From mark at ubuntu.com Fri Jul 1 05:35:37 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Fri, 1 Jul 2016 06:35:37 +0100 Subject: listing all the interfaces / sockets? In-Reply-To: <28575.1467311913@parc.com> References: <24070.1467302757@parc.com> <57754895.3090604@ubuntu.com> <28575.1467311913@parc.com> Message-ID: <57760129.1060104@ubuntu.com> On 30/06/16 19:38, Bill Janssen wrote: > >> 'snap interfaces -a' might >> for example show you available but unused slots / plugs. > $ snap interfaces -a > error: unknown flag `a' > Yeah, it doesn't exist yet, I was pointing the way to a "tell me about all the *possible* interfaces" patch on snapd for you or someone else who feels inspired :) So someone who wanted to know about the interfaces could do something like this (rather than trawl the source): $ snap interfaces -a Interface Description camera Provides access to a V4L2 video camera cups-control Allows printing ... x11 Steals all the things Mark From vasilisc777 at gmail.com Fri Jul 1 07:50:03 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Fri, 1 Jul 2016 10:50:03 +0300 Subject: libGL Message-ID: <577620AB.6060202@gmail.com> Please explain this behavior. I create test poligon in VirtualBox - Ubuntu 16.04, MATE, Lubuntu, Xubuntu, GNOME, Kubuntu. the gtk3-audio-app is installed and works well. Only in Kubuntu the error appears Couldn't open libGL.so.1: libGL.so.1: cannot open shared object file: No such file or directory I shall add the line libgl1-mesa-glx in stage-packages? -- Best regards, vasilisc From neil at themcphails.uk Fri Jul 1 11:02:54 2016 From: neil at themcphails.uk (Neil McPhail) Date: Fri, 1 Jul 2016 12:02:54 +0100 Subject: Should hidden files and directories be treated differently by the home plug Message-ID: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> Dear all, I'm trying to snappify my fork of the silversearcher-ag package. This is an application which works like a recursive "grep", but much faster, and with awareness of git/mercurial repository formats and ignore files. It has to parse ".gitignore" files, for example, to know what to search and what not to search. I have attached an early draft of my snapcraft.yaml file. This doesn't seem possible with the current set of snapd interfaces. The "home" plug excludes access to any "dot"/hidden files or directories. This appears to be an attempt to prevent access to things like GPG and SSH keys, which is admirable. Unfortunately, it has the side effect of blocking legitimate access to the vast majority of innocent hidden files. I think it is wrong to conflate "hidden" files with "sensitive or secure" files in this way. The dot-prefix was never intended to add security; merely to hide ugly files and directories. My $HOME has not been arranged with the expectation that my dotfiles contain my sensitive information and my plainfiles are public. Access to my sensitive files is controlled by standard permissions. Indeed, SSH will baulk if my keyfile is world-readable. On the other hand, if Boris Johnson managed to get me to install his malicious .snap which would search out and destroy "boris_with_strawberry_jam_and_poodle.jpg" from my $HOME/Pictures directory (which is accessible via the "home" plug even although it it chmodded 600) he would be back in the Tory leadership race in minutes. Perhaps a better way would be to run "home" plugged apps as a different user, who has been added to the real user's primary group? Thus 600-modded files would remain inaccessible but innocent dotfiles could be manipulated via group permissions. What do you all think? NMP -------------- next part -------------- A non-text attachment was scrubbed... Name: snapcraft.yaml Type: application/x-yaml Size: 423 bytes Desc: not available URL: From janssen at parc.com Fri Jul 1 15:03:08 2016 From: janssen at parc.com (Bill Janssen) Date: Fri, 01 Jul 2016 08:03:08 -0700 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> Message-ID: <26085.1467385388@parc.com> Neil McPhail wrote: > This doesn't seem possible with the current set of snapd interfaces. The > "home" plug excludes access to any "dot"/hidden files or directories. Oh, heck. I need those, too. Bill From fcole90 at gmail.com Fri Jul 1 15:17:13 2016 From: fcole90 at gmail.com (Fabio Colella) Date: Fri, 1 Jul 2016 17:17:13 +0200 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: <26085.1467385388@parc.com> References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> <26085.1467385388@parc.com> Message-ID: This would indeed solve some confinement issues with my snap ubuntukylin-icon-theme too (access to and evntual creation of the .icons folder) On 1 July 2016 at 17:03, Bill Janssen wrote: > Neil McPhail wrote: > > > This doesn't seem possible with the current set of snapd interfaces. The > > "home" plug excludes access to any "dot"/hidden files or directories. > > Oh, heck. I need those, too. > > Bill > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gustavo at niemeyer.net Sat Jul 2 08:25:31 2016 From: gustavo at niemeyer.net (Gustavo Niemeyer) Date: Sat, 2 Jul 2016 05:25:31 -0300 Subject: All MongoDB releases, everywhere In-Reply-To: References: Message-ID: Another curiosity around this: the mup IRC and Telegram bot we use to help on our development-related conversations on the #snappy channel is now living its life inside a snap. It's a bit confined there, but he's fine with that apparently. On Thu, Jun 30, 2016 at 5:34 PM, Gustavo Niemeyer wrote: > Hello snapcrafters, > > Hot off the presses, all MongoDB releases in snap form: > > http://blog.labix.org/2016/06/30/all-mongodb-releases-everywhere > > Please drop me a note if you have any questions. > > > gustavo @ http://niemeyer.net > -- gustavo @ http://niemeyer.net -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Sat Jul 2 12:19:24 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Sat, 02 Jul 2016 14:19:24 +0200 Subject: All MongoDB releases, everywhere In-Reply-To: References: Message-ID: <1467461964.14415.1.camel@anubis> hi, Am Donnerstag, den 30.06.2016, 17:34 -0300 schrieb Gustavo Niemeyer: > Hello snapcrafters, > > > Hot off the presses, all MongoDB releases in snap form: > > > http://blog.labix.org/2016/06/30/all-mongodb-releases-everywhere > > > Please drop me a note if you have any questions. > how about turning that into snapcraft parts too ? :) ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From mutlaqja at ikarustech.com Sat Jul 2 18:40:58 2016 From: mutlaqja at ikarustech.com (Jasem Mutlaq) Date: Sat, 2 Jul 2016 21:40:58 +0300 Subject: Problems creating snap package for KStars Message-ID: Hello, I wanted to create a single snap package for KStars ( http://edu.kde.org/kstars) including all dependencies. I started off from Krita's snapcraft.yaml file and customized it to fit KStars. However, since KStars leverages several KDE components (KNewStuff, kinit, kded5..etc) and also includes INDI server, drivers, and tools which communicate with devices for control and data acquisition, it is quite complex compared to other snaps I've seen so far. At first, I ran it in strict confinement mode and used all the plugs necessary for it to work, but ran into several issues and used it in devmode since. The snap is created on my primary machine (64bit Kubuntu 16.04), and all testing is done on a fresh Ubuntu Mate 16.04 Virtual Machine. Thus far, I ran into the follow issues: 1. Cannot read from usr/share/indi. It keeps referring to $SNAP_DATA and $SNAP_USER_DATA but INDI is a 3rd party program that installs its data files to /usr/share/data, so how do I deal with that exactly? Where is $SNAP_DATA located with respect the root FS? 2. Cannot play audio notifications: org.kde.knotifications: Audio notification requested, but sound file from notifyrc file was not found, aborting audio notification 3. Breeze icons installed but not loaded up. As indicated in the YAML file, the breeze icon theme is installed, but it is not loading up. Tried modifying qt5-launch file (export QT_QPA_PLATFORMTHEME=kde) tried changing QT_SELECT (snappy-qt5 vs qt5) but the icons never load up. 4. Download New Data (KNewStuff3) loads up providers, show download list, and when selecting a component to download, it downloads it but fails to install any file. In strict confinement mode, it fails to load the providers.xml file 5. dbus calls fail in strict confinement mode. 6. Keep getting this message: XmbTextListToTextProperty result code -2 XmbTextListToTextProperty result code -2 XmbTextListToTextProperty result code -2 XmbTextListToTextProperty result code -2 7. Other apparmor problems (see log below) Here is a link to the files: 1. snapcraft.yaml: http://www.indilib.org/jdownloads/snap/snapcraft.yaml 2. qt5-lunch: http://www.indilib.org/jdownloads/snap/qt5-lunch 3. AppArmor log: http://www.indilib.org/jdownloads/snap/apparmor.txt As you can see from the AppArmor log, there are a lot of calls been made among all the various components, libraries, file system..etc. Even when an executable like indi_simulator_ccd is accessing a dependent library, it says: = AppArmor = Time: Jul 2 20:30:33 Log: apparmor="ALLOWED" operation="open" profile="snap.kstars.kstars//null-/snap/kstars/x2/usr/bin/indiserver//null-/snap/kstars/x5/usr/bin/indi_simulator_ccd" name="/snap/kstars/x5/usr/lib/x86_64-linux-gnu/libjpeg.so.8.0.2" pid=4015 comm="indi_simulator_" requested_mask="r" denied_mask="r" fsuid=1000 ouid=0 File: /snap/kstars/x5/usr/lib/x86_64-linux-gnu/libjpeg.so.8.0.2 (read) Suggestion: * adjust program to read necessary files from $SNAP, $SNAP_DATA or $SNAP_USER_DATA Again, not sure how to adjust program to read from dependent library. -- Best Regards, Jasem Mutlaq -------------- next part -------------- An HTML attachment was scrubbed... URL: From didrocks at ubuntu.com Mon Jul 4 06:27:58 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Mon, 4 Jul 2016 08:27:58 +0200 Subject: Problems creating snap package for KStars In-Reply-To: References: Message-ID: <577A01EE.7000605@ubuntu.com> Le 02/07/2016 20:40, Jasem Mutlaq a écrit : > Hello, Hey Jasem! > > I wanted to create a single snap package for KStars > (http://edu.kde.org/kstars) including all dependencies. I started off > from Krita's snapcraft.yaml file and customized it to fit KStars. > However, since KStars leverages several KDE components (KNewStuff, > kinit, kded5..etc) and also includes INDI server, drivers, and tools > which communicate with devices for control and data acquisition, it is > quite complex compared to other snaps I've seen so far. Waow, you are attacking a great beast! Glad you are trying to figure out how we could snap a component with so many system/session KDE components! Good project :) > > At first, I ran it in strict confinement mode and used all the plugs > necessary for it to work, but ran into several issues and used it in > devmode since. The snap is created on my primary machine (64bit > Kubuntu 16.04), and all testing is done on a fresh Ubuntu Mate 16.04 > Virtual Machine. I would advise you to always start with devmode. Then, once your snap works reasonably well, you can switch the confinement on and look at missing plugs, errors and I so on. > > Thus far, I ran into the follow issues: > > 1. Cannot read from usr/share/indi. It keeps referring to $SNAP_DATA > and $SNAP_USER_DATA but INDI is a 3rd party program that installs its > data files to /usr/share/data, so how do I deal with that exactly? > Where is $SNAP_DATA located with respect the root FS? Most of the apps have either some runtime env variable or some package build configuration (like --prefix or --datadir). One of the issue we are discussing with upstream is that there is nothing that ensure you those paths will stay the same. Gustavo (CCed here) mentioned maybe some declarative open() redirection in libc would be in order. To ensure directly to your question, SNAP_DATA is currently located at /var/snap// (current/ is a symlink to current_snap_rev directory). You can have a quick look at those env variables by installing the hello-world snap and run: $ hello-world.env > > 2. Cannot play audio notifications: org.kde.knotifications: Audio > notification requested, but sound file from notifyrc file was not > found, aborting audio notification We can have a look at that if you have this issue with the pulseaudio plug, but let's quick out confinement for now. > > 3. Breeze icons installed but not loaded up. As indicated in the YAML > file, the breeze icon theme is installed, but it is not loading up. > Tried modifying qt5-launch file (export QT_QPA_PLATFORMTHEME=kde) > tried changing QT_SELECT (snappy-qt5 vs qt5) but the icons never load up. Interesting, we'll announce a new launcher in a couple of days, I have hope that this one supports theme in a better way that the current qt5-launch one. Stay tuned to this ML! > > 4. Download New Data (KNewStuff3) loads up providers, show download > list, and when selecting a component to download, it downloads it but > fails to install any file. In strict confinement mode, it fails to > load the providers.xml file I wonder where it tries to install them, is that by any chance in $SNAP? (and so, in a read only area?) > > 5. dbus calls fail in strict confinement mode. I'm ccing Jamie here, he may have a look at them if they make sense to add to any interface. > > 6. Keep getting this message: > > XmbTextListToTextProperty result code -2 > XmbTextListToTextProperty result code -2 > XmbTextListToTextProperty result code -2 > XmbTextListToTextProperty result code -2 I confirm that one with other Qt projects. Didn't find a way personally to fix this, maybe some other have more chance? > > 7. Other apparmor problems (see log below) > > Here is a link to the files: > > 1. snapcraft.yaml: http://www.indilib.org/jdownloads/snap/snapcraft.yaml > 2. qt5-lunch: http://www.indilib.org/jdownloads/snap/qt5-lunch > 3. AppArmor log: http://www.indilib.org/jdownloads/snap/apparmor.txt > > > As you can see from the AppArmor log, there are a lot of calls been > made among all the various components, libraries, file system..etc. > Even when an executable like indi_simulator_ccd is accessing a > dependent library, it says: > > = AppArmor = > Time: Jul 2 20:30:33 > Log: apparmor="ALLOWED" operation="open" > profile="snap.kstars.kstars//null-/snap/kstars/x2/usr/bin/indiserver//null-/snap/kstars/x5/usr/bin/indi_simulator_ccd" > name="/snap/kstars/x5/usr/lib/x86_64-linux-gnu/libjpeg.so.8.0.2" > pid=4015 comm="indi_simulator_" requested_mask="r" denied_mask="r" > fsuid=1000 ouid=0 > File: /snap/kstars/x5/usr/lib/x86_64-linux-gnu/libjpeg.so.8.0.2 (read) > Suggestion: > * adjust program to read necessary files from $SNAP, $SNAP_DATA or > $SNAP_USER_DATA > > Again, not sure how to adjust program to read from dependent library. > Same, needing Jamie's feedback on those :) Cheers, Didier From pawel.stolowski at canonical.com Mon Jul 4 09:24:30 2016 From: pawel.stolowski at canonical.com (Pawel Stolowski) Date: Mon, 4 Jul 2016 11:24:30 +0200 Subject: Problems creating snap package for KStars In-Reply-To: <577A01EE.7000605@ubuntu.com> References: <577A01EE.7000605@ubuntu.com> Message-ID: <577A2B4E.9090400@canonical.com> Hi Didier, >> Thus far, I ran into the follow issues: >> >> 1. Cannot read from usr/share/indi. It keeps referring to $SNAP_DATA >> and $SNAP_USER_DATA but INDI is a 3rd party program that installs its >> data files to /usr/share/data, so how do I deal with that exactly? >> Where is $SNAP_DATA located with respect the root FS? > Most of the apps have either some runtime env variable or some package > build configuration (like --prefix or --datadir). One of the issue we > are discussing with upstream is that there is nothing that ensure you > those paths will stay the same. Gustavo (CCed here) mentioned maybe some > declarative open() redirection in libc would be in order. > That's a very interesting point and something I've asked myself too while battling with similar issues (which I still haven't solved; the app I'm working on uses hardcoded /usr/share/... paths too; if only patching was streamlined in snapcraft, that would at least help in some cases). I think it's not just open(), but access(), *stat(), opendir()... And I think it will get tricky. And then it only helps with apps using libc (ok, probably a majority is libc-based). Isn't this something that chroot() solves (which of course affects all the plugs etc)? Cheers, Pawel From scarlett.gately.clark at gmail.com Mon Jul 4 13:10:54 2016 From: scarlett.gately.clark at gmail.com (Scarlett Clark) Date: Mon, 4 Jul 2016 06:10:54 -0700 Subject: Problems creating snap package for KStars In-Reply-To: <577A2B4E.9090400@canonical.com> References: <577A01EE.7000605@ubuntu.com> <577A2B4E.9090400@canonical.com> Message-ID: I already started this: https://github.com/snappy-packages/applications-kstars Scarlett On Mon, Jul 4, 2016 at 2:24 AM, Pawel Stolowski < pawel.stolowski at canonical.com> wrote: > Hi Didier, > >> Thus far, I ran into the follow issues: >>> >>> 1. Cannot read from usr/share/indi. It keeps referring to $SNAP_DATA >>> and $SNAP_USER_DATA but INDI is a 3rd party program that installs its >>> data files to /usr/share/data, so how do I deal with that exactly? >>> Where is $SNAP_DATA located with respect the root FS? >>> >> Most of the apps have either some runtime env variable or some package >> build configuration (like --prefix or --datadir). One of the issue we >> are discussing with upstream is that there is nothing that ensure you >> those paths will stay the same. Gustavo (CCed here) mentioned maybe some >> declarative open() redirection in libc would be in order. >> >> > That's a very interesting point and something I've asked myself too while > battling with similar issues (which I still haven't solved; the app I'm > working on uses hardcoded /usr/share/... paths too; if only patching was > streamlined in snapcraft, that would at least help in some cases). > I think it's not just open(), but access(), *stat(), opendir()... And I > think it will get tricky. And then it only helps with apps using libc (ok, > probably a majority is libc-based). Isn't this something that chroot() > solves (which of course affects all the plugs etc)? > > Cheers, > Pawel > > > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.planella at ubuntu.com Mon Jul 4 16:28:00 2016 From: david.planella at ubuntu.com (David Planella) Date: Mon, 4 Jul 2016 18:28:00 +0200 Subject: Shaping up universal snaps Message-ID: Hi all, Here is a call for participation those of us who have been involved in were really looking forward to publish :) Following the announcement of snaps being supported across a range of key Linux distributions [1], the development teams working on snaps and Snapcraft are making universal snaps one of the main topics of their next sprint in Heidelberg, Germany, from 18-22 July. Snappy sprints are face-to-face events where multiple teams working on snap technologies, including Ubuntu founder Mark Shuttleworth, get together to plan, design and develop their next release and longer term roadmap. After the initial positive reception amongst initial adopters, tech media and wider open source community, continuous improvement of the snap user and developer experience is a major focus. A number of upstreams, contributors and developers of leading open source projects such as Debian, Elementary OS, Fedora, KDE, Kubuntu, MATE or VLC have already confirmed participation at the sprint to collaborate on better distro-agnostic snap support. At this point, we'd like to extend this invitation to contributors of other projects to influence the roadmap and work together on shaping up the universal snaps story. If you are interested in participating, we have a limited amount of seats to financially sponsor travel and accommodation for contributors of upstreams, distros or desktop projects who are willing to actively work towards this goal. If the answer is yes, feel free to apply for participation and sponsorship to the Heidelberg snappy sprint: >> http://goo.gl/forms/yfGXoK8Aqv0TEiMJ3 Please note that a sprint is not a tech conference: it is a set of focused working and planning sessions where the snappy Engineering team execute work items and plan the next iteration of snapd and Snapcraft. Attendees will be expected to actively participate in discussions and decision making and be willing to take work items where appropriate. Also do note that while all contributions are valuable, we have a limited capacity to sponsor participants and we cannot support everyone. As such, sponsorship will be subject to review and final confirmation. Once the requests are in, we will review all of the applicants and contact you as soon as possible to let you know if your request for sponsorship has been approved. It will be a great chance to build together app distribution across platforms and we’ll be looking forward to working with you! [1] https://insights.ubuntu.com/2016/06/14/universal-snap-packages-launch-on-multiple-linux-distros Source: http://developer.ubuntu.com/en/blog/2016/07/04/shaping-universal-snaps -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.calle at canonical.com Mon Jul 4 18:51:51 2016 From: david.calle at canonical.com (=?UTF-8?Q?David_Call=c3=a9?=) Date: Mon, 4 Jul 2016 20:51:51 +0200 Subject: Snap! Community weekly update: W26 Message-ID: <577AB047.5010406@canonical.com> Here is an update on the highlights in the snap community for the past week. We are releasing this one a bit late on schedule because we really wanted to include an invitation for contributors in the community to the Heidelberg snappy sprint, from july 18th to 22nd in Germany! ## Heidelberg sprint, 18-22 July A face-to-face event this summer to meet and share your experience with snap developers, early adopters and contributors. You are invited to contribute shaping the future of snaps, plan the next features of snapd and snapcraft along developers of leading open source projects, such as Debian, KDE, MATE, etc. Learn more and apply for sponsorship . ## Let’s Play Snapcraft Alan Pope has launched a video series called "Let’s Play Snapcraft" and it’s quite a treat. In this first video, Alan is snapping OpenTTD, the Transport Tycoon Deluxe open remake. It’s 20 minutes long and covers the basics of Snapcraft,enjoy ! ## Playpen * The Playpen is steadily growing and now containsexamples for 32 snaps * We have started anUpstreaming wiki page, with guidelines to get snaps to production quality and submit them to upstreams that have not started on their own snaps. * Thenext Playpen is tomorrow , we will continue our documentation effort and start working on the upstreaming aspect of it * Some snaps from the Playpen are starting to use theGTK/Qt desktop helpers created by Didier Roche, that take care of integrating themes, icons, and much more in your desktop snaps. Stay tuned for more on this topic! If you are having issues while making snaps, check if they are referenced on the newKnown issues wiki page, and make sure you come talk to experimented snappers onthe Snappy Playpen chat . ## Community highlights * The shout out of the week goes to Zack Casey! In the span of a few days, this fantastic snapper has landedBlender ,Electrum (bitcoin wallet) andLiteIDE (Go IDE) in the store. This is simply amazing and we are looking forward to his next snap! * Brian Douglass hasenabled snaps distribution through the OpenStore, an alternative app store initially targeted at Ubuntu phones and tablets. If you are looking for alternative ways of playing with the system, this is a great starting point. * Popescu Sorin is making demos of snaps running in Unity8 (as well as other bleeding edge things). Here is a video of theBlender snap, with its dedicated X window completely confined since Unity8 itself is running on Mir. * Daniel Wood createda CAD app for the#ConvergeMePlease contest distributed as a snap ## Reading list * The Snapcraft Parts Ecosystem , in-depth introduction to an important new feature in Snapcraft 2.12 by Sergio Schvezov * Snapcraft 2.12: an ecosystem of parts, qmake and gulp , the Snapcraft 2.12 release log with examples * All MongoDB releases, everywhere , do you like MongoDB? Do you like being able to test and run several versions, including bleeding edge features? Gustavo Niemeyer got you covered * On writing my first snaps, and keeping my passwords secure everywhere , KeepassX and kpcli have been snapped, here is how, a great write up by Leo Arias * snap try: The quick way to package snaps , Jamie Bennett introduces you to `snap try`, the easiest way to turn a snap directory into an mounted, editable, snap * SwitchDev in Ubuntu-Core? Yes Please! , a follow-up on custom kernel snapping by Lucas Williams This is a list of news we are compiling weekly, but make sure you subscribe to the following social channels for daily updates:twitter ,google+ ,facebook . Cheers, David -------------- next part -------------- An HTML attachment was scrubbed... URL: From w.andrew.keech at gmail.com Tue Jul 5 08:05:09 2016 From: w.andrew.keech at gmail.com (Andrew Keech) Date: Tue, 5 Jul 2016 01:05:09 -0700 Subject: Future of Launchpad and the Snap Store Message-ID: <577B6A35.8040308@gmail.com> hey everyone, so wonderful to see the incredible progress all the time. I've been whittling away at a snap for `gimp-edge` for a while, but feel it's been overdue to bring in the maintainer of the ppa where I currently get my builds from. Having been in touch with Thorsten now, he's got some unanswered questions about the possible future relationships or integrations between Launchpad and the Snap Store. Right now he wants to start a new ppa just for delivering the snap as a separate project from the traditional gimp-edge deb. Are there any best practice ideas that exist? Is there some kind of plan for how the Store and Launchpad might work together, or if they'd exist separately? thanks very much -Andy From cjwatson at ubuntu.com Tue Jul 5 12:06:43 2016 From: cjwatson at ubuntu.com (Colin Watson) Date: Tue, 5 Jul 2016 13:06:43 +0100 Subject: Future of Launchpad and the Snap Store In-Reply-To: <577B6A35.8040308@gmail.com> References: <577B6A35.8040308@gmail.com> Message-ID: <20160705120642.GB30483@riva.ucam.org> On Tue, Jul 05, 2016 at 01:05:09AM -0700, Andrew Keech wrote: > I've been whittling away at a snap for `gimp-edge` for a while, but feel > it's been overdue to bring in the maintainer of the ppa where I currently > get my builds from. Having been in touch with Thorsten now, he's got some > unanswered questions about the possible future relationships or integrations > between Launchpad and the Snap Store. Right now he wants to start a new ppa > just for delivering the snap as a separate project from the traditional > gimp-edge deb. > > Are there any best practice ideas that exist? Is there some kind of plan for > how the Store and Launchpad might work together, or if they'd exist > separately? The plan is definitely for them to be able to work together, and indeed quite a bit of this is already in place. The elevator pitch here is that Launchpad handles code hosting and building, taking advantage of the multi-architecture build farm we already have in place, while the store handles publishing and distribution. What you can do today: * Push a Bazaar or Git branch to Launchpad containing a snapcraft.yaml file at its top level, or import such a branch hosted elsewhere into Launchpad's Bazaar code hosting * Create a snap package in Launchpad, starting from a Bazaar or Git branch as above * Manually request builds of that snap package in Launchpad * Configure the snap package in Launchpad to automatically push completed builds to the store * Manually publish the package in the store to whatever channels you want What you'll be able to do soon (pending code review and deployment): * More obvious configuration for building snap packages for architectures other than the defaults of amd64 and i386 * Configure Launchpad to automatically build a snap package any time the branch containing its snapcraft.yaml changes * Configure Launchpad to automatically tell the store to publish a snap package to a given set of channels after pushing completed builds What you'll be able to do later: * Git-to-Git code imports, which will be more robust than Git-to-Bazaar and will cope with some cases that Git-to-Bazaar can't handle; this will make things easier for people who e.g. want to host their code primarily on GitHub * Create a code import right from the "new snap package" form on Launchpad * Discover all of this from the store, rather than being unconditionally guided through uploading packages manually Regarding integration with PPAs, snap packages aren't contained in PPAs as such. However, you can either have Launchpad build the code that goes into the snap package from scratch, or you can have it assemble a snap package based on existing .debs in the Ubuntu archive and/or a PPA, depending on what you put in your snapcraft.yaml; so the two approaches can work together if you want, and this may make sense if you already have good automation for doing edge builds in a PPA. Does that help with those unanswered questions? -- Colin Watson [cjwatson at ubuntu.com] From jamie at canonical.com Tue Jul 5 16:15:02 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Tue, 05 Jul 2016 11:15:02 -0500 Subject: Problems creating snap package for KStars In-Reply-To: <577A01EE.7000605@ubuntu.com> References: <577A01EE.7000605@ubuntu.com> Message-ID: <1467735302.10258.19.camel@canonical.com> On Mon, 2016-07-04 at 08:27 +0200, Didier Roche wrote: > Le 02/07/2016 20:40, Jasem Mutlaq a écrit : > > > >  > > 5. dbus calls fail in strict confinement mode. > I'm ccing Jamie here, he may have a look at them if they make sense to > add to any interface. I suspect we'll need a transition interface similar to 'unity7' for KDE apps since KDE, like all traditional Linux desktops, has a different trust model (where everything running in your session is trusted) than snappy (where apps are considered untrusted). As Didier said, get your snap running in --devmode, then file bugs at https://bugs.launchpad.net/snappy/+filebug adding the 'snapd- interface' tag and we can work through what is needed. That said, I predict KDE apps needing kinit and all the various KDE IPC services and how they (auto)start each other are going to present a real challenge to have any meaningful security policy (based on experience with profiling KDE apps and your apparmor logs) and require thoughtful design. We'll know more once there is a working devmode snap and a bug is filed for it to work in strict mode. > > 7. Other apparmor problems (see log below) > > > > Here is a link to the files: > > > > 1. snapcraft.yaml: http://www.indilib.org/jdownloads/snap/snapcraft.yaml > > 2. qt5-lunch: http://www.indilib.org/jdownloads/snap/qt5-lunch > > 3. AppArmor log: http://www.indilib.org/jdownloads/snap/apparmor.txt > > > > > > As you can see from the AppArmor log, there are a lot of calls been > > made among all the various components, libraries, file system..etc. > > Even when an executable like indi_simulator_ccd is accessing a > > dependent library, it says: > > > > = AppArmor = > > Time: Jul  2 20:30:33 > > Log: apparmor="ALLOWED" operation="open" > > profile="snap.kstars.kstars//null-/snap/kstars/x2/usr/bin/indiserver//null- > > /snap/kstars/x5/usr/bin/indi_simulator_ccd" > > name="/snap/kstars/x5/usr/lib/x86_64-linux-gnu/libjpeg.so.8.0.2" > > pid=4015 comm="indi_simulator_" requested_mask="r" denied_mask="r" > > fsuid=1000 ouid=0 > > File: /snap/kstars/x5/usr/lib/x86_64-linux-gnu/libjpeg.so.8.0.2 (read) > > Suggestion: > > * adjust program to read necessary files from $SNAP, $SNAP_DATA or > > $SNAP_USER_DATA > > > > Again, not sure how to adjust program to read from dependent library. > > > Same, needing Jamie's feedback on those :) This file should be allowed by the policy since it is in your $SNAP directory. I'm a bit puzzled as to why this is showing up in the log. I will note that the profile in question is 'snap.kstars.kstars//null- /snap/kstars/x2/usr/bin/indiserver//null- /snap/kstars/x5/usr/bin/indi_simulator_ccd' which is the complain-mode profile name for the indi_simulator_ccd process that was launched by indiserver which was started by something under the snap.kstars.kstars profile. Notice the 'x2' for the snap's revision for indieserver and the 'x5' for the snap's revision for indi_simulator_ccd. It seems that that snap was updated from 'x2' to 'x5' while indieserver was still running? Regardless, can you file a bug with detailed steps on how to reproduce and we can work