_PATH=$SNAP/etc/fonts/config.d [07:06] export FONTCONFIG_FILE=$SNAP/etc/fonts/fonts.conf === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [07:24] hey hey [07:25] *wave* [07:27] 'sup. [07:34] hey pbek [07:34] good moring, dholbach ^_^ [07:34] how are things? :-) [07:35] hello people :) I have a problem with snappy playpen... [07:35] hikiko, what is it? [07:36] hello hikiko :) [07:36] dholbach, I made my first snap and snapcraft cleanbuild returns this error: http://pastebin.com/HEDwtuzD [07:36] hi tsimonq2 :) [07:36] how are you? [07:36] dholbach: testing, testing... :) [07:36] hm is it possible that snapcraft is picking up deps from host? [07:37] ld deps [07:37] great hikiko, 2:30 AM, just about to go to bed ;) [07:37] wow [07:37] so night :P [07:37] hikiko, I don't know what to do about it - it was asked here: http://askubuntu.com/questions/787258/internal-server-error-when-retrieving-files-from-the-archive-in-lxd [07:37] good night tsimonq2 [07:37] night dholbach :) [07:37] tsimonq2, time to get some rest! :) what are you doing here!! [07:37] good night! [07:38] sleep well, tsimonq2 ;) [07:38] qengho: strace found that the program accesses 3500 files with 'font' in the path [07:38] dholbach, I had similar issues in the past when I was trying to apt-get update from those repos I had to use mirrors [07:39] dholbach, is it possible to change archive.ubuntu.com.. to uk.archive or de somehow? [07:39] I'm feeding them to dpkg -S automatically now, lets see how many packages we end up with [07:39] tzzzz, sourceforge is down, bad if source files are hosted there for the snaps... :/ [07:40] dholbach, ping [07:42] liuxg, pong [07:42] hikiko, does --enable-geoip work there? [07:42] this will use my country's mirror right? [07:43] try it - it's the only idea I have right now [07:43] ok :) thanks dholbach!! [07:45] dholbach, the syntax should be: snapcraft cleanbuild --enable-geoip? [07:45] dholbach, today, I met a very strange problem. I have a snap app at https://github.com/liu-xiao-guo/rssreader_snap. if I add "network" into the plug, I do not have any AppArmor warnings, but the app does not get up (no UI). If I remove "network" into the plug, the app can be running, but with an error/warning like http://paste.ubuntu.com/19253339/. [07:45] I don't know - maybe it just works with "build"? [07:45] liuxg, I have never heard of anything like this before [07:46] dholbach, the UI of the app is documented at http://blog.csdn.net/ubuntutouch/article/details/51894469 [07:46] liuxg, which other plugs does it use? [07:46] liuxg, does "snappy-debug.security scanlog" say anything if you start the app? [07:46] dholbach, you may find it at https://github.com/liu-xiao-guo/rssreader_snap/blob/master/snapcraft.yaml [07:47] dholbach, do I need to run it separately in another terminal after the app is running? [07:48] separately in another terminal and start it before you start the app [07:50] dholbach, OK. I will do it. what pacakge should I install for snappy-debug.security? [07:52] snap install snappy-debug [07:53] dholbach, OK. thanks [07:54] PR snapd#1537 opened: tests: add env command to test-snapd-tools [07:55] dholbach, http://paste.ubuntu.com/19253833/, this is the message.. I do not know how to interpret it [07:56] liuxg, maybe it needs "network" and "network-bind"? [07:56] I'm just guessing [07:56] I'm not an expert on this one [07:57] dholbach, that solved my issue but now I have a new one... (sorry) I have some compile errors, what can I do in this case? fix them send the patch to the developers and try again in a few days? They forgot to include a file somewhere... [07:57] hikiko, or push to your own branch for now and use that as a basis for building [07:57] dholbach, in fact, it is just a rss reader app. I did not use it to pull any data yet (just use local static images). The app works well for the phone. [07:58] (and at the same time get them to include it if it's a more general issue) [07:58] dholbach, what is preferable? [07:58] liuxg, I have no idea - try it [07:58] hikiko, I don't know how that particular project operates, but I would send the patch (or PR or whatever they use) and use a local branch for continuing to snap - that way you don't get blocked for a few days waiting for an answer [07:59] hexchat [07:59] dholbach, the thing is that if I add network, the UI does not show up at all.. Anyway, I will try to add network-bind to see how it goes. [07:59] ok :) [07:59] thanks dholbach [07:59] anytime [08:11] PR snapd#1538 opened: snap-exec: fix silly off-by-one error [08:14] PR snapd#1537 closed: tests: add env command to test-snapd-tools === ant_ is now known as Guest47305 [08:40] PR snapd#1539 opened: tests: improve snap run symlink tests [08:41] PR snapd#1536 closed: tests: stop using hello-world.echo in the tests [09:04] zyga: https://launchpad.net/ubuntu/+source/snap-confine [09:05] jamiebennett, mvo: https://launchpad.net/ubuntu/+source/snap-confine has 1.0.36-1 now [09:05] jamiebennett, mvo: want me to tack ~16.04 on the end of the version number and upload it to the xenial queue? [09:15] mwhudson: I'm doing 1.0.36.1 now :/ [09:16] zyga: ok [09:17] zyga: i made some packaging changes beyond what is in master btw [09:17] (renaming the apparmor profile to usr.lib.snapd.snap-confine mostly) [09:18] mwhudson: thanks for doing that [09:18] mwhudson: do you have an idea on how I could use debian packaging in spread tests? [09:18] zyga: how do the spread tests work? [09:18] mwhudson: look at spread.yaml and spread-tests/**/task.yaml [09:18] mwhudson: you have a pre/post/execute script [09:18] zyga: but i guess git-recipe-builder or something? [09:18] mwhudson: you allocate some VMs [09:18] mwhudson: send all the code there [09:19] mwhudson: and run the scripts [09:19] zyga: spread.yaml where? [09:19] oh, master [09:19] (was in debian branch) [09:20] yep, sorry [09:22] zyga: for the package building stuff, basically you want a launchpad git recipe i think [09:23] take one git tree, transplant debian/ from some other git tree into it (that's https://launchpad.net/git-build-recipe) [09:24] then some fluff around guessing where the upstream tarball is or making it into a native package [09:24] can't remember where that code is [09:24] but cjwatson will know :) [09:26] git-build-recipe [09:26] oh, you already mentioned that [09:26] what code are you looking for? [09:26] all the upstream tarball etc. stuff is in git-build-recipe too ... [09:27] there's a tiny bit of glue in lp:launchpad-buildd (buildrecipe) to call it with suitable arguments [09:36] cjwatson: ah ok, good to know [09:36] zyga: use git-build-recipe, problem solved! [09:40] mwhudson: yep, I'll try to do that [09:41] zyga: this is a recipe that's a bit like the sort of thing you're doing, fwiw https://code.launchpad.net/~gophers/+recipe/golang-tip [09:45] sergiusens, can you reply to http://askubuntu.com/questions/797031/how-do-you-exclude-files-from-wiki-part-used-in-snapcraft-yaml# later on? [09:45] ^ or anyone else really [09:48] hey so it looks like snapd autopkgtests now fail because "snap install hello-world" doesn't provide hello-world.echo any more [09:48] is that plausible? [09:49] false alarm, maybe [09:55] mwhudson: should be unlikely, maybe PATH does not contain /snap? [10:07] Hello. How do I make image that will use custom store instead of the stock one? [10:09] PR snapd#1540 opened: spread.yaml, tests: replace hello-world with test-snapd-tools [10:14] dholbach, ping :) [10:14] I've fixed my snap [10:14] hikiko, pong :) [10:14] awesome [10:14] but [10:14] I am not sure I am using git correctly :p [10:15] dholbach, does git push fork master updates my pull request as well? [10:16] it seems so [10:16] yes, https://github.com/ubuntu/snappy-playpen/pull/171 (and https://reviewable.io/reviews/ubuntu/snappy-playpen/171#-) look like it :) [10:16] PR ubuntu/snappy-playpen#171: Add hexchat snap [10:17] ok :) sorry it's a bit messy, it's my first snap :) [10:17] no worries [10:17] take all the time you need :) [10:20] dholbach, should I add all the packages that are in build-depends of apt-cache show [10:20] :s/show/showsrc [10:20] yep, it should be a good start [10:21] if we want we can still trim down the list of build-packages later on [10:21] and for example some packages have a version like: [10:21] if you want to test the build in lxd locally, you can try "snapcraft cleanbuild" [10:21] debhelper (>= 9.0.0) [10:21] drop the version - snapcraft has no equivalent of this - it will use whatever is in xenial right now [10:21] yeah that was my question :D [10:21] debhelper (just as an example) is likely not required [10:22] ok let me fix it :) thanks a lot dholbach [10:22] thank YOU! [10:36] zyga: after i've used devtool refresh-bits i often end up in a situation where i can't run the original snapd again, i think because original snapd quits with "error: cannot downgrade: snapd is too old for the current system state" - is there any way to force this