[00:54] <dougb> I am needing guidance/help with enabling the built-in serial ports, other than ttyO0, on the BeagleBone Black, using Snappy.
[01:01] <wililupy> So I'm taking the latest build of snappy 16.04 for a spin, and I need to change some things in the /etc directory, particularly the motd since it says to still use the 'snappy' command even though it is snap now. Also, more importantly, I need to change /etc/default/grub.d/50-system-image.cfg so that the right tty is referenced and I get a console screen (Its supposed to be tty0 not tty1 for my device. Any ideas how to make
[01:01] <wililupy>  these changes?
[02:21] <sergiusens> elopio hey, mind giving 2.8.4 a spin? Should be in the archives
[08:45] <ogra_> sigh ... no mvo ?
[08:46] <davmor2> ogra_: you mean you broke mvo too, man you're evil
[08:46] <ogra_> no, but the snappy builds are broken with the final kernels that landed tonight
[08:48] <pedronis> ogra_: he is off today and tomorrow (not sure about Monday)
[08:48] <ogra_> damned
[08:52] <dpm> dholbach, davidcalle, popey, I started playing with moving snappy-playpen to github, and the list of apps last night. I created a repo for testing purposes on my personal account, which we can then move to ubuntu-core or any other suitable place
[08:53] <dpm> dholbach, davidcalle, popey -> https://github.com/dplanella/snappy-playpen - and also, check out 'sudo snap install notes-dpm', freshly added to the playpen and store ;)
[08:55] <davidcalle> dpm: nice!
[09:01] <sergiusens> good morning folks
[09:02] <popey> dpm: ooh
[09:02] <sergiusens> ogra_ heh mvo likes to sleep in the morning :-P
[09:02] <davidcalle> dpm: wow, custom plugin, desktop file and works flawlessly!
[09:03] <sergiusens> dpm github seems to be taking over ;-)
[09:03] <dpm> yeah, I'm now sold on plugins :)
[09:04] <ogra_> sergiusens, thats news to me :)
[09:04] <sergiusens> dpm makes things easy, you might be our guest speaker on the 26 ;-)
[09:04] <dpm> :)
[09:10] <dpm> sergiusens, now that I've got you here... on https://github.com/dplanella/snappy-playpen/blob/master/notes/snapcraft.yaml#L33 I had an issue whereby snapcraft's copy plugin wouldn't take a path relative to the YAML file location. Strangely enough, L32 works. Is this a bug in the plugin, or something I did wrong?
[09:11] <dholbach> dpm, cool
[09:13] <sergiusens> dpm let me see
[09:14] <sergiusens> dpm so the `notes` part has no install rule?
[09:23] <sergiusens> dpm maybe try and remove this line https://github.com/dplanella/snappy-playpen/blob/master/notes/parts/plugins/x-qmake.py#L34 and here call super().build() at the end of method instead
[09:23] <sergiusens> I don't see you calling `make install`
[09:26] <sergiusens> dpm if you still want to use the copy plugin like this; the "relative path" is relative to "source" but you didn't specify that key, so it is relative to "TOPDIR" so it would be `parts/notes/build/...`
[09:30] <ppisati> ...
[09:30] <ppisati> Starting new HTTPS connection (1): search.apps.ubuntu.com
[09:30] <ppisati> "GET /api/v1/search?q=package_name%3A%22ubuntu-core%22&fields=download_url%2Canon_download_url%2Cdownload_sha512 HTTP/1.1" 500 None
[09:30] <ppisati> search results {'oops-id': 'OOPS-084b4881d5044f92c480531034deb772'}
[09:30] <dpm> sergiusens, yeah, there isn't an install target in the Makefile, I just had to move the binary around
[09:30] <ppisati> i was able to build a kernel, and then it started to error out this way
[09:33] <dpm> sergiusens, and re: the copy plugin. That was exactly how I understood it, however specifying `parts/notes/build/...` did not work, that's why I had to add the '../../', but strangely on L32 it does work relative to TOPDIR
[09:38]  * Gunther_ is courious how to add an interface Slot so hat his app snap is able to access a specific devise in /dev ...
[09:44] <sergiusens> Gunther_ talk or write to zyga when becomes available
[09:45] <sergiusens> Gunther_ he said he'd write a blog post sometime this week or Monday with the process
[09:45] <Gunther_> sergiusens: ok thanks - I am currently completely blocked because of the remove of old-security ....
[09:45] <sergiusens> dpm heh, as a hack it works then; ideally I'd propose a PR so they get an `install` rule :-)
[09:47] <dpm> sergiusens, yeah, I was actually thinking of doing that, but they've got very simple qmake cross-platform build, and I didn't want to get into learning about win32 and osx...