this out there. Thanks! -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From evan.dandrea at canonical.com Tue Jul 5 20:13:33 2016 From: evan.dandrea at canonical.com (Evan Dandrea) Date: Tue, 05 Jul 2016 20:13:33 +0000 Subject: Handling config Message-ID: Hi, Snapd does not presently have support for getting and setting individual snap configurations. We needed to this for the Juju charm of Cassandra (set `edition: apache-snap`), so I've come up with three simple helpers: - config-get - config-set - env-get These save us from having to hardcode paths in the charm. In the case of env-get, it's useful to pull out SNAP_DATA when constructing a config file that embeds absolute paths. Your feedback is of course welcome, and do feel free to pull these into your own snaps as needed. -------------- next part -------------- An HTML attachment was scrubbed... URL: From w.andrew.keech at gmail.com Wed Jul 6 15:05:37 2016 From: w.andrew.keech at gmail.com (Andrew Keech) Date: Wed, 6 Jul 2016 08:05:37 -0700 Subject: Snapcraft Digest, Vol 3, Issue 7 In-Reply-To: References: Message-ID: <577D1E41.7030004@gmail.com> hi Colin, yeah, thanks a ton that's great news! Is there somewhere specific that I can keep up with the Launchpad developments regarding relevant snap integration features? -Andy On 2016-07-06 05:00 AM, snapcraft-request at lists.ubuntu.com wrote: > Date: Tue, 5 Jul 2016 13:06:43 +0100 > From: Colin Watson > To: snapcraft at lists.ubuntu.com > Subject: Re: Future of Launchpad and the Snap Store > Message-ID: <20160705120642.GB30483 at riva.ucam.org> > Content-Type: text/plain; charset=us-ascii > > On Tue, Jul 05, 2016 at 01:05:09AM -0700, Andrew Keech wrote: >> > I've been whittling away at a snap for `gimp-edge` for a while, but feel >> > it's been overdue to bring in the maintainer of the ppa where I currently >> > get my builds from. Having been in touch with Thorsten now, he's got some >> > unanswered questions about the possible future relationships or integrations >> > between Launchpad and the Snap Store. Right now he wants to start a new ppa >> > just for delivering the snap as a separate project from the traditional >> > gimp-edge deb. >> > >> > Are there any best practice ideas that exist? Is there some kind of plan for >> > how the Store and Launchpad might work together, or if they'd exist >> > separately? > The plan is definitely for them to be able to work together, and indeed > quite a bit of this is already in place. The elevator pitch here is > that Launchpad handles code hosting and building, taking advantage of > the multi-architecture build farm we already have in place, while the > store handles publishing and distribution. > > What you can do today: > > * Push a Bazaar or Git branch to Launchpad containing a snapcraft.yaml > file at its top level, or import such a branch hosted elsewhere into > Launchpad's Bazaar code hosting > * Create a snap package in Launchpad, starting from a Bazaar or Git > branch as above > * Manually request builds of that snap package in Launchpad > * Configure the snap package in Launchpad to automatically push > completed builds to the store > * Manually publish the package in the store to whatever channels you > want > > What you'll be able to do soon (pending code review and deployment): > > * More obvious configuration for building snap packages for > architectures other than the defaults of amd64 and i386 > * Configure Launchpad to automatically build a snap package any time > the branch containing its snapcraft.yaml changes > * Configure Launchpad to automatically tell the store to publish a snap > package to a given set of channels after pushing completed builds > > What you'll be able to do later: > > * Git-to-Git code imports, which will be more robust than Git-to-Bazaar > and will cope with some cases that Git-to-Bazaar can't handle; this > will make things easier for people who e.g. want to host their code > primarily on GitHub > * Create a code import right from the "new snap package" form on > Launchpad > * Discover all of this from the store, rather than being > unconditionally guided through uploading packages manually > > Regarding integration with PPAs, snap packages aren't contained in PPAs > as such. However, you can either have Launchpad build the code that > goes into the snap package from scratch, or you can have it assemble a > snap package based on existing .debs in the Ubuntu archive and/or a PPA, > depending on what you put in your snapcraft.yaml; so the two approaches > can work together if you want, and this may make sense if you already > have good automation for doing edge builds in a PPA. > > Does that help with those unanswered questions? > > -- Colin Watson [cjwatson at ubuntu.com] From mark at ubuntu.com Wed Jul 6 17:13:03 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Wed, 6 Jul 2016 18:13:03 +0100 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> <26085.1467385388@parc.com> Message-ID: <577D3C1F.4030700@ubuntu.com> In principle I think the way to do this is to have the snap declare some specific .files that it wants to see, as parameters on the interface. This would trigger a one-time review, but assuming the request is sane, those would then be granted. The silversearcher example, though, is more interesting because it's explicitly wanting to search dotfiles that belong to other apps. Mark From cjwatson at ubuntu.com Wed Jul 6 17:57:06 2016 From: cjwatson at ubuntu.com (Colin Watson) Date: Wed, 6 Jul 2016 18:57:06 +0100 Subject: Snapcraft Digest, Vol 3, Issue 7 In-Reply-To: <577D1E41.7030004@gmail.com> References: <577D1E41.7030004@gmail.com> Message-ID: <20160706175706.GB25131@riva.ucam.org> On Wed, Jul 06, 2016 at 08:05:37AM -0700, Andrew Keech wrote: > Is there somewhere specific that I can keep up with the Launchpad > developments regarding relevant snap integration features? It's fallen by the wayside a bit due to other pressures, but I hope to get back to posting updates on blog.launchpad.net soon, and talking about snap integration there is certainly a high priority. -- Colin Watson [cjwatson at ubuntu.com] From neil at themcphails.uk Wed Jul 6 21:41:37 2016 From: neil at themcphails.uk (Neil McPhail) Date: Wed, 6 Jul 2016 22:41:37 +0100 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: <577D3C1F.4030700@ubuntu.com> References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> <26085.1467385388@parc.com> <577D3C1F.4030700@ubuntu.com> Message-ID: <577D7B11.70709@themcphails.uk> On 06/07/16 18:13, Mark Shuttleworth wrote: > In principle I think the way to do this is to have the snap declare some > specific .files that it wants to see, as parameters on the interface. > This would trigger a one-time review, but assuming the request is sane, > those would then be granted. > > The silversearcher example, though, is more interesting because it's > explicitly wanting to search dotfiles that belong to other apps. > > Mark > If a solution could be found for silversearcher, it may become a class solution for useful snaps of "git" and other build tools. I'm sure we need a more generic solution for dotfiles than exclusion+whitelisting. NMP From mark at ubuntu.com Wed Jul 6 21:46:54 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Wed, 6 Jul 2016 17:46:54 -0400 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: <577D7B11.70709@themcphails.uk> References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> <26085.1467385388@parc.com> <577D3C1F.4030700@ubuntu.com> <577D7B11.70709@themcphails.uk> Message-ID: <577D7C4E.9000206@ubuntu.com> On 06/07/16 17:41, Neil McPhail wrote: > On 06/07/16 18:13, Mark Shuttleworth wrote: >> In principle I think the way to do this is to have the snap declare some >> specific .files that it wants to see, as parameters on the interface. >> This would trigger a one-time review, but assuming the request is sane, >> those would then be granted. >> >> The silversearcher example, though, is more interesting because it's >> explicitly wanting to search dotfiles that belong to other apps. >> >> Mark >> > If a solution could be found for silversearcher, it may become a class > solution for useful snaps of "git" and other build tools. I'm sure we > need a more generic solution for dotfiles than exclusion+whitelisting. Why would a git snap need to see dotfiles other than its own? Code trees are not in .foo, they are in places like src/foo/ and dotfiles inside those would of course be readable using the home interface. Mark From neil at themcphails.uk Wed Jul 6 21:52:23 2016 From: neil at themcphails.uk (Neil McPhail) Date: Wed, 6 Jul 2016 22:52:23 +0100 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: <577D7C4E.9000206@ubuntu.com> References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> <26085.1467385388@parc.com> <577D3C1F.4030700@ubuntu.com> <577D7B11.70709@themcphails.uk> <577D7C4E.9000206@ubuntu.com> Message-ID: <577D7D97.3010600@themcphails.uk> On 06/07/16 22:46, Mark Shuttleworth wrote: > On 06/07/16 17:41, Neil McPhail wrote: >> On 06/07/16 18:13, Mark Shuttleworth wrote: >>> In principle I think the way to do this is to have the snap declare some >>> specific .files that it wants to see, as parameters on the interface. >>> This would trigger a one-time review, but assuming the request is sane, >>> those would then be granted. >>> >>> The silversearcher example, though, is more interesting because it's >>> explicitly wanting to search dotfiles that belong to other apps. >>> >>> Mark >>> >> If a solution could be found for silversearcher, it may become a class >> solution for useful snaps of "git" and other build tools. I'm sure we >> need a more generic solution for dotfiles than exclusion+whitelisting. > Why would a git snap need to see dotfiles other than its own? > > Code trees are not in .foo, they are in places like src/foo/ and > dotfiles inside those would of course be readable using the home interface. > > Mark Not everyone organises their git repos without dotfiles ;) For example, I'm currently hacking on the neverball-touch repo which has a ".build" file and ".tx" directory in the root. A snapped version of git, for example, wouldn't see them. NMP From neil at themcphails.uk Wed Jul 6 22:06:13 2016 From: neil at themcphails.uk (Neil McPhail) Date: Wed, 6 Jul 2016 23:06:13 +0100 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: <577D7D97.3010600@themcphails.uk> References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> <26085.1467385388@parc.com> <577D3C1F.4030700@ubuntu.com> <577D7B11.70709@themcphails.uk> <577D7C4E.9000206@ubuntu.com> <577D7D97.3010600@themcphails.uk> Message-ID: <577D80D5.602@themcphails.uk> On 06/07/16 22:52, Neil McPhail wrote: > On 06/07/16 22:46, Mark Shuttleworth wrote: >> On 06/07/16 17:41, Neil McPhail wrote: >>> On 06/07/16 18:13, Mark Shuttleworth wrote: >>>> In principle I think the way to do this is to have the snap declare >>>> some >>>> specific .files that it wants to see, as parameters on the interface. >>>> This would trigger a one-time review, but assuming the request is >>>> sane, >>>> those would then be granted. >>>> >>>> The silversearcher example, though, is more interesting because it's >>>> explicitly wanting to search dotfiles that belong to other apps. >>>> >>>> Mark >>>> >>> If a solution could be found for silversearcher, it may become a class >>> solution for useful snaps of "git" and other build tools. I'm sure we >>> need a more generic solution for dotfiles than exclusion+whitelisting. >> Why would a git snap need to see dotfiles other than its own? >> >> Code trees are not in .foo, they are in places like src/foo/ and >> dotfiles inside those would of course be readable using the home >> interface. >> >> Mark > Not everyone organises their git repos without dotfiles ;) > > For example, I'm currently hacking on the neverball-touch repo which > has a ".build" file and ".tx" directory in the root. A snapped version > of git, for example, wouldn't see them. > > NMP > Sorry, ignore the noise. You are correct, of course. I hadn't appreciated dotfiles in subdirectories are treated differently by the home interface. Apologies NMP From mark at ubuntu.com Wed Jul 6 22:14:22 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Wed, 6 Jul 2016 18:14:22 -0400 Subject: Should hidden files and directories be treated differently by the home plug In-Reply-To: <577D80D5.602@themcphails.uk> References: <3663cbceae49a8736d26c4df884ab230.squirrel@email.themcphails.uk> <26085.1467385388@parc.com> <577D3C1F.4030700@ubuntu.com> <577D7B11.70709@themcphails.uk> <577D7C4E.9000206@ubuntu.com> <577D7D97.3010600@themcphails.uk> <577D80D5.602@themcphails.uk> Message-ID: <577D82BE.5060508@ubuntu.com> No worries, it's always useful to dig in and test assumptions. Mark From mark at ubuntu.com Fri Jul 8 08:31:16 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Fri, 8 Jul 2016 04:31:16 -0400 Subject: Iterating on a skeleton snap with 'snap try' Message-ID: <577F64D4.1000705@ubuntu.com> There was a question on AskUbuntu today [0] about rapid iteration on snaps, avoiding the squashfs compression and remounting. As Daniel answered, this is what 'snap try' is for. A snap is a filesystem that is mounted so it looks like a directory with all of the app files (and dependencies etc) underneath a single directory. 'snap try' lets you actually use a normal directory as a snap. So you make a directory, put all the bits in that directory, then 'try' it. Now you can edit the snap directly just by changing things in that directory, live, and see the immediate consequences of those changes. I wonder if we shouldn't have a "sample snap" directory which snapcraft can create, to walk people through this? Something like: $ snapcraft try-example Unpacking into ./try-example/ ... Now you can 'snap try ./try-example' After the 'snap try' command, the directory is mounted as a snap. You can edit the contents of the try-example/ directory to live-edit the try-example snap. Start with 'snap list' and the 'try-example' command. I think all we need is a patch to snapcraft which includes the try-example snap contents and that error message, and some checking to make sure we can write the directory and are not over-writing a directory. Thoughts? Anybody want to have a go at that? NB there is currently a bug in snapd if you *remove* the directory while 'try'ing it, to be fixed in the next release. Mark [0] http://askubuntu.com/questions/795882/how-can-i-iterate-more-quickly-when-creating-a-snap -------------- next part -------------- An HTML attachment was scrubbed... URL: From xiaoguo.liu at canonical.com Fri Jul 8 08:49:12 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Fri, 8 Jul 2016 16:49:12 +0800 Subject: no such package 'qtubuntu-desktop' when building Ubuntu-clock snap example Message-ID: Hi I remember I successfully compiled the Ubuntu-clock example long time ago. Today, I tried to build it again using the latest snapcraft to build. However, I got an error like the following? What could be the problem for it? I am building it on 16.04, and my snapcraft versin is 2.12. I also got the same problem when build the sample app webcam-webui, and I reported at https://bugs.launchpad.net/developer-ubuntu-com/+bug/1600122. liuxg at liuxg:~/snappy/desktop/snappy-playpen/ubuntu-clock-app$ snapcraft "confinement" property not specified: defaulting to "strict" Preparing to pull clock Get:1 http://ppa.launchpad.net/ubuntu-sdk-team/ppa/ubuntu xenial InRelease [17.5 kB] Get:2 http://archive.linux.duke.edu/ubuntu xenial InRelease [247 kB] Get:3 http://ppa.launchpad.net/ubuntu-sdk-team/staging/ubuntu xenial InRelease [17.5 kB] ... Get:18 http://archive.linux.duke.edu/ubuntu xenial/restricted amd64 Packages [8,344 B] Get:19 http://archive.linux.duke.edu/ubuntu xenial/restricted Translation-en [2,908 B] Get:20 http://archive.linux.duke.edu/ubuntu xenial/restricted amd64 DEP-11 Metadata [186 B] Get:21 http://archive.linux.duke.edu/ubuntu xenial/multiverse amd64 Packages [144 kB] Get:22 http://archive.linux.duke.edu/ubuntu xenial/multiverse Translation-en [106 kB] Get:23 http://archive.linux.duke.edu/ubuntu xenial/multiverse amd64 DEP-11 Metadata [63.8 kB] Get:24 http://archive.linux.duke.edu/ubuntu xenial/multiverse DEP-11 64x64 Icons [230 kB] Get:25 http://archive.linux.duke.edu/ubuntu xenial-updates/main Sources [90.6 kB] Get:26 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse Sources [728 B] Get:27 http://archive.linux.duke.edu/ubuntu xenial-updates/main amd64 Packages [248 kB] Get:28 http://archive.linux.duke.edu/ubuntu xenial-updates/main Translation-en [98.1 kB] Get:29 http://archive.linux.duke.edu/ubuntu xenial-updates/main amd64 DEP-11 Metadata [186 kB] Get:30 http://archive.linux.duke.edu/ubuntu xenial-updates/main DEP-11 64x64 Icons [150 kB] Get:31 http://archive.linux.duke.edu/ubuntu xenial-updates/restricted amd64 DEP-11 Metadata [157 B] Get:32 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse amd64 Packages [1,176 B] Get:33 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse Translation-en [628 B] Get:34 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse amd64 DEP-11 Metadata [158 B] Get:35 http://archive.linux.duke.edu/ubuntu xenial-backports/main Sources [752 B] Get:36 http://archive.linux.duke.edu/ubuntu xenial-backports/main amd64 Packages [676 B] Get:37 http://archive.linux.duke.edu/ubuntu xenial-backports/main Translation-en [524 B] Get:38 http://archive.linux.duke.edu/ubuntu xenial-backports/main amd64 DEP-11 Metadata [197 B] Get:39 http://archive.linux.duke.edu/ubuntu xenial-backports/restricted amd64 DEP-11 Metadata [194 B] Get:40 http://archive.linux.duke.edu/ubuntu xenial-backports/multiverse amd64 DEP-11 Metadata [194 B] Get:41 http://archive.linux.duke.edu/ubuntu xenial-security/main Sources [30.5 kB] Get:42 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse Sources [728 B] Get:43 http://archive.linux.duke.edu/ubuntu xenial-security/main amd64 Packages [117 kB] Get:44 http://archive.linux.duke.edu/ubuntu xenial-security/main Translation-en [44.8 kB] Get:45 http://archive.linux.duke.edu/ubuntu xenial-security/main amd64 DEP-11 Metadata [44.0 kB] Get:46 http://archive.linux.duke.edu/ubuntu xenial-security/main DEP-11 64x64 Icons [50.8 kB] Get:47 http://archive.linux.duke.edu/ubuntu xenial-security/restricted amd64 DEP-11 Metadata [158 B] Get:48 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse amd64 Packages [1,176 B] Get:49 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse Translation-en [628 B] Get:50 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse amd64 DEP-11 Metadata [200 B] Fetched 6,174 kB in 6s (452 kB/s) Error downloading stage packages for part 'clock': no such package 'qtubuntu-desktop' Thanks & best regards, XiaoGuo -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From didrocks at ubuntu.com Fri Jul 8 09:17:12 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Fri, 8 Jul 2016 11:17:12 +0200 Subject: no such package 'qtubuntu-desktop' when building Ubuntu-clock snap example In-Reply-To: References: Message-ID: <577F6F98.40804@ubuntu.com> Le 08/07/2016 10:49, XiaoGuo Liu a écrit : > Hi > > I remember I successfully compiled the Ubuntu-clock example long time > ago. Today, I tried to build it again using the latest snapcraft to > build. However, I got an error like the following? What could be the > problem for it? I am building it on 16.04, and my snapcraft versin is > 2.12. I also got the same problem when build the sample app > webcam-webui, and I reported > at https://bugs.launchpad.net/developer-ubuntu-com/+bug/1600122. > > liuxg at liuxg:~/snappy/desktop/snappy-playpen/ubuntu-clock-app$ snapcraft > "confinement" property not specified: defaulting to "strict" > Preparing to pull clock > Get:1 http://ppa.launchpad.net/ubuntu-sdk-team/ppa/ubuntu xenial > InRelease [17.5 kB] > Get:2 http://archive.linux.duke.edu/ubuntu xenial InRelease [247 kB] > > Get:3 http://ppa.launchpad.net/ubuntu-sdk-team/staging/ubuntu xenial > InRelease [17.5 kB] > ... > Get:18 http://archive.linux.duke.edu/ubuntu xenial/restricted amd64 > Packages [8,344 B] > Get:19 http://archive.linux.duke.edu/ubuntu xenial/restricted > Translation-en [2,908 B] > Get:20 http://archive.linux.duke.edu/ubuntu xenial/restricted amd64 > DEP-11 Metadata [186 B] > Get:21 http://archive.linux.duke.edu/ubuntu xenial/multiverse amd64 > Packages [144 kB] > Get:22 http://archive.linux.duke.edu/ubuntu xenial/multiverse > Translation-en [106 kB] > Get:23 http://archive.linux.duke.edu/ubuntu xenial/multiverse amd64 > DEP-11 Metadata [63.8 kB] > Get:24 http://archive.linux.duke.edu/ubuntu xenial/multiverse DEP-11 > 64x64 Icons [230 kB] > Get:25 http://archive.linux.duke.edu/ubuntu xenial-updates/main > Sources [90.6 kB] > Get:26 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > Sources [728 B] > Get:27 http://archive.linux.duke.edu/ubuntu xenial-updates/main amd64 > Packages [248 kB] > Get:28 http://archive.linux.duke.edu/ubuntu xenial-updates/main > Translation-en [98.1 kB] > Get:29 http://archive.linux.duke.edu/ubuntu xenial-updates/main amd64 > DEP-11 Metadata [186 kB] > Get:30 http://archive.linux.duke.edu/ubuntu xenial-updates/main DEP-11 > 64x64 Icons [150 kB] > Get:31 http://archive.linux.duke.edu/ubuntu xenial-updates/restricted > amd64 DEP-11 Metadata [157 B] > Get:32 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > amd64 Packages [1,176 B] > Get:33 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > Translation-en [628 B] > Get:34 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > amd64 DEP-11 Metadata [158 B] > Get:35 http://archive.linux.duke.edu/ubuntu xenial-backports/main > Sources [752 B] > Get:36 http://archive.linux.duke.edu/ubuntu xenial-backports/main > amd64 Packages [676 B] > Get:37 http://archive.linux.duke.edu/ubuntu xenial-backports/main > Translation-en [524 B] > Get:38 http://archive.linux.duke.edu/ubuntu xenial-backports/main > amd64 DEP-11 Metadata [197 B] > Get:39 http://archive.linux.duke.edu/ubuntu > xenial-backports/restricted amd64 DEP-11 Metadata [194 B] > Get:40 http://archive.linux.duke.edu/ubuntu > xenial-backports/multiverse amd64 DEP-11 Metadata [194 B] > Get:41 http://archive.linux.duke.edu/ubuntu xenial-security/main > Sources [30.5 kB] > Get:42 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > Sources [728 B] > Get:43 http://archive.linux.duke.edu/ubuntu xenial-security/main amd64 > Packages [117 kB] > Get:44 http://archive.linux.duke.edu/ubuntu xenial-security/main > Translation-en [44.8 kB] > Get:45 http://archive.linux.duke.edu/ubuntu xenial-security/main amd64 > DEP-11 Metadata [44.0 kB] > Get:46 http://archive.linux.duke.edu/ubuntu xenial-security/main > DEP-11 64x64 Icons [50.8 kB] > Get:47 http://archive.linux.duke.edu/ubuntu xenial-security/restricted > amd64 DEP-11 Metadata [158 B] > Get:48 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > amd64 Packages [1,176 B] > Get:49 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > Translation-en [628 B] > Get:50 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > amd64 DEP-11 Metadata [200 B] > Fetched 6,174 kB in 6s (452 kB/s) > > Error downloading stage packages for part 'clock': no such package > 'qtubuntu-desktop' Hey, >From this output, it seems that the failure is in pulling the "qtubuntu-desktop" package. This package still exists in xenial and is in the universe pocket. This is failing for you because it seems that universe isn't enabled in your system sources.list. Do you mind checking this? "apt-cache policy qtubuntu-desktop" should returns some packages to you Cheers, Didier From xiaoguo.liu at canonical.com Fri Jul 8 09:44:42 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Fri, 8 Jul 2016 17:44:42 +0800 Subject: no such package 'qtubuntu-desktop' when building Ubuntu-clock snap example In-Reply-To: <577F6F98.40804@ubuntu.com> References: <577F6F98.40804@ubuntu.com> Message-ID: Hi Didier, Thanks. it resolved my problem. Best regards, XiaoGuo On Fri, Jul 8, 2016 at 5:17 PM, Didier Roche wrote: > Le 08/07/2016 10:49, XiaoGuo Liu a écrit : > > Hi > > > > I remember I successfully compiled the Ubuntu-clock example long time > > ago. Today, I tried to build it again using the latest snapcraft to > > build. However, I got an error like the following? What could be the > > problem for it? I am building it on 16.04, and my snapcraft versin is > > 2.12. I also got the same problem when build the sample app > > webcam-webui, and I reported > > at https://bugs.launchpad.net/developer-ubuntu-com/+bug/1600122. > > > > liuxg at liuxg:~/snappy/desktop/snappy-playpen/ubuntu-clock-app$ snapcraft > > "confinement" property not specified: defaulting to "strict" > > Preparing to pull clock > > Get:1 http://ppa.launchpad.net/ubuntu-sdk-team/ppa/ubuntu xenial > > InRelease [17.5 kB] > > Get:2 http://archive.linux.duke.edu/ubuntu xenial InRelease [247 kB] > > > > Get:3 http://ppa.launchpad.net/ubuntu-sdk-team/staging/ubuntu xenial > > InRelease [17.5 kB] > > ... > > Get:18 http://archive.linux.duke.edu/ubuntu xenial/restricted amd64 > > Packages [8,344 B] > > Get:19 http://archive.linux.duke.edu/ubuntu xenial/restricted > > Translation-en [2,908 B] > > Get:20 http://archive.linux.duke.edu/ubuntu xenial/restricted amd64 > > DEP-11 Metadata [186 B] > > Get:21 http://archive.linux.duke.edu/ubuntu xenial/multiverse amd64 > > Packages [144 kB] > > Get:22 http://archive.linux.duke.edu/ubuntu xenial/multiverse > > Translation-en [106 kB] > > Get:23 http://archive.linux.duke.edu/ubuntu xenial/multiverse amd64 > > DEP-11 Metadata [63.8 kB] > > Get:24 http://archive.linux.duke.edu/ubuntu xenial/multiverse DEP-11 > > 64x64 Icons [230 kB] > > Get:25 http://archive.linux.duke.edu/ubuntu xenial-updates/main > > Sources [90.6 kB] > > Get:26 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > > Sources [728 B] > > Get:27 http://archive.linux.duke.edu/ubuntu xenial-updates/main amd64 > > Packages [248 kB] > > Get:28 http://archive.linux.duke.edu/ubuntu xenial-updates/main > > Translation-en [98.1 kB] > > Get:29 http://archive.linux.duke.edu/ubuntu xenial-updates/main amd64 > > DEP-11 Metadata [186 kB] > > Get:30 http://archive.linux.duke.edu/ubuntu xenial-updates/main DEP-11 > > 64x64 Icons [150 kB] > > Get:31 http://archive.linux.duke.edu/ubuntu xenial-updates/restricted > > amd64 DEP-11 Metadata [157 B] > > Get:32 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > > amd64 Packages [1,176 B] > > Get:33 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > > Translation-en [628 B] > > Get:34 http://archive.linux.duke.edu/ubuntu xenial-updates/multiverse > > amd64 DEP-11 Metadata [158 B] > > Get:35 http://archive.linux.duke.edu/ubuntu xenial-backports/main > > Sources [752 B] > > Get:36 http://archive.linux.duke.edu/ubuntu xenial-backports/main > > amd64 Packages [676 B] > > Get:37 http://archive.linux.duke.edu/ubuntu xenial-backports/main > > Translation-en [524 B] > > Get:38 http://archive.linux.duke.edu/ubuntu xenial-backports/main > > amd64 DEP-11 Metadata [197 B] > > Get:39 http://archive.linux.duke.edu/ubuntu > > xenial-backports/restricted amd64 DEP-11 Metadata [194 B] > > Get:40 http://archive.linux.duke.edu/ubuntu > > xenial-backports/multiverse amd64 DEP-11 Metadata [194 B] > > Get:41 http://archive.linux.duke.edu/ubuntu xenial-security/main > > Sources [30.5 kB] > > Get:42 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > > Sources [728 B] > > Get:43 http://archive.linux.duke.edu/ubuntu xenial-security/main amd64 > > Packages [117 kB] > > Get:44 http://archive.linux.duke.edu/ubuntu xenial-security/main > > Translation-en [44.8 kB] > > Get:45 http://archive.linux.duke.edu/ubuntu xenial-security/main amd64 > > DEP-11 Metadata [44.0 kB] > > Get:46 http://archive.linux.duke.edu/ubuntu xenial-security/main > > DEP-11 64x64 Icons [50.8 kB] > > Get:47 http://archive.linux.duke.edu/ubuntu xenial-security/restricted > > amd64 DEP-11 Metadata [158 B] > > Get:48 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > > amd64 Packages [1,176 B] > > Get:49 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > > Translation-en [628 B] > > Get:50 http://archive.linux.duke.edu/ubuntu xenial-security/multiverse > > amd64 DEP-11 Metadata [200 B] > > Fetched 6,174 kB in 6s (452 kB/s) > > > > Error downloading stage packages for part 'clock': no such package > > 'qtubuntu-desktop' > > Hey, > > From this output, it seems that the failure is in pulling the > "qtubuntu-desktop" package. This package still exists in xenial and is > in the universe pocket. > This is failing for you because it seems that universe isn't enabled in > your system sources.list. Do you mind checking this? "apt-cache policy > qtubuntu-desktop" should returns some packages to you > > Cheers, > Didier > > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From daniel.holbach at canonical.com Fri Jul 8 12:44:47 2016 From: daniel.holbach at canonical.com (Daniel Holbach) Date: Fri, 8 Jul 2016 14:44:47 +0200 Subject: AskUbuntu tag changes Message-ID: <577FA03F.2030900@canonical.com> Hello everybody, after some discussion on meta.askubuntu.com Marco and Jorge helped implement the following changes http://meta.askubuntu.com/a/15740/94326 - Rename snap-package to snap, easy to remember and covers everything from the snaps themselves to snapcraft, this allows the core team to subscribe to just one tag so they can answer questions. - Alias snapcraft to snap. - Leave snappy aliased to ubuntu-core, which is a pure snappy system. This means you now have only two tags to monitor: snap and ubuntu-core. :-) Have a great day, Daniel From sergio.schvezov at canonical.com Fri Jul 8 15:32:31 2016 From: sergio.schvezov at canonical.com (Sergio Schvezov) Date: Fri, 8 Jul 2016 12:32:31 -0300 Subject: Iterating on a skeleton snap with 'snap try' In-Reply-To: <577F64D4.1000705@ubuntu.com> References: <577F64D4.1000705@ubuntu.com> Message-ID: <577FC78F.1050302@ubuntu.com> El 08/07/16 a las 05:31, Mark Shuttleworth escribió: > > There was a question on AskUbuntu today [0] about rapid iteration on > snaps, avoiding the squashfs compression and remounting. > > As Daniel answered, this is what 'snap try' is for. > > A snap is a filesystem that is mounted so it looks like a directory with > all of the app files (and dependencies etc) underneath a single > directory. 