downgrade? [10:37] joc_: you could backup and restore /var/lib/snapd/state.json [10:38] zyga: so make a backup after a fresh install and use that in future to downgrade? [10:39] joc_: perhaps tie this into setup and restore steps? [10:39] yeah, could be useful [10:49] Hi. Is it possible to make system journal log to tmpfs instead of disk? [10:49] on snappy on rpi === chihchun is now known as chihchun_afk === hikiko is now known as hikiko|ln === chihchun_afk is now known as chihchun [11:14] PR snapd#1538 closed: snap-exec: fix silly off-by-one error [11:16] PR snapd#1535 closed: overlord: switch snapstate.Update to use ListRefresh (aka /snaps/metadata) [11:27] hmmm, so I uploaded an snap to the store in the beta channel, but "snap install --channel=beta libreoffice" does find anything. hints? [11:29] --devmode ? [11:30] (might be (falsely ?) implied by beta) [11:30] ogra: snap is _not_ devmode anymore [11:31] Sweet5hark, nope, but the store might require --devmode if you use other channels than stable ... not sure [11:32] ogra: soo, uhmm, when I have a confined app in a beta channel we cant install it confined? that sounds very wrong. [11:32] it definitely installs here with "sudo snap install --channel=beta libreoffice --devmode" [11:32] yes, thats why i said "falsely" ... definitely a store bug [11:33] ogra: meh. that doesnt install the right stuff. [11:33] i could understand edge to enforce --devmode ... but the staging channel for stable should not need it [11:34] oh, wait ... that channel might actually be "candidate" [11:34] ogra: ah! [11:34] ogra: the 5.2.0.2 I uploaded still has "manual review pending" [11:34] try if candidate works without devmode [11:34] oh ? it should tell you why [11:35] in the review details [11:35] sborovkov, not easily ... we use syslog and the configs are readonly iirc [11:35] ogra: I guess someone still needs to look under each bit if there is some even visual basic macro hidden by me there .... [11:36] oh you evil guy :) [11:36] Sweet5hark: Chipaca might help understanding what is happening there, might depend on which snapd version you have as well [11:36] jokes aside, the erros should usually give you some detailed info [11:37] btw that store UI is hugely misleading: it says both "Package status is Published" and "Status: Manual Review Pending" on the same page. :/ [11:37] (on the myapps page that is) [11:37] Sweet5hark, but for different revisions i guess [11:38] my internet isn't that fast, does it make sense to set up a local apt chache server when you build snaps with a lot of stage packages? Or will snapcraft bypass apt proxy settings? [11:39] ogra: I dont think its rejected or errored yet. At least "pending" does sound like "you suck" to me. Rather like "Im still looking at all the wonderful bytes you threw over the fence". [11:39] if you dont use cleanbuild it uses your local sources,.list iirc [11:39] so if that points to a package proxy the build shoudl too [11:40] Sweet5hark, well, if you click on the revision you can see it chug along [11:41] ogra: hohum, it says "reserved interface 'bluez' for vetted applications only security-snap-v2_app_plug_safe (libreoffice, bluez)" [11:41] Sweet5hark, oh, then you need someone to manually approve [11:42] (i think) [11:42] ogra: thanks, I might set up a proxy in that case [11:42] ogra: speaking teutonic, this is all bohemian villages to me ;) [11:42] ogra: ok, who do I need to get drunk? [11:43] trijntje, sudo snap install packageproxy ... then point to localhost:9999 [11:43] ;) [11:43] qengho: I added 58 font packages to my snap, which were accessed by the app when it runs outside the snap, but i still get the exact same error trace [11:43] Sweet5hark, i dont really know, beuno was always whom i harassed in that case ... but he is gone ... all others i wiould know are on vacation [11:44] Sweet5hark, usually popey or jdstrand ... perhaps tyhicks can chime in as security team person to nod it off [11:44] ogra: perfect, ill give it a go [11:45] Sweet5hark, out of curiosity, why does libreoffice need direct BT access ? [11:47] ogra: well, I havent tested it yet. but it will need bluetooth in the end for https://play.google.com/store/apps/details?id=org.libreoffice.impressremote [11:47] ah ! === hikiko|ln is now known as hikiko [11:47] yeah, that makes sense (also for physical pointers) [11:49] I'm trying to create my first snap, but I have this weird problem where all the parts of the snap appear to work, but using the snap