[09:48] <sergiusens> dpm since your custom plugin is not really specific to qmake (since you add the qt5 config file), it might be better to just do it in there.
[09:48] <dpm> ack
[09:49] <dpm> sergiusens, so what do you think about adding the qt5 config file, is that the right place and way to do it?
[09:50] <sergiusens> dpm at the end of build, just do `os.makedirs(os.path.join(self.installdir, 'bin')); os.link(os.path.join(self.builddir, 'bin', 'Notes'), os.path.join(self.installdir, 'bin', 'Notes'))
[09:50] <dpm> ok, will give it a go
[09:51] <sergiusens> dpm for qt5config I'd just create a part for it, with the config file in some source tree and put the part on the wiki (then everyone can consume from it)
[09:52] <dpm> sergiusens, yeah, the issue is that the file would then have hardcoded arch-specific paths
[09:53] <sergiusens> dpm right, forgot about that; a file and a Makefile then ;-)
[09:54] <dpm> sergiusens, but it seems a lot more self-contained to do it in the plugin code, doesn't it?
[09:55] <sergiusens> dpm yeah; the problem is; what if you want qmake but not this? what if you use cmake? what if, for unknown reasons you have a qmake project that uses GTK?
[09:56] <sergiusens> dpm those are the questions I ask myself which make me just doubt adding this as an official plugin. I would of done it already if I had these questions answered
[09:56] <sergiusens> I am still hoping to dream of a solution, wake up and implement :-)
[09:56] <dpm> sergiusens, how would I solve this with a Makefile, though? Could it not be solved with passing configflags to the plugin?
[10:25] <asac> Gunther_: couldnt you go back to the previous build until the rework has settled in?
[10:29] <sergiusens> dpm for some reason I cannot login to wiki.ubuntu.com and add this part, so here you go https://github.com/sergiusens/qt5conf
[10:30] <sergiusens> dpm copy paste the "part" and change source to this github repo
[10:30] <sergiusens> dpm or tell me how to log into the wiki, or if you have access, please add my part :-)
[10:31] <dpm> sergiusens, what was your exact issue logging in? Timeout?
[10:32] <sergiusens> dpm after oauth it never comes back; I didn't wait long enough; but did wait at least 30s to 60s
[10:32] <dpm> surprisingly, after the experiences in the last few days, I could login after a few seconds. In any case, where in the wiki does the part go?
[10:34] <dpm> sergiusens, ^
[10:38] <dpm> ok, found https://wiki.ubuntu.com/Snappy/Parts but I've no idea what the syntax is nor how to use it
[10:39] <dpm> I'll give it a go adding it
[10:39] <sergiusens> dpm: easy; let me do it for you in a pastebin
[10:41] <sergiusens> dpm http://pastebin.ubuntu.com/15962652/
[10:41] <sergiusens> dpm then in your parts definition, for notes, just add `after: [qt5conf]`
[10:42] <sergiusens> dpm that's all you need; no need to define the part locally
[10:43] <dpm> sergiusens, is there a reason why you removed `build-packages: [dpkg-dev]` on the paste to the wiki?
[10:43] <sergiusens> dpm err, my mistake :-)
[10:44] <dpm> sergiusens, ok, https://wiki.ubuntu.com/Snappy/Parts
[10:44] <sergiusens> dpm I am thinking of a feature to come in snapcraft where we can point to origins and it will consume the snapcraft.yaml from there
[10:45] <sergiusens> dpm I wish there was something more lightweight than dpkg-dev to get the arch triplet (we got rid of the dep in snapcraft and rolled our own, I might make a shell callable for that or export the envvars'
[10:51] <Gunther_> snap interfaces lists interesting things for me : network-control and network-observe. For example I would like to restart or reconfigure a running daemon (in snap) if a network interface has changed. Are there examples how to use those slots?
[11:04] <dpm> sergiusens, I just tried the wiki part, but it seems to fail here: http://pastebin.ubuntu.com/15962859
[11:06] <sergiusens> dpm did you remove your qt5 config logic from the qmake plug?
[11:06] <sergiusens> I need to run for a bit, will bbiab
[11:06] <dpm> sergiusens, I did
[11:06] <dpm> ok, np
[11:50] <ogra_> ppisati, !!!
[11:51] <ogra_> Setting up linux-firmware-raspi2 (1.20151118+b70b451-0ubuntu1) ...
[11:51] <ogra_> cp: cannot create regular file '/boot/firmware/bootcode.bin': No such file or directory
[11:51] <ogra_> dpkg: error processing package linux-firmware-raspi2 (--configure):
[11:51] <ogra_>  subprocess installed post-installation script returned error exit status 1
[11:51] <ogra_> Setting up linux-image-raspi2 (4.4.0.1009.9) ...