'snap try' lets you actually use a normal directory as a > snap. So you make a directory, put all the bits in that directory, then > 'try' it. Now you can edit the snap directly just by changing things in > that directory, live, and see the immediate consequences of those changes. > > I wonder if we shouldn't have a "sample snap" directory which snapcraft > can create, to walk people through this? > > Something like: > > $ snapcraft try-example > Unpacking into ./try-example/ ... Now you can 'snap try ./try-example' > > After the 'snap try' command, the directory is mounted as a snap. > You can edit the contents of the try-example/ directory to live-edit the > try-example snap. Start with 'snap list' and the 'try-example' command. > > I think all we need is a patch to snapcraft which includes the try-example > snap contents and that error message, and some checking to make sure we can > write the directory and are not over-writing a directory. > > Thoughts? Anybody want to have a go at that? Our `snapcraft init` provides a "nothing" snap that builds out of the box and gives you instructions on what to do next (run snapcraft). We can extend this to tell the user that the snap can be "tried" by doing x and y as well. Counter thoughts? :-) Cheers Sergio -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: OpenPGP digital signature URL: From michael.vogt at canonical.com Fri Jul 8 15:42:03 2016 From: michael.vogt at canonical.com (Michael Vogt) Date: Fri, 8 Jul 2016 17:42:03 +0200 Subject: New snapd updates process Message-ID: <20160708154203.GA4702@bod> Hello, we will enable an additional way to ship snapd updates soon. With the next update of the "ubuntu-core" snap we will automatically run snapd from this image [1]. This allows quicker and more seamless updates for snapd. Which of course means you get the bugfixes and new features quicker :) Additionally we plan to use the channels mechanism to give you more choices on what version of snapd to run. We will have the latest daily version of snapd available in the "edge" channel. We also plan to pre-release all our stable releases in the "candidate" channel first so that users can use/test them early. This means with: $ snap refresh --candidate ubuntu-core you can start tracking the candidate channel and get early access to the releases. To disable this feature (e.g. if you need to ensure you are on a particular version for some reason), you can set SNAP_REEXEC=0 in your environment. You also need to tweak the systemd service: # mkdir /etc/systemd/system/snapd.service.d/ # cat > /etc/systemd/system/snapd.service.d/local.conf < References: <577F64D4.1000705@ubuntu.com> <577FC78F.1050302@ubuntu.com> Message-ID: <577FCD92.40103@ubuntu.com> On 08/07/16 11:32, Sergio Schvezov wrote: > Our `snapcraft init` provides a "nothing" snap that builds out of the > box and gives you instructions on what to do next (run snapcraft). We > can extend this to tell the user that the snap can be "tried" by doing x > and y as well. > > Counter thoughts? :-) Only that a "nothing" snap wouldn't show off many of the things you'd want to live-edit in a 'try' situation. Mark -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From mark at ubuntu.com Fri Jul 8 16:02:14 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Fri, 8 Jul 2016 12:02:14 -0400 Subject: New snapd updates process In-Reply-To: <20160708154203.GA4702@bod> References: <20160708154203.GA4702@bod> Message-ID: <577FCE86.4030803@ubuntu.com> On 08/07/16 11:42, Michael Vogt wrote: > Additionally we plan to use the channels mechanism to give you more > choices on what version of snapd to run. We will have the latest daily > version of snapd available in the "edge" channel. We also plan to > pre-release all our stable releases in the "candidate" channel first > so that users can use/test them early. > > This means with: > > $ snap refresh --candidate ubuntu-core > > you can start tracking the candidate channel and get early access to > the releases. Oh, I love this, great call! Mark From ogra at ubuntu.com Fri Jul 8 16:07:14 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 08 Jul 2016 18:07:14 +0200 Subject: New snapd updates process In-Reply-To: <577FCE86.4030803@ubuntu.com> References: <20160708154203.GA4702@bod> <577FCE86.4030803@ubuntu.com> Message-ID: <1467994034.30394.4.camel@ubuntu.com> hi, Am Freitag, den 08.07.2016, 12:02 -0400 schrieb Mark Shuttleworth: > On 08/07/16 11:42, Michael Vogt wrote: > > > > Additionally we plan to use the channels mechanism to give you more > > choices on what version of snapd to run. We will have the latest > > daily > > version of snapd available in the "edge" channel. We also plan to > > pre-release all our stable releases in the "candidate" channel > > first > > so that users can use/test them early. > > > > This means with: > > > > $ snap refresh --candidate ubuntu-core > > > > you can start tracking the candidate channel and get early access > > to > > the releases. > Oh, I love this, great call! so the decision is to not build ubuntu core images from the archive anymore then but use PPAs ?   (happy to do the final adjustments to the builds, i just want the final confirmation. that discussion went on so long (and silent for so long) that i want to be absolutely sure before switching the build scripts) ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From gustavo.niemeyer at canonical.com Fri Jul 8 16:48:08 2016 From: gustavo.niemeyer at canonical.com (Gustavo Niemeyer) Date: Fri, 8 Jul 2016 13:48:08 -0300 Subject: New snapd updates process In-Reply-To: <1467994034.30394.4.camel@ubuntu.com> References: <20160708154203.GA4702@bod> <577FCE86.4030803@ubuntu.com> <1467994034.30394.4.camel@ubuntu.com> Message-ID: The idea is still to build the core images from the archive mostly, with a blessed stable PPA for snap tooling only. (as we discussed before) On Fri, Jul 8, 2016 at 1:07 PM, Oliver Grawert wrote: > hi, > Am Freitag, den 08.07.2016, 12:02 -0400 schrieb Mark Shuttleworth: > > On 08/07/16 11:42, Michael Vogt wrote: > > > > > > Additionally we plan to use the channels mechanism to give you more > > > choices on what version of snapd to run. We will have the latest > > > daily > > > version of snapd available in the "edge" channel. We also plan to > > > pre-release all our stable releases in the "candidate" channel > > > first > > > so that users can use/test them early. > > > > > > This means with: > > > > > > $ snap refresh --candidate ubuntu-core > > > > > > you can start tracking the candidate channel and get early access > > > to > > > the releases. > > Oh, I love this, great call! > > so the decision is to not build ubuntu core images from the archive > anymore then but use PPAs ? > > (happy to do the final adjustments to the builds, i just want the final > confirmation. that discussion went on so long (and silent for so long) > that i want to be absolutely sure before switching the build scripts) > > > ciao > oli > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- gustavo @ http://niemeyer.net -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Fri Jul 8 16:53:47 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 08 Jul 2016 18:53:47 +0200 Subject: New snapd updates process In-Reply-To: References: <20160708154203.GA4702@bod> <577FCE86.4030803@ubuntu.com> <1467994034.30394.4.camel@ubuntu.com> Message-ID: <1467996827.30394.18.camel@ubuntu.com> hi, Am Freitag, den 08.07.2016, 13:48 -0300 schrieb Gustavo Niemeyer: > > The idea is still to build the core images from the archive mostly, > with a blessed stable PPA for snap tooling only. > > (as we discussed before) > good,  do we have the stable PPA set up already ? i assume this is a job leo will do ?  (note that we already build the edge ubuntu-core snaps from Qa's daily edge PPA,  i need to set up a second build job for stable that uses the stable PPA as input once someone can give me the PPA url) also note that we are still blocked by  https://bugs.launchpad.net/software-center-agent/+bug/1580371 for auto-submitting our builds to the store. ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From michael.vogt at canonical.com Sat Jul 9 08:34:19 2016 From: michael.vogt at canonical.com (Michael Vogt) Date: Sat, 9 Jul 2016 10:34:19 +0200 Subject: ANN: snapd 2.0.10 available in 16.04 Message-ID: <20160709083419.GB2304@bod> Snapd 2.0.10 available in 16.04 ================================ The snappy team is very happy to announce that the 2.0.10 release is now available in 16.04 via "xenial-updates". The 2.0.10 release contains a number of improvements and fixes over the previous 2.0.9 release that was availalbe before. The highlights: - add support for snap {install,refresh} --{edge,beta,candidate,stable} shorthand - add mpris interface - add the camera interface - add optical-drive interface - allow gvfs shares in home and temporarily allow socketcall by default (LP: #1592901, LP: #1594675) - unversionized data directories availalbe via the SNAP_COMMON and SNAP_USER_COMMON directories - deal better with removed `snap try` dirs - various interface policy updates - fixes towards running snapd inside lxc - make `snap change ` show task progress - auto-connect home only if running on classic A more verbose changelog is available https://github.com/snapcore/snapd/blob/master/debian/changelog and the full details can be found here: https://github.com/snapcore/snapd/tree/2.0.10 We hope you like it as much as we do. If you find any issues, please let us know via: http://bugs.launchpad.net/snappy Thanks! Michael (on behalf of the snappy team) From mark at ubuntu.com Sat Jul 9 11:21:26 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Sat, 9 Jul 2016 07:21:26 -0400 Subject: snapd in lubuntu In-Reply-To: <576A8FF9.3040907@gmail.com> References: <576A8FF9.3040907@gmail.com> Message-ID: <5780DE36.8010402@ubuntu.com> On 22/06/16 09:17, Alekseenko Vasilii wrote: > I create virtual machines (Xubuntu, Kubuntu, Ubuntu GNOME, Lubuntu, > MATE) and only Lubuntu 16.04 displays an error. > > URL screenshot http://itmages.ru/image/view/4456402/27529211 > > In Lubuntu 16.04 languagetool wants open file for write > /home/vasilisc/.config/LanguageTool/gui.state > > BUT this impossible. Alekseenko, did you figure this out? I didn't see a follow-up on the list. Mark From vasilisc777 at gmail.com Mon Jul 11 05:12:57 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Mon, 11 Jul 2016 08:12:57 +0300 Subject: languagetool in lubuntu 16.04 Message-ID: <57832AD9.4050703@gmail.com> 09.07.2016 14:21, Mark Shuttleworth пишет: > On 22/06/16 09:17, Alekseenko Vasilii wrote: >> I create virtual machines (Xubuntu, Kubuntu, Ubuntu GNOME, Lubuntu, >> MATE) and only Lubuntu 16.04 displays an error. >> >> URL screenshot http://itmages.ru/image/view/4456402/27529211 >> >> In Lubuntu 16.04 languagetool wants open file for write >> /home/vasilisc/.config/LanguageTool/gui.state >> >> BUT this impossible. > > > Alekseenko, did you figure this out? I didn't see a follow-up on the list. > > Mark > the problem is actual. -- $ LANG=C sudo apt-cache policy snapd snapd: Installed: 2.0.10 Candidate: 2.0.10 Version table: *** 2.0.10 500 500 http://ru.archive.ubuntu.com/ubuntu xenial-updates/main amd64 Packages 100 /var/lib/dpkg/status 2.0.2 500 500 http://ru.archive.ubuntu.com/ubuntu xenial/main amd64 Packages $ lsb_release -a No LSB modules are available. Distributor ID: Ubuntu Description: Ubuntu 16.04 LTS Release: 16.04 Codename: xenial $ languagetool java.io.FileNotFoundException: /home/vasilisc/.config/LanguageTool/gui.state (No such file or directory) at java.io.FileOutputStream.open0(Native Method) at java.io.FileOutputStream.open(FileOutputStream.java:270) at java.io.FileOutputStream.(FileOutputStream.java:213) at java.io.FileOutputStream.(FileOutputStream.java:162) -- Best regards, vasilisc From xiaoguo.liu at canonical.com Mon Jul 11 08:32:33 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 11 Jul 2016 16:32:33 +0800 Subject: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system Message-ID: Hi, Today, I tried to snap a Ubuntu phone qmake project into a snappy application. When I run it I get a failure like the following: liuxg at liuxg:~/snappy/desktop/snaptest$ snaptest-app.test ln: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system (process:7180): Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale. Gtk-Message: Failed to load module "overlay-scrollbar" Gtk-Message: Failed to load module "gail" Gtk-Message: Failed to load module "atk-bridge" Gtk-Message: Failed to load module "unity-gtk-module" Gtk-Message: Failed to load module "canberra-gtk-module" XmbTextListToTextProperty result code -2 XmbTextListToTextProperty result code -2 May I know whether this is a real problem or not? Thanks & best regards, XiaoGuo -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From vasilisc777 at gmail.com Mon Jul 11 08:37:29 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Mon, 11 Jul 2016 11:37:29 +0300 Subject: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system In-Reply-To: References: Message-ID: <57835AC9.2030509@gmail.com> 11.07.2016 11:32, XiaoGuo Liu пишет: > Hi, > > Today, I tried to snap a Ubuntu phone qmake project into a snappy > application. When I run it I get a failure like the following: > > liuxg at liuxg:~/snappy/desktop/snaptest$ snaptest-app.test > ln: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system > > (process:7180): Gtk-WARNING **: Locale not supported by C library. > Using the fallback 'C' locale. > Gtk-Message: Failed to load module "overlay-scrollbar" > Gtk-Message: Failed to load module "gail" > Gtk-Message: Failed to load module "atk-bridge" > Gtk-Message: Failed to load module "unity-gtk-module" > Gtk-Message: Failed to load module "canberra-gtk-module" > XmbTextListToTextProperty result code -2 > XmbTextListToTextProperty result code -2 > > > May I know whether this is a real problem or not? > > Thanks & best regards, > XiaoGuo > > -- > XiaoGuo, Liu (刘晓国） > Mobile: +86-13911181302 > > use this https://developer.ubuntu.com/en/blog/2016/07/06/announcing-new-snap-desktop-launchers/ -- Best regards, vasilisc From didrocks at ubuntu.com Mon Jul 11 10:21:16 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Mon, 11 Jul 2016 12:21:16 +0200 Subject: languagetool in lubuntu 16.04 In-Reply-To: <57832AD9.4050703@gmail.com> References: <57832AD9.4050703@gmail.com> Message-ID: <5783731C.7090203@ubuntu.com> Le 11/07/2016 07:12, Vasilisc a écrit : > 09.07.2016 14:21, Mark Shuttleworth пишет: >> On 22/06/16 09:17, Alekseenko Vasilii wrote: >>> I create virtual machines (Xubuntu, Kubuntu, Ubuntu GNOME, Lubuntu, >>> MATE) and only Lubuntu 16.04 displays an error. >>> >>> URL screenshot http://itmages.ru/image/view/4456402/27529211 >>> >>> In Lubuntu 16.04 languagetool wants open file for write >>> /home/vasilisc/.config/LanguageTool/gui.state >>> >>> BUT this impossible. >> Alekseenko, did you figure this out? I didn't see a follow-up on the > list. >> Mark >> > the problem is actual. We discussed with jdstrand something similar with IntelliJ IDEA based IDE family (and so, Android Studio) trying to unconditionnally open something under /home//.IDEAxxx. Even the home plug doesn't give you access to create that directory (as it's hidden). >From what I got, it seems the global conscensus was to open a bugand a PR to upstream so that they don't harcode /home/ paths, but respect $HOME that is set by snap-confine. I'm adding Jamie so that he can confirm I didn't get him wrong Meanwhile, devmode should work for you, doesn't it? Cheers, Didier -------------- next part -------------- An HTML attachment was scrubbed... URL: From xiaoguo.liu at canonical.com Mon Jul 11 14:03:21 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 11 Jul 2016 22:03:21 +0800 Subject: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system In-Reply-To: <57835AC9.2030509@gmail.com> References: <57835AC9.2030509@gmail.com> Message-ID: Hi Vasilisc, Thanks. In fact, I have used the trick introduced there. Please find my code at: https://github.com/liu-xiao-guo/snaptest_working/blob/master/snapcraft.yaml The result I got was actually from the project. Thanks & best regards, XiaoGuo On Mon, Jul 11, 2016 at 4:37 PM, Vasilisc wrote: > 11.07.2016 11:32, XiaoGuo Liu пишет: > > Hi, >> >> Today, I tried to snap a Ubuntu phone qmake project into a snappy >> application. When I run it I get a failure like the following: >> >> liuxg at liuxg:~/snappy/desktop/snaptest$ snaptest-app.test >> ln: failed to create symbolic link >> '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system >> >> (process:7180): Gtk-WARNING **: Locale not supported by C library. >> Using the fallback 'C' locale. >> Gtk-Message: Failed to load module "overlay-scrollbar" >> Gtk-Message: Failed to load module "gail" >> Gtk-Message: Failed to load module "atk-bridge" >> Gtk-Message: Failed to load module "unity-gtk-module" >> Gtk-Message: Failed to load module "canberra-gtk-module" >> XmbTextListToTextProperty result code -2 >> XmbTextListToTextProperty result code -2 >> >> >> May I know whether this is a real problem or not? >> >> Thanks & best regards, >> XiaoGuo >> >> -- >> XiaoGuo, Liu (刘晓国） >> Mobile: +86-13911181302 >> >> >> use this > > https://developer.ubuntu.com/en/blog/2016/07/06/announcing-new-snap-desktop-launchers/ > > > -- > Best regards, > vasilisc > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Mon Jul 11 14:10:38 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Mon, 11 Jul 2016 16:10:38 +0200 Subject: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system In-Reply-To: References: Message-ID: <1468246238.18433.27.camel@ubuntu.com> hi, Am Montag, den 11.07.2016, 16:32 +0800 schrieb XiaoGuo Liu: > Hi, > > Today, I tried to snap a Ubuntu phone qmake project into a snappy > application. When I run it I get a failure like the following: > > liuxg at liuxg:~/snappy/desktop/snaptest$ snaptest-app.test  > ln: failed to create symbolic link '/home/liuxg/snap/snaptest- > app/x2/.themes/themes': Read-only file system use a wrapper script and add: cd $SNAP_USER_DATA  before the exec command, so that SNAP_USER_DATA becomes your actual workdir... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From ogra at ubuntu.com Mon Jul 11 14:15:02 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Mon, 11 Jul 2016 16:15:02 +0200 Subject: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system In-Reply-To: <1468246238.18433.27.camel@ubuntu.com> References: <1468246238.18433.27.camel@ubuntu.com> Message-ID: <1468246502.18433.29.camel@ubuntu.com> hi, Am Montag, den 11.07.2016, 16:10 +0200 schrieb Oliver Grawert: > hi, > Am Montag, den 11.07.2016, 16:32 +0800 schrieb XiaoGuo Liu: > > > > Hi, > > > > Today, I tried to snap a Ubuntu phone qmake project into a snappy > > application. When I run it I get a failure like the following: > > > > liuxg at liuxg:~/snappy/desktop/snaptest$ snaptest-app.test  > > ln: failed to create symbolic link '/home/liuxg/snap/snaptest- > > app/x2/.themes/themes': Read-only file system > use a wrapper script and add: > > cd $SNAP_USER_DATA  > > before the exec command, so that SNAP_USER_DATA becomes your actual > workdir... oops, I forgot ...  https://github.com/ogra1/laidout has an example  where i worked around the same problem  .... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From xiaoguo.liu at canonical.com Mon Jul 11 14:17:39 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 11 Jul 2016 22:17:39 +0800 Subject: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system In-Reply-To: <1468246502.18433.29.camel@ubuntu.com> References: <1468246238.18433.27.camel@ubuntu.com> <1468246502.18433.29.camel@ubuntu.com> Message-ID: Hi Oliver, Many thanks! I will definitely have a try. Best regards, XiaoGuo On Mon, Jul 11, 2016 at 10:15 PM, Oliver Grawert wrote: > hi, > Am Montag, den 11.07.2016, 16:10 +0200 schrieb Oliver Grawert: > > hi, > > Am Montag, den 11.07.2016, 16:32 +0800 schrieb XiaoGuo Liu: > > > > > > Hi, > > > > > > Today, I tried to snap a Ubuntu phone qmake project into a snappy > > > application. When I run it I get a failure like the following: > > > > > > liuxg at liuxg:~/snappy/desktop/snaptest$ snaptest-app.test > > > ln: failed to create symbolic link '/home/liuxg/snap/snaptest- > > > app/x2/.themes/themes': Read-only file system > > use a wrapper script and add: > > > > cd $SNAP_USER_DATA > > > > before the exec command, so that SNAP_USER_DATA becomes your actual > > workdir... > > > oops, I forgot ... > > https://github.com/ogra1/laidout has an example where i worked around > the same problem .... > > ciao > oli > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From xiaoguo.liu at canonical.com Mon Jul 11 14:45:03 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 11 Jul 2016 22:45:03 +0800 Subject: failed to create symbolic link '/home/liuxg/snap/snaptest-app/x2/.themes/themes': Read-only file system In-Reply-To: References: <1468246238.18433.27.camel@ubuntu.com> <1468246502.18433.29.camel@ubuntu.com> Message-ID: Hi Oliver, I just tried your method. Now, it is the same problem as before. My new code is at: https://github.com/liu-xiao-guo/snaptest_wrapper Thanks & best regards, XiaoGuo On Mon, Jul 11, 2016 at 10:17 PM, XiaoGuo Liu wrote: > Hi Oliver, > > Many thanks! I will definitely have a try. > > Best regards, > XiaoGuo > > On Mon, Jul 11, 2016 at 10:15 PM, Oliver Grawert wrote: > >> hi, >> Am Montag, den 11.07.2016, 16:10 +0200 schrieb Oliver Grawert: >> > hi, >> > Am Montag, den 11.07.2016, 16:32 +0800 schrieb XiaoGuo Liu: >> > > >> > > Hi, >> > > >> > > Today, I tried to snap a Ubuntu phone qmake project into a snappy >> > > application. When I run it I get a failure like the following: >> > > >> > > liuxg at liuxg:~/snappy/desktop/snaptest$ snaptest-app.test >> > > ln: failed to create symbolic link '/home/liuxg/snap/snaptest- >> > > app/x2/.themes/themes': Read-only file system >> > use a wrapper script and add: >> > >> > cd $SNAP_USER_DATA >> > >> > before the exec command, so that SNAP_USER_DATA becomes your actual >> > workdir... >> >> >> oops, I forgot ... >> >> https://github.com/ogra1/laidout has an example where i worked around >> the same problem .... >> >> ciao >> oli >> -- >> Snapcraft mailing list >> Snapcraft at lists.ubuntu.com >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/snapcraft >> >> > > > -- > XiaoGuo, Liu (刘晓国） > Mobile: +86-13911181302 > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From silver.bullet at zoho.com Tue Jul 12 00:56:53 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Tue, 12 Jul 2016 02:56:53 +0200 Subject: Is there a German howto/documentation for snapcraft? Message-ID: <20160712025653.225b637b@utnubu> Hi, I don't understand how to build a snap. Is there a German howto/documentation for snapcraft? Here is what I tried [1]. I guess I confused stage-packages [1] with build-packages, anyway, were do I need to add the build-packages? Perhaps like that: parts: qjackctl: source: qjackctl-0.4.2-r13-g5a8f67fc build-packages: [libasound2-dev, libjack-dev, libfoo-dev, libbar-dev] configflags: CFLAGS = -Wall -g export QTDIR=/usr/share/qt5 export QT_SELECT=qt5 [snip] Is there a way to depend on JACK1 or JACK2 outside of the snap? IOW to use the snap with JACK of the regular Ubuntu install? As you can see a script generates my snapcraft.yaml. Is it possible to clone and get the version number within the snapcraft.yaml? Regards, Ralf [1] [weremouse at moonstudio qjackctl]$ ls -hAl total 16K -rwxrwxr-x 1 weremouse weremouse 1.2K Jul 12 01:57 make_qjackctl-git_yaml.sh drwxrwxr-x 5 weremouse weremouse 4.0K Jul 12 01:58 qjackctl-0.4.2-r13-g5a8f67fc drwxrwxr-x 3 weremouse weremouse 4.0K Jul 12 01:58 setup -rw-rw-r-- 1 weremouse weremouse 806 Jul 12 01:58 snapcraft.yaml [weremouse at moonstudio qjackctl]$ snapcraft snap Issues while validating snapcraft.yaml: The 'stage-packages' property does not match the required schema: ['libasound2-dev', 'libjack-dev', 'qtbase5-dev', 'qttools5-dev-tools', 'libqt5x11extras5-dev', 'libx11-dev', 'jackd2', 'libasound2', 'libc6', 'libgcc1', 'libjack-jackd2-0', 'libqt5core5a', 'libqt5dbus5', 'libqt5gui5', 'libqt5widgets5', 'libqt5x11extras5', 'libqt5xml5', 'libstdc++6', 'libstdc++6', 'libx11-6'] has non-unique elements [weremouse at moonstudio qjackctl]$ head -n5 make_qjackctl-git_yaml.sh #!/bin/dash yaml() { cat< ../snapcraft.yaml exit [weremouse at moonstudio qjackctl]$ cat snapcraft.yaml name: qjackctl-git version: 0.4.2-r13-g5a8f67fc summary: QjackCtl description: QjackCtl offers a Qt front-end for the JACK low-latency audio server. confinement: strict apps: qjackctl: command: bin/qjackctl parts: qjackctl: source: qjackctl-0.4.2-r13-g5a8f67fc configflags: CFLAGS = -Wall -g export QTDIR=/usr/share/qt5 export QT_SELECT=qt5 stage-packages: - libasound2-dev - libjack-dev - qtbase5-dev - qttools5-dev-tools - libqt5x11extras5-dev - libx11-dev - jackd2 - libasound2 - libc6 - libgcc1 - libjack-jackd2-0 - libqt5core5a - libqt5dbus5 - libqt5gui5 - libqt5widgets5 - libqt5x11extras5 - libqt5xml5 - libstdc++6 - libstdc++6 - libx11-6 [weremouse at moonstudio qjackctl]$ ls -Ggh setup/gui/ total 0 lrwxrwxrwx 1 58 Jul 12 01:58 qjackctl.png -> ../../qjackctl-0.4.2-r13-g5a8f67fc/src/images/qjackctl.png From silver.bullet at zoho.com Tue Jul 12 01:45:37 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Tue, 12 Jul 2016 03:45:37 +0200 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <20160712025653.225b637b@utnubu> References: <20160712025653.225b637b@utnubu> Message-ID: <20160712034537.66a46895@utnubu> FWIW I'm neither familiar with Python nor with Go. [weremouse at moonstudio qjackctl]$ ls -Ggh total 4.0K -rwxrwxr-x 1 1.2K Jul 12 03:37 make_qjackctl-git_yaml.sh [weremouse at moonstudio qjackctl]$ ./make_qjackctl-git_yaml.sh Cloning into 'qjackctl'... remote: Counting objects: 9091, done. remote: Compressing objects: 100% (3050/3050), done. remote: Total 9091 (delta 7058), reused 7765 (delta 6003) Receiving objects: 100% (9091/9091), 21.66 MiB | 697.00 KiB/s, done. Resolving deltas: 100% (7058/7058), done. Checking connectivity... done. [weremouse at moonstudio qjackctl]$ snapcraft pull Issue while loading plugin: properties failed to load for qjackctl: Additional properties are not allowed ('configflags' was unexpected) [weremouse at moonstudio qjackctl]$ ls -Ggh total 16K -rwxrwxr-x 1 1.2K Jul 12 03:37 make_qjackctl-git_yaml.sh drwxrwxr-x 5 4.0K Jul 12 03:38 qjackctl-0.4.2-r13-g5a8f67fc drwxrwxr-x 3 4.0K Jul 12 03:38 setup -rw-rw-r-- 1 825 Jul 12 03:38 snapcraft.yaml [weremouse at moonstudio qjackctl]$ cat snapcraft.yaml name: qjackctl-git version: 0.4.2-r13-g5a8f67fc summary: QjackCtl description: QjackCtl offers a Qt front-end for the JACK low-latency audio server. confinement: strict apps: qjackctl: command: bin/qjackctl parts: qjackctl: plugin: make source: qjackctl-0.4.2-r13-g5a8f67fc build-packages: - libasound2-dev - libjack-dev - qtbase5-dev - qttools5-dev-tools - libqt5x11extras5-dev - libx11-dev configflags: CFLAGS = -Wall -g export QTDIR=/usr/share/qt5 export QT_SELECT=qt5 stage-packages: - jackd2 - libasound2 - libc6 - libgcc1 - libjack-jackd2-0 - libqt5core5a - libqt5dbus5 - libqt5gui5 - libqt5widgets5 - libqt5x11extras5 - libqt5xml5 - libstdc++6 - libx11-6 From mark at ubuntu.com Tue Jul 12 02:29:00 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Mon, 11 Jul 2016 19:29:00 -0700 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <20160712025653.225b637b@utnubu> References: <20160712025653.225b637b@utnubu> Message-ID: <578455EC.7000901@ubuntu.com> On 11/07/16 17:56, Ralf Mardorf wrote: > I guess I confused stage-packages [1] with build-packages, anyway, were > do I need to add the build-packages? The difference between these is "what do you need installed on the build system" and "what do you want inserted into the snap"? For example, if you install headers on the system when building the app, these allow you to compile the app that's going into the snap. Since you don't want them in the snap itself, you put them in build-packages not stage-packages. That way they are installed on the build system ("apt-get foo") but not pushed into the snap itself after the app is built. > Is there a way to depend on JACK1 or JACK2 outside of the snap? IOW to > use the snap with JACK of the regular Ubuntu install? Connections between the snap and the host Linux need to be described by a plug/slot/interface. I know there are audio-related interfaces (MPRIS was in snapd 2.0.10) but I don't think there is a JACK specific one. If you are feeling adventurous, you may want to look at the snapd code for the pulse audio interfaces and cargo-cult that into a patch for JACK support. In short, you'll have a little bit of yaml in your snap.yaml that says to snapd "this snap uses JACK", and when snapd sees that it will open up the pathways for your snap to talk to JACK when installed. Mark From silver.bullet at zoho.com Tue Jul 12 05:40:52 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Tue, 12 Jul 2016 07:40:52 +0200 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <578455EC.7000901@ubuntu.com> References: <20160712025653.225b637b@utnubu> <578455EC.7000901@ubuntu.com> Message-ID: <20160712074052.484f0c7d@utnubu> On Mon, 11 Jul 2016 19:29:00 -0700, Mark Shuttleworth wrote: >On 11/07/16 17:56, Ralf Mardorf wrote: >> I guess I confused stage-packages [1] with build-packages, anyway, >> were do I need to add the build-packages? > >The difference between these is "what do you need installed on the >build system" and "what do you want inserted into the snap"? For >example, [snip] Thank you Mark, maybe I'll feel adventurous at a more suitable time. Regards, Ralf From ogra at ubuntu.com Tue Jul 12 11:00:57 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Tue, 12 Jul 2016 13:00:57 +0200 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <20160712025653.225b637b@utnubu> References: <20160712025653.225b637b@utnubu> Message-ID: <1468321257.20308.66.camel@anubis> hi, Am Dienstag, den 12.07.2016, 02:56 +0200 schrieb Ralf Mardorf: > Hi, > > I don't understand how to build a snap. > > Is there a German howto/documentation for snapcraft? > > Here is what I tried [1]. > > I guess I confused stage-packages [1] with build-packages, anyway, were > do I need to add the build-packages? > > Perhaps like that: > > parts: > qjackctl: > source: qjackctl-0.4.2-r13-g5a8f67fc > build-packages: > [libasound2-dev, libjack-dev, libfoo-dev, libbar-dev] > configflags: > CFLAGS = -Wall -g > export QTDIR=/usr/share/qt5 > export QT_SELECT=qt5 > [snip] you did not define a "plugin:" parameter ... looking at the qjackctl source it seems that uses autotools, so you could pick that plugin to get more info about possible parameters: snapcraft help autotools you could also use the "make" plugin and have your Makefile do the necessary steps https://github.com/ogra1/upnp-server is an example where i use two subdirs with makefiles and the make plugin for git clone, configure and friends (you could even have it run the steps you do in your shell script (apart from the snapcraft.yaml creation indeed)) ... > > Is there a way to depend on JACK1 or JACK2 outside of the snap? IOW to > use the snap with JACK of the regular Ubuntu install? > as mark mentioned, you would need a jack interface. for dependencies: your snap needs to be self contained, so if there are runtime library dependencies you need to include them in stage-packages (likewise for build time dependencies and build-packages). > As you can see a script generates my snapcraft.yaml. Is it possible to > clone and get the version number within the snapcraft.yaml? > i dont think that is possible yet but i remember someone filed a bug about the other way around: https://bugs.launchpad.net/snapcraft/+bug/1588336 while you currently work top down starting from a shellscript and then going to snapcraft, try to think the other way around, snapcraft (and snapcraft.yaml) are the toplevel authoritative command and config. you can call shellscripts from the make command to generate and patch things then. when i re-work your snapcraft.yaml a little like below and add a Makefile for the build and cloning (even further below) i get a proper build ... (the snapcraft.yaml still has some boilerplate summary and description from "snapcraft init" ... ) root at anubis:/qjackctl# cat snapcraft.yaml name: qjackctl version: 0.1 summary: This is my-snap's summary # 79 char long summary description: This is my-snap's description # a longer description for the snap confinement: devmode # use "strict" to enforce system access only via declared interfaces parts: qjctlbuild: source: ./ plugin: make build-packages: - autoconf - automake - autopoint - autotools-dev - gcc - gettext - git - libasound2-dev - libjack-dev - qt5-default - qt5-qmake - qtbase5-dev - qttools5-dev-tools - libqt5x11extras5-dev - libx11-dev stage-packages: - jackd2 - libasound2 - libc6 - libgcc1 - libjack-jackd2-0 - libqt5core5a - libqt5dbus5 - libqt5gui5 - libqt5widgets5 - libqt5x11extras5 - libqt5xml5 - libstdc++6 - libx11-6 after: [desktop/qt5] root at anubis:/qjackctl# cat Makefile CFLAGS = -Wall -g all: git clone git://git.code.sf.net/p/qjackctl/code qjackctl (cd qjackctl; ./autogen.sh; ./configure; make) install: echo "foo" root at anubis:/qjackctl# have fun :) ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From david.planella at ubuntu.com Tue Jul 12 11:09:16 2016 From: david.planella at ubuntu.com (David Planella) Date: Tue, 12 Jul 2016 13:09:16 +0200 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <20160712025653.225b637b@utnubu> References: <20160712025653.225b637b@utnubu> Message-ID: Hi Ralf, In addition to the replies you've already gotten here, we're running a weekly event called the snappy playpen, where snap pioneers help each other getting their apps snapped. It's on today all day (across European and some US timezones), you'll be most welcome to join us! You'll find out more here: https://daniel.holba.ch/blog/2016/07/snappy-playpen-event-tomorrow Cheers, David. On Tue, Jul 12, 2016 at 2:56 AM, Ralf Mardorf wrote: > Hi, > > I don't understand how to build a snap. > > Is there a German howto/documentation for snapcraft? > > Here is what I tried [1]. > > I guess I confused stage-packages [1] with build-packages, anyway, were > do I need to add the build-packages? > > Perhaps like that: > > parts: > qjackctl: > source: qjackctl-0.4.2-r13-g5a8f67fc > build-packages: > [libasound2-dev, libjack-dev, libfoo-dev, libbar-dev] > configflags: > CFLAGS = -Wall -g > export QTDIR=/usr/share/qt5 > export QT_SELECT=qt5 > [snip] > > Is there a way to depend on JACK1 or JACK2 outside of the snap? IOW to > use the snap with JACK of the regular Ubuntu install? > > As you can see a script generates my snapcraft.yaml. Is it possible to > clone and get the version number within the snapcraft.yaml? > > Regards, > Ralf > > [1] > [weremouse at moonstudio qjackctl]$ ls -hAl > total 16K > -rwxrwxr-x 1 weremouse weremouse 1.2K Jul 12 01:57 > make_qjackctl-git_yaml.sh > drwxrwxr-x 5 weremouse weremouse 4.0K Jul 12 01:58 > qjackctl-0.4.2-r13-g5a8f67fc > drwxrwxr-x 3 weremouse weremouse 4.0K Jul 12 01:58 setup > -rw-rw-r-- 1 weremouse weremouse 806 Jul 12 01:58 snapcraft.yaml > [weremouse at moonstudio qjackctl]$ snapcraft snap > Issues while validating snapcraft.yaml: The 'stage-packages' property does > not match the required schema: ['libasound2-dev', 'libjack-dev', > 'qtbase5-dev', 'qttools5-dev-tools', 'libqt5x11extras5-dev', 'libx11-dev', > 'jackd2', 'libasound2', 'libc6', 'libgcc1', 'libjack-jackd2-0', > 'libqt5core5a', 'libqt5dbus5', 'libqt5gui5', 'libqt5widgets5', > 'libqt5x11extras5', 'libqt5xml5', 'libstdc++6', 'libstdc++6', 'libx11-6'] > has non-unique elements > [weremouse at moonstudio qjackctl]$ head -n5 make_qjackctl-git_yaml.sh > #!