itself gives a bunch of java errors [11:49] http://pastebin.com/Tcy8KvyX [11:50] dmesg only shows a bunch of apparmor ALLOW lines, and including 58 font packages in the snap did not solve this issue [11:50] trijntje, sis you try using -jar instead of -cp in your launcher (i think thats what i asked last yesterday) [11:50] s/sis/did/ [11:51] i thinnk the first one is java itself ... the second error you have is clearly fonts related ... [11:52] * ogra goes for a break [11:53] popey, jdstrand, tyhicks: Anyone in for nodding off the libreoffice snap? [11:54] ogra: I'll give that a go now, thanks. I must have missed that yesterday === ehbello_ is now known as ehbello [12:09] dholbach, when I run snapcraft clean and snapcraft build locally I see no errors and then I push to git and there are still unsolved dependencies :s [12:11] hikiko, that's because you very likely have lots of local build-dependencies already installed :) [12:12] and the build in travis uses a clean ubuntu container [12:12] so you need to specify each and every package in there [12:12] hello hello [12:12] Sweet5hark, what's up [12:13] Sweet5hark, doing a quick scan of the backlog: you have a non-devmode snap in the beta channel that only installs when you specify --devmode? [12:13] Chipaca: could you nod off the "libreoffice.canonical" package in the store? it hangs there because of wanting bluez. [12:14] Chipaca: https://myapps.developer.ubuntu.com/dev/click-apps/5179/rev/2/ [12:14] I don't know what nodding off something is in this context, but I don't have admin on the server so probably no :-) [12:15] Chipaca: well, it says "pending manual review", so hmm ... [12:16] hikiko, you can mimic that behaviour with "snapcraft cleanbuild", that builds locally in a container as well [12:17] ogra, I know but cleanbuild fails at downloading because it can't find some repos [12:17] Sweet5hark, noise][ , ev or maybe jdstrand can help [12:17] hikiko, oh ? [12:17] Sweet5hark, but what was the thing about beta and devmode? [12:18] ogra give me some time to run cleanbuild and paste the error [12:18] hikiko, do you use any remote parts like the desktop launchers ? iirc thats a known issue ... repos shoulld be fine though [12:18] I pasted it here before but I can't find the link [12:19] ogra, maybe, how can I test that? [12:19] ogra, it's my first snap I might have done things accidentally... [12:19] well, you would know if you added something "desktop/..." to your snapcraft.yaml i assume :) [12:19] oh no I didn't [12:19] ah, then it shouldnnt use the launchers that make it fail [12:20] norml deb repos should just work === chihchun is now known as chihchun_afk [12:21] ogra, I wonder if the problem is unrelated to snap, because sometimes apt-get update fails as well if I use the original repos instead of mirrors, if I run snapcraft --enable-geoip there's no problem but that's not equivalent to cleanbuild :/ [12:22] yeah, weird [12:25] ogra: meh. So no easy solution? Cause all this writes are going to be killing SD card on RPI very fast. [12:26] i doubt that, unless you have an app that dos excessive writing [12:26] (fox that app then :) ) [12:26] *fix [12:27] usually there isnt much log activity [12:28] App does not do a lot of writing. Devices uptime is 24 hours a day though... And we really want to minimize the writes since Sd cards dying is like the most common issue when using RPI. On raspbian we were just logging to tmpfs. Oh well. [12:28] when we still had "snappy config" you cuold modify some parameters ... i guess once that feature comes back we will add something for this again ... but i dont know where that stands ... perhaps Chipaca knows [12:28] config is kyrofa these days :-D [12:28] ah [12:28] * Chipaca leveled up his deflection skills [12:28] * ogra needs to get the wiringin his brain fixed for that evemntually :) [12:29] *wiring [12:29] ogra, you have the deflection skills of a horny buffalo :-D [12:29] lol [12:34] Hello again, How do I build image that uses private store? [12:39] sborovkov, you mean with a private kernel and gadget ? i dont think you can yet ... ubuntu-image will support that once it replaced ubuntu-device-flash [12:40] (i know it was possible in 15.04, but not yet for the 16 series due to architectural changes) [12:43] https://paste.ubuntu.com/19269020/ ogra [12:43] that was the error [12:43] well [12:43] 500 Internal Server Error [12:43] i blame IS [12:44] ogra which are the steps of cleanbuild? maybe I can follow them manually [12:44] or the infrastructure or some such ... that should be a temporary issuue [12:44] good news :D [12:45] instead of "snapcraft" you run "snapcraft cleanbuild" to build your package :) [12:45] thats all ... it wll magically create an lxc container and build in there [12:46] ogra: oh, that's interesting. Right now I magically create an lxc container myself and do all my snapcraft thingies in there, then lxc pull the snap out onto my system afterward. :) [12:46] nhaines, well, it has its drawbacks ... like remote snapcraft parts doo not work atm [12:47] (so you cant build anything that uses the desktop launchers) [12:47] should be fixed with the next snapcraft i think [12:49] ogra: slightly disappointing, but luckily I'm just trying (and always failing) to work on a little personal project that only needs debian packages, and also the good news is that there's always a new snapcraft around the corner. :D [12:49] yeah, sergiusens is tireless :) [12:51] PR snapcraft#651 opened: Re-based feature/gradle-plugin [12:54] ogra: do you have any suspicion on when the RPi2 will get an all-snap image? [12:54] nhaines, not yet, no ... we're waiting for ubuntu-image ... which waits for some snpd and store features in turn [12:54] *snapd [12:55] Okay. My RPi2 is sitting in its little Ubuntu case, just waiting. :D Although it's not a half bad retrogaming machine until then, either. [12:55] Full bad desktop machine (Ubuntu MATE is impressive, but mouse lag drives me crazy on any computer.) [12:57] ogra: I mean. I can build custom ones manually. Just need to unsquashsf current one and make changes and squashfs again. We have private store. 'screenly'. I want image in which it's turned on. So I can install snap from there [12:58] sborovkov, right, as i said, i dont think the 16 series supports that yet ... [12:59] since it will change the gadget and kernel snaps to use assertions, whichin turn defin the store your image uses [12:59] (as i understand it) [12:59] ogra: Is it possible to connect to private store from console at least? I can work with that [12:59] Oh [13:00] ubuntu-image will support tht once it is there ... but currently we are waiting for the store and snapd to have full support for assertions [13:07] I have an app with the interfaces at http://paste.ubuntu.com/19270412/. when I run my app, I still get the some errors like http://paste.ubuntu.com/19270507/. However "snappy-debug.security scanlog" does not give me much info at http://paste.ubuntu.com/19270592/ what could the cause of it? thanks [13:10] liuxg, well, the -control and -manage interfaces obviously arent autoconnect ones, you need to manually connect them [13:10] (seems -observe too, which i find a bit weird, but it might allow access to the MAc address or so) [13:11] ogra, thanks. May I know how to connect them? In fact, the app is just a very simple app, it uses network manager to create a cache. I think network manager might be the one needed.. [13:12] you use snap connect :) [13:12] ogra, what I am currently doing is just try to add all of them. after that, I can isolate the problem. where can I find an example for it? I need to do it in terminal, right? [13:13] calling the command without options gives you an example [13:15] something like: sudo snap connect foobar:network-control ubuntu-core:network-control [13:16] ogra, thanks! I will have a try! [13:17] I'm still messing with my first snap, now I'm getting apparmor DENIED in dmesg http://pastebin.com/1BEBRS3G [13:18] Even though I believe I have given the snap permission to acces home [13:18] plugs: [x11, network, network-bind, home] [13:18] trijntje, no dotfiles ... just home data [13:18] ogra, many thanks for your help. You are right. now the error disappears. I used the command "sudo snap connect rssreader-app:network-manager ubuntu-core:network-manager" to do it. [13:19] you need to make sure to point your XDG dirs to the SNAP_USER_DATA dir [13:19] (i gve you the wrappper of jtiledownloadeer yesterday, didnt i ? that should have the needed vars) [13:20] liuxg, now the million dollar question ... wh does your rss reader need to manage your network connection ? :) [13:20] *why [13:20] ogra: yes, looking them up now. Thanks for all the hand holding, I'm going to be so happy when I finally get my first snap running [13:20] ogra, the source code of the project is at https://github.com/liu-xiao-guo/rssreader_snap [13:20] trijntje, me too :) [13:21] ogra, in the file