[11:51] <ogra_> Errors were encountered while processing:
[11:51] <ogra_>  linux-firmware-raspi2
[11:51] <ogra_> what is that ?
[11:51] <ogra_> (breaks the image builds)
[12:52] <svij> hm, I snapcrafted a package, installed it, removed it. But "snap list" still list it? Do I have to do something else next to "snap remove <snap>"?
[12:52] <svij> oh now it's gone
[12:52] <svij> weird.
[12:58] <jcastro> svij: I noticed that yesterday too, there's a bit of a delay from when it gets removed and when it stops being listed on that list
[12:58] <svij> jcastro: ah okay
[12:58] <svij> and the docs on https://developer.ubuntu.com/en/desktop/get-started/ are (already?) wrong.
[12:58] <svij> had to install ubuntu-calculator-app without ".ubuntucoredev"
[12:59] <svij> also, running ubuntu-calculator-app.calculator only returns a "command not found"?
[12:59] <davidcalle> svij: hmmm
[12:59] <dpm> jcastro, on the "a new app a day" spirit, I uploaded a new app for you, check out notes-dpm ;)
[12:59] <davidcalle> Youare right about ubuntucoredev
[12:59] <jcastro> dpm: already installing it. :D
[13:00] <dpm> :)
[13:00] <jcastro> slow dl though, I hope the snappy store is on the same infra as the rest of our stuff
[13:00] <popey> http://popey.mooo.com/mirror/clicks/graph_snaps.png it's making a difference to the graph already :)
[13:00] <dpm> svij, davidcalle, I'll update the get-started page, thanks for the heads up
[13:00] <popey> jcastro: there's issues, they're firefighting atm
[13:00] <svij> dpm: thx!
[13:00] <jcastro> popey: ack
[13:00]  * svij still doesn't know how to start the calculator…
[13:01] <jcastro> it should just show up in your dash
[13:01] <svij> I'm using kde
[13:01] <svij> will that also show up there?
[13:01] <jcastro> I believe so
[13:01] <jcastro> XDG_DATA_DIRS=/usr/share/ubuntu:/usr/share/gnome:/usr/local/share/:/usr/share/:/var/lib/snapd/desktop
[13:01] <davidcalle> svij: should be
[13:02] <svij> echo $XDG_DATA_DIRS
[13:02] <jcastro> I think all DEs use that variable. Though, I was logged in so long I never picked up that change when they changed it a few days ago, I had to log out and back in
[13:02] <svij> /usr/share//usr/share/xsessions/plasma:/usr/local/share/:/usr/share/
[13:02] <jcastro> try logging out and back in
[13:02] <svij> ah
[13:03] <dpm> svij, davidcalle, updated https://developer.ubuntu.com/en/desktop/get-started/?edit_off#setting-it-all-up
[13:05] <popey> https://blog.mozilla.org/futurereleases/2016/04/21/firefox-default-browser-for-linux-users-ubuntu-new-snap-format-coming-soon/
[13:05] <popey> hello!
[13:05] <davidcalle> dpm: another thing to note down somewhere (that I'm very happy about): if a command matches the snap name, you don't need to $ <snap>.<command> to run it, just $ <snap>
[13:05] <kyrofa> Good morning
[13:06] <davidcalle> Morning kyrofa
[13:06] <jcastro> popey: wow so Mozilla itself will publish right into the store?
[13:06] <davidcalle> popey: fireworks!
[13:07] <popey> i guess so
[13:08] <dpm> davidcalle, ah, nice. I should probably change ubuntu-calculator-app.calculator to ubuntu-calculator-app.ubuntu-calculator-app then
[13:08] <dz0ny_> svij: fresh xenial works as described on https://developer.ubuntu.com/en/desktop/get-started/
[13:08] <jcastro> dpm: I get QXcbConnection: Could not connect to display :0
[13:08] <jcastro> Aborted ore dumped)
[13:08] <jcastro> with the notes app
[13:08] <dpm> argh
[13:09] <davidcalle> jcastro: it was all a trap jcastro, now you help debugging :)
[13:09] <jcastro> heh
[13:09] <dpm> jcastro, I've seen that happening for some apps, but it tended to go away as snapd changes were landing
[13:09] <dpm> so I've never figured out what caused it
[13:09] <dpm> davidcalle, does the notes app work for you?