/bin/dash > > yaml() { > cat< name: qjackctl-git > [weremouse at moonstudio qjackctl]$ tail -n12 make_qjackctl-git_yaml.sh > EOF > } > > git clone git://git.code.sf.net/p/qjackctl/code qjackctl > cd qjackctl > qjackctl_version=$(git describe --abbrev=8|cut -f2 -d"l"|sed -r > 's/^.//;s/_/./g;s/([^-]*-g)/r\1/') > mv ../qjackctl ../qjackctl-$qjackctl_version > mkdir -p ../setup/gui > ln -s ../../qjackctl-$qjackctl_version/src/images/qjackctl.png > ../setup/gui/qjackctl.png > yaml > ../snapcraft.yaml > > exit > [weremouse at moonstudio qjackctl]$ cat snapcraft.yaml > name: qjackctl-git > version: 0.4.2-r13-g5a8f67fc > summary: QjackCtl > description: QjackCtl offers a Qt front-end for the JACK low-latency audio > server. > confinement: strict > > apps: > qjackctl: > command: bin/qjackctl > > parts: > qjackctl: > source: qjackctl-0.4.2-r13-g5a8f67fc > configflags: > CFLAGS = -Wall -g > export QTDIR=/usr/share/qt5 > export QT_SELECT=qt5 > stage-packages: > - libasound2-dev > - libjack-dev > - qtbase5-dev > - qttools5-dev-tools > - libqt5x11extras5-dev > - libx11-dev > - jackd2 > - libasound2 > - libc6 > - libgcc1 > - libjack-jackd2-0 > - libqt5core5a > - libqt5dbus5 > - libqt5gui5 > - libqt5widgets5 > - libqt5x11extras5 > - libqt5xml5 > - libstdc++6 > - libstdc++6 > - libx11-6 > [weremouse at moonstudio qjackctl]$ ls -Ggh setup/gui/ > total 0 > lrwxrwxrwx 1 58 Jul 12 01:58 qjackctl.png -> > ../../qjackctl-0.4.2-r13-g5a8f67fc/src/images/qjackctl.png > > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Tue Jul 12 11:16:30 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Tue, 12 Jul 2016 13:16:30 +0200 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <1468321257.20308.66.camel@anubis> References: <20160712025653.225b637b@utnubu> <1468321257.20308.66.camel@anubis> Message-ID: <1468322190.20308.69.camel@anubis> hi, Am Dienstag, den 12.07.2016, 13:00 +0200 schrieb Oliver Grawert: .... and one thing i forgot ... you do not need interfaces initially, you can always install your snap with the --devmode parameter which gives it as much access as any deb package has, without any interfaces at all (such --devmode snaps can never go to the stable channel in the store though, but you can put them into the "edge" or "beta" channels until an interface exists) ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From mihamina-rakotomandimby at rktmb.org Tue Jul 12 11:49:52 2016 From: mihamina-rakotomandimby at rktmb.org (Mihamina RAKOTOMANDIMBY) Date: Tue, 12 Jul 2016 14:49:52 +0300 Subject: resident package database update Message-ID: <20160712144952.41b59d22@andriamanitra> Hi all, Snap can install packages on several distribution: - When installing on Debian (+derivatives) is the dpkg database updated? - When installing on RedHat/OpenSuse (+derivatives) is the rpm database updated? Is there a plan for that or is it just no go? Thank you. From ogra at ubuntu.com Tue Jul 12 12:02:51 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Tue, 12 Jul 2016 14:02:51 +0200 Subject: resident package database update In-Reply-To: <20160712144952.41b59d22@andriamanitra> References: <20160712144952.41b59d22@andriamanitra> Message-ID: <1468324971.20308.73.camel@anubis> hi, Am Dienstag, den 12.07.2016, 14:49 +0300 schrieb Mihamina RAKOTOMANDIMBY: > Hi all, > > Snap can install packages on several distribution: > - When installing on Debian (+derivatives) is the dpkg database updated? > - When installing on RedHat/OpenSuse (+derivatives) is the rpm database > updated? > > Is there a plan for that or is it just no go? what would you expect there to be added ? (they are widely different package formats (and concepts) each using their own DB and meta data) ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From mihamina-rakotomandimby at rktmb.org Tue Jul 12 12:31:55 2016 From: mihamina-rakotomandimby at rktmb.org (Mihamina RAKOTOMANDIMBY) Date: Tue, 12 Jul 2016 15:31:55 +0300 Subject: resident package database update In-Reply-To: <1468324971.20308.73.camel@anubis> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> Message-ID: <20160712153155.60927ca3@andriamanitra> On Tue, 12 Jul 2016 14:02:51 +0200 Oliver Grawert wrote: > > Snap can install packages on several distribution: > > - When installing on Debian (+derivatives) is the dpkg database > > updated? > > - When installing on RedHat/OpenSuse (+derivatives) is the rpm > > database updated? > > > > Is there a plan for that or is it just no go? > > what would you expect there to be added ? > (they are widely different package formats (and concepts) each using > their own DB and meta data) I just expect integration to the existing system. If a package is added via Snap I find it fair to be added to the existing package database. Yes it would cover a wide basis, but it makes sense to me. Doesn't it? (I'm really asking, from my POV) From ogra at ubuntu.com Tue Jul 12 12:41:15 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Tue, 12 Jul 2016 14:41:15 +0200 Subject: resident package database update In-Reply-To: <20160712153155.60927ca3@andriamanitra> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> Message-ID: <1468327275.18433.45.camel@ubuntu.com> hi, Am Dienstag, den 12.07.2016, 15:31 +0300 schrieb Mihamina RAKOTOMANDIMBY: > On Tue, 12 Jul 2016 14:02:51 +0200 > Oliver Grawert wrote: > > > > > > > > > Snap can install packages on several distribution: > > > - When installing on Debian (+derivatives) is the dpkg database > > > updated? > > > - When installing on RedHat/OpenSuse (+derivatives) is the rpm > > > database updated? > > > > > > Is there a plan for that or is it just no go?   > > what would you expect there to be added ? > > (they are widely different package formats (and concepts) each > > using > > their own DB and meta data) > I just expect integration to the existing system. > > If a package is added via Snap I find it fair to be added to the > existing package database. > > Yes it would cover a wide basis, but it makes sense to me. Doesn't > it? snaps are not like debs or rpms, you can have a ton of different binaries and libs inside ... and these binaries and libs might not even be named like the content of the rpm/deb ... how would you translate that into a rpm/deb db entry ? snaps can also be built from debs/rpms or upstream git trees tarballs or whatever else upstream source exists, if my snap ships libfoo, libfoo can not be seen from the normal system (only the binary using libfoo inside teh snap can see it), and libfoo could be the last git trunk version vs some stable release the distro provides. you can surely have the interconnection between package db's on a very high level (like we do in gnome-software in ubuntu for example), but i doubt it makes much sense to mix up packaging systems that are conceptual so extremely different. ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From silver.bullet at zoho.com Tue Jul 12 13:15:47 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Tue, 12 Jul 2016 15:15:47 +0200 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <1468322190.20308.69.camel@anubis> References: <20160712025653.225b637b@utnubu> <1468321257.20308.66.camel@anubis> <1468322190.20308.69.camel@anubis> Message-ID: <20160712151547.76729ba5@utnubu> Thank you David and Oli, I'm short in time, but flagged Oli's first reply with a color label (12 Jul 2016 13:00:57 +0200) and I'll start with it, if I have the time to do so. IIUC I could add a "git checkout 1234567", to get commit 1234567, or anything else between the open and the closed bracket. all: git clone git://git.code.sf.net/p/qjackctl/code qjackctl (cd qjackctl; git checkout 1234567; ./autogen.sh; ./configure; make) Regards, Ralf From ogra at ubuntu.com Tue Jul 12 13:20:03 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Tue, 12 Jul 2016 15:20:03 +0200 Subject: Is there a German howto/documentation for snapcraft? In-Reply-To: <20160712151547.76729ba5@utnubu> References: <20160712025653.225b637b@utnubu> <1468321257.20308.66.camel@anubis> <1468322190.20308.69.camel@anubis> <20160712151547.76729ba5@utnubu> Message-ID: <1468329603.18433.51.camel@ubuntu.com> hi, Am Dienstag, den 12.07.2016, 15:15 +0200 schrieb Ralf Mardorf: > Thank you David and Oli, > > I'm short in time, but flagged Oli's first reply with a color label > (12 Jul 2016 13:00:57 +0200) and I'll start with it, if I have the > time > to do so. > > IIUC I could add a "git checkout 1234567", to get commit 1234567, or > anything else between the open and the closed bracket. > > all: > git clone git://git.code.sf.net/p/qjackctl/code qjackctl > (cd qjackctl; git checkout 1234567; ./autogen.sh; ./configure; > make) yup, that would work ... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From mihamina-rakotomandimby at rktmb.org Tue Jul 12 13:29:19 2016 From: mihamina-rakotomandimby at rktmb.org (Mihamina RAKOTOMANDIMBY) Date: Tue, 12 Jul 2016 16:29:19 +0300 Subject: resident package database update In-Reply-To: <1468327275.18433.45.camel@ubuntu.com> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> <1468327275.18433.45.camel@ubuntu.com> Message-ID: <20160712162919.5815af04@andriamanitra> On Tue, 12 Jul 2016 14:41:15 +0200 Oliver Grawert wrote: > > > > > > > > > > > Snap can install packages on several distribution: > > > > - When installing on Debian (+derivatives) is the dpkg database > > > > updated? > > > > - When installing on RedHat/OpenSuse (+derivatives) is the rpm > > > > database updated? > > > > > > > > Is there a plan for that or is it just no go?   > > > what would you expect there to be added ? > > > (they are widely different package formats (and concepts) each > > > using > > > their own DB and meta data) > > I just expect integration to the existing system. > > > > If a package is added via Snap I find it fair to be added to the > > existing package database. > > > > Yes it would cover a wide basis, but it makes sense to me. Doesn't > > it? > > snaps are not like debs or rpms, you can have a ton of different > binaries and libs inside ... OK, I got it. > and these binaries and libs might not > even be named like the content of the rpm/deb ... how would you > translate that into a rpm/deb db entry ? I though about another layer of naming correspondance. But I agree it would be PITA. Thank you for your answers. :-) From seth.arnold at canonical.com Tue Jul 12 20:51:34 2016 From: seth.arnold at canonical.com (Seth Arnold) Date: Tue, 12 Jul 2016 13:51:34 -0700 Subject: resident package database update In-Reply-To: <20160712153155.60927ca3@andriamanitra> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> Message-ID: <20160712205134.GB24443@hunt> On Tue, Jul 12, 2016 at 03:31:55PM +0300, Mihamina RAKOTOMANDIMBY wrote: > I just expect integration to the existing system. > > If a package is added via Snap I find it fair to be added to the > existing package database. I can understand the desire but this would lead to an incredible increase in complexity -- users would then expect to be able to use e.g. dpkg --purge or rpm -e to remove these packages, and then snapd would need to be made aware of changes made via other tools, and concurrent operations kicked off with different tools would be a disaster to try to coordinate. They are different types of packages, they live in different portions of the filesystem namespace, and they have different purposes, so I think they should be kept separate. Thanks -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 473 bytes Desc: not available URL: From silver.bullet at zoho.com Tue Jul 12 22:30:02 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Wed, 13 Jul 2016 00:30:02 +0200 Subject: resident package database update In-Reply-To: <20160712205134.GB24443@hunt> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> <20160712205134.GB24443@hunt> Message-ID: <20160713003002.732485ae@utnubu> On Tue, 12 Jul 2016 13:51:34 -0700, Seth Arnold wrote: >I can understand the desire but this would lead to an incredible >increase in complexity Not necessarily. For Arch Linux a developer introduced a tool for Flatpak that should become more or less equal to the official package management. IIUC the developer also was thinking about integration into the official package manager, but I guess there is no acceptance. What I'm tying to say is, that it is possible to integrate snaps, Flatpak, Nix/Guix or what ever else to the already existing package management of a distro. Imagine you even could write a simple wrapper that handles both, the distro's package management and snaps. I'm very skeptic regarding the idea of the snap, Flatak etc. approach. Developers from upstream and some PPAs already provide installs to /opt, that perfectly integrate to the rest of the install, but that don't use shared libs, instead they ship with all the required dependencies. IMO the /opt approach is better. However, "Snaps don’t intrinsically depend on the Ubuntu store, that’s just what snapd does today, and we expect people will have different stores for their snaps in future." - http://snapcraft.io/ Assumed there should be acceptance for e.g. snaps, then distros most likely will provide their own tools. Not necessarily a "store", but much likely an integration to the already existing package management. This would not mix e.g. deb, rpm, pacman with snaps, just the management of both, the classic package management and the snaps could be handled by one tool. For Arch Linux there are already tools available, that handle official packages providing binaries with the official tool to build from source, so that building software from source from a user repository, could be done by the same commands, that are used to install binaries from official repositories. Actually such tools are just wrappers. The difference to integrating snaps just is, that those tools build regular packages, but a wrapper also could handle different formats. 2 Cents, Ralf From silver.bullet at zoho.com Tue Jul 12 23:03:14 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Wed, 13 Jul 2016 01:03:14 +0200 Subject: resident package database update In-Reply-To: <20160713003002.732485ae@utnubu> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> <20160712205134.GB24443@hunt> <20160713003002.732485ae@utnubu> Message-ID: <20160713010314.242158b8@utnubu> PS: Imagine you are searching packages by description. What would be complex, if one tool could search all snaps and deb packages and the output would show the result for the search of both, snaps and deps? It's easy to provide. I guess this is what users expect and I fear it will lead to most uneducated computer users, since this "mobile phone approach" has got pros and cons. IMO loss of knowledge unfortunately is one of the cons, but that's off-topic. Regarding the topic, one command could handle deb repos and a snap shop, the same command could install a snap or a deb, search bot data bases etc.. From sergio.schvezov at canonical.com Tue Jul 12 23:14:49 2016 From: sergio.schvezov at canonical.com (Sergio Schvezov) Date: Tue, 12 Jul 2016 20:14:49 -0300 Subject: resident package database update In-Reply-To: <20160713010314.242158b8@utnubu> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> <20160712205134.GB24443@hunt> <20160713003002.732485ae@utnubu> <20160713010314.242158b8@utnubu> Message-ID: <578579E9.2040707@ubuntu.com> El 12/07/16 a las 20:03, Ralf Mardorf escribió: > PS: > > Imagine you are searching packages by description. What would be > complex, if one tool could search all snaps and deb packages and the > output would show the result for the search of both, snaps and deps? > > It's easy to provide. In Ubuntu, there's a GUI for this (Ubuntu Software, based out of Gnome Software). > > I guess this is what users expect and I fear it will lead to most > uneducated computer users, since this "mobile phone approach" has got > pros and cons. IMO loss of knowledge unfortunately is one of the cons, > but that's off-topic. With such GUI it shouldn't be a problem. > Regarding the topic, one command could handle deb repos and a snap > shop, the same command could install a snap or a deb, search bot data > bases etc.. So it feels this is about searching and installing and more about aggregating than sharing the same database? -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: OpenPGP digital signature URL: From silver.bullet at zoho.com Tue Jul 12 23:31:52 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Wed, 13 Jul 2016 01:31:52 +0200 Subject: resident package database update In-Reply-To: <578579E9.2040707@ubuntu.com> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> <20160712205134.GB24443@hunt> <20160713003002.732485ae@utnubu> <20160713010314.242158b8@utnubu> <578579E9.2040707@ubuntu.com> Message-ID: <20160713013152.20e9d138@utnubu> On Tue, 12 Jul 2016 20:14:49 -0300, Sergio Schvezov wrote: >So it feels this is about searching and installing and more about >aggregating than sharing the same database? They share virtual the same data base ;). If two management systems would share the same database, they would become a new, third package management ;). Users might not care about what happens under the hood, especially those who prefer a GUI over command line tools to maintain their installs. GUIs don't make life always easier, since command line tools could be used in combination with other command line tools and handy bashisms, so tailored installs could be easier realized by using command line. From superlinkx at gmail.com Wed Jul 13 05:52:50 2016 From: superlinkx at gmail.com (Steven Holms) Date: Wed, 13 Jul 2016 05:52:50 +0000 Subject: Including Postgresql in a Snap App Message-ID: Has anyone had success in setting up a postgresql server within a snap? I've been reading over the docs, and figured I'd try adding postgresql as a stage-package, but now I'm not sure how to get it up and running within my snap. I think I'm on the right track, I just don't know how to configure the "app" portion of it so that the daemon will run. My app is pretty simple. It's a go server that needs access to postgresql for it's backend. I know I can communicate with an external postgresql database, but was hoping to get it all into one package. Any help is appreciated. -- Steven Holms, Web Developer -------------- next part -------------- An HTML attachment was scrubbed... URL: From vasilisc777 at gmail.com Wed Jul 13 06:16:06 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Wed, 13 Jul 2016 09:16:06 +0300 Subject: Why lubuntu so behaves? Message-ID: <5785DCA6.4080105@gmail.com> Why lubuntu so behaves? I create virtual machines (Xubuntu, Kubuntu, Ubuntu GNOME, Lubuntu, MATE) and only Lubuntu 16.04 displays an error. ------------------------ ** (java:2607): WARNING **: Error retrieving accessibility bus address: org.freedesktop.DBus.Error.ServiceUnknown: The name org.a11y.Bus was not provided by any .service files Failed to create secure directory (/run/user/1000/pulse): Permission denied javax.sound.midi.MidiUnavailableException: Can not open line --- LANG=C sudo apt-cache policy snapd snapd: Installed: 2.0.10 Candidate: 2.0.10 Version table: *** 2.0.10 500 500 http://ru.archive.ubuntu.com/ubuntu xenial-updates/main amd64 Packages 100 /var/lib/dpkg/status 2.0.2 500 500 http://ru.archive.ubuntu.com/ubuntu xenial/main amd64 Packages --- $ lsb_release -a No LSB modules are available. Distributor ID: Ubuntu Description: Ubuntu 16.04 LTS Release: 16.04 Codename: xenial I won't test a snap package in a lubuntu any more. -- Best regards, vasilisc From james.jesudason at canonical.com Wed Jul 13 08:23:01 2016 From: james.jesudason at canonical.com (James Jesudason) Date: Wed, 13 Jul 2016 09:23:01 +0100 Subject: Including Postgresql in a Snap App In-Reply-To: References: Message-ID: On 13 July 2016 at 06:52, Steven Holms wrote: > Has anyone had success in setting up a postgresql server within a snap? > I've been reading over the docs, and figured I'd try adding postgresql as a > stage-package, but now I'm not sure how to get it up and running within my > snap. I think I'm on the right track, I just don't know how to configure > the "app" portion of it so that the daemon will run. > > My app is pretty simple. It's a go server that needs access to postgresql > for it's backend. I know I can communicate with an external postgresql > database, but was hoping to get it all into one package. > > Any help is appreciated. > > I've done something like this for a project that I'm working on, though the architecture is slightly different to what you are suggesting. I have a Go web service that uses the network and network-bind plugs and then I'm using the postgresql snap from the store. So, I use two separate snaps and then use a command on my web service snap to set the config file (so it knows how to connect to postgresql). After that, a simple restart of the web service snap and everything is connected. Go web service: https://github.com/ubuntu-core/identity-vault Web service snap files: https://github.com/ubuntu-core/serial-vault-snap (these files will soon be moved to the web service source repo) As a bonus, there is also a juju 2.0 charm that deploys the web service: https://github.com/ubuntu-core/serial-vault-charm Even if you are not going to use the charm for deployment, it is helpful as it walks you through the commands that you need to get config file set up and pushed to the snap. Kudos to Evan Dandrea as I just copied his approach for config from the Jenkins Snap [1]. With this approach, it works just as well if the snaps are on the same machine or on separate juju units. James [1] https://github.com/evandandrea/jenkins-snap -------------- next part -------------- An HTML attachment was scrubbed... URL: From mabnhdev at gmail.com Wed Jul 13 10:45:07 2016 From: mabnhdev at gmail.com (MikeB) Date: Wed, 13 Jul 2016 06:45:07 -0400 Subject: Snap daemonization ordering and manual interface connections Message-ID: I'm working on a snapping up OpenSwitch which is essentially a network operating system (NOS) for white box network switches. The OpenSwitch NOS is made up of many daemons (snapcraft apps) that control network hardware and networking protocols. Many of the apps have plugs for network-control, network-manager, and network-observe. These plugs are not automatically connected by snapd when the snap is installed - they must be manually connected using the 'snap connect' command. The problem I'm running into is that daemonization of most of the apps fails because the plugs are not connected at the time snapd starts up the daemons and everything falls apart. I understand why you want manual connections for some "critcal" interfaces. However, if you're going to require manual intervention, the snapd should be smart enough to wait for the required connections before starting up any daemons that have plugs for those "critical" interfaces. Perhaps even nagging about unconnected plugs in the snap. A somewhat related issue is that I really think snapd needs to support the concept of ordering daemonization much like systemd's before/after clauses - otherwise, you force developers to create awkward wrappers around daemons to create ad-hoc ordering. Regards, Mike -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Wed Jul 13 10:54:41 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Wed, 13 Jul 2016 12:54:41 +0200 Subject: resident package database update In-Reply-To: <20160713013152.20e9d138@utnubu> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> <20160712205134.GB24443@hunt> <20160713003002.732485ae@utnubu> <20160713010314.242158b8@utnubu> <578579E9.2040707@ubuntu.com> <20160713013152.20e9d138@utnubu> Message-ID: <1468407281.20308.81.camel@anubis> hi, Am Mittwoch, den 13.07.2016, 01:31 +0200 schrieb Ralf Mardorf: > On Tue, 12 Jul 2016 20:14:49 -0300, Sergio Schvezov wrote: > >So it feels this is about searching and installing and more about > >aggregating than sharing the same database? ... > > GUIs don't make life always easier, since command line tools could be > used in combination with other command line tools and handy bashisms, > so tailored installs could be easier realized by using command line. so there could be a console UI for the GUI app (or simply a tool that uses the appstream database that gnome-sw provides) ;) what sergio was saying is that it doesnt need the aggregated DB but a tool that can use both DBs since the focus is on search and install ... there is no need to mangle the deb or rpm DBs for this feature (you not only need to patch the db handling but also modify a lot inside dpkg and rpm to make that happen). ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From ogra at ubuntu.com Wed Jul 13 10:58:08 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Wed, 13 Jul 2016 12:58:08 +0200 Subject: resident package database update In-Reply-To: <20160713003002.732485ae@utnubu> References: <20160712144952.41b59d22@andriamanitra> <1468324971.20308.73.camel@anubis> <20160712153155.60927ca3@andriamanitra> <20160712205134.GB24443@hunt> <20160713003002.732485ae@utnubu> Message-ID: <1468407488.20308.83.camel@anubis> hi, Am Mittwoch, den 13.07.2016, 00:30 +0200 schrieb Ralf Mardorf: > "Snaps don’t intrinsically depend on the Ubuntu store, that’s just what > snapd does today, and we expect people will have different stores for > their snaps in future." - http://snapcraft.io/ > and for the record: ogra at styx:~$ sudo snap install snapstore-example 42.55 MB / 42.55 MB [=====================================================================================================================] 100.00 % 3.87 MB/s Name Version Rev Developer Notes snapstore-example 0.3 4 noise - ogra at styx:~$ (in case you ever want to run your own store :) ) ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From vasilisc777 at gmail.com Wed Jul 13 12:04:04 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Wed, 13 Jul 2016 15:04:04 +0300 Subject: is this bug? Message-ID: <57862E34.2000005@gmail.com> In snapcraft.yaml I request plugs: [network, network-bind, x11, home, unity7, pulseaudio, mount-observe, gsettings, locale-control] After upload TuxGuitar in Ubuntu Store I see $ snap list Name Version Rev Developer Notes tuxguitar-vs 1.3.2-snap1 1 vs - ubuntu-core 16.04+20160531.11-56 122 canonical - $ snap interfaces Slot Plug :gsettings tuxguitar-vs :home tuxguitar-vs :network tuxguitar-vs :network-bind tuxguitar-vs :pulseaudio tuxguitar-vs :unity7 tuxguitar-vs :x11 tuxguitar-vs - tuxguitar-vs:locale-control - tuxguitar-vs:mount-observe Not autoconnect locale-control mount-observe User must manually in CLI connect mount-observe for example? -- Best regards, vasilisc From zygmunt.krynicki at canonical.com Wed Jul 13 12:19:23 2016 From: zygmunt.krynicki at canonical.com (Zygmunt Krynicki) Date: Wed, 13 Jul 2016 14:19:23 +0200 Subject: is this bug? In-Reply-To: <57862E34.2000005@gmail.com> References: <57862E34.2000005@gmail.com> Message-ID: <951C96A3-4054-46EF-856D-988D951E507B@canonical.com> > Wiadomość napisana przez Vasilisc w dniu 13.07.2016, o godz. 14:04: > > In snapcraft.yaml I request > plugs: [network, network-bind, x11, home, unity7, pulseaudio, mount-observe, gsettings, locale-control] > > After upload TuxGuitar in Ubuntu Store I see > > $ snap list > Name Version Rev Developer Notes > tuxguitar-vs 1.3.2-snap1 1 vs - > ubuntu-core 16.04+20160531.11-56 122 canonical - > > > $ snap interfaces > Slot Plug > :gsettings tuxguitar-vs > :home tuxguitar-vs > :network tuxguitar-vs > :network-bind tuxguitar-vs > :pulseaudio tuxguitar-vs > :unity7 tuxguitar-vs > :x11 tuxguitar-vs > - tuxguitar-vs:locale-control > - tuxguitar-vs:mount-observe > > Not autoconnect > locale-control > mount-observe > > User must manually in CLI connect mount-observe for example? Yes, because those plugs are more sensitive and grant more power. Thanks ZK From sergio.schvezov at canonical.com Wed Jul 13 15:44:23 2016 From: sergio.schvezov at canonical.com (Sergio Schvezov) Date: Wed, 13 Jul 2016 12:44:23 -0300 Subject: ANN: snapcraft 2.12 is now available Message-ID: <578661D7.5050104@ubuntu.com> Hello snapcrafters! We are pleased to announce the release of version 2.12.1 of snapcraft: https://launchpad.net/snapcraft/+milestone/2.12.1 While this is a bugfix release it is worth of mention as there have been some awaited fixes rolled in among other goodies: - pkg-config issues have been sorted, there is now one common sysroot. - snapcraft register error scenarios have nicer messages with some end user help on the next steps that can be taken from such errors. - There are progress bars mostly everywhere now, so no more staring and waiting for things to happen. - cleanbuild-ing with cloud parts should be sorted. - A new store endpoint for tracking the status of pushed snaps is now used which provides feedback straight on the cli to know if it is all good or manual action needs to be taken (e.g.; in the case of needing manual review). - the python3 plugin now creates entry points if defined. - the go plugin should now work with in tree snapcraft.yaml - the qmake no does in source builds instead of out of tree ones. This release has seen many contributions from outside of the snapcraft core team, so we want to give a shout out to these folks, here's a team thank you for: - Simon Quigley - Colin Watson - Jonathan Cave To consume the latest snapcraft release on Xenial Xerus (16.04), we suggest you install the snapcraft package from the Ubuntu Archives: sudo apt update sudo apt install snapcraft To get the source for this release check it out at https://github.com/ubuntu-core/snapcraft/releases/tag/2.12.1 A great place to collaborate and discuss features, bugs and ideas on snapcraft is snapcraft at lists.ubuntu.com mailing list or directly in the #snappy channel on irc.freenode.net. To file bugs, please go to https://bugs.launchpad.net/snapcraft/+filebug. Happy snapcrafting, - Sergio and the team -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: OpenPGP digital signature URL: From didrocks at ubuntu.com Wed Jul 13 16:04:44 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Wed, 13 Jul 2016 09:04:44 -0700 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: References: Message-ID: <5786669C.1090307@ubuntu.com> Le 13/07/2016 03:45, MikeB a écrit : > I'm working on a snapping up OpenSwitch which is essentially a network > operating system (NOS) for white box network switches. > > The OpenSwitch NOS is made up of many daemons (snapcraft apps) that > control network hardware and networking protocols. > > Many of the apps have plugs for network-control, network-manager, and > network-observe. These plugs are not automatically connected by snapd > when the snap is installed - they must be manually connected using the > 'snap connect' command. > > The problem I'm running into is that daemonization of most of the apps > fails because the plugs are not connected at the time snapd starts up > the daemons and everything falls apart. > > I understand why you want manual connections for some "critcal" > interfaces. However, if you're going to require manual intervention, > the snapd should be smart enough to wait for the required connections > before starting up any daemons that have plugs for those "critical" > interfaces. Perhaps even nagging about unconnected plugs in the snap. > > A somewhat related issue is that I really think snapd needs to support > the concept of ordering daemonization much like systemd's before/after > clauses - otherwise, you force developers to create awkward wrappers > around daemons to create ad-hoc ordering. Hey Mike, I agree with both statements and proposal. We could have smarter systemd unit generated for those use cases. Do you mind opening them as 2 bugs on https://bugs.launchpad.net/snappy/+filebug? Cheers, Didier PS: I guess giving a way for reusing parts of a systemd units syntax