https://github.com/liu-xiao-guo/rssreader_snap/blob/master/src/rssreader/main.cpp, it uses NetworkManager to create a cache for the app. [13:22] ogra, I think normally it should not need that at all [13:22] wow, weird ... why does it need NM for creating a disk cache [13:22] yeah [13:23] ogra, could you please take a look at MyNetworkAccessManagerFactory::create(QObject *parent) function. it caches the pictures if any during the network connection. [13:23] i'm really bad at C++, but i can take a look indeed :) [13:23] ogra, in phone app, only network is needed. I do not know why it needs to have network manager in the snap app. [13:23] ogra, OK. thanks for your help indeed! [13:24] hi [13:24] is it mandatory that the "system-boot" partition be named like this? [13:24] system-boot / writable [13:24] I know that writable is a requirement and the label is used [13:26] liuxg, well, on the phone there are less fine grained interfaces ... once the phone uses snappy it will be similar [13:27] ysionneau, [13:27] ogra, yeah, I agree with you. that is the ideal situation. the app was orignally a phone app :) [13:27] ogra@styx:~/Devel/packages/initramfs-tools-ubuntu-core-0.7.43+ppa6$ grep system-boot scripts/ubuntu-core-rootfs [13:27] boot_partition=$(findfs LABEL="system-boot" 2>/dev/null || :) [13:27] tmpboot_mnt="/tmpmnt_system-boot" [13:27] yes ... [13:29] ysionneau, note that this might change with ubuntu-image though ... not sure if we can keep the partition names wiith that [13:30] are there known issues with snap login not working? [13:30] worked yesterday for me, today not so much [13:32] wierd, stopped my vpn and things work. nevermind [13:35] A little lost (sorry) but I'm looking for some official collaboration from Canonical for a few forthcoming educational events, workshops. Anyone know who the contact would be for embedded/IoT snappy education? [13:36] thibautr_, ^^ ? [13:41] ogra: thanks [13:46] mvo, Hi Michael! [13:48] mvo, I had some issues running the new classic, so I just filed a bug. [13:48] pcoca: hello, what is the bugnumber? [13:49] mvo, #1602693 [13:49] Bug #1602693: classic.create error: chroot: failed to run command ‘/var/lib/classic/enable.sh’: No such file or directory [13:49] heh [13:49] pcoca: what version of ubuntu-core do you have installed? [13:49] just answered [13:49] you missed switching ubuntu-core to edge [13:50] pcoca: what ogra said is also my theory [13:50] (as the version in the bug tells ... i only saw the version after asking :) ) [13:50] ogra, mvo: thanks! === chihchun_afk is now known as chihchun [13:51] PR snapd#1457 closed: snapstate: drop revisions after "current" on refresh [13:51] mvo, ogra, I added the comment and changed to bug to invalid. [13:52] PR snapd#1444 closed: many: allow refresh to locally installed snaps (with data copy) [13:53] pcoca, cool, thanks ("incomplete" would have auto-closed it in 60 days anyway :) ) [13:53] pcoca: your welcome, thanks for testing this new feature, much appreciated [13:54] +1 [13:57] PR snapd#1507 closed: cmd: add buy command [14:15] PR snapcraft#651 closed: Re-based feature/gradle-plugin [14:23] dholbach answered [14:24] thanks sergiusens! [14:24] aha! [14:28] PR snapd#1506 closed: store/auth: add helper for the macaroon refresh endpoint [14:37] sergiusens, should http://paste.ubuntu.com/19276737/ work? [14:37] I'm getting this: Issues while validating snapcraft.yaml: The 'parts' property does not match the required schema: Additional properties are not allowed ('desktop/qt5' was unexpected) [14:50] dholbach yeah, you found a bug [14:50] thanks sergiusens - shall I file it? [14:50] dholbach in theory, the part/subparts thing was not supposed to be something we end user exposed (it was for dependencies to statisfy the main part) [14:50] dholbach yeah, I guess we need to support it [14:51] ok, thanks - I'll file a bug [14:53] https://bugs.launchpad.net/snapcraft/+bug/1602728 [14:53] Bug #1602728: Redefining elements of remote parts does not work [14:53] thanks [15:36] PR snapd#1541 opened: snap-exec: add proper integration test for snap-exec [15:39] hey ogra. wanted to say thanks for pointing me to the nethack in your repository for example of locale installation and initialization. helped me move forward in my work on postgresql package. [15:39] I'm now looking at this error message: [15:39] 2016-07-13 15:32:13 GMT [480-1] LOG: invalid value for parameter "log_timezone": "PST" [15:39] which I understand now is most likely due to missing timezone, much