[13:10] <davidcalle> dpm: yep
[13:10] <dpm> phew
[13:10] <svij> /var/lib/snapd/desktop is now in XDG_DATA_DIRS but I still get a "command not found" on ubuntu-calculator-app.calculator
[13:12] <svij> dz0ny_: I've just updated vom 15.10
[13:14] <svij> hm, ok, starting calculator from the kde-menu works. Now I still don't know what misses to start my own snap, which I just snapcrafted… (cli tool)
[13:15] <jcastro> svij: cool, good to know the GUI bits work at least
[13:16] <dz0ny_> http://i.imgur.com/gWBFB13.png
[13:16] <dpm> svij, strange that the menu works but not the CLI... For your own snap, can you put it on a branch somewhere for other people to look at it and perhaps help you?
[13:16] <dz0ny_> works without problems on ubuntu gnome flavor
[13:16] <dpm> dz0ny_, \o/
[13:16] <dz0ny_> and in fish shell :D
[13:16] <svij> dpm: I think the snap is fine, I cant run the calculator from the command line either
[13:17] <dz0ny_> svij: maybe broken bashrc?
[13:17] <dpm> svij, perhaps something to do with how your terminal client handles PATH?
[13:17] <dz0ny_> works from bash too
[13:18] <kalikiana> svij: You'll want to make sure your PATH contains /snap/bin
[13:18] <svij> kalikiana: aha! It doesn't
[13:18] <svij> so the question is: Why?
[13:19] <kalikiana> If you have a custom shell config like me it's obvious :-)
[13:19] <svij> well, I was using zsh, same for bash.
[13:19] <ogra_> svij, /etc/profile.d/apps-bin-path.sh sets it
[13:19] <ogra_> from the snapd package
[13:20] <svij> ah okay
[13:20] <svij> it works now
[13:20] <svij> thanks kalikiana
[13:20] <kalikiana> svij: I override PATH completely to be in charge of the order of evyerthing I put in there. So regardless of the shell it won't pick anything up from the system.
[13:20] <ogra_> how evil
[13:20] <ogra_> :)
[13:20] <kalikiana> :-D
[13:21] <ogra_> you dont trust your packagers, eh ?
[13:21] <svij> ogra_: he's that evil, he doesn't even trust shoes! :)
[13:21] <dpm> kalikiana, now that I've got you here.... how do we solve the LocalStorage AppArmor denial in http://i.imgur.com/gWBFB13.png ?
[13:21] <ogra_> svij, thats fine ... keeps you grounded :)
[13:22] <svij> ogra_: hehe
[13:23] <kalikiana> ogra_: Not sure though why I started doing this originally. It's been a long time that anything changed PATH in a meaningful way. I might try what breaks (or not) if I only extend it.
[13:24] <kalikiana> dpm: Lemme check that here. Now that I've got the snaps working
[13:25] <dpm> kalikiana, awesome, thanks. That's the last bit we need to get the calculator and clock apps fully working. Right now they cannot write to their databases, so they're more for showcasing purposes than everyday use. If we can fix that, it will open the door for the rest of the core apps to be snappyfied and available on the desktop
[13:25] <dz0ny_> webdm seems broken thou
[13:25] <dz0ny_> curl http://localhost:4200/api/v2/packages/                                                                        čet 21 apr 2016 15:24:54 CEST
[13:25] <dz0ny_> "Error: cannot list snaps: not found"
[13:26] <dz0ny_> aka snappyd
[13:26] <dpm> kalikiana, also note I tried this: http://bazaar.launchpad.net/~dpm/ubuntu-calculator-app/snap-all-things/view/head:/calculator.wrapper#L42 on the wrapper that's called to set the env variables before starting the calculator, but it didn't seem to work
[13:30] <kalikiana> dpm: What's snappy-qt5?
[13:30] <dpm> kalikiana, it's a config file that needs to be shipped for qt-selector to work and use qt5. Otherwise Qt apps don't start
[13:32] <dpm> kalikiana, `cat /snap/ubuntu-calculator-app/current/etc/xdg/qtchooser/snappy-qt5.conf`- I don't know much about it other than you need it. Mirv is the expert on qt selector
[13:34] <kalikiana> dpm: Are you mixing up XDG_DATA_DIRS with XDG_DATA_HOME by any chance? LocalStorage uses _DATA_HOME and that matches the error message, which tries to use $SNAP which is not writable
[13:36] <kalikiana> Ah so the snappy-qt5 just makes it look inside the snap, so it's not yet another platform plugin
[13:37] <Gunther_> Anyone seeing something like that? /snap/bin/trionet.trionet-srv  /dev/pts/ptmx does not exist. errmsg: No such file or directory
[13:37] <sborovkov> Hello, is there some documentation on how to set read/write access with new security permissions? (I had it set that with security-override before, but that seems to not work anymore).