would go a long way. -------------- next part -------------- An HTML attachment was scrubbed... URL: From kyle.fazzari at canonical.com Wed Jul 13 16:37:30 2016 From: kyle.fazzari at canonical.com (Kyle Fazzari) Date: Wed, 13 Jul 2016 12:37:30 -0400 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: References: Message-ID: Hey Mike, On Jul 13, 2016 3:46 AM, "MikeB" wrote: > I understand why you want manual connections for some "critcal" interfaces. However, if you're going to require manual intervention, the snapd should be smart enough to wait for the required connections before starting up any daemons that have plugs for those "critical" interfaces. Perhaps even nagging about unconnected plugs in the snap. I understand where you're coming from here. However, I can imagine use-cases where such a daemon can actually run successfully without one of its plugs (particularly when seccomp starts using ERRNO), and have extra functionality if it's connected. If snapd didn't start services until all their plugs were connected, such a scenario would be impossible. Also, keep in mind that the user can disconnect interfaces at any time, so ideally daemons would be able to handle such things (though I realize that's not always the case). Just my two cents. Kyle -------------- next part -------------- An HTML attachment was scrubbed... URL: From mabnhdev at gmail.com Wed Jul 13 17:54:08 2016 From: mabnhdev at gmail.com (MikeB) Date: Wed, 13 Jul 2016 13:54:08 -0400 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: References: Message-ID: ​Yes, I see your point. Perhaps a concept of optional vs required is needed. Mike​ On Wed, Jul 13, 2016 at 12:37 PM, Kyle Fazzari wrote: > Hey Mike, > > On Jul 13, 2016 3:46 AM, "MikeB" wrote: > > I understand why you want manual connections for some "critcal" > interfaces. However, if you're going to require manual intervention, the > snapd should be smart enough to wait for the required connections before > starting up any daemons that have plugs for those "critical" interfaces. > Perhaps even nagging about unconnected plugs in the snap. > > I understand where you're coming from here. However, I can imagine > use-cases where such a daemon can actually run successfully without one of > its plugs (particularly when seccomp starts using ERRNO), and have extra > functionality if it's connected. If snapd didn't start services until all > their plugs were connected, such a scenario would be impossible. Also, keep > in mind that the user can disconnect interfaces at any time, so ideally > daemons would be able to handle such things (though I realize that's not > always the case). > > Just my two cents. > > Kyle > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mark at ubuntu.com Wed Jul 13 18:25:59 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Wed, 13 Jul 2016 11:25:59 -0700 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: References: Message-ID: <578687B7.9040308@ubuntu.com> In the context of a snap running on Ubuntu Core, the all-snap version of Ubuntu, the gadget snap (which defines behaviors specific to that particular device) would describe additional auto-connection rules. So in your case, a switch that had certified your OpenSwitch NOS would specify that, when installed, OpenSwitch plugs and slots be connected appropriately. Think if it as that "the user OR the device manufacturer can take this decision" and the gadget snap is how the device manufacturer would do so. Would that be useful? Mark On 13/07/16 10:54, MikeB wrote: > ​Yes, I see your point. > > Perhaps a concept of optional vs required is needed. > > Mike​ > > On Wed, Jul 13, 2016 at 12:37 PM, Kyle Fazzari > > wrote: > > Hey Mike, > > On Jul 13, 2016 3:46 AM, "MikeB" > wrote: > > I understand why you want manual connections for some "critcal" > interfaces. However, if you're going to require manual > intervention, the snapd should be smart enough to wait for the > required connections before starting up any daemons that have > plugs for those "critical" interfaces. Perhaps even nagging about > unconnected plugs in the snap. > > I understand where you're coming from here. However, I can imagine > use-cases where such a daemon can actually run successfully > without one of its plugs (particularly when seccomp starts using > ERRNO), and have extra functionality if it's connected. If snapd > didn't start services until all their plugs were connected, such a > scenario would be impossible. Also, keep in mind that the user can > disconnect interfaces at any time, so ideally daemons would be > able to handle such things (though I realize that's not always the > case). > > Just my two cents. > > Kyle > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mabnhdev at gmail.com Wed Jul 13 18:44:16 2016 From: mabnhdev at gmail.com (MikeB) Date: Wed, 13 Jul 2016 14:44:16 -0400 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: <578687B7.9040308@ubuntu.com> References: <578687B7.9040308@ubuntu.com> Message-ID: It it very useful for deploying on a target in the future. Not very helpful to developing/testing/debugging on system without the gadget snap today. Perhaps, if a snap is installed in --devmode, then all interfaces are auto-connected. I would think that if a snap can be installed in --devmode on a system, then you are already opening up the system to all sorts of "dangers". Mike On Wed, Jul 13, 2016 at 2:25 PM, Mark Shuttleworth wrote: > > In the context of a snap running on Ubuntu Core, the all-snap version of > Ubuntu, the gadget snap (which defines behaviors specific to that > particular device) would describe additional auto-connection rules. > > So in your case, a switch that had certified your OpenSwitch NOS would > specify that, when installed, OpenSwitch plugs and slots be connected > appropriately. > > Think if it as that "the user OR the device manufacturer can take this > decision" and the gadget snap is how the device manufacturer would do so. > > Would that be useful? > > Mark > > > On 13/07/16 10:54, MikeB wrote: > > ​Yes, I see your point. > > Perhaps a concept of optional vs required is needed. > > Mike​ > > On Wed, Jul 13, 2016 at 12:37 PM, Kyle Fazzari > wrote: > >> Hey Mike, >> >> On Jul 13, 2016 3:46 AM, "MikeB" < mabnhdev at gmail.com> >> wrote: >> > I understand why you want manual connections for some "critcal" >> interfaces. However, if you're going to require manual intervention, the >> snapd should be smart enough to wait for the required connections before >> starting up any daemons that have plugs for those "critical" interfaces. >> Perhaps even nagging about unconnected plugs in the snap. >> >> I understand where you're coming from here. However, I can imagine >> use-cases where such a daemon can actually run successfully without one of >> its plugs (particularly when seccomp starts using ERRNO), and have extra >> functionality if it's connected. If snapd didn't start services until all >> their plugs were connected, such a scenario would be impossible. Also, keep >> in mind that the user can disconnect interfaces at any time, so ideally >> daemons would be able to handle such things (though I realize that's not >> always the case). >> >> Just my two cents. >> >> Kyle >> > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From luke.yelavich at canonical.com Thu Jul 14 03:06:07 2016 From: luke.yelavich at canonical.com (Luke Yelavich) Date: Thu, 14 Jul 2016 13:06:07 +1000 Subject: The optical-drive interface, and read/write. Message-ID: <20160714030607.GA19183@buffalo> Hey folks. I am working on a cdparanoia snap. CDParanoia requires read/write access to the optical drive, From what I have gathered from the cdparanoia code, read/write access is required for detection of the drive/kernel interface, i.e SG IO, etc, and extraction/reading can be done read-only. I am wondering whether it is possible to add read/write support to the optical-drive interface for a selection of ioctls, I think only one at this point... If so, could someone point me to a document that explains the syntax for aparmor etc to implement this? I'll be of course happy to propose a merge to snapd for the optical-drive interface once I have identified the ioctls required by cdparanoia. I think this is worth doing because lots of software uses cdparanoia for reading/extraction, both the cdparanoia library and the command-line utility, gstreamer included. Of course, if folks feel that another interface is a better solution, I'm up for that as well. Thanks in advance. Luke From mark at ubuntu.com Thu Jul 14 10:04:45 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Thu, 14 Jul 2016 03:04:45 -0700 Subject: The optical-drive interface, and read/write. In-Reply-To: <20160714030607.GA19183@buffalo> References: <20160714030607.GA19183@buffalo> Message-ID: <578763BD.5090502@ubuntu.com> Hi Luke I think it's https://en.wikipedia.org/wiki/Seccomp seccomp-bpf that you're looking for, and snapd can configure that just as it does apparmor, based on the plug/slot interfaces declared by the snaps. Jamie will know more. Mark On 13/07/16 20:06, Luke Yelavich wrote: > Hey folks. > I am working on a cdparanoia snap. CDParanoia requires read/write access to the optical drive, From what I have gathered from the cdparanoia code, read/write access is required for detection of the drive/kernel interface, i.e SG IO, etc, and extraction/reading can be done read-only. > > I am wondering whether it is possible to add read/write support to the optical-drive interface for a selection of ioctls, I think only one at this point... If so, could someone point me to a document that explains the syntax for aparmor etc to implement this? I'll be of course happy to propose a merge to snapd for the optical-drive interface once I have identified the ioctls required by cdparanoia. > > I think this is worth doing because lots of software uses cdparanoia for reading/extraction, both the cdparanoia library and the command-line utility, gstreamer included. > > Of course, if folks feel that another interface is a better solution, I'm up for that as well. > > Thanks in advance. > > Luke > From jamie at canonical.com Thu Jul 14 14:41:59 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Thu, 14 Jul 2016 09:41:59 -0500 Subject: The optical-drive interface, and read/write. In-Reply-To: <578763BD.5090502@ubuntu.com> References: <20160714030607.GA19183@buffalo> <578763BD.5090502@ubuntu.com> Message-ID: <1468507319.6103.11.camel@canonical.com> On Thu, 2016-07-14 at 03:04 -0700, Mark Shuttleworth wrote: > Hi Luke > > I think it's https://en.wikipedia.org/wiki/Seccomp seccomp-bpf that > you're looking for, and snapd can configure that just as it does > apparmor, based on the plug/slot interfaces declared by the snaps. Jamie > will know more. > Yes, seccomp via seccomp arg filtering is where we should be able to do this. That feature landed in upstream snap-confine a little while ago and AIUI the snappy team is working on landing that in 16.04. Luke, please file a bug and add the 'snapd-interface' tag with what you need and a member of the interfaces team can take a look. Thanks! > Mark > > On 13/07/16 20:06, Luke Yelavich wrote: > > > > Hey folks. > > I am working on a cdparanoia snap. CDParanoia requires read/write access to > > the optical drive, From what I have gathered from the cdparanoia code, > > read/write access is required for detection of the drive/kernel interface, > > i.e SG IO, etc, and extraction/reading can be done read-only. > > > > I am wondering whether it is possible to add read/write support to the > > optical-drive interface for a selection of ioctls, I think only one at this > > point... If so, could someone point me to a document that explains the > > syntax for aparmor etc to implement this? I'll be of course happy to propose > > a merge to snapd for the optical-drive interface once I have identified the > > ioctls required by cdparanoia. > > > > I think this is worth doing because lots of software uses cdparanoia for > > reading/extraction, both the cdparanoia library and the command-line > > utility, gstreamer included. > > > > Of course, if folks feel that another interface is a better solution, I'm up > > for that as well. > > > > Thanks in advance. > > > > Luke > > > > -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From evan.dandrea at canonical.com Thu Jul 14 15:27:12 2016 From: evan.dandrea at canonical.com (Evan Dandrea) Date: Thu, 14 Jul 2016 15:27:12 +0000 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: <578687B7.9040308@ubuntu.com> References: <578687B7.9040308@ubuntu.com> Message-ID: On Wed, 13 Jul 2016 at 13:27 Mark Shuttleworth wrote: > > In the context of a snap running on Ubuntu Core, the all-snap version of > Ubuntu, the gadget snap (which defines behaviors specific to that > particular device) would describe additional auto-connection rules. > The non-Ubuntu Core case is interesting. The user must take action to connect the interface, but currently receives no suggestion to do so. As well, the snap cannot signal that it shouldn't be started until its interfaces are connected (in the case of a daemon). I believe this is one of the things hooks set out to resolve: https://bugs.launchpad.net/ubuntu/+source/snapd/+bug/1586465 -------------- next part -------------- An HTML attachment was scrubbed... URL: From gustavo.niemeyer at canonical.com Thu Jul 14 16:45:58 2016 From: gustavo.niemeyer at canonical.com (Gustavo Niemeyer) Date: Thu, 14 Jul 2016 13:45:58 -0300 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: References: <578687B7.9040308@ubuntu.com> Message-ID: This is indeed lacking, and we have a good plan on how to support it. Coming soon. We have some roadmap discussions next week, so let's see where this sits. On Jul 14, 2016 12:29 PM, "Evan Dandrea" wrote: > On Wed, 13 Jul 2016 at 13:27 Mark Shuttleworth wrote: > >> >> In the context of a snap running on Ubuntu Core, the all-snap version of >> Ubuntu, the gadget snap (which defines behaviors specific to that >> particular device) would describe additional auto-connection rules. >> > > The non-Ubuntu Core case is interesting. The user must take action to > connect the interface, but currently receives no suggestion to do so. As > well, the snap cannot signal that it shouldn't be started until its > interfaces are connected (in the case of a daemon). > > I believe this is one of the things hooks set out to resolve: > https://bugs.launchpad.net/ubuntu/+source/snapd/+bug/1586465 > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mark at ubuntu.com Thu Jul 14 18:25:17 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Thu, 14 Jul 2016 11:25:17 -0700 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: References: <578687B7.9040308@ubuntu.com> Message-ID: <5787D90D.3070807@ubuntu.com> On 14/07/16 09:45, Gustavo Niemeyer wrote: > > This is indeed lacking, and we have a good plan on how to support it. > > Coming soon. We have some roadmap discussions next week, so let's see > where this sits. > > Great topic for the sprint in Heidelberg indeed! We could easily prompt the user appropriately, and design for CLI and GUI experiences shaped by the snap publisher. Mark From gustavo.niemeyer at canonical.com Thu Jul 14 18:27:28 2016 From: gustavo.niemeyer at canonical.com (Gustavo Niemeyer) Date: Thu, 14 Jul 2016 15:27:28 -0300 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: <5787D90D.3070807@ubuntu.com> References: <578687B7.9040308@ubuntu.com> <5787D90D.3070807@ubuntu.com> Message-ID: Yeah, and it's also easy to have a field that defines readiness based on an expression, which means we can then evaluate the expression and tell the user what's lacking. You may recall that conversation from the $PLACE sprint.. :) On Thu, Jul 14, 2016 at 3:25 PM, Mark Shuttleworth wrote: > On 14/07/16 09:45, Gustavo Niemeyer wrote: > > > > This is indeed lacking, and we have a good plan on how to support it. > > > > Coming soon. We have some roadmap discussions next week, so let's see > > where this sits. > > > > > > Great topic for the sprint in Heidelberg indeed! > > We could easily prompt the user appropriately, and design for CLI and > GUI experiences shaped by the snap publisher. > > Mark > -- gustavo @ http://niemeyer.net -------------- next part -------------- An HTML attachment was scrubbed... URL: From mark at ubuntu.com Thu Jul 14 18:40:01 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Thu, 14 Jul 2016 11:40:01 -0700 Subject: Snap daemonization ordering and manual interface connections In-Reply-To: References: <578687B7.9040308@ubuntu.com> <5787D90D.3070807@ubuntu.com> Message-ID: <5787DC81.9070106@ubuntu.com> On 14/07/16 11:27, Gustavo Niemeyer wrote: > > You may recall that conversation from the $PLACE sprint.. :) > It's nice to see those conversations bearing fruit :) From luke.yelavich at canonical.com Thu Jul 14 22:01:36 2016 From: luke.yelavich at canonical.com (Luke Yelavich) Date: Fri, 15 Jul 2016 08:01:36 +1000 Subject: The optical-drive interface, and read/write. In-Reply-To: <1468507319.6103.11.camel@canonical.com> References: <20160714030607.GA19183@buffalo> <578763BD.5090502@ubuntu.com> <1468507319.6103.11.camel@canonical.com> Message-ID: <20160714220136.GA6103@buffalo> On Fri, Jul 15, 2016 at 12:41:59AM AEST, Jamie Strandboge wrote: > Yes, seccomp via seccomp arg filtering is where we should be able to do this. > That feature landed in upstream snap-confine a little while ago and AIUI the > snappy team is working on landing that in 16.04. > > Luke, please file a bug and add the 'snapd-interface' tag with what you need and > a member of the interfaces team can take a look. Will do, is this against snapd in LP, or somewhere else? Luke From gustavo.niemeyer at canonical.com Fri Jul 15 03:31:00 2016 From: gustavo.niemeyer at canonical.com (Gustavo Niemeyer) Date: Fri, 15 Jul 2016 00:31:00 -0300 Subject: The optical-drive interface, and read/write. In-Reply-To: <20160714220136.GA6103@buffalo> References: <20160714030607.GA19183@buffalo> <578763BD.5090502@ubuntu.com> <1468507319.6103.11.camel@canonical.com> <20160714220136.GA6103@buffalo> Message-ID: Hey Luke, That's at https://launchpad.net/snappy please. On Thu, Jul 14, 2016 at 7:01 PM, Luke Yelavich wrote: > On Fri, Jul 15, 2016 at 12:41:59AM AEST, Jamie Strandboge wrote: > > Yes, seccomp via seccomp arg filtering is where we should be able to do > this. > > That feature landed in upstream snap-confine a little while ago and AIUI > the > > snappy team is working on landing that in 16.04. > > > > Luke, please file a bug and add the 'snapd-interface' tag with what you > need and > > a member of the interfaces team can take a look. > > Will do, is this against snapd in LP, or somewhere else? > > Luke > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -- gustavo @ http://niemeyer.net -------------- next part -------------- An HTML attachment was scrubbed... URL: From luke.yelavich at canonical.com Fri Jul 15 03:58:55 2016 From: luke.yelavich at canonical.com (Luke Yelavich) Date: Fri, 15 Jul 2016 13:58:55 +1000 Subject: The optical-drive interface, and read/write. In-Reply-To: References: <20160714030607.GA19183@buffalo> <578763BD.5090502@ubuntu.com> <1468507319.6103.11.camel@canonical.com> <20160714220136.GA6103@buffalo> Message-ID: <20160715035855.GA24738@buffalo> On Fri, Jul 15, 2016 at 01:31:00PM AEST, Gustavo Niemeyer wrote: > Hey Luke, > > That's at https://launchpad.net/snappy please. Thanks. Luke From vasilisc777 at gmail.com Fri Jul 15 08:26:45 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Fri, 15 Jul 2016 11:26:45 +0300 Subject: java + gtk3 Message-ID: <57889E45.6000101@gmail.com> My snapcraft.yaml java-files: plugin: jdk source: . project-files: plugin: copy files: run.sh: usr/bin/run.sh opt/* : /opt/ after: [desktop/gtk3] Raise error Parts 'desktop/gtk3' and 'java-files' have the following file paths in common which have different contents: usr/share/pkgconfig/shared-mime-info.pc ------------------- $ apt-cache policy snapd snapd: Installed: 2.0.10 Candidate: 2.0.10 Version table: *** 2.0.10 500 500 http://fi.archive.ubuntu.com/ubuntu xenial-updates/main amd64 Packages 100 /var/lib/dpkg/status 2.0.2 500 500 http://fi.archive.ubuntu.com/ubuntu xenial/main amd64 Packages I use apt-cacher-ng for saving time. -- Best regards, vasilisc From ogra at ubuntu.com Fri Jul 15 08:33:49 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 15 Jul 2016 10:33:49 +0200 Subject: java + gtk3 In-Reply-To: <57889E45.6000101@gmail.com> References: <57889E45.6000101@gmail.com> Message-ID: <1468571629.20308.98.camel@anubis> hi, Am Freitag, den 15.07.2016, 11:26 +0300 schrieb Vasilisc: > Raise error > > Parts 'desktop/gtk3' and 'java-files' have the following file paths in > common which have different contents: thats sounds like: https://bugs.launchpad.net/bugs/1600238 ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From cjwatson at ubuntu.com Fri Jul 15 15:45:00 2016 From: cjwatson at ubuntu.com (Colin Watson) Date: Fri, 15 Jul 2016 16:45:00 +0100 Subject: Future of Launchpad and the Snap Store In-Reply-To: <20160705120642.GB30483@riva.ucam.org> References: <577B6A35.8040308@gmail.com> <20160705120642.GB30483@riva.ucam.org> Message-ID: <20160715154500.GA660@riva.ucam.org> On Tue, Jul 05, 2016 at 01:06:43PM +0100, Colin Watson wrote: > * More obvious configuration for building snap packages for > architectures other than the defaults of amd64 and i386 This is now in place on the "Create a new snap package" page. > * Configure Launchpad to automatically build a snap package any time > the branch containing its snapcraft.yaml changes This is now in place. You'll find the necessary options on the "Create a new snap package" page, or you can edit an existing one. -- Colin Watson [cjwatson at ubuntu.com] From manik at canonical.com Fri Jul 15 19:32:38 2016 From: manik at canonical.com (Manik Taneja) Date: Fri, 15 Jul 2016 12:32:38 -0700 Subject: Future of Launchpad and the Snap Store In-Reply-To: <20160715154500.GA660@riva.ucam.org> References: <577B6A35.8040308@gmail.com> <20160705120642.GB30483@riva.ucam.org> <20160715154500.GA660@riva.ucam.org> Message-ID: +ogra On Fri, Jul 15, 2016 at 8:45 AM, Colin Watson wrote: > On Tue, Jul 05, 2016 at 01:06:43PM +0100, Colin Watson wrote: > > * More obvious configuration for building snap packages for > > architectures other than the defaults of amd64 and i386 > > This is now in place on the "Create a new snap package" page. > do we support armhf and aarch64 as well now? > > > * Configure Launchpad to automatically build a snap package any time > > the branch containing its snapcraft.yaml changes > > This is now in place. You'll find the necessary options on the "Create > a new snap package" page, or you can edit an existing one. > thanks for the detailed write up on all the capabilities we have enabled so far. > > -- > Colin Watson [cjwatson at ubuntu.com] > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From manik at canonical.com Fri Jul 15 19:42:19 2016 From: manik at canonical.com (Manik Taneja) Date: Fri, 15 Jul 2016 12:42:19 -0700 Subject: Future of Launchpad and the Snap Store In-Reply-To: <577B6A35.8040308@gmail.com> References: <577B6A35.8040308@gmail.com> Message-ID: On Tue, Jul 5, 2016 at 1:05 AM, Andrew Keech wrote: > hey everyone, so wonderful to see the incredible progress all the time. > welcome to snapcraft! > > I've been whittling away at a snap for `gimp-edge` for a while, but feel > it's been overdue to bring in the maintainer of the ppa where I currently > get my builds from. Having been in touch with Thorsten now, he's got some > unanswered questions about the possible future relationships or > integrations between Launchpad and the Snap Store. Right now he wants to > start a new ppa just for delivering the snap as a separate project from the > traditional gimp-edge deb. > > Are there any best practice ideas that exist? Is there some kind of plan > for how the Store and Launchpad might work together, or if they'd exist > separately? > > colin has already shared an excellent write-up on hosting/building with launchpad and publishing/promotion/distribution with the store. can we help you with landing the pkg in the store and making it available for general consumption? david, can you please work with andrew and help connect the dots between andrew/thorsten and the owner of the gimp snap work that's underway.. -------------- next part -------------- An HTML attachment was scrubbed... URL: From cjwatson at ubuntu.com Fri Jul 15 20:27:36 2016 From: cjwatson at ubuntu.com (Colin Watson) Date: Fri, 15 Jul 2016 21:27:36 +0100 Subject: Future of Launchpad and the Snap Store In-Reply-To: References: <577B6A35.8040308@gmail.com> <20160705120642.GB30483@riva.ucam.org> <20160715154500.GA660@riva.ucam.org> Message-ID: <20160715202736.GF16099@riva.ucam.org> On Fri, Jul 15, 2016 at 12:32:38PM -0700, Manik Taneja wrote: > On Fri, Jul 15, 2016 at 8:45 AM, Colin Watson <[1]cjwatson at ubuntu.com> > wrote: > > On Tue, Jul 05, 2016 at 01:06:43PM +0100, Colin Watson wrote: > >  * More obvious configuration for building snap packages for > >    architectures other than the defaults of amd64 and i386 > > This is now in place on the "Create a new snap package" page. > > do we support armhf and aarch64 as well now? We have done for several months, but it's now a bit more obvious that we do. -- Colin Watson [cjwatson at ubuntu.com] From dale_f_beaudoin at hotmail.com Sat Jul 16 23:15:59 2016 From: dale_f_beaudoin at hotmail.com (Dale Beaudoin) Date: Sat, 16 Jul 2016 23:15:59 +0000 Subject: contact info for 3rd party snaps Message-ID: Hi All, A member of ubuntuforums asks: "Where, if anywhere, is the contact info for 3rd party snaps? I don't see it anywhere after installing a snap. " Thanks and regards.. Ubuntu Development Version Testing -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.planella at canonical.com Sun Jul 17 12:41:56 2016 From: david.planella at canonical.com (David Planella) Date: Sun, 17 Jul 2016 14:41:56 +0200 Subject: Future of Launchpad and the Snap Store In-Reply-To: References: <577B6A35.8040308@gmail.com> Message-ID: On Fri, Jul 15, 2016 at 9:42 PM, Manik Taneja wrote: > > > On Tue, Jul 5, 2016 at 1:05 AM, Andrew Keech > wrote: > >> hey everyone, so wonderful to see the incredible progress all the time. >> > welcome to snapcraft! > >> >> I've been whittling away at a snap for `gimp-edge` for a while, but feel >> it's been overdue to bring in the maintainer of the ppa where I currently >> get my builds from. Having been in touch with Thorsten now, he's got some >> unanswered questions about the possible future relationships or >> integrations between Launchpad and the Snap Store. Right now he wants to >> start a new ppa just for delivering the snap as a separate project from the >> traditional gimp-edge deb. >> >> Are there any best practice ideas that exist? Is there some kind of plan >> for how the Store and Launchpad might work together, or if they'd exist >> separately? >> >> colin has already shared an excellent write-up on hosting/building with > launchpad and publishing/promotion/distribution with the store. can we help > you with landing the pkg in the store and making it available for general > consumption? > > david, can you please work with andrew and help connect the dots between > andrew/thorsten and the owner of the gimp snap work that's underway.. > Hi Manik, Andrew himself has been working on the GIMP snap as part of the snappy playpen: https://github.com/wandrewkeech/snappy-playpen/tree/gimp-git Thanks to his feedback one of the main blockers in building GIMP (and other snaps) [1] was fixed as part of the recent Snapcraft 2.12.1 release. After that, a new build uncovered another issue, described here [2]. I added a GIMP snap hacking session for the Heidelberg sprint next week, and I'd encourage everyone who can help making the GIMP snap happen (either remotely or at the event) to try Andrew's snap code and contribute with feedback or fixes to complete the build. Thanks! Cheers, David. [1] https://bugs.launchpad.net/snapcraft/+bug/1595243 [2] https://github.com/ubuntu/snappy-playpen/issues/174#issuecomment-232703467 -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.calle at canonical.com Mon Jul 18 12:21:24 2016 From: david.calle at canonical.com (=?UTF-8?Q?David_Call=c3=a9?