like the problem with locale [15:46] I'm wondering if I can get away with just including a tzdata to stage-packages.. [15:49] iliv, there i a "timezone-control" interface your snap could perhaps use (though what you really want is rather something to read the TZ, not to control it) [15:50] i assumewe are missing an interface here [15:50] tyhicks, ^^ shouldn't we have a timezone-read interface additionally to -control ? [15:53] ogra: yes, I'd think so [15:53] iliv, mind filing bug for this? [15:54] sure. however, I don't know where to do so. [15:55] see the channell topic :) [15:55] and add the "snapd-interface" tag to the bug [15:55] so it shows up in the right list [16:09] ogra, tyhicks done https://bugs.launchpad.net/snappy/+bug/1602752 [16:09] Bug #1602752: Add support for timezone-read interface [16:10] in the gadget snap, you can specify a set of snaps to be already installed in the generated by UDF ubuntu image [16:10] Bug #1602752 opened: Add support for timezone-read interface [16:10] can you specify that one snap has to be installed in devmode ? [16:14] ysionneau, yes, you can use confinement: devmode in snapcraft.yml, which is a per snap configuration setting. you can also use snap install package.snap --devmode [16:18] iliv, thanks !! [16:23] iliv: how can I set devmode in the snapcraft.yaml ? === JanC is now known as Guest81961 [16:25] PR snapd#1533 closed: many: add authcontext to handle user updates to state [16:34] ysionneau, add after description: this "confinement: devmode" (without quotes) [16:35] how does one rebuild a package after adding only plugs in snapcraft.yml without actually rebuilding any of the parts? [16:38] thx iliv [16:40] ysionneau, be sure to install your package with: snap install package.snap --devmode. I may be wrong but I think you need both confinement: devmode and a package installed with the --devmode flag. [16:41] I know that just snap install --devmode is enough [16:41] but question is, is confinement: devmode enough when it is listed as a package in gadget snap ? [16:41] so that it is automatically installed ... in devmode [16:42] that is something I'm not really sure about [16:43] somebody else here should be able chime in and give a better answer === chihchun is now known as chihchun_afk [17:02] I am trying to create snap for go application that needs "GO15VENDOREXPERIMENT=1" environment variable. How can that be done? [17:12] PR snapcraft#650 closed: feature/maven-targets w/ re-based maven.py [17:15] PR snapcraft#637 closed: Do not clean before running the snaps tests === mup_ is now known as mup [18:17] so if I have `daemon: simple` in my app entry, then it should install as a system service when it gets installed right? [18:18] jacekn [18:18] jacekn: the only way that I know how, is to add a shell script to setup the environment [18:20] jacekn: also, I dont know go, but most languages have the ability to set their own environment, for example in python you can do os.environ["GO15VENDOREXPERIMENT"] = "1" not sure if thats what you want or not [18:22] is it no longer possible to run snap commands from inside a snap? I have a test that is inside a snap, and even something as basic as 'snap list' fails with: [18:22] error: cannot list snaps: cannot unmarshal: json: cannot unmarshal string into Go value of type int [18:22] or is this just a bug in the version of snap it uses? I also noticed that /usr/bin/snap that it sees is different from the one I have on the host system [18:23] or maybe it uses the one from ubuntu-core? [18:27] ah, yes. If I copy the snap binary from the host system, it just works. [18:37] do snappy daemons log stdout anywhere? [18:45] nvm, found it [18:51] PR snapcraft#613 closed: parser - Don't allow duplicate wiki entries [19:12] PR snapcraft#652 opened: Demo/gradle [19:15] PR snapcraft#653 opened: Implement `snapcraft push` [19:15] elopio josepht https://github.com/snapcore/snapcraft/pull/653 [19:15] PR snapcraft#653: Implement `snapcraft push` [19:51] PR snapcraft#571 closed: Add setuptools tests_require [20:06] PR snapcraft#655 opened: Gradle plugin fix for command invocation [20:31] hey there, i just installed snappy core on my raspberry pi. now i have to login but i don't know username nor password :( [20:35] luaf: ubuntu:ubuntu if I'm recalling correctly [20:37] im considering running core inside of docker, inside of core [20:37] core inseption [20:38] core inception [22:51] PR snapcraft#656 opened: Use requirements files in travis tests