[13:37] <Gunther_> the same snap is working flawlessly in xenial. The error is happening in "allsnaps"
[13:41] <dpm> kalikiana, I had tried setting _DATA_HOME, but that seems to then break fontconfig (apps show up with no fonts). See http://bazaar.launchpad.net/~dpm/ubuntu-calculator-app/snap-all-things/view/head:/calculator.wrapper#L46
[13:44] <sergiusens> dpm works here, albeit I am trying directly http://pastebin.ubuntu.com/15965473/
[13:44] <sergiusens> I'll give your project a go
[13:47] <kalikiana> dpm: Hrm. I don't know much about fontconfig. What about XDG_CACHE_HOME?
[13:48] <kalikiana> That should be honored by fontconfig according to a quick duck
[13:49] <jcastro> dpm: notes worked on my other laptop though, weird.
[13:57] <dougb> How do you enable the available UARTs / serial ports on the BeagleBone Black with Snappy?
[14:08] <Gunther_> The latest changes in snapcraft/sources.py make git not to use --depth 1 :(
[14:08] <Gunther_> at least when i try to make a clean kernel build
[14:22] <qengho> Can I branch the Store automated tests and suggest others?
[14:27] <dpm> sergiusens, it worked in the end. It turned out I had to do a full `snapcraft clean` instead of just `snapcraft clean -- step build`. I've updated the code, what do you think now of the YAML file and plugin code? https://github.com/dplanella/snappy-playpen/tree/master/notes - I keep thinking there must be a better way of doing the notes.wrapper thing rather than having a huge wrapper file.
[14:27] <dpm> kalikiana, I don't know anything about fontconfig, either unfortunately
[14:28] <beuno> qengho, you can, yes: lp:click-reviewers-tools
[14:28] <beuno> qengho, what do you have in mind?
[14:28] <dpm> jcastro, hm, strange, not sure what could cause it not to start in your other machine :/
[14:30] <dpm> beuno, for apps that we don't own, do you think the convention of using $APPNAME-$MYNAME to avoid naming conflicts is a sensible one to recommend? E.g. `notes-dpm`
[14:30] <beuno> dpm, sure
[14:31] <dpm> ok, thanks
[14:33] <jdstrand> beuno: can someone sync the review tools to the store? (added slots tests)
[14:33] <jdstrand> asac, dholbach: fyi, uploaded review tools 0.42 to have 'snap-review' symlink
[14:33] <beuno> jdstrand, probably!
[14:33] <beuno> cc/ roadmr
[14:34] <jdstrand> roadmr: it isn't an emergency. it could be next week if needed
[14:34] <sergiusens> dpm in your plugin all that can go in `env` ideally we can patch qmlscene to have an argument to rebase dirs for a snap
[14:34] <beuno> deployments are a bit thin today, given the release
[14:34]  * jdstrand nods
[14:34] <sergiusens> dpm you don't need this `export LD_LIBRARY_PATH=$SNAP/usr/lib/$ARCH:$LD_LIBRARY_PATH`
[14:34] <dholbach> jdstrand, nice one!
[14:35] <dpm> sergiusens, I tried to put it in env as the qmlplugin did, but that seemed to break the build for some reason (qmake could not be found)
[14:35] <sergiusens> dpm beuno is `.` allowed again?
[14:35] <jdstrand> sergiusens: do you know what SNAP_LIBRARY_PATH is about? eg: 'SNAP_LIBRARY_PATH=/var/lib/snapd/lib/gl:'
[14:35] <beuno> sergiusens, in a package name?  no
[14:35] <sergiusens> dpm yes, of course :-)
[14:35] <jdstrand> sergiusens: I'm a bit concerned about the trailing colon
[14:36] <sergiusens> jdstrand it was what was mentioned on some list; the extra colon was something niemeyer suggested
[14:36] <jdstrand> niemeyer: hi! I'm concerned about the trailing colon with 'SNAP_LIBRARY_PATH=/var/lib/snapd/lib/gl:'. Is there a reason why I shouldn't be concerned?
[14:36] <jdstrand> sergiusens: thanks
[14:37] <sergiusens> dpm another suggestion is, let me add something
[14:38] <niemeyer> jdstrand: Yeah, this is a path list, supposed to be used as such
[14:38] <roadmr> beuno, jdstrand ; will do
[14:38] <niemeyer> jdstrand: The fact it has a single item there is coincidental
[14:38] <jdstrand> niemeyer: where and how is it used? a trailing colon often evaluates to the current direct which can be an attack vector
[14:38] <niemeyer> jdstrand: The column is there precisely so people don't misunderstand it, and thus don't have broken code in a short while
[14:38] <jdstrand> directory*
[14:39] <asac> jdstrand: pretty awesome!
[14:39] <asac> thanks
[14:39] <jdstrand> np
[14:39] <asac> jdstrand: will that be SRU i guess?