=) Date: Mon, 18 Jul 2016 14:21:24 +0200 Subject: Snap! Community weekly update: W28 Message-ID: <578CC9C4.1020607@canonical.com> Here is an update on the highlights in the snap community for the past week. This is a special one: starting today, a lot of snappy developers, 3rd party app upstreams, distros leaders and contributors are together for a week-long sprint in Heidelberg, Germany! ## Heidelberg sprint, day 1 Here is an outline of all discussions planned for this first day in sunny Heidelberg: Snaps: * Content interfaces: allowing snaps to share content and libraries * Snaps update control * Defining stores * Connecting interfaces in a GUI * GNOME runtime snap: a snap exposing GNOME services to other snaps * Snapweb (formerly known as WebDM) design review Security: * SELinux policy module for snaps * Improving the cross-distro security story Core: * ubuntu-image future and planning * Boot variables changes Snapcraft: * Builds in LXD by default for Snapcraft * Extending Snapcraft plugins * Desktop launcher parts review and future * Snapcraft store integration UX review General: * Docs publishing and templates * Introduction to Snappy terminology The rest of the week is also packed with more discussions and hacking sessions ! You can find the sprint agenda on this Trello board . ## Playpen You are starting to expect this, but we have a lot of new great examples in the Playpen! Thanks to this amazing community for allowing us to build expertise together, the repo now contains 44 examples, including the new: * ResidualVM, by LyzardKing * TexWorks, by LyzardKing * jTileDownloader, by dplanella * MirageOS, by elopio * Docker Compose, by elopio * QOwnNotes, by pbek * click-parser, by bhdouglass * 2048, by dholbach * The list of examples in the README has also been updated with tags, to let you see at a glance which technologies/plugins/snapcraft features are used by each example. * The snaps upstreaming initiative is going strong as well, and more info has been added to the getting started upstreaming page. ## Community highlights * The Open Build Service by OpenSuse is getting Snapcraft support . * Titania , an X3D editor and browser, is getting snapped upstream . * QDriverStation , a driving station for competition robots is getting snapped upstream . * The School of Engineering and Architecture of Fribourg has a prominent telecom tower on top of its main building, covered with leds, and has created a piece of software to display messages on it. Unfortunately, I haven't found a good picture of it running, but there is a snap for it ! ## Reading list * Snapd 2.0.10: new media interfaces, channel switching * Daniel Holbach is giving you a lot of ways to contribute to snaps if you are not a coder * Didier Roche introduces you to the new desktop launchers parts for Snapcraft * Zygmunt Krynicki is giving cross-distro updates: snapd on Arch has moved to the main repo and on Fedora, you can use a COPR to get snapd 2.0.10 * Kevin Gunn is giving you an introduction to snapping mir apps on Unity8 classic This is a list of news we are compiling weekly, but make sure you subscribe to the following social channels for daily updates:twitter ,google+ ,facebook . Cheers, David -------------- next part -------------- An HTML attachment was scrubbed... URL: From daniel.holbach at canonical.com Mon Jul 18 15:57:01 2016 From: daniel.holbach at canonical.com (Daniel Holbach) Date: Mon, 18 Jul 2016 17:57:01 +0200 Subject: Snappy Playpen event tomorrow! Message-ID: <578CFC4D.4060302@canonical.com> Hello everybody, if you want to get started snapping, tomorrow is a good time. Here's the announcement of the event tomorrow: https://daniel.holba.ch/blog/2016/07/snappy-playpen-event-tomorrow-2/ See you there! :-) Have a great day, Daniel From woodrow.shen at canonical.com Tue Jul 19 04:30:59 2016 From: woodrow.shen at canonical.com (Woodrow Shen) Date: Tue, 19 Jul 2016 12:30:59 +0800 Subject: ubuntu-core snap update status for armhf Message-ID: Hi all, I have a project to port ubuntu-core 16 to roseapple-pi ( https://github.com/xapp-le/SnappyUbuntuCore), and right now I check `snap find` to see all available snaps from store, but ubuntu-core snap is still 120 rev. (16.04+20160531.12-01). I wonder if we can make its rev. for armhf synchronize with amd64 ? It's great to get the latest version for development. Enjoy snap, Woodrow -- Snapcraft mailing list Snapcraft at lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft -------------- next part -------------- An HTML attachment was scrubbed... URL: From patrizio at bekerle.com Tue Jul 19 04:39:49 2016 From: patrizio at bekerle.com (Patrizio Bekerle) Date: Tue, 19 Jul 2016 06:39:49 +0200 Subject: Qt theming troubles under KDE Message-ID: Hi everyone! I’m having troubles with getting my native Qt theme with the `qownnotes` snap under KDE (I’m using kubuntu with KDE Neon). QOwnNotes doesn’t use the native KDE theme, like colors and widget styles. That’s the snapcraft I’m using: https://git.launchpad.net/~pbek/qownnotes-snap/tree/ Does anyone have the same problems? It seems to work under Unity, like Daniel H. tested here: https://github.com/ubuntu/snappy-playpen/pull/181 Cheers, Patrizio -------------- next part -------------- An HTML attachment was scrubbed... URL: From didrocks at ubuntu.com Tue Jul 19 07:15:33 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Tue, 19 Jul 2016 09:15:33 +0200 Subject: ubuntu-core snap update status for armhf In-Reply-To: References: Message-ID: <578DD395.8060402@ubuntu.com> Le 19/07/2016 06:30, Woodrow Shen a écrit : > Hi all, > > I have a project to port ubuntu-core 16 to roseapple-pi > (https://github.com/xapp-le/SnappyUbuntuCore), and right now I check > `snap find` to see all available snaps from store, but ubuntu-core > snap is still 120 rev. (16.04+20160531.12-01). I wonder if we can make > its rev. for armhf synchronize with amd64 ? It's great to get > the latest version for development. Hey Woodrow, (Note that it's the version "16.04+20160531.12-01" which is important, amd64 and armhf will always have different revisions as "120" is just a number autoincremented with each snap upload.) I know that Michael wanted to do a new ubuntu-core release sync in the stable channel, Michael, do you have any idea for when that's going to happen? Cheers, Didier From didrocks at ubuntu.com Tue Jul 19 07:17:44 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Tue, 19 Jul 2016 09:17:44 +0200 Subject: Qt theming troubles under KDE In-Reply-To: References: Message-ID: <578DD418.3020703@ubuntu.com> Le 19/07/2016 06:39, Patrizio Bekerle a écrit : > Hi everyone! Hey Patrizio! > > I’m having troubles with getting my native Qt theme with the > `qownnotes` snap under KDE (I’m using kubuntu with KDE Neon). > QOwnNotes doesn’t use the native KDE theme, like colors and widget styles. > That’s the snapcraft I’m > using: https://git.launchpad.net/~pbek/qownnotes-snap/tree/ > > > Does anyone have the same problems? Yeah, we are currently using some workarounds in the various desktop launchers for the lack of a real "theme" interface. One way meawhile to circumvent that is to ship the KDE Neon theme for now with your snap (in the stage-packages list for example) and that should do it with your snap until we have this interface. I hope that answers your question. Cheers, Didier -------------- next part -------------- An HTML attachment was scrubbed... URL: From woodrow.shen at canonical.com Tue Jul 19 08:40:29 2016 From: woodrow.shen at canonical.com (Woodrow Shen) Date: Tue, 19 Jul 2016 16:40:29 +0800 Subject: ubuntu-core snap update status for armhf In-Reply-To: <578DD395.8060402@ubuntu.com> References: <578DD395.8060402@ubuntu.com> Message-ID: Hi Didier, Thanks your information, Yes, the version is the major factor what I think, and current ver. is 16.04+20160714.10-11 for amd64, and afaik lots of interfaces are also applied. However, ubuntu-core edge channel doesn't keep up with amd64, so I wish armhf can be promoted as well. Thanks, Woodrow On Tue, Jul 19, 2016 at 3:15 PM, Didier Roche wrote: > Le 19/07/2016 06:30, Woodrow Shen a écrit : > > Hi all, > > > > I have a project to port ubuntu-core 16 to roseapple-pi > > (https://github.com/xapp-le/SnappyUbuntuCore), and right now I check > > `snap find` to see all available snaps from store, but ubuntu-core > > snap is still 120 rev. (16.04+20160531.12-01). I wonder if we can make > > its rev. for armhf synchronize with amd64 ? It's great to get > > the latest version for development. > > Hey Woodrow, > > (Note that it's the version "16.04+20160531.12-01" which is important, > amd64 and armhf will always have different revisions as "120" is just a > number autoincremented with each snap upload.) > > I know that Michael wanted to do a new ubuntu-core release sync in the > stable channel, Michael, do you have any idea for when that's going to > happen? > > Cheers, > Didier > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -- Woodrow Shen Software Engineer, Canonical ltd. UES | CE | PC & Core, Taipei -------------- next part -------------- An HTML attachment was scrubbed... URL: From vasilisc777 at gmail.com Tue Jul 19 10:30:03 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Tue, 19 Jul 2016 13:30:03 +0300 Subject: locale Message-ID: <578E012B.30901@gmail.com> The java-app in snap package can't open file if path contain non-English letters. How it is correct to work with a locale? snapcraft.yaml integration: plugin: nil stage-packages: - ttf-wqy-zenhei #china - fonts-kacst #arabic - locales - locales-all .... I try this java -Dsun.jnu.encoding="UTF-8" -Dfile.encoding="UTF-8" .. but it not work. -- Best regards, vasilisc From vasilisc777 at gmail.com Tue Jul 19 13:27:15 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Tue, 19 Jul 2016 16:27:15 +0300 Subject: fix locale issue Message-ID: <578E2AB3.8000506@gmail.com> I found this solution --- export I18NPATH=$SNAP/usr/share/i18n export LOCPATH=$SNAP_USER_DATA export LC_ALL=$LANG LANG1=$(echo $LANG | cut -f1 -d.) ENC=UTF-8 LOC="$LANG" # generate a locale so we get properly working charsets and graphics if [ ! -e $SNAP_USER_DATA/$LOC ]; then $SNAP/usr/bin/localedef --prefix=$SNAP_USER_DATA -f $ENC -i $LANG1 $SNAP_USER_DATA/$LOC fi -- Best regards, vasilisc From cjwatson at ubuntu.com Tue Jul 19 13:40:30 2016 From: cjwatson at ubuntu.com (Colin Watson) Date: Tue, 19 Jul 2016 14:40:30 +0100 Subject: Future of Launchpad and the Snap Store In-Reply-To: <20160705120642.GB30483@riva.ucam.org> References: <577B6A35.8040308@gmail.com> <20160705120642.GB30483@riva.ucam.org> Message-ID: <20160719134016.GA23570@riva.ucam.org> On Tue, Jul 05, 2016 at 01:06:43PM +0100, Colin Watson wrote: > * Configure Launchpad to automatically tell the store to publish a snap > package to a given set of channels after pushing completed builds This is now in place too. You'll find the necessary options on the "Create a new snap package" page, or you can edit an existing one. (The presentation on the snap's index page is a little ugly, but https://code.launchpad.net/~cjwatson/launchpad/snap-channel-nicer-display/+merge/300366 will fix that.) -- Colin Watson [cjwatson at ubuntu.com] From alejandro.vera at gmail.com Tue Jul 19 23:17:36 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Tue, 19 Jul 2016 20:17:36 -0300 Subject: Liferea Message-ID: Hi guys, first time in this list I am trying to snap liferea, to learn and to use it. I think I am almost ready, but when I run my snap I have an error: (process:4122): Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale. Gtk-Message: Failed to load module "canberra-gtk-module" Gtk-Message: Failed to load module "canberra-gtk-module" ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 (unable to open database file) ** ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit (db)) I think that the error is that it is trying to create the database inside the snap but it is read-only. Is there a way I can use the real HOME as HOME so liferea can create the database? Or is there a common way to solve this kind of problem? I am a spanish speaker, sorry for my english -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From chris.wayne at canonical.com Wed Jul 20 01:04:46 2016 From: chris.wayne at canonical.com (Chris Wayne) Date: Tue, 19 Jul 2016 21:04:46 -0400 Subject: Liferea In-Reply-To: References: Message-ID: Hi Alejandro, Try using the home interface (described here: http://snapcraft.io/docs/reference/interfaces) More info on how to use interdaces is found here: http://snapcraft.io/create/ Thanks Chris On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera wrote: > Hi guys, first time in this list > > I am trying to snap liferea, to learn and to use it. > > I think I am almost ready, but when I run my snap I have an error: > > (process:4122): Gtk-WARNING **: Locale not supported by C library. > Using the fallback 'C' locale. > Gtk-Message: Failed to load module "canberra-gtk-module" > Gtk-Message: Failed to load module "canberra-gtk-module" > > ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 > (unable to open database file) > ** > ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit (db)) > > > I think that the error is that it is trying to create the database inside > the snap but it is read-only. > > Is there a way I can use the real HOME as HOME so liferea can create the > database? Or is there a common way to solve this kind of problem? > > I am a spanish speaker, sorry for my english > > -- > Alejandro Vera > http://www.recicleta.cl > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Wed Jul 20 03:05:34 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 00:05:34 -0300 Subject: Liferea In-Reply-To: References: Message-ID: Thanks Chris, but I am already using it plugs: [home, x11, unity7, gsettings] Maybe the problem is other. But I do not know how to know... On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne wrote: > Hi Alejandro, > > Try using the home interface (described here: > http://snapcraft.io/docs/reference/interfaces) More info on how to use > interdaces is found here: http://snapcraft.io/create/ > > Thanks > Chris > > On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera > wrote: > >> Hi guys, first time in this list >> >> I am trying to snap liferea, to learn and to use it. >> >> I think I am almost ready, but when I run my snap I have an error: >> >> (process:4122): Gtk-WARNING **: Locale not supported by C library. >> Using the fallback 'C' locale. >> Gtk-Message: Failed to load module "canberra-gtk-module" >> Gtk-Message: Failed to load module "canberra-gtk-module" >> >> ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 >> (unable to open database file) >> ** >> ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit (db)) >> >> >> I think that the error is that it is trying to create the database inside >> the snap but it is read-only. >> >> Is there a way I can use the real HOME as HOME so liferea can create the >> database? Or is there a common way to solve this kind of problem? >> >> I am a spanish speaker, sorry for my english >> >> -- >> Alejandro Vera >> http://www.recicleta.cl >> >> -- >> Snapcraft mailing list >> Snapcraft at lists.ubuntu.com >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/snapcraft >> >> > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From chris.wayne at canonical.com Wed Jul 20 03:09:02 2016 From: chris.wayne at canonical.com (Chris Wayne) Date: Tue, 19 Jul 2016 23:09:02 -0400 Subject: Liferea In-Reply-To: References: Message-ID: Did you connect the plugs/slots by running "sudo snap connect liferea:home ubuntu-core:home"? On Tue, Jul 19, 2016 at 11:05 PM, Alejandro Vera wrote: > Thanks Chris, but I am already using it > > plugs: [home, x11, unity7, gsettings] > > Maybe the problem is other. But I do not know how to know... > > On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne > wrote: > >> Hi Alejandro, >> >> Try using the home interface (described here: >> http://snapcraft.io/docs/reference/interfaces) More info on how to use >> interdaces is found here: http://snapcraft.io/create/ >> >> Thanks >> Chris >> >> On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera > > wrote: >> >>> Hi guys, first time in this list >>> >>> I am trying to snap liferea, to learn and to use it. >>> >>> I think I am almost ready, but when I run my snap I have an error: >>> >>> (process:4122): Gtk-WARNING **: Locale not supported by C library. >>> Using the fallback 'C' locale. >>> Gtk-Message: Failed to load module "canberra-gtk-module" >>> Gtk-Message: Failed to load module "canberra-gtk-module" >>> >>> ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 >>> (unable to open database file) >>> ** >>> ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit (db)) >>> >>> >>> I think that the error is that it is trying to create the database >>> inside the snap but it is read-only. >>> >>> Is there a way I can use the real HOME as HOME so liferea can create the >>> database? Or is there a common way to solve this kind of problem? >>> >>> I am a spanish speaker, sorry for my english >>> >>> -- >>> Alejandro Vera >>> http://www.recicleta.cl >>> >>> -- >>> Snapcraft mailing list >>> Snapcraft at lists.ubuntu.com >>> Modify settings or unsubscribe at: >>> https://lists.ubuntu.com/mailman/listinfo/snapcraft >>> >>> >> > > > -- > Alejandro Vera > http://www.recicleta.cl > -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Wed Jul 20 03:14:03 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 00:14:03 -0300 Subject: Liferea In-Reply-To: References: Message-ID: Do i have to do it by hand? It is not made at install time? On Tue, Jul 19, 2016 at 11:09 PM, Chris Wayne wrote: > Did you connect the plugs/slots by running "sudo snap connect liferea:home > ubuntu-core:home"? > > On Tue, Jul 19, 2016 at 11:05 PM, Alejandro Vera > wrote: > >> Thanks Chris, but I am already using it >> >> plugs: [home, x11, unity7, gsettings] >> >> Maybe the problem is other. But I do not know how to know... >> >> On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne >> wrote: >> >>> Hi Alejandro, >>> >>> Try using the home interface (described here: >>> http://snapcraft.io/docs/reference/interfaces) More info on how to use >>> interdaces is found here: http://snapcraft.io/create/ >>> >>> Thanks >>> Chris >>> >>> On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera < >>> alejandro.vera at gmail.com> wrote: >>> >>>> Hi guys, first time in this list >>>> >>>> I am trying to snap liferea, to learn and to use it. >>>> >>>> I think I am almost ready, but when I run my snap I have an error: >>>> >>>> (process:4122): Gtk-WARNING **: Locale not supported by C library. >>>> Using the fallback 'C' locale. >>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>> >>>> ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 >>>> (unable to open database file) >>>> ** >>>> ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit (db)) >>>> >>>> >>>> I think that the error is that it is trying to create the database >>>> inside the snap but it is read-only. >>>> >>>> Is there a way I can use the real HOME as HOME so liferea can create >>>> the database? Or is there a common way to solve this kind of problem? >>>> >>>> I am a spanish speaker, sorry for my english >>>> >>>> -- >>>> Alejandro Vera >>>> http://www.recicleta.cl >>>> >>>> -- >>>> Snapcraft mailing list >>>> Snapcraft at lists.ubuntu.com >>>> Modify settings or unsubscribe at: >>>> https://lists.ubuntu.com/mailman/listinfo/snapcraft >>>> >>>> >>> >> >> >> -- >> Alejandro Vera >> http://www.recicleta.cl >> > > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Wed Jul 20 03:16:29 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 00:16:29 -0300 Subject: Liferea In-Reply-To: References: Message-ID: snap interfaces Slot Plug :gsettings liferea :home liferea :unity7 liferea :x11 liferea,x11-apps On Tue, Jul 19, 2016 at 11:14 PM, Alejandro Vera wrote: > Do i have to do it by hand? It is not made at install time? > > > On Tue, Jul 19, 2016 at 11:09 PM, Chris Wayne > wrote: > >> Did you connect the plugs/slots by running "sudo snap connect >> liferea:home ubuntu-core:home"? >> >> On Tue, Jul 19, 2016 at 11:05 PM, Alejandro Vera < >> alejandro.vera at gmail.com> wrote: >> >>> Thanks Chris, but I am already using it >>> >>> plugs: [home, x11, unity7, gsettings] >>> >>> Maybe the problem is other. But I do not know how to know... >>> >>> On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne >>> wrote: >>> >>>> Hi Alejandro, >>>> >>>> Try using the home interface (described here: >>>> http://snapcraft.io/docs/reference/interfaces) More info on how to use >>>> interdaces is found here: http://snapcraft.io/create/ >>>> >>>> Thanks >>>> Chris >>>> >>>> On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera < >>>> alejandro.vera at gmail.com> wrote: >>>> >>>>> Hi guys, first time in this list >>>>> >>>>> I am trying to snap liferea, to learn and to use it. >>>>> >>>>> I think I am almost ready, but when I run my snap I have an error: >>>>> >>>>> (process:4122): Gtk-WARNING **: Locale not supported by C library. >>>>> Using the fallback 'C' locale. >>>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>>> >>>>> ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 >>>>> (unable to open database file) >>>>> ** >>>>> ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit (db)) >>>>> >>>>> >>>>> I think that the error is that it is trying to create the database >>>>> inside the snap but it is read-only. >>>>> >>>>> Is there a way I can use the real HOME as HOME so liferea can create >>>>> the database? Or is there a common way to solve this kind of problem? >>>>> >>>>> I am a spanish speaker, sorry for my english >>>>> >>>>> -- >>>>> Alejandro Vera >>>>> http://www.recicleta.cl >>>>> >>>>> -- >>>>> Snapcraft mailing list >>>>> Snapcraft at lists.ubuntu.com >>>>> Modify settings or unsubscribe at: >>>>> https://lists.ubuntu.com/mailman/listinfo/snapcraft >>>>> >>>>> >>>> >>> >>> >>> -- >>> Alejandro Vera >>> http://www.recicleta.cl >>> >> >> > > > -- > Alejandro Vera > http://www.recicleta.cl > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.chen at canonical.com Wed Jul 20 03:25:56 2016 From: david.chen at canonical.com (David Chen) Date: Wed, 20 Jul 2016 11:25:56 +0800 Subject: mpv snap Message-ID: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> Hi, I am trying to create an mpv-demo snap on snappy for media playback, here is the current status: 1. mpv-demo snap is functional, can be tested by installing mir-server-demo and mpv-demo from edge channel with --devmode. 2. Video only, no audio. I would like to know if there is any suggestion on how to make audio work? I tried pulseaudio snap from link below, doesn't work. Also noticed that in classic ubuntu, there is a pulseaudio interface, but not available in snappy ubuntu core. http://anonscm.debian.org/cgit/pkg-pulseaudio/pulseaudio.git/tree/snapcraft.yaml?h=ubuntu-snappy Thanks and Regards, -David -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Wed Jul 20 03:43:04 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 00:43:04 -0300 Subject: Liferea In-Reply-To: References: Message-ID: I am learning a lot :D HOME is well defined. It is /home/myuser/snap/liferea/x4 and has write permission. But now I know that I have to set XDG_DATA_DIR to that folder. How is the best way to set environmetn variables in a snap? I am using the desktop laucher from gtk3 part. On Tue, Jul 19, 2016 at 11:16 PM, Alejandro Vera wrote: > snap interfaces > Slot Plug > :gsettings liferea > :home liferea > :unity7 liferea > :x11 liferea,x11-apps > > > On Tue, Jul 19, 2016 at 11:14 PM, Alejandro Vera > wrote: > >> Do i have to do it by hand? It is not made at install time? >> >> >> On Tue, Jul 19, 2016 at 11:09 PM, Chris Wayne >> wrote: >> >>> Did you connect the plugs/slots by running "sudo snap connect >>> liferea:home ubuntu-core:home"? >>> >>> On Tue, Jul 19, 2016 at 11:05 PM, Alejandro Vera < >>> alejandro.vera at gmail.com> wrote: >>> >>>> Thanks Chris, but I am already using it >>>> >>>> plugs: [home, x11, unity7, gsettings] >>>> >>>> Maybe the problem is other. But I do not know how to know... >>>> >>>> On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne >>> > wrote: >>>> >>>>> Hi Alejandro, >>>>> >>>>> Try using the home interface (described here: >>>>> http://snapcraft.io/docs/reference/interfaces) More info on how to >>>>> use interdaces is found here: http://snapcraft.io/create/ >>>>> >>>>> Thanks >>>>> Chris >>>>> >>>>> On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera < >>>>> alejandro.vera at gmail.com> wrote: >>>>> >>>>>> Hi guys, first time in this list >>>>>> >>>>>> I am trying to snap liferea, to learn and to use it. >>>>>> >>>>>> I think I am almost ready, but when I run my snap I have an error: >>>>>> >>>>>> (process:4122): Gtk-WARNING **: Locale not supported by C library. >>>>>> Using the fallback 'C' locale. >>>>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>>>> >>>>>> ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 >>>>>> (unable to open database file) >>>>>> ** >>>>>> ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit >>>>>> (db)) >>>>>> >>>>>> >>>>>> I think that the error is that it is trying to create the database >>>>>> inside the snap but it is read-only. >>>>>> >>>>>> Is there a way I can use the real HOME as HOME so liferea can create >>>>>> the database? Or is there a common way to solve this kind of problem? >>>>>> >>>>>> I am a spanish speaker, sorry for my english >>>>>> >>>>>> -- >>>>>> Alejandro Vera >>>>>> http://www.recicleta.cl >>>>>> >>>>>> -- >>>>>> Snapcraft mailing list >>>>>> Snapcraft at lists.ubuntu.com >>>>>> Modify settings or unsubscribe at: >>>>>> https://lists.ubuntu.com/mailman/listinfo/snapcraft >>>>>> >>>>>> >>>>> >>>> >>>> >>>> -- >>>> Alejandro Vera >>>> http://www.recicleta.cl >>>> >>> >>> >> >> >> -- >> Alejandro Vera >> http://www.recicleta.cl >> > > > > -- > Alejandro Vera > http://www.recicleta.cl > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Wed Jul 20 03:52:14 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 00:52:14 -0300 Subject: Liferea In-Reply-To: References: Message-ID: This is my snapcraft.yaml On Tue, Jul 19, 2016 at 11:43 PM, Alejandro Vera wrote: > I am learning a lot :D > > HOME is well defined. It is /home/myuser/snap/liferea/x4 and has write > permission. But now I know that I have to set XDG_DATA_DIR to that folder. > > How is the best way to set environmetn variables in a snap? I am using the > desktop laucher from gtk3 part. > > On Tue, Jul 19, 2016 at 11:16 PM, Alejandro Vera > wrote: > >> snap interfaces >> Slot Plug >> :gsettings liferea >> :home liferea >> :unity7 liferea >> :x11 liferea,x11-apps >> >> >> On Tue, Jul 19, 2016 at 11:14 PM, Alejandro Vera < >> alejandro.vera at gmail.com> wrote: >> >>> Do i have to do it by hand? It is not made at install time? >>> >>> >>> On Tue, Jul 19, 2016 at 11:09 PM, Chris Wayne >> > wrote: >>> >>>> Did you connect the plugs/slots by running "sudo snap connect >>>> liferea:home ubuntu-core:home"? >>>> >>>> On Tue, Jul 19, 2016 at 11:05 PM, Alejandro Vera < >>>> alejandro.vera at gmail.com> wrote: >>>> >>>>> Thanks Chris, but I am already using it >>>>> >>>>> plugs: [home, x11, unity7, gsettings] >>>>> >>>>> Maybe the problem is other. But I do not know how to know... >>>>> >>>>> On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne < >>>>> chris.wayne at canonical.com> wrote: >>>>> >>>>>> Hi Alejandro, >>>>>> >>>>>> Try using the home interface (described here: >>>>>> http://snapcraft.io/docs/reference/interfaces) More info on how to >>>>>> use interdaces is found here: http://snapcraft.io/create/ >>>>>> >>>>>> Thanks >>>>>> Chris >>>>>> >>>>>> On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera < >>>>>> alejandro.vera at gmail.com> wrote: >>>>>> >>>>>>> Hi guys, first time in this list >>>>>>> >>>>>>> I am trying to snap liferea, to learn and to use it. >>>>>>> >>>>>>> I think I am almost ready, but when I run my snap I have an error: >>>>>>> >>>>>>> (process:4122): Gtk-WARNING **: Locale not supported by C library. >>>>>>> Using the fallback 'C' locale. >>>>>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>>>>> Gtk-Message: Failed to load module "canberra-gtk-module" >>>>>>> >>>>>>> ** (liferea:4122): WARNING **: Unexpected status on SQL execution: >>>>>>> 14 (unable to open database file) >>>>>>> ** >>>>>>> ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit >>>>>>> (db)) >>>>>>> >>>>>>> >>>>>>> I think that the error is that it is trying to create the database >>>>>>> inside the snap but it is read-only. >>>>>>> >>>>>>> Is there a way I can use the real HOME as HOME so liferea can create >>>>>>> the database? Or is there a common way to solve this kind of problem? >>>>>>> >>>>>>> I am a spanish speaker, sorry for my english >>>>>>> >>>>>>> -- >>>>>>> Alejandro Vera >>>>>>> http://www.recicleta.cl >>>>>>> >>>>>>> -- >>>>>>> Snapcraft mailing list >>>>>>> Snapcraft at lists.ubuntu.com >>>>>>> Modify settings or unsubscribe at: >>>>>>> https://lists.ubuntu.com/mailman/listinfo/snapcraft >>>>>>> >>>>>>> >>>>>> >>>>> >>>>> >>>>> -- >>>>> Alejandro Vera >>>>> http://www.recicleta.cl >>>>> >>>> >>>> >>> >>> >>> -- >>> Alejandro Vera >>> http://www.recicleta.cl >>> >> >> >> >> -- >> Alejandro Vera >> http://www.recicleta.cl >> > > > > -- > Alejandro Vera > http://www.recicleta.cl > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: snapcraft.yaml Type: application/x-yaml Size: 1279 bytes Desc: not available URL: From didrocks at ubuntu.com Wed Jul 20 06:45:25 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Wed, 20 Jul 2016 08:45:25 +0200 Subject: mpv snap In-Reply-To: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> Message-ID: <578F1E05.1040102@ubuntu.com> Le 20/07/2016 05:25, David Chen a écrit : > > Hi, > Hey David, > > I am trying to create an mpv-demo snap on snappy for media playback, > here is the current status: > > 1. mpv-demo snap is functional, can be tested by installing > mir-server-demo and mpv-demo from edge channel with --devmode. > > 2. Video only, no audio. > > I would like to know if there is any suggestion on how to make audio work? > > I tried pulseaudio snap from link below, doesn't work. Also noticed > that in classic ubuntu, there is a pulseaudio interface, but not > available in snappy ubuntu core. > > http://anonscm.debian.org/cgit/pkg-pulseaudio/pulseaudio.git/tree/snapcraft.yaml?h=ubuntu-snappy > We already have some mpv in the snappy playpen, do you mind contributing your enhancements there? (https://github.com/ubuntu/snappy-playpen/tree/master/mpv) I think that indeed those media interfaces are only available in classic ubuntu for now, but they will surely be available on any capable ubuntu core device once we start having series 16 images. Cheers, Didier -------------- next part -------------- An HTML attachment was scrubbed... URL: From didrocks at ubuntu.com Wed Jul 20 06:51:43 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Wed, 20 Jul 2016 08:51:43 +0200 Subject: Liferea In-Reply-To: References: Message-ID: <578F1F7F.8050301@ubuntu.com> Le 20/07/2016 05:43, Alejandro Vera a écrit : > I am learning a lot :D Excellent :) > > HOME is well defined. It is /home/myuser/snap/liferea/x4 and has write > permission. But now I know that I have to set XDG_DATA_DIR to that > folder. This is weird, XDG_DATA_DIR isn't part of the freedesktop spec (https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html). The desktop launcher though already set XDG_DATA_DIRS (note the finale S) to what is expected due to $HOME redirection. Are you sure this is really the issue you are seeing? You can always try to set a wrapper script after desktop-launch in your snap $PATH to expert and add/remove variables. I know that Sergio is going to bring support for runtime env variable from snapcraft.yaml directly in the near future. > > How is the best way to set environmetn variables in a snap? I am using > the desktop laucher from gtk3 part. > > On Tue, Jul 19, 2016 at 11:16 PM, Alejandro Vera > > wrote: > > snap interfaces > Slot Plug > :gsettings liferea > :home liferea > :unity7 liferea > :x11 liferea,x11-apps > > > On Tue, Jul 19, 2016 at 11:14 PM, Alejandro Vera > wrote: > > Do i have to do it by hand? It is not made at install time? > > > On Tue, Jul 19, 2016 at 11:09 PM, Chris Wayne > > > wrote: > > Did you connect the plugs/slots by running "sudo snap > connect liferea:home ubuntu-core:home"? > > On Tue, Jul 19, 2016 at 11:05 PM, Alejandro Vera > wrote: > > Thanks Chris, but I am already using it > > plugs: [home, x11, unity7, gsettings] > > Maybe the problem is other. But I do not know how to > know... > > On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne > wrote: > > Hi Alejandro, > > Try using the home interface (described > here: http://snapcraft.io/docs/reference/interfaces) > More info on how to use interdaces is found > here: http://snapcraft.io/create/ > > Thanks > Chris > > On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera > wrote: > > Hi guys, first time in this list > > I am trying to snap liferea, to learn and to > use it. > > I think I am almost ready, but when I run my > snap I have an error: > > (process:4122): Gtk-WARNING **: Locale not > supported by C library. > Using the fallback 'C' locale. > Gtk-Message: Failed to load module > "canberra-gtk-module" > Gtk-Message: Failed to load module > "canberra-gtk-module" > > ** (liferea:4122): WARNING **: Unexpected > status on SQL execution: 14 (unable to open > database file) > ** > ERROR:db.c:729:db_init: assertion failed: > (sqlite3_get_autocommit (db)) > > > I think that the error is that it is trying to > create the database inside the snap but it is > read-only. > > Is there a way I can use the real HOME as HOME > so liferea can create the database? Or is > there a common way to solve this kind of problem? > > I am a spanish speaker, sorry for my english > > -- > Alejandro Vera > http://www.recicleta.cl > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > > > > > -- > Alejandro Vera > http://www.recicleta.cl > > > > > > -- > Alejandro Vera > http://www.recicleta.cl > > > > > -- > Alejandro Vera > http://www.recicleta.cl > > > > > -- > Alejandro Vera > http://www.recicleta.cl > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ted at ubuntu.com Wed Jul 20 07:05:38 2016 From: ted at ubuntu.com (Ted Gould) Date: Wed, 20 Jul 2016 09:05:38 +0200 Subject: Liferea In-Reply-To: References: Message-ID: <1468998337.6084.16.camel@ubuntu.com> The home interface doesn't allow you to use directories with a "." as the first character. This is to avoid having applications able to read your private keys and other important information. You should change the place that Liferea uses to write it's configuration and databases to be the snap user writable areas, and I think in the end you might not need the home interface at all! The environment variable that gives the location of writable area for the application is $SNAP_USER_DATA. If you want to know any of the variables that is typically given to a snap you can install the hello- world snap and use "hello-world.env" to see them. Good luck! Ted On Wed, 2016-07-20 at 00:05 -0300, Alejandro Vera wrote: > Thanks Chris, but I am already using it > >    plugs: [home, x11, unity7, gsettings] > > Maybe the problem is other. But I do not know how to know... > > On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne > om> wrote: > > Hi Alejandro, > > > > Try using the home interface (described here: http://snapcraft.io/d > > ocs/reference/interfaces) More info on how to use interdaces is > > found here: http://snapcraft.io/create/ > >   > > Thanks > > Chris > > > > On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera > > il.com> wrote: > > > Hi guys, first time in this list > > > > > > I am trying to snap liferea, to learn and to use it. > > > > > > I think I am almost ready, but when I run my snap I have an > > > error: > > > > > > (process:4122): Gtk-WARNING **: Locale not supported by C > > > library. > > >     Using the fallback 'C' locale. > > > Gtk-Message: Failed to load module "canberra-gtk-module" > > > Gtk-Message: Failed to load module "canberra-gtk-module" > > > > > > ** (liferea:4122): WARNING **: Unexpected status on SQL > > > execution: 14 (unable to open database file) > > > ** > > > ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit > > > (db)) > > > > > > > > > I think that the error is that it is trying to create the > > > database inside the snap but it is read-only. > > > > > > Is there a way I can use the real HOME as HOME so liferea can > > > create the database? Or is there a common way to solve this kind > > > of problem? > > > > > > I am a spanish speaker, sorry for my english  > > > > > > --  > > > Alejandro Vera > > > http://www.recicleta.cl > > > > > > -- > > > Snapcraft mailing list > > > Snapcraft at lists.ubuntu.com > > > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailm > > > an/listinfo/snapcraft > > > > > > > > > > > --  > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/l > istinfo/snapcraft -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From alejandro.vera at gmail.com Wed Jul 20 14:52:48 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 11:52:48 -0300 Subject: Liferea In-Reply-To: <1468998337.6084.16.camel@ubuntu.com> References: <1468998337.6084.16.camel@ubuntu.com> Message-ID: Liferea uses this environment variables (from man) $XDG_CONFIG_DIR/liferea/feedlist.opml Contains the current list of subscriptions $XDG_CONFIG_DIR/liferea/liferea.css Stylesheet that can be used to override default HTML style $XDG_DATA_DIR/liferea/liferea.db The first two I think are standar and the last one is from liferea... I think then that the right solution is to set the three of them to the SNAP_USER_DATA and remove the home interface :D what do you think? On Wed, Jul 20, 2016 at 3:05 AM, Ted Gould wrote: > > The home interface doesn't allow you to use directories with a "." as the > first character. This is to avoid having applications able to read your > private keys and other important information. > > You should change the place that Liferea uses to write it's configuration > and databases to be the snap user writable areas, and I think in the end > you might not need the home interface at all! > > The environment variable that gives the location of writable area for the > application is $SNAP_USER_DATA. If you want to know any of the variables > that is typically given to a snap you can install the hello-world snap and > use "hello-world.env" to see them. > > Good luck! > Ted > > On Wed, 2016-07-20 at 00:05 -0300, Alejandro Vera wrote: > > Thanks Chris, but I am already using it > > plugs: [home, x11, unity7, gsettings] > > Maybe the problem is other. But I do not know how to know... > > On Tue, Jul 19, 2016 at 9:04 PM, Chris Wayne > wrote: > > Hi Alejandro, > > Try using the home interface (described here: > http://snapcraft.io/docs/reference/interfaces) More info on how to use > interdaces is found here: http://snapcraft.io/create/ > > Thanks > Chris > > On Tue, Jul 19, 2016 at 7:17 PM, Alejandro Vera > wrote: > > Hi guys, first time in this list > > I am trying to snap liferea, to learn and to use it. > > I think I am almost ready, but when I run my snap I have an error: > > (process:4122): Gtk-WARNING **: Locale not supported by C library. > Using the fallback 'C' locale. > Gtk-Message: Failed to load module "canberra-gtk-module" > Gtk-Message: Failed to load module "canberra-gtk-module" > > ** (liferea:4122): WARNING **: Unexpected status on SQL execution: 14 > (unable to open database file) > ** > ERROR:db.c:729:db_init: assertion failed: (sqlite3_get_autocommit (db)) > > > I think that the error is that it is trying to create the database inside > the snap but it is read-only. > > Is there a way I can use the real HOME as HOME so liferea can create the > database? Or is there a common way to solve this kind of problem? > > I am a spanish speaker, sorry for my english > > -- > Alejandro Vera > http://www.recicleta.cl > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > > > > > > > -- > Snapcraft mailing listSnapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft > > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Wed Jul 20 14:57:43 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Wed, 20 Jul 2016 16:57:43 +0200 Subject: Liferea In-Reply-To: References: <1468998337.6084.16.camel@ubuntu.com> Message-ID: <1469026663.23385.1.camel@ubuntu.com> hi, Am Mittwoch, den 20.07.2016, 11:52 -0300 schrieb Alejandro Vera: > Liferea uses this environment variables (from man) > >       $XDG_CONFIG_DIR/liferea/feedlist.opml             Contains the > current list of subscriptions >        $XDG_CONFIG_DIR/liferea/liferea.css              Stylesheet > that can be used to override default HTML style >        $XDG_DATA_DIR/liferea/liferea.db > > The first two I think are standar and the last one is from liferea... > > I think then that the right solution is to set the three of them to > the SNAP_USER_DATA and remove the home interface :D > > what do you think? it is definitely what i would do (in a separate wrapper script), just do not forget that you need to mkdir the subdir if it doesn't exist ... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From alejandro.vera at gmail.com Wed Jul 20 14:59:33 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 11:59:33 -0300 Subject: Liferea In-Reply-To: <1469026663.23385.1.camel@ubuntu.com> References: <1468998337.6084.16.camel@ubuntu.com> <1469026663.23385.1.camel@ubuntu.com> Message-ID: oooo.. I didnt think about it mkdir -p $XDG_CONFIG_DIR/liferea/ that?? On Wed, Jul 20, 2016 at 10:57 AM, Oliver Grawert wrote: > hi, > Am Mittwoch, den 20.07.2016, 11:52 -0300 schrieb Alejandro Vera: > > Liferea uses this environment variables (from man) > > > > $XDG_CONFIG_DIR/liferea/feedlist.opml Contains the > > current list of subscriptions > > $XDG_CONFIG_DIR/liferea/liferea.css Stylesheet > > that can be used to override default HTML style > > $XDG_DATA_DIR/liferea/liferea.db > > > > The first two I think are standar and the last one is from liferea... > > > > I think then that the right solution is to set the three of them to > > the SNAP_USER_DATA and remove the home interface :D > > > > what do you think? > > it is definitely what i would do (in a separate wrapper script), just > do not forget that you need to mkdir the subdir if it doesn't exist ... > > ciao > oli > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From jamie at canonical.com Wed Jul 20 17:40:53 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Wed, 20 Jul 2016 12:40:53 -0500 Subject: mpv snap In-Reply-To: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> Message-ID: <1469036453.20475.9.camel@canonical.com> On Wed, 2016-07-20 at 11:25 +0800, David Chen wrote: > Also noticed > that in classic ubuntu, there is a pulseaudio interface, but not > available in snappy ubuntu core. > On classic the interface provides access to the user session's pulseaudio. On an all snaps system (eg, Ubuntu Core) pulseaudio doesn't exist on the system and a pulseaudio snap must be installed. Once installed, the interface is available for other snaps to connect. There is a PR that is very close to landing that implements the all-snaps part of the slot side of the pulseaudio interface and I suspect that once it has landed a pulseaudio snap that slots pulseaudio will be uploaded very soon after. In other words, what you are seeing is expected right now and soon all-snaps images will be able to use pulseaudio too. -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From alejandro.vera at gmail.com Wed Jul 20 23:25:33 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 20:25:33 -0300 Subject: Liferea In-Reply-To: References: <1468998337.6084.16.camel@ubuntu.com> <1469026663.23385.1.camel@ubuntu.com> Message-ID: if a snap cant read folder with '.' in the first char, how can they use their config in .config??? On Wed, Jul 20, 2016 at 10:59 AM, Alejandro Vera wrote: > oooo.. I didnt think about it > > mkdir -p $XDG_CONFIG_DIR/liferea/ > > that?? > > On Wed, Jul 20, 2016 at 10:57 AM, Oliver Grawert wrote: > >> hi, >> Am Mittwoch, den 20.07.2016, 11:52 -0300 schrieb Alejandro Vera: >> > Liferea uses this environment variables (from man) >> > >> > $XDG_CONFIG_DIR/liferea/feedlist.opml Contains the >> > current list of subscriptions >> > $XDG_CONFIG_DIR/liferea/liferea.css Stylesheet >> > that can be used to override default HTML style >> > $XDG_DATA_DIR/liferea/liferea.db >> > >> > The first two I think are standar and the last one is from liferea... >> > >> > I think then that the right solution is to set the three of them to >> > the SNAP_USER_DATA and remove the home interface :D >> > >> > what do you think? >> >> it is definitely what i would do (in a separate wrapper script), just >> do not forget that you need to mkdir the subdir if it doesn't exist ... >> >> ciao >> oli >> -- >> Snapcraft mailing list >> Snapcraft at lists.ubuntu.com >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/snapcraft >> >> > > > -- > Alejandro Vera > http://www.recicleta.cl > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Thu Jul 21 00:00:52 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 21:00:52 -0300 Subject: Liferea In-Reply-To: References: <1468998337.6084.16.camel@ubuntu.com> <1469026663.23385.1.camel@ubuntu.com> Message-ID: Hi guys. Thanks for all the help I've created a launcher and setted all the variables, but the database is created in the SNAP/.local/share/lifera folder... The weird thing is that the file is created and populated, but then liferea cant read it... I created a fork of playpen with my files. If someone can helps me it would be great... https://github.com/almejo/snappy-playpen/tree/liferea On Wed, Jul 20, 2016 at 7:25 PM, Alejandro Vera wrote: > if a snap cant read folder with '.' in the first char, how can they use > their config in .config??? > > On Wed, Jul 20, 2016 at 10:59 AM, Alejandro Vera > wrote: > >> oooo.. I didnt think about it >> >> mkdir -p $XDG_CONFIG_DIR/liferea/ >> >> that?? >> >> On Wed, Jul 20, 2016 at 10:57 AM, Oliver Grawert wrote: >> >>> hi, >>> Am Mittwoch, den 20.07.2016, 11:52 -0300 schrieb Alejandro Vera: >>> > Liferea uses this environment variables (from man) >>> > >>> > $XDG_CONFIG_DIR/liferea/feedlist.opml Contains the >>> > current list of subscriptions >>> > $XDG_CONFIG_DIR/liferea/liferea.css Stylesheet >>> > that can be used to override default HTML style >>> > $XDG_DATA_DIR/liferea/liferea.db >>> > >>> > The first two I think are standar and the last one is from liferea... >>> > >>> > I think then that the right solution is to set the three of them to >>> > the SNAP_USER_DATA and remove the home interface :D >>> > >>> > what do you think? >>> >>> it is definitely what i would do (in a separate wrapper script), just >>> do not forget that you need to mkdir the subdir if it doesn't exist ... >>> >>> ciao >>> oli >>> -- >>> Snapcraft mailing list >>> Snapcraft at lists.ubuntu.com >>> Modify settings or unsubscribe at: >>> https://lists.ubuntu.com/mailman/listinfo/snapcraft >>> >>> >> >> >> -- >> Alejandro Vera >> http://www.recicleta.cl >> > > > > -- > Alejandro Vera > http://www.recicleta.cl > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Thu Jul 21 02:21:31 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Wed, 20 Jul 2016 23:21:31 -0300 Subject: git Message-ID: Hi guys I am trying to snap another app. This time is git :D Buy my problem is that git uses ssh when cloning. But ssh wants to add a key to my known_hosts and loading my key. I know that snapcraft blocks snaps to read and write folders with '.' in the begining. Soo... some advice? thanks in advance!! :D -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From xiaoguo.liu at canonical.com Thu Jul 21 05:03:43 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Thu, 21 Jul 2016 13:03:43 +0800 Subject: How to use "snap run" command? Message-ID: Hi, I just installed the hello-world app from the snap store. I found that snap has a command "run". liuxg at liuxg:~$ snap run -h Usage: snap [OPTIONS] run [run-OPTIONS] Run the given snap command with the right confinement and environment Application Options: --version print the version and exit Help Options: -h, --help Show this help message [run command arguments] : the snap (e.g. hello-world) or application to run (e.g. hello-world.env) When I run it like: "snap run hello-world", but it does not run at all. What is the correct way to use the command? Thanks & best regards, XiaoGuo -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From mwinter at opensourcerouting.org Thu Jul 21 10:14:02 2016 From: mwinter at opensourcerouting.org (Martin Winter) Date: Thu, 21 Jul 2016 03:14:02 -0700 Subject: package contains external symlinks: lib64/ld-linux-x86-64.so.2 lint-snap-v2_external_symlinks Message-ID: <41CE2D43-70A3-44CA-AEFE-8E6A5C780D05@opensourcerouting.org> Hi there, I’m currently failing on my package with this error. I have the libc6 included as a stage package, but trying to find the correct linker option to actually use the library included in snap instead of the system wide one. Anyone knows how to get gcc/ld to use the correct library? - Martin From loic.minier at ubuntu.com Thu Jul 21 10:49:11 2016 From: loic.minier at ubuntu.com (=?UTF-8?B?TG/Dr2MgTWluaWVy?=) Date: Thu, 21 Jul 2016 12:49:11 +0200 Subject: package contains external symlinks: lib64/ld-linux-x86-64.so.2 lint-snap-v2_external_symlinks In-Reply-To: <41CE2D43-70A3-44CA-AEFE-8E6A5C780D05@opensourcerouting.org> References: <41CE2D43-70A3-44CA-AEFE-8E6A5C780D05@opensourcerouting.org> Message-ID: This is tricky; three options: 1) patch the files to use your ld-linux; patchelf is supposed to be able to do this, but when I tried --set-interpreter on a simple binary I got this error: cannot find section .interp You could still patch it in ugly ways. 2) you can wrap your commands in shell scripts which run the linker; you have a stable interface in the form of POSIX shell script /bin/sh; it goes like this: ./copy-of-ld --inhibit-cache ./foo 3) you can set the interpreter at build time; this is supposedly done with -Wl,--dynamic-linker=foo, but by default your toolchain probably uses linker scripts which will set a different linker. You may dump the default linker script and override it (gcc -T) with your own fork. In any case I'd be interest to read about why you want to avoid the ld-linux dep and how you implement it in the end. :-) Cheers, - Loïc Minier On Thu, Jul 21, 2016 at 12:14 PM, Martin Winter < mwinter at opensourcerouting.org> wrote: > Hi there, > > I’m currently failing on my package with this error. > > I have the libc6 included as a stage package, but trying to find the > correct linker option > to actually use the library included in snap instead of the system wide > one. > > Anyone knows how to get gcc/ld to use the correct library? > > - Martin > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From tom at analytical-labs.com Thu Jul 21 11:38:51 2016 From: tom at analytical-labs.com (Tom Barber) Date: Thu, 21 Jul 2016 12:38:51 +0100 Subject: Maven plugin for snap definition generation Message-ID: Hi guys, Figured I should put this on here for posterity: https://github.com/buggtb/snappy-maven-plugin For those of you not at my talk last night, its a maven plugin to generate and build snapcraft definitions inside an existing Maven build cycle. Its still very rough around the edges, but I'll put it to use on my own projects over the next few weeks and iron out the kinks and niggles. To all of those in Germany, thanks for the invite and for introducing me to the world of Snaps! Cheers Tom -------------- Director Meteorite.bi - Saiku Analytics Founder Tel: +44(0)5603641316 (Thanks to the Saiku community we reached our Kickstart goal, but you can always help by sponsoring the project ) -------------- next part -------------- An HTML attachment was scrubbed... URL: From didrocks at ubuntu.com Thu Jul 21 11:56:00 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Thu, 21 Jul 2016 13:56:00 +0200 Subject: Maven plugin for snap definition generation In-Reply-To: References: Message-ID: <5790B850.7040301@ubuntu.com> Le 21/07/2016 13:38, Tom Barber a écrit : > Hi guys, > > Figured I should put this on here for > posterity: https://github.com/buggtb/snappy-maven-plugin Hey Tom, Excellent work! > > For those of you not at my talk last night, its a maven plugin to > generate and build snapcraft definitions inside an existing Maven > build cycle. Its still very rough around the edges, but I'll put it to > use on my own projects over the next few weeks and iron out the kinks > and niggles. > > To all of those in Germany, thanks for the invite and for introducing > me to the world of Snaps! Same for us, was really nice talking to and meeting you. Very nice lightning talk! Great to see this code out. Let's keep us posted as you are polishing it. Cheers, Didier From mwinter at opensourcerouting.org Thu Jul 21 12:48:22 2016 From: mwinter at opensourcerouting.org (Martin Winter) Date: Thu, 21 Jul 2016 05:48:22 -0700 Subject: package contains external symlinks: lib64/ld-linux-x86-64.so.2 lint-snap-v2_external_symlinks In-Reply-To: References: <41CE2D43-70A3-44CA-AEFE-8E6A5C780D05@opensourcerouting.org> Message-ID: On 21 Jul 2016, at 3:49, Loïc Minier wrote: > This is tricky; three options: > 1) patch the files to use your ld-linux; patchelf is supposed to be able to > do this, but when I tried --set-interpreter on a simple binary I got this > error: > cannot find section .interp > > You could still patch it in ugly ways. > > 2) you can wrap your commands in shell scripts which run the linker; you > have a stable interface in the form of POSIX shell script /bin/sh; it goes > like this: > ./copy-of-ld --inhibit-cache ./foo Runtime isn’t the issue. The problem is the ubuntu store failing the snap in the automated tests after the upload. > 3) you can set the interpreter at build time; this is supposedly done with > -Wl,--dynamic-linker=foo, but by default your toolchain probably uses > linker scripts which will set a different linker. You may dump the default > linker script and override it (gcc -T) with your own fork. > > In any case I'd be interest to read about why you want to avoid the > ld-linux dep and how you implement it in the end. :-) Not sure I understand your question. But the snap upload is the main one complaining and failing the the snap in the automated test because of the reference. I thought that libc6 would be an exception and doesn’t need to be added to the snap. I thought this should be a common issue from other snap’s (and expected known solutions). Or how do other C projects solve this? - Martin > On Thu, Jul 21, 2016 at 12:14 PM, Martin Winter < > mwinter at opensourcerouting.org> wrote: > >> Hi there, >> >> I’m currently failing on my package with this error. >> >> I have the libc6 included as a stage package, but trying to find the >> correct linker option >> to actually use the library included in snap instead of the system wide >> one. >> >> Anyone knows how to get gcc/ld to use the correct library? >> >> - Martin >> >> -- >> Snapcraft mailing list >> Snapcraft at lists.ubuntu.com >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/snapcraft >> From jamie at canonical.com Thu Jul 21 13:05:00 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Thu, 21 Jul 2016 08:05:00 -0500 Subject: package contains external symlinks: lib64/ld-linux-x86-64.so.2 lint-snap-v2_external_symlinks In-Reply-To: References: <41CE2D43-70A3-44CA-AEFE-8E6A5C780D05@opensourcerouting.org> Message-ID: <1469106300.20475.46.camel@canonical.com> On Thu, 2016-07-21 at 05:48 -0700, Martin Winter wrote: > > On 21 Jul 2016, at 3:49, Loïc Minier wrote: > > > > > This is tricky; three options: > > 1) patch the files to use your ld-linux; patchelf is supposed to be able to > > do this, but when I tried --set-interpreter on a simple binary I got this > > error: > > cannot find section .interp > > > > You could still patch it in ugly ways. > > > > 2) you can wrap your commands in shell scripts which run the linker; you > > have a stable interface in the form of POSIX shell script /bin/sh; it goes > > like this: > > ./copy-of-ld --inhibit-cache ./foo > Runtime isn’t the issue. The problem is the ubuntu store failing the snap > in the automated tests after the upload. > There is a bug in the review tools which are supposed to allow external symlinks to libc libs but this particular one's matching rule didn't catch this lib when it should have.  Please request a manual review in the store and I can take a look at the snap and fix the tools. -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From mwinter at opensourcerouting.org Thu Jul 21 17:26:36 2016 From: mwinter at opensourcerouting.org (Martin Winter) Date: Thu, 21 Jul 2016 10:26:36 -0700 Subject: snap interface (plugs) matching Message-ID: <6A351D99-F925-43A5-AC35-599F6070513E@opensourcerouting.org> I’m running into a weird issue here while trying to get the correct plugs added and they don’t seem to get detected/parsed correctly: My snapcraft.yaml looks like this (extract for a simple process only): […] apps: ospf6d: command: bin/ospf6d-service daemon: simple plugs: - firewall-control - network - network-bind - network-control - network-observe […] But when I run the program, I still get the following error: (with snappy-debug.security scanlog quagga) = AppArmor = Time: Jul 21 10:13:38 Log: apparmor="DENIED" operation="create" profile="snap.quagga.ospf6d" pid=20622 comm="ospf6d" family="inet6" sock_type="raw" protocol=89 requested_mask="create" denied_mask="create" Suggestion: * add one of 'firewall-control, network-control, network-observe' to 'plugs' Looking at the interfaces with the snap command, I see the following: # snap interfaces Slot Plug :camera - :cups-control - :firewall-control - :gsettings - :home - :locale-control - :log-observe snappy-debug :modem-manager - :mount-observe - :network quagga :network-bind quagga :network-control - :network-manager - :network-observe - :opengl - :optical-drive - :ppp - :pulseaudio - :snapd-control - :system-observe - :timeserver-control - :timezone-control - :unity7 - :x11 - - quagga:firewall-control - quagga:network-control - quagga:network-observe Question: Why is firewall-control / network-control / network-observe not correctly detected? It looks like it gets prefixed by “quagga:” and not assigned the correct Slot. (This is Ubuntu 16.04, Snapcraft 2.12, snap 2.0.10) - Martin From jamie at canonical.com Thu Jul 21 17:51:24 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Thu, 21 Jul 2016 12:51:24 -0500 Subject: snap interface (plugs) matching In-Reply-To: <6A351D99-F925-43A5-AC35-599F6070513E@opensourcerouting.org> References: <6A351D99-F925-43A5-AC35-599F6070513E@opensourcerouting.org> Message-ID: <1469123484.20475.67.camel@canonical.com> On Thu, 2016-07-21 at 10:26 -0700, Martin Winter wrote: > I’m running into a weird issue here while trying to get the correct  > plugs added and they don’t seem to get detected/parsed correctly: > > My snapcraft.yaml looks like this (extract for a simple process only): > > […] > apps: >     ospf6d: >         command: bin/ospf6d-service >         daemon: simple >         plugs: >             - firewall-control >             - network >             - network-bind >             - network-control >             - network-observe > […] > > But when I run the program, I still get the following error: > (with snappy-debug.security scanlog quagga) > > = AppArmor = > Time: Jul 21 10:13:38 > Log: apparmor="DENIED" operation="create" profile="snap.quagga.ospf6d"  > pid=20622 comm="ospf6d" family="inet6" sock_type="raw" protocol=89  > requested_mask="create" denied_mask="create" > Suggestion: > * add one of 'firewall-control, network-control, network-observe' to  > 'plugs' > > Looking at the interfaces with the snap command, I see the following: > > # snap interfaces > Slot                 Plug > :camera              - > :cups-control        - > :firewall-control    - > :gsettings           - > :home                - > :locale-control      - > :log-observe         snappy-debug > :modem-manager       - > :mount-observe       - > :network             quagga > :network-bind        quagga > :network-control     - > :network-manager     - > :network-observe     - > :opengl              - > :optical-drive       - > :ppp                 - > :pulseaudio          - > :snapd-control       - > :system-observe      - > :timeserver-control  - > :timezone-control    - > :unity7              - > :x11                 - > -                    quagga:firewall-control > -                    quagga:network-control > -                    quagga:network-observe > > > Question: > > Why is firewall-control / network-control / network-observe not  > correctly detected? It looks like it gets prefixed by “quagga:” and  > not assigned the correct Slot. > > (This is Ubuntu 16.04, Snapcraft 2.12, snap 2.0.10) > They are detected but not automatically connected because firewall-control, network-control and network-observe give privileged access to the system. After install, you should do: $ sudo snap connect quagga:firewall-control ubuntu-core:firewall-control $ sudo snap connect quagga:network-control ubuntu-core:network-control $ sudo snap connect quagga:network-observe ubuntu-core:network-observe Once done, this will be remembered on upgrades (but not remove/install). AIUI the snappy team is discussing how to make this easier and discoverable. This was also discussed a bit here: https://lists.ubuntu.com/archives/snapcraft/2016-July /000416.html Side note for those interested in cross-distro: AIUI, at some point 'ubuntu- core' will both not be required in the command (ie, use ':firewall-control') and 'ubuntu-core' will be renamed to not include 'ubuntu' (ie, use ':firewall- control'). -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From zygmunt.krynicki at canonical.com Thu Jul 21 17:52:48 2016 From: zygmunt.krynicki at canonical.com (Zygmunt Krynicki) Date: Thu, 21 Jul 2016 19:52:48 +0200 Subject: snap interface (plugs) matching In-Reply-To: <6A351D99-F925-43A5-AC35-599F6070513E@opensourcerouting.org> References: <6A351D99-F925-43A5-AC35-599F6070513E@opensourcerouting.org> Message-ID: <0FEBD60C-471A-4D16-B072-E3D7B6FCF1DC@canonical.com> > Wiadomość napisana przez Martin Winter w dniu 21.07.2016, o godz. 19:26: > > I’m running into a weird issue here while trying to get the correct plugs added and they don’t seem to get detected/parsed correctly: > > My snapcraft.yaml looks like this (extract for a simple process only): > > […] > apps: > ospf6d: > command: bin/ospf6d-service > daemon: simple > plugs: > - firewall-control > - network > - network-bind > - network-control > - network-observe > […] > > But when I run the program, I still get the following error: > (with snappy-debug.security scanlog quagga) > > = AppArmor = > Time: Jul 21 10:13:38 > Log: apparmor="DENIED" operation="create" profile="snap.quagga.ospf6d" pid=20622 comm="ospf6d" family="inet6" sock_type="raw" protocol=89 requested_mask="create" denied_mask="create" > Suggestion: > * add one of 'firewall-control, network-control, network-observe' to 'plugs' > > Looking at the interfaces with the snap command, I see the following: > > # snap interfaces > Slot Plug > :camera - > :cups-control - > :firewall-control - > :gsettings - > :home - > :locale-control - > :log-observe snappy-debug > :modem-manager - > :mount-observe - > :network quagga > :network-bind quagga > :network-control - > :network-manager - > :network-observe - > :opengl - > :optical-drive - > :ppp - > :pulseaudio - > :snapd-control - > :system-observe - > :timeserver-control - > :timezone-control - > :unity7 - > :x11 - > - quagga:firewall-control > - quagga:network-control > - quagga:network-observe > > > Question: > > Why is firewall-control / network-control / network-observe not correctly detected? It looks like it gets prefixed by “quagga:” and not assigned the correct Slot. > Those interfaces do not auto-connect. You need to explicitly connect them with: $ snap connect quagga:firewall-control ubuntu-core:firewall-control Best regards ZK From nicholas.skaggs at canonical.com Thu Jul 21 18:23:19 2016 From: nicholas.skaggs at canonical.com (Nicholas Skaggs) Date: Thu, 21 Jul 2016 14:23:19 -0400 Subject: Liferea In-Reply-To: <1469026663.23385.1.camel@ubuntu.com> References: <1468998337.6084.16.camel@ubuntu.com> <1469026663.23385.1.camel@ubuntu.com> Message-ID: But what happens if I'm transitioning from Liferea installed a deb to the snap -- everything I expect is in $HOME. I get we can convince the app to look elsewhere to write files, but as a user, what happens to my config files when I transition? Liferea isn't going to read them afaict. How about when I uninstall the snap? Re-install my OS? (keeping only $HOME) Nicholas On 07/20/2016 10:57 AM, Oliver Grawert wrote: > hi, > Am Mittwoch, den 20.07.2016, 11:52 -0300 schrieb Alejandro Vera: >> Liferea uses this environment variables (from man) >> >> $XDG_CONFIG_DIR/liferea/feedlist.opml Contains the >> current list of subscriptions >> $XDG_CONFIG_DIR/liferea/liferea.css Stylesheet >> that can be used to override default HTML style >> $XDG_DATA_DIR/liferea/liferea.db >> >> The first two I think are standar and the last one is from liferea... >> >> I think then that the right solution is to set the three of them to >> the SNAP_USER_DATA and remove the home interface :D >> >> what do you think? > it is definitely what i would do (in a separate wrapper script), just > do not forget that you need to mkdir the subdir if it doesn't exist ... > > ciao > oli > > From nicholas.skaggs at canonical.com Thu Jul 21 18:27:11 2016 From: nicholas.skaggs at canonical.com (Nicholas Skaggs) Date: Thu, 21 Jul 2016 14:27:11 -0400 Subject: Liferea In-Reply-To: References: <1468998337.6084.16.camel@ubuntu.com> <1469026663.23385.1.camel@ubuntu.com> Message-ID: <845f4767-9859-ecdc-9f2c-393393477d06@canonical.com> /me finds a home interface and answers his own questions. http://snapcraft.io/docs/reference/interfaces Sorry for the noise. Nicholas On 07/21/2016 02:23 PM, Nicholas Skaggs wrote: > But what happens if I'm transitioning from Liferea installed a deb to > the snap -- everything I expect is in $HOME. I get we can convince the > app to look elsewhere to write files, but as a user, what happens to > my config files when I transition? Liferea isn't going to read them > afaict. How about when I uninstall the snap? Re-install my OS? > (keeping only $HOME) > > > Nicholas > > On 07/20/2016 10:57 AM, Oliver Grawert wrote: >> hi, >> Am Mittwoch, den 20.07.2016, 11:52 -0300 schrieb Alejandro Vera: >>> Liferea uses this environment variables (from man) >>> >>> $XDG_CONFIG_DIR/liferea/feedlist.opml Contains the >>> current list of subscriptions >>> $XDG_CONFIG_DIR/liferea/liferea.css Stylesheet >>> that can be used to override default HTML style >>> $XDG_DATA_DIR/liferea/liferea.db >>> >>> The first two I think are standar and the last one is from liferea... >>> >>> I think then that the right solution is to set the three of them to >>> the SNAP_USER_DATA and remove the home interface :D >>> >>> what do you think? >> it is definitely what i would do (in a separate wrapper script), just >> do not forget that you need to mkdir the subdir if it doesn't exist ... >> >> ciao >> oli >> >> > From jamie at canonical.com Thu Jul 21 21:36:22 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Thu, 21 Jul 2016 16:36:22 -0500 Subject: languagetool in lubuntu 16.04 In-Reply-To: <5783731C.7090203@ubuntu.com> References: <57832AD9.4050703@gmail.com> <5783731C.7090203@ubuntu.com> Message-ID: <1469136982.20475.89.camel@canonical.com> On Mon, 2016-07-11 at 12:21 +0200, Didier Roche wrote: > Le 11/07/2016 07:12, Vasilisc a écrit : > > > > > 09.07.2016 14:21, Mark Shuttleworth пишет: > > > > > > On 22/06/16 09:17, Alekseenko Vasilii wrote: > > > > > > > > I create virtual machines (Xubuntu, Kubuntu, Ubuntu GNOME, Lubuntu, > > > > MATE) and only Lubuntu 16.04 displays an error. > > > > > > > > URL screenshot http://itmages.ru/image/view/4456402/27529211 > > > > > > > > In Lubuntu 16.04 languagetool wants open file for write > > > > /home/vasilisc/.config/LanguageTool/gui.state > > > > > > > > BUT this impossible. > > > Alekseenko, did you figure this out? I didn't see a follow-up on the > > list. > > > > > > Mark > > > > > the problem is actual. > We discussed with jdstrand something similar with IntelliJ IDEA based > IDE family (and so, Android Studio) trying to unconditionnally open > something under /home//.IDEAxxx. Even the home plug doesn't give > you access to create that directory (as it's hidden). > > From what I got, it seems the global conscensus was to open a bugand a > PR to upstream so that they don't harcode /home/ paths, but > respect $HOME that is set by snap-confine. I'm adding Jamie so that he > can confirm I didn't get him wrong  This is the consensus, yes, and this is a good thing in general for IntelliJ IDEA to do since /home/ is not guaranteed to actually be the user's HOME (eg, nfs mounts, odd partitioning, etc). However, a related conversation came up again on IRC today about openstack hard- coding /etc for config directories and so I updated the preloading bug[1] for the most up to date info as I understand Gustavo's thoughts