[14:39] <asac> or in main pocket?
[14:39] <jdstrand> I'll get it into 16.04 somehow, yes
[14:39] <niemeyer> jdstrand: LD_LIBRARY_PATH
[14:39] <jdstrand> right
[14:40] <jdstrand> that is the problematic one
[14:40] <jdstrand> a trailing colon in LD_LIBRARY_PATH means add your cwd to LD_LIBRARY_PATH
[14:40] <roadmr> jdstrand: you want c-r-t r634, right?
[14:40] <niemeyer> jdstrand: That's nuts :/
[14:41] <dpm> sergiusens, sure, feel free to do a pull request directly
[14:41] <qengho> Any empty string means ".". A trailing, or internal, leading empty path.
[14:42] <niemeyer> We'll fix that for the SRU, and include a second path there
[14:42] <jdstrand> I can file a bug
[14:42] <niemeyer> jdstrand: Thanks!
[14:42] <qengho> jdstrand: Would you be happy with  something:/dev/null  ?
[14:42] <jdstrand> roadmr: let's make it r635 since that is what I just uploaded
[14:42] <roadmr> jdstrand: ok :)
[14:44] <jdstrand> qengho: I'd prefer simply not ending with a colon
[14:52] <jdstrand> roadmr: thanks!
[14:52] <jdstrand> niemeyer: fyi, bug #1573082
[14:52] <jdstrand> niemeyer: note, because apparmor protects us, this is more of a best practice thing and therefore not an emergency
[14:53] <niemeyer> jdstrand: Thanks, glad to fix nevertheless, and to learn about the wart
[14:53] <qengho> beuno: I was looking for a place to complain about invalid or deprecated variables in snapcraft "command:" yaml, but that's deep in wrapper shell scripts by the time it gets to the store, and snapcraft doesn't have sanity checks yet.
[15:01] <sergiusens> dpm https://github.com/dplanella/snappy-playpen/pull/1
[15:02] <ogra_> beuno, (or jdstrand ), mind approving  https://myapps.developer.ubuntu.com/dev/click-apps/4284/rev/12/ and https://myapps.developer.ubuntu.com/dev/click-apps/4739/rev/6/ ?
[15:02] <ogra_> these are the final 16.04 kernels
[15:02]  * ogra_ cant belive it took two days to get them into the store
[15:03] <jdstrand> beuno: I'll take a look
[15:04] <jdstrand> ogra_: huh. I'm fine to approve but do we expect symlinks outside of the snap with kernels?
[15:04] <ogra_> (note, the dangling symlink is fine on the dragonboard ... it is needed(
[15:04] <jdstrand> ogra_: can you request a manual review for https://myapps.developer.ubuntu.com/dev/click-apps/4739/rev/6/?
[15:04] <ogra_> jdstrand, it is the only way to have a persistent MAC with the android driver that kernel uses
[15:05] <jdstrand> interesting. ok, I'll update the review tools
[15:05] <ogra_> requested
[15:06] <ogra_> the driver requires the macaddr0 file under /lib/firmware/wlan ... sice we dont have that writable this is a link to /run/macaddr0 ... which gets populated on boot from the initrd
[15:07] <jdstrand> ogra_: it looks like xenial-preinstalled-core-amd64.kernel.snap has lib/modules/4.4.0-21-generic/build
[15:07] <ogra_> (generating a MAC based on the board serial number)
[15:07] <jdstrand> ogra_: that will dangle, no?
[15:07] <ogra_> yeah
[15:07] <ogra_> thats from the deb ... it would link to the headers if we installed them
[15:08] <jdstrand> ogra_: I'm inclined to adjust the test to continue to error if lib/modules/ but otherwise not. I could also only error if not lib/firmware
[15:08] <jdstrand> ogra_: do you have an opinion?
[15:09] <ogra_> hmm, is there any security risk ?
[15:09] <jdstrand> no, not unless you already had control of the device, but it is untidy
[15:10] <ogra_> i'd just let them dangle
[15:10] <ogra_> we do that on desktop installs too :)
[15:10] <ogra_> (until you install the headers)
[15:10] <jdstrand> why not create the symlink when you install the headers? (rhetorical)
[15:10] <jdstrand> ok, I'll just allow the symlinks
[15:11] <ogra_> dunno, i didnt package the deb :)
[15:11] <ogra_> i can wipe the build one during snap creation in livecd-rootfs ...
[15:11] <ogra_> but thats an extra hack to maintain
[15:13] <MichaelTunnell> I thought Snappy didn't need root to install packages. It appears as though it does?
[15:13] <jdstrand> I'll just allow it for now. if someone asks you to remove the build symlink, then ping me and I'll adjust the tools
[15:13] <ogra_> ok
[15:16] <dpm> sergiusens, thanks! otp now, but will merge once we're done
[15:19] <jdstrand> ogra_: fyi, both snaps approved
[15:20]  * ogra_ hugs jdstrand 
[15:20] <ogra_> phew ... that was a heavy one
[15:20] <jdstrand> roadmr: if you haven't already started the pull, can you use r636 instead for the kernel symlink fix discussed above?
[15:20]  * jdstrand hugs ogra_ 
[15:20] <roadmr> jdstrand: sure
[15:20] <jdstrand> ogra_: happy release day :)
[15:21] <jdstrand> roadmr: thanks! :)
[15:21] <roadmr> jdstrand: (you're just in time btw, I was just about to push my changes ;)
[15:21] <jdstrand> heh
[15:29] <jdstrand> niemeyer: happy release day to you too :)
[15:29] <jdstrand> I imagine you're feeling like this day would never arrive :)
[15:30] <niemeyer_> jdstrand: Yeah, time did compress a little bit in the last couple of months :)
[15:30] <niemeyer_> jdstrand: Thanks!
[15:30] <niemeyer_> It's been a fun ride
[15:30]  * ogra_ must admit he had better release days 
[15:31] <ogra_> (finding in the morning that our images explode after archive changes wasnt really fun )
[15:32] <jdstrand> yikes
[15:32] <roadmr>  /o\ ugh!
[15:32] <ogra_> well, all sorted now ...
[15:33] <ogra_> but unexopected after i spent weeks with mvo getting everything sorted in advance
[15:33] <jdstrand> best laid plans...
[15:33] <ogra_> yep :)
[15:36] <dpm> sergiusens, thanks! Where does qt5-launch come from? I saw you replaced the notes.wrapper, but I don't quite understand with what
[15:54] <davidcalle> dpm: qt5-launch and qt5conf -> https://wiki.ubuntu.com/Snappy/Parts
[16:00] <dpm> davidcalle, aha, thanks! it was a bit hidden
[16:02] <davidcalle> There should be a snapcraft list parts command
[16:05] <davidcalle> There should be an IRC bot to turn IRC feature requests into bug reports
[16:20] <sergiusens> dpm it is in my branch
[16:20] <sergiusens> dpm made sense since qt5-config sets stuff up relevant to the wrapper itself
[16:21] <sergiusens> davidcalle we have a plan of something like apt for parts
[16:21] <dpm> sergiusens, I'll give it all a go in a few minutes, thanks!
[16:21] <davidcalle> sergiusens: nice :)
[16:28] <wililupy> Is ubuntu-device-flash broken? I keep getting the error expected gadget snaps: snap not found for --gadget=canonical-pc.
[16:29] <ali1234> did the snappy tour get updated for 16.04?
[16:30] <ali1234> also, any news on pi 3 support?
[16:30] <ogra_> wililupy, get the latest u-d-f from http://people.canonical.com/~mvo/all-snaps/ and use "16" instead of "rolling"
[16:31] <ogra_> (the series got renamed)
[16:31] <wililupy> Thanks ogra_
[16:34] <qengho> (I'm not sure that^ image is good. I get weird errors. I assumed a recent bug.)
[16:35] <qengho> (ubuntu-core-launcher[1885]: seccomp_load failed with -22  //  ubuntu-core-launcher[1885]: aborting. errmsg: Invalid argument)
[16:35] <ogra_> qengho, what arch
[16:35] <qengho> rpi
[16:35] <ogra_> update your kernel snap
[16:36] <ogra_> the final kernel only landed 1h ago
[16:39] <roadmr> hey folks, trying to build an image i get "expected 3 partitions but found 0" - here's the command and full output http://paste.ubuntu.com/15968587/ any ideas?
[16:39] <ogra_> roadmr, using the latest u-d-f from http://people.canonical.com/~mvo/all-snaps/ ?
[16:40] <roadmr> ogra_: should be, I used zyga's script to download. Am double-checking now
[16:41] <ogra_> "gadget configuration" ?
[16:41] <ogra_> you have trailing trash there
[16:41] <ogra_> oh
[16:41] <ogra_> no, u-d-f has a bug :P
[16:42]  * roadmr pictures himself dragging a bag of trash with a long string :)
[16:42] <ogra_> heh
[16:43] <ogra_> anyway, i built quite a few images today with that u-d-f
[16:43] <ogra_> so technically it should work :)
[16:43] <roadmr> ogra_: yes, I have the apr-19 u-d-f from all-snaps
[16:43] <roadmr> ogra_: am I missing some command maybe? I'm running this inside an lxc too, may that affect things?
[16:43] <ogra_> theer were two ;)
[16:43] <roadmr> the 21:25 one
[16:44] <ogra_> k
[16:44] <ogra_> yeah, perhaps ... i run it natively here
[16:44] <ogra_> and also on a xenial host
[16:44] <roadmr> yes, xenial here. I'll try natively!
[17:11] <dpm> ogra_, sergiusens, kyrofa, would you happen to know the answer to http://askubuntu.com/questions/760057/will-a-sideloaded-snap-get-upgraded-to-a-newer-version-in-the-store ? And if so, would you mind answering there?
[17:11] <dpm> or actually beuno might know too ^
[17:12] <kyrofa> dpm sure
[17:12] <dpm> awesome, thanks!
[17:13] <dpm> "You've earned the 'Popular Question' badge (question with 1000 views) for 'What is snapcraft?'"
[17:14] <dpm> nice :)
[17:14]  * dpm goes for dinner
[17:26] <roadmr> ogra_: ok, so run natively, u-d-f works. So it doesn't seem to like being run inside a container :( in case someone else asks. Thanks!
[17:38] <kyrofa> roadmr, indeed, I've run into the same problem. Runs fine inside vbox though
[17:45] <dougb_> I need some guidance for how to enable the available UARTs / serial ports on the BeagleBone Black
[17:47] <kyrofa> dougb_, you probably want ogra_
[17:49] <dougb_> ogra, do you have any advice on enabling serial ports in Snappy?
[17:52] <dougb_> ogra_, do you have any advice on enabling serial ports in Snappy?
[18:07] <dpm> kgunn, would you happen to know what could be causing this OpenGL error when launching snaps? -> http://askubuntu.com/q/759647/9781
[18:24] <kgunn> dpm-afk: lemme read
[18:33] <kgunn> dpm-afk: so i "answered" as i don't have enough kwan to comment :)
[18:59] <kgunn> attente: ping
[19:10] <attente> kgunn: hi
[20:41] <roadmr> hey folks, which plug does my snap need in order to be able to run ssh-keygen? I just want to generate a keypair in $SNAP_DATA
[20:44] <qengho> roadmr: You have to include your own openssh.
[20:44] <roadmr> qengho: that's inconvenient ;) but ok, guess that's snappy life! I'll do it.
[20:46] <qengho> roadmr: In your snapcraft part, add  stage-packages: [ openssh-client ]
[20:47] <roadmr> qengho: yay :)
[20:47] <roadmr> qengho: where'll the binary live? $SNAP/bin/ssh-keygen?
[20:48] <roadmr> (well I'll find out soon enough, building snap now)
[20:48] <qengho> $SNAP/usr/bin/ssh-keygen -- Normal path with $SNAP prepended.
[20:48] <roadmr> oh! qengho thanks! /me tweaks script and rebuilds
[20:51] <roadmr> qengho: it works \o/ thanks so much for your help!
[20:51] <qengho> roadmr: Happy Release Day!
[20:53] <qengho> Okay, install a snap. Refresh it, get a new version. The apparmor profile in /var/lib/snapd/apparmor/profiles should be updated then, right?
[21:01] <sergiusens> elopio mind looking at https://github.com/ubuntu-core/snapcraft/pull/480/files
[21:01] <sergiusens> qengho yeah
[21:02] <sergiusens> qengho I am seeing an issue that if I sideload twice the binary I call would still be pointing to the older snap
[21:06] <qengho> sergiusens: no need to sideload.
[21:06] <qengho> https://bugs.launchpad.net/snappy/+bug/1573247
[21:07] <qengho> sergiusens: Oh, you mean the wrapper script is the older one?
[21:43] <sergiusens> qengho yeah
[21:45]  * jdstrand wonders what this means:
[21:45] <jdstrand> $ sudo snap install ./2048-qml_1_amd64.snap --devmode
[21:45] <jdstrand> error: change finished in status "Hold" with no error message
[21:58] <skypce> hello
[21:58] <skypce> how can install a snap package?
[21:58] <skypce> i was installed ubuntu 16.04 in virtualbox
[21:59] <skypce> i want test the feature
[22:01] <qengho> skypce: "sudo apt install snapd; sudo snap install tor-middle-relay"
[22:02] <skypce> great, thank you qengho
[22:02] <skypce> i believe that it is enabled by default
[22:06] <qengho> skypce: did it work? "ps ax |grep tor-"
[22:10] <skypce> qengho, excelent
[22:10] <skypce> i was testing vlc
[22:11] <skypce> it is a probability that snapd will be in trusty?
[22:11] <qengho> No.
[22:12] <skypce> ahh, will be great
[22:12] <skypce> thank you qengho