on it. Once this feature is implemented, it could be used to address IntelliJ IDEA and any snap where the upstream is hard-coding paths in incompatible ways. I know it is late in the week, but this would be a great 10-15 minute discussion at the sprint to get people aligned on this bug. [1] https://bugs.launchpad.net/snapcraft/+bug/1577514 -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From xiaoguo.liu at canonical.com Fri Jul 22 01:52:56 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Fri, 22 Jul 2016 09:52:56 +0800 Subject: Liferea In-Reply-To: <845f4767-9859-ecdc-9f2c-393393477d06@canonical.com> References: <1468998337.6084.16.camel@ubuntu.com> <1469026663.23385.1.camel@ubuntu.com> <845f4767-9859-ecdc-9f2c-393393477d06@canonical.com> Message-ID: I think we might need to update the table. Some new interfaces are not included on the page. For example, the network-manager is not there. The document at https://github.com/snapcore/snapd/blob/master/docs/interfaces.md is not updated as well. liuxg at liuxg:~$ snap interfaces Slot Plug :camera - :cups-control - :firewall-control - :gsettings - :home telegram-sergiusens :locale-control - :log-observe - :modem-manager - :mount-observe - :network telegram-sergiusens :network-bind telegram-sergiusens :network-control - :network-manager - :network-observe - :opengl - :optical-drive - :ppp - :pulseaudio - :snapd-control - :system-observe - :timeserver-control - :timezone-control - :unity7 telegram-sergiusens :x11 - On Fri, Jul 22, 2016 at 2:27 AM, Nicholas Skaggs < nicholas.skaggs at canonical.com> wrote: > /me finds a home interface and answers his own questions. > > http://snapcraft.io/docs/reference/interfaces > > Sorry for the noise. > > Nicholas > > > > > On 07/21/2016 02:23 PM, Nicholas Skaggs wrote: > >> But what happens if I'm transitioning from Liferea installed a deb to the >> snap -- everything I expect is in $HOME. I get we can convince the app to >> look elsewhere to write files, but as a user, what happens to my config >> files when I transition? Liferea isn't going to read them afaict. How about >> when I uninstall the snap? Re-install my OS? (keeping only $HOME) >> >> >> Nicholas >> >> On 07/20/2016 10:57 AM, Oliver Grawert wrote: >> >>> hi, >>> Am Mittwoch, den 20.07.2016, 11:52 -0300 schrieb Alejandro Vera: >>> >>>> Liferea uses this environment variables (from man) >>>> >>>> $XDG_CONFIG_DIR/liferea/feedlist.opml Contains the >>>> current list of subscriptions >>>> $XDG_CONFIG_DIR/liferea/liferea.css Stylesheet >>>> that can be used to override default HTML style >>>> $XDG_DATA_DIR/liferea/liferea.db >>>> >>>> The first two I think are standar and the last one is from liferea... >>>> >>>> I think then that the right solution is to set the three of them to >>>> the SNAP_USER_DATA and remove the home interface :D >>>> >>>> what do you think? >>>> >>> it is definitely what i would do (in a separate wrapper script), just >>> do not forget that you need to mkdir the subdir if it doesn't exist ... >>> >>> ciao >>> oli >>> >>> >>> >> > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From jamie at canonical.com Fri Jul 22 04:51:49 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Thu, 21 Jul 2016 23:51:49 -0500 Subject: Liferea In-Reply-To: References: <1468998337.6084.16.camel@ubuntu.com> <1469026663.23385.1.camel@ubuntu.com> <845f4767-9859-ecdc-9f2c-393393477d06@canonical.com> Message-ID: <1469163109.6847.1.camel@canonical.com> On Fri, 2016-07-22 at 09:52 +0800, XiaoGuo Liu wrote: > I think we might need to update the table. Some new interfaces are not > included on the page. For example, the network-manager is not there. The > document at https://github.com/snapcore/snapd/blob/master/docs/interfaces.md > is not updated as well. > This is related to https://bugs.launchpad.net/bugs/1570618 in that I have a PR for that bug that also adds all the interfaces. -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From mwinter at opensourcerouting.org Fri Jul 22 09:58:21 2016 From: mwinter at opensourcerouting.org (Martin Winter) Date: Fri, 22 Jul 2016 02:58:21 -0700 Subject: GPL 2+ license on Ubuntu Store Message-ID: I’ve noticed that the Ubunut Store for Snaps only supports GPLv2 and GPLv3 for the license selection. Is it possible to add “GPLv2 or later” to it? The project I’m working on is under such a license and they are pretty common. I do not want to limit the endusers anyway further and take the GPLv2 or v3 choice away from them. - Martin Winter From matt.aguirre at gmail.com Fri Jul 22 20:04:29 2016 From: matt.aguirre at gmail.com (Matthew Aguirre) Date: Fri, 22 Jul 2016 16:04:29 -0400 Subject: Snapcraft + Gradle and Maven In-Reply-To: References: Message-ID: <1469217869.7718.1@smtp.gmail.com> I have added a new plugin to snapcraft for any that are interested. It's the ability to build a Java project that uses gradle (more exactly, one that uses the gradlew wrapper). I have a quick write up of it here: http://tros.org/dev/2016/07/13/gradle-in-snapcraft.html with a demo example in my snapcraft fork: https://github.com/ZenHarbinger/snapcraft/tree/demo/gradle/demos/gradle Essentially, the plugin acts similarly to the maven plugin. - It builds the Java jar/war artifact - Searches for jar/war files in the 'build/libs' output directory that is generated during the 'jar' gradle task. - The output is put in the snap's 'jar' or 'war' directory depending on the type of the output. Maven: Also, I have added the ability to specify maven-targets in the maven plugin. This allows a developer to have a composite project file that references child pom.xml files. Specifying the maven-targets will copy the jar/war files from the target directories of the child projects into the snap. Example: parts: algolink-mvn: plugin: maven source: . maven-targets: - algolink-gui - algolink-cli Which allows a user to package the targets of algolink-gui and algolink-cli into the snap into directories of the target name. Thanks! -- Matt -------------- next part -------------- An HTML attachment was scrubbed... URL: From pengliu.mail at gmail.com Sat Jul 23 02:53:50 2016 From: pengliu.mail at gmail.com (Peng Liu) Date: Fri, 22 Jul 2016 21:53:50 -0500 Subject: Need systemctl to run ubuntu-device-flash? Message-ID: Hi Everyone, I am trying to build the snappy ubuntu image with the ubuntu-device-flash command downloaded from https://people.canonical.com/~mvo/all-snaps/ Unfortunately, it reports error: #need '/bin/systemctl to work I am using Ubuntu 14.04 LTS so systemd is not installed by default. Do I need to upgrade to 15.04 or newer version in order to build a snappy ubuntu image? Thanks. Peng -------------- next part -------------- An HTML attachment was scrubbed... URL: From xiaoguo.liu at canonical.com Mon Jul 25 01:11:33 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 25 Jul 2016 09:11:33 +0800 Subject: A question about using --devmode Message-ID: Hi, If I use "--devmode" when installing a snap application, do I have to define "confinement" in snapcraft.yaml to be ""devmode" exactly. What happens if I defined "confinement" to be "strict"? The combination is as follows: - confinement: strict - installation: --devmode - confinement: devmode - installation: --devmode Thanks & best regards, XiaoGuo -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From michael.nelson at canonical.com Mon Jul 25 01:20:55 2016 From: michael.nelson at canonical.com (Michael Nelson) Date: Mon, 25 Jul 2016 01:20:55 +0000 Subject: A question about using --devmode In-Reply-To: References: Message-ID: On Mon, Jul 25, 2016 at 11:12 AM XiaoGuo Liu wrote: > Hi, > > If I use "--devmode" when installing a snap application, do I have to > define "confinement" in snapcraft.yaml to be ""devmode" exactly. What > happens if I defined "confinement" to be "strict"? > > The combination is as follows: > > - confinement: strict > - installation: --devmode > Using confinement: strict tells the store that it's OK to return this package as a result of `snap find` for users to install. Which means, if it doesn't work unless you use --devmode, you might have unhappy users :/ > > - confinement: devmode > - installation: --devmode > Using confinement: devmode tells the store that this package should *not* be included in results of `snap find`. You (and any users who know) will still be able to `snap install --devmode` it, but it's not provided to general users via snap find/install. Hope that helps > > Thanks & best regards, > XiaoGuo > > -- > XiaoGuo, Liu (刘晓国） > Mobile: +86-13911181302 > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From xiaoguo.liu at canonical.com Mon Jul 25 01:46:23 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 25 Jul 2016 09:46:23 +0800 Subject: A question about using --devmode In-Reply-To: References: Message-ID: Hi Michael, Thanks for your explanation. Now, I get the idea of it though it is a little bit confusing initially. Best regards, XiaoGuo On Mon, Jul 25, 2016 at 9:20 AM, Michael Nelson < michael.nelson at canonical.com> wrote: > On Mon, Jul 25, 2016 at 11:12 AM XiaoGuo Liu > wrote: > >> Hi, >> >> If I use "--devmode" when installing a snap application, do I have to >> define "confinement" in snapcraft.yaml to be ""devmode" exactly. What >> happens if I defined "confinement" to be "strict"? >> >> The combination is as follows: >> >> - confinement: strict >> - installation: --devmode >> > > Using confinement: strict tells the store that it's OK to return this > package as a result of `snap find` for users to install. Which means, if it > doesn't work unless you use --devmode, you might have unhappy users :/ > > > >> >> - confinement: devmode >> - installation: --devmode >> > > Using confinement: devmode tells the store that this package should *not* > be included in results of `snap find`. You (and any users who know) will > still be able to `snap install --devmode` it, but it's not provided to > general users via snap find/install. > > Hope that helps > >> >> Thanks & best regards, >> XiaoGuo >> >> -- >> XiaoGuo, Liu (刘晓国） >> Mobile: +86-13911181302 >> -- >> Snapcraft mailing list >> Snapcraft at lists.ubuntu.com >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/snapcraft >> > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From didrocks at ubuntu.com Mon Jul 25 06:12:10 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Mon, 25 Jul 2016 08:12:10 +0200 Subject: Need systemctl to run ubuntu-device-flash? In-Reply-To: References: Message-ID: <5795ADBA.7010705@ubuntu.com> Le 23/07/2016 04:53, Peng Liu a écrit : > Hi Everyone, > > I am trying to build the snappy ubuntu image with the > ubuntu-device-flash command downloaded from > https://people.canonical.com/~mvo/all-snaps/ > > > Unfortunately, it reports error: > #need '/bin/systemctl to work > > I am using Ubuntu 14.04 LTS so systemd is not installed by default. Do > I need to upgrade to 15.04 or newer version in order to build a snappy > ubuntu image? > Hey Peng, As ubuntu-device-flash (soon to be renamed IIRC) and other tools are using some plateform features to create new images, indeed, you really need to be on the ubuntu version you are targetting to build the image as of now. I would recommend you using 16.04 LTS desktop consequently. I don't know if we have any up to date instructions on how to build an image, but maybe Oliver (CCed) will have some guidance? Cheers, Didier From xiaoguo.liu at canonical.com Mon Jul 25 07:28:55 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 25 Jul 2016 15:28:55 +0800 Subject: Revert command is not supported Message-ID: Hi， I just found the following command from http://snapcraft.io/ $ sudo snap revert hello-world Name Version Rev Developer Notes hello-world 6.0 25 canonical - -- When I tried to run the above command, it turned to be that "revert" is not a valid command. What is the correct command to revert an installation? Thanks & best regards, XiaoGuo XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From gustavo.niemeyer at canonical.com Mon Jul 25 07:40:12 2016 From: gustavo.niemeyer at canonical.com (Gustavo Niemeyer) Date: Mon, 25 Jul 2016 09:40:12 +0200 Subject: Revert command is not supported In-Reply-To: References: Message-ID: Oops, that was a mistake on our side, sorry. The revert command is landing on snapd 2.11, which should be out later this week. On Jul 25, 2016 9:30 AM, "XiaoGuo Liu" wrote: > Hi， > > I just found the following command from http://snapcraft.io/ > > $ sudo snap revert hello-world > Name Version Rev Developer Notes > hello-world 6.0 25 canonical - > > > -- > > > > > When I tried to run the above command, it turned to be that "revert" is > not a valid command. What is the correct command to revert an installation? > > Thanks & best regards, > XiaoGuo > > XiaoGuo, Liu (刘晓国） > Mobile: +86-13911181302 > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Mon Jul 25 07:52:42 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Mon, 25 Jul 2016 09:52:42 +0200 Subject: Need systemctl to run ubuntu-device-flash? In-Reply-To: References: Message-ID: <1469433162.12092.2.camel@anubis> hi, Am Freitag, den 22.07.2016, 21:53 -0500 schrieb Peng Liu: > Hi Everyone, > > > I am trying to build the snappy ubuntu image with the > ubuntu-device-flash command downloaded from > https://people.canonical.com/~mvo/all-snaps/ > > > > Unfortunately, it reports error: > #need '/bin/systemctl to work > > > > I am using Ubuntu 14.04 LTS so systemd is not installed by default. Do > I need to upgrade to 15.04 or newer version in order to build a snappy > ubuntu image? can you put the full log on paste.ubuntu.com ? i havent used it on 14.04 for a while (due to other issues with kpartx on my 14.04 machine), it definitely has not needed systemd support in the past when i did though... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From ogra at ubuntu.com Mon Jul 25 07:57:12 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Mon, 25 Jul 2016 09:57:12 +0200 Subject: A question about using --devmode In-Reply-To: References: Message-ID: <1469433432.12092.6.camel@anubis> hi, Am Montag, den 25.07.2016, 09:11 +0800 schrieb XiaoGuo Liu: > Hi, > > > If I use "--devmode" when installing a snap application, do I have to > define "confinement" in snapcraft.yaml to be ""devmode" exactly. > What happens if I defined "confinement" to be "strict"? > only packages with "confinement: strict" are allowed into the stable channel. devmode snaps can only go to edge, beta and candiadate and as michael pointed out "snap find" currently only searches the stable channel (i belive this is about to get a switch for searching non-stable channels though). ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From xiaoguo.liu at canonical.com Mon Jul 25 08:51:23 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 25 Jul 2016 16:51:23 +0800 Subject: A question about using --devmode In-Reply-To: <1469433432.12092.6.camel@anubis> References: <1469433432.12092.6.camel@anubis> Message-ID: Hi Oliver, Thanks for your reply. What happens if we do the following: - confinement: devmode - installation: do not use "--devmode" Will this be OK? Will the app be running in confinement? Thanks & best regards, XiaoGuo On Mon, Jul 25, 2016 at 3:57 PM, Oliver Grawert wrote: > hi, > Am Montag, den 25.07.2016, 09:11 +0800 schrieb XiaoGuo Liu: > > Hi, > > > > > > If I use "--devmode" when installing a snap application, do I have to > > define "confinement" in snapcraft.yaml to be ""devmode" exactly. > > What happens if I defined "confinement" to be "strict"? > > > only packages with "confinement: strict" are allowed into the stable > channel. devmode snaps can only go to edge, beta and candiadate and as > michael pointed out "snap find" currently only searches the stable > channel (i belive this is about to get a switch for searching non-stable > channels though). > > ciao > oli > > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.lenton at canonical.com Mon Jul 25 08:55:11 2016 From: john.lenton at canonical.com (John Lenton) Date: Mon, 25 Jul 2016 10:55:11 +0200 Subject: A question about using --devmode In-Reply-To: References: Message-ID: On 25 July 2016 at 03:11, XiaoGuo Liu wrote: > If I use "--devmode" when installing a snap application, do I have to define > "confinement" in snapcraft.yaml to be ""devmode" exactly. What happens if I > defined "confinement" to be "strict"? while --devmode is required for snaps that have confinement:devmode, it can also be specified for confinement:strict, in which case you'll have a non-confined (and soon non-auto-updating) snap that could be confined. Useful if you're seeing weird behaviour in the snap and wondering if it's because of a confinement bug, but not for much otherwise. The reverse is not true; if you have confinement:devmode and you don't specify --devmode (or, in 2.11, override it with --jailmode), you'll get an error telling you as much. HTH, From john.lenton at canonical.com Mon Jul 25 08:57:51 2016 From: john.lenton at canonical.com (John Lenton) Date: Mon, 25 Jul 2016 10:57:51 +0200 Subject: A question about using --devmode In-Reply-To: References: <1469433432.12092.6.camel@anubis> Message-ID: On 25 July 2016 at 10:51, XiaoGuo Liu wrote: > Hi Oliver, > > Thanks for your reply. What happens if we do the following: > > - confinement: devmode > - installation: do not use "--devmode" > > Will this be OK? No, this will abort with an error. > Will the app be running in confinement? No. In 2.11 (coming soon!) you'll be able to specify --jailmode which would do this, but will (in 2.11) break on refresh. That is, auto-refresh of a --jailmode will fail. That particular case will be addressed down the road a bit. From xiaoguo.liu at canonical.com Mon Jul 25 08:59:55 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 25 Jul 2016 16:59:55 +0800 Subject: A question about using --devmode In-Reply-To: References: Message-ID: Hi John, Thanks. Just now, I tried with confinement: devmode, but I could still install without "--devmode". It does not complain anything there :). The result was that the app ran in confined mode. Thanks & best regards, XiaoGuo On Mon, Jul 25, 2016 at 4:55 PM, John Lenton wrote: > On 25 July 2016 at 03:11, XiaoGuo Liu wrote: > > If I use "--devmode" when installing a snap application, do I have to > define > > "confinement" in snapcraft.yaml to be ""devmode" exactly. What happens > if I > > defined "confinement" to be "strict"? > > while --devmode is required for snaps that have confinement:devmode, > it can also be specified for confinement:strict, in which case you'll > have a non-confined (and soon non-auto-updating) snap that could be > confined. Useful if you're seeing weird behaviour in the snap and > wondering if it's because of a confinement bug, but not for much > otherwise. > > The reverse is not true; if you have confinement:devmode and you don't > specify --devmode (or, in 2.11, override it with --jailmode), you'll > get an error telling you as much. > > HTH, > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.lenton at canonical.com Mon Jul 25 09:06:51 2016 From: john.lenton at canonical.com (John Lenton) Date: Mon, 25 Jul 2016 11:06:51 +0200 Subject: A question about using --devmode In-Reply-To: References: Message-ID: On 25 July 2016 at 10:59, XiaoGuo Liu wrote: > Thanks. Just now, I tried with confinement: devmode, but I could still > install without "--devmode". It does not complain anything there :). The > result was that the app ran in confined mode. Oooh! that's a bug, I think :-) From john.lenton at canonical.com Mon Jul 25 09:17:21 2016 From: john.lenton at canonical.com (John Lenton) Date: Mon, 25 Jul 2016 11:17:21 +0200 Subject: A question about using --devmode In-Reply-To: References: Message-ID: On 25 July 2016 at 11:06, John Lenton wrote: > Oooh! that's a bug, I think :-) just to confirm, this is currently fixed in master and will be in 2.11. If you sideloaded a snap that specifies confinement:devmode, it'll get picked up by 2.11 as if you had specified --jailmode, and not specifying that or --devmode is an error, as expected. \o/ From xiaoguo.liu at canonical.com Mon Jul 25 14:00:28 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Mon, 25 Jul 2016 22:00:28 +0800 Subject: A question about using --devmode In-Reply-To: References: Message-ID: Hi John, Thanks for your confirmation. Then I will be waiting for that. Best regards, XiaoGuo On Mon, Jul 25, 2016 at 5:17 PM, John Lenton wrote: > On 25 July 2016 at 11:06, John Lenton wrote: > > Oooh! that's a bug, I think :-) > > just to confirm, this is currently fixed in master and will be in > 2.11. If you sideloaded a snap that specifies confinement:devmode, > it'll get picked up by 2.11 as if you had specified --jailmode, and > not specifying that or --devmode is an error, as expected. \o/ > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From leo.arias at canonical.com Mon Jul 25 18:07:30 2016 From: leo.arias at canonical.com (Leo Arias) Date: Mon, 25 Jul 2016 12:07:30 -0600 Subject: locale In-Reply-To: <578E012B.30901@gmail.com> References: <578E012B.30901@gmail.com> Message-ID: <57965562.8000301@canonical.com> Hello! Can you please report a bug in https://bugs.launchpad.net/snapcraft ? Remember to attach your snapcraft.yaml and any other relevant files so we can reproduce it. Thanks! -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: OpenPGP digital signature URL: From pengliu.mail at gmail.com Mon Jul 25 22:48:05 2016 From: pengliu.mail at gmail.com (Peng Liu) Date: Mon, 25 Jul 2016 17:48:05 -0500 Subject: Need systemctl to run ubuntu-device-flash? In-Reply-To: <5795ADBA.7010705@ubuntu.com> References: <5795ADBA.7010705@ubuntu.com> Message-ID: Hi Didier and Oliver, Thanks for the information. I guess the reason must be that I am using the ubuntu-device-flash download from https://people.canonical.com/~mvo/all-snaps/ubuntu-device-flash By the way I am trying to build the ubuntu-core image version 16.04. Thanks again. Peng On Mon, Jul 25, 2016 at 1:12 AM, Didier Roche wrote: > Le 23/07/2016 04:53, Peng Liu a écrit : > > Hi Everyone, > > > > I am trying to build the snappy ubuntu image with the > > ubuntu-device-flash command downloaded from > > https://people.canonical.com/~mvo/all-snaps/ > > > > > > Unfortunately, it reports error: > > #need '/bin/systemctl to work > > > > I am using Ubuntu 14.04 LTS so systemd is not installed by default. Do > > I need to upgrade to 15.04 or newer version in order to build a snappy > > ubuntu image? > > > Hey Peng, > > As ubuntu-device-flash (soon to be renamed IIRC) and other tools are > using some plateform features to create new images, indeed, you really > need to be on the ubuntu version you are targetting to build the image > as of now. > I would recommend you using 16.04 LTS desktop consequently. I don't know > if we have any up to date instructions on how to build an image, but > maybe Oliver (CCed) will have some guidance? > > Cheers, > Didier > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From pengliu.mail at gmail.com Tue Jul 26 02:40:44 2016 From: pengliu.mail at gmail.com (Peng Liu) Date: Mon, 25 Jul 2016 21:40:44 -0500 Subject: Which version of "snappy" support snap.yaml? Message-ID: Hi All, I am confused about the tools to build snap packages. There are three different flavors as far I know: package.yaml, snap.yaml, and snapcraft.yaml. The snap.yaml was landed at Jan 2016, and I think snapcraft.yaml is the new version (2.x?). Which tool (snappy, snapcraft) and which version should I use to build a snap with snap.yaml? Is there any document explaining that? Thanks. Peng -------------- next part -------------- An HTML attachment was scrubbed... URL: From xiaoguo.liu at canonical.com Tue Jul 26 04:08:40 2016 From: xiaoguo.liu at canonical.com (XiaoGuo Liu) Date: Tue, 26 Jul 2016 12:08:40 +0800 Subject: Which version of "snappy" support snap.yaml? In-Reply-To: References: Message-ID: as far as I know, snapcraft.yaml is the only one recommended for building a snap project. snap.yaml is generated during the build of a snap project. It can be found in the "meta" directory of "prime". package.yaml was used in the early days of building a snap application. It was used in a form like "snappy build". On Tue, Jul 26, 2016 at 10:40 AM, Peng Liu wrote: > Hi All, > > I am confused about the tools to build snap packages. > There are three different flavors as far I know: package.yaml, snap.yaml, > and snapcraft.yaml. > > The snap.yaml was landed at Jan 2016, and I think snapcraft.yaml is the > new version (2.x?). Which tool (snappy, snapcraft) and which version should > I use to build a snap with snap.yaml? > > Is there any document explaining that? > > Thanks. > > Peng > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- XiaoGuo, Liu (刘晓国） Mobile: +86-13911181302 -------------- next part -------------- An HTML attachment was scrubbed... URL: From leo.arias at canonical.com Tue Jul 26 04:36:29 2016 From: leo.arias at canonical.com (Leo Arias) Date: Mon, 25 Jul 2016 22:36:29 -0600 Subject: git In-Reply-To: References: Message-ID: <5796E8CD.20301@canonical.com> Hello! On 2016-07-20 20:21, Alejandro Vera wrote: > Hi guys > > I am trying to snap another app. This time is git :D > > Buy my problem is that git uses ssh when cloning. But ssh wants to add a > key to my known_hosts and loading my key. The public repositories of git servers also provide an https address for anonymous cloning. For example: https://github.com/snapcore/snapcraft/blob/master/demos/godd/snapcraft.yaml#L16 pura vida. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: OpenPGP digital signature URL: From gustavo.niemeyer at canonical.com Tue Jul 26 04:56:37 2016 From: gustavo.niemeyer at canonical.com (Gustavo Niemeyer) Date: Tue, 26 Jul 2016 06:56:37 +0200 Subject: git In-Reply-To: References: Message-ID: Hi Alejandro, We need a new interface specifically for accessing the ssh keys. Should be fairly easy to put it in place. How's your git snap going? On Thu, Jul 21, 2016 at 4:21 AM, Alejandro Vera wrote: > Hi guys > > I am trying to snap another app. This time is git :D > > Buy my problem is that git uses ssh when cloning. But ssh wants to add a > key to my known_hosts and loading my key. > > I know that snapcraft blocks snaps to read and write folders with '.' in > the begining. > > Soo... some advice? > > thanks in advance!! :D > > > > -- > Alejandro Vera > http://www.recicleta.cl > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- gustavo @ http://niemeyer.net -------------- next part -------------- An HTML attachment was scrubbed... URL: From sergio.schvezov at canonical.com Tue Jul 26 06:58:32 2016 From: sergio.schvezov at canonical.com (Sergio Schvezov) Date: Tue, 26 Jul 2016 08:58:32 +0200 Subject: Which version of "snappy" support snap.yaml? In-Reply-To: References: Message-ID: <1dfcfacb-1819-9f63-fa97-5d24c08f5acf@ubuntu.com> El 26/07/16 a las 04:40, Peng Liu escribió: > Hi All, > > I am confused about the tools to build snap packages. > There are three different flavors as far I know: package.yaml, > snap.yaml, and snapcraft.yaml. > > The snap.yaml was landed at Jan 2016, and I think snapcraft.yaml is the > new version (2.x?). Which tool (snappy, snapcraft) and which version > should I use to build a snap with snap.yaml? > > Is there any document explaining that? package.yaml is supported only on 15.04 snap.yaml is the new format introduced in series 16. snapcraft.yaml is used by snapcraft and generates such snap.yaml when run on a series 16 supported system (likely xenial) with snapcraft 2.x and a valid package.yaml when run on a 15.04 supported system with snapcraft 1.x. Cheers Sergio -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: OpenPGP digital signature URL: From fcole90 at gmail.com Tue Jul 26 08:26:24 2016 From: fcole90 at gmail.com (Fabio Colella) Date: Tue, 26 Jul 2016 10:26:24 +0200 Subject: git In-Reply-To: References: Message-ID: How can we create new interfaces? I would like to make one to access .icons and .theme folders. On 26 July 2016 at 06:56, Gustavo Niemeyer wrote: > Hi Alejandro, > > We need a new interface specifically for accessing the ssh keys. > > Should be fairly easy to put it in place. > > How's your git snap going? > > On Thu, Jul 21, 2016 at 4:21 AM, Alejandro Vera > wrote: > >> Hi guys >> >> I am trying to snap another app. This time is git :D >> >> Buy my problem is that git uses ssh when cloning. But ssh wants to add a >> key to my known_hosts and loading my key. >> >> I know that snapcraft blocks snaps to read and write folders with '.' in >> the begining. >> >> Soo... some advice? >> >> thanks in advance!! :D >> >> >> >> -- >> Alejandro Vera >> http://www.recicleta.cl >> >> -- >> Snapcraft mailing list >> Snapcraft at lists.ubuntu.com >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/snapcraft >> >> > > > -- > gustavo @ http://niemeyer.net > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From zygmunt.krynicki at canonical.com Tue Jul 26 08:41:08 2016 From: zygmunt.krynicki at canonical.com (Zygmunt Krynicki) Date: Tue, 26 Jul 2016 10:41:08 +0200 Subject: git In-Reply-To: References: Message-ID: <6615E957-1C75-4B89-959E-6AFD45A3924C@canonical.com> > Wiadomość napisana przez Fabio Colella w dniu 26.07.2016, o godz. 10:26: > > How can we create new interfaces? I would like to make one to access .icons and .theme folders. Please get started with http://www.zygoon.pl/2016/04/anatomy-of-snappy-interface.html I am working on a new article that dives deeper into the process but this should help you get started Best regards ZK From fcole90 at gmail.com Tue Jul 26 08:47:34 2016 From: fcole90 at gmail.com (Fabio Colella) Date: Tue, 26 Jul 2016 10:47:34 +0200 Subject: git In-Reply-To: <6615E957-1C75-4B89-959E-6AFD45A3924C@canonical.com> References: <6615E957-1C75-4B89-959E-6AFD45A3924C@canonical.com> Message-ID: Thank you very much On 26 July 2016 at 10:41, Zygmunt Krynicki wrote: > > > Wiadomość napisana przez Fabio Colella w dniu > 26.07.2016, o godz. 10:26: > > > > How can we create new interfaces? I would like to make one to access > .icons and .theme folders. > > Please get started with > http://www.zygoon.pl/2016/04/anatomy-of-snappy-interface.html > > I am working on a new article that dives deeper into the process but this > should help you get started > > Best regards > ZK -------------- next part -------------- An HTML attachment was scrubbed... URL: From vasilisc777 at gmail.com Tue Jul 26 11:08:52 2016 From: vasilisc777 at gmail.com (Vasilisc) Date: Tue, 26 Jul 2016 14:08:52 +0300 Subject: qt5 and external urls Message-ID: <4cbe3283-ba8b-1770-801b-d9b809de650d@gmail.com> qt5-app wants to open a URL. virtual bool QGenericUnixServices::openUrl(const QUrl&): Unable to detect a web browser to launch What shall I add to the snapcraft.yaml to fix a problem? -- Best regards, vasilisc From didrocks at ubuntu.com Tue Jul 26 13:11:22 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Tue, 26 Jul 2016 15:11:22 +0200 Subject: qt5 and external urls In-Reply-To: <4cbe3283-ba8b-1770-801b-d9b809de650d@gmail.com> References: <4cbe3283-ba8b-1770-801b-d9b809de650d@gmail.com> Message-ID: <5797617A.6010703@ubuntu.com> Le 26/07/2016 13:08, Vasilisc a écrit : > qt5-app wants to open a URL. > > virtual bool QGenericUnixServices::openUrl(const QUrl&): Unable to > detect a web browser to launch > > What shall I add to the snapcraft.yaml to fix a problem? > Hey, Michael, I thought we did ship a fake xdg-open in ubuntu core after the Vancouver sprint so that Qt apps or electron-based one like telegram were able to open urls on the system, is the issue the fact that the ubuntu core snap with that functionality didn't ship yet? Cheers, Didier From matt.aguirre at gmail.com Tue Jul 26 13:15:31 2016 From: matt.aguirre at gmail.com (Matthew Aguirre) Date: Tue, 26 Jul 2016 09:15:31 -0400 Subject: qt5 and external urls In-Reply-To: <5797617A.6010703@ubuntu.com> References: <4cbe3283-ba8b-1770-801b-d9b809de650d@gmail.com> <5797617A.6010703@ubuntu.com> Message-ID: <1469538931.5366.0@smtp.gmail.com> Is there a way that this will work with Java too? Using the: Desktop.getDesktop().browse(uri); I figured it was a containment thing, but if there is an xdg-open for some apps, will this extend to Java somehow? -- Matt On Tue, Jul 26, 2016 at 9:11 AM, Didier Roche wrote: > Le 26/07/2016 13:08, Vasilisc a écrit : >> qt5-app wants to open a URL. >> >> virtual bool QGenericUnixServices::openUrl(const QUrl&): Unable to >> detect a web browser to launch >> >> What shall I add to the snapcraft.yaml to fix a problem? >> > > Hey, > > Michael, I thought we did ship a fake xdg-open in ubuntu core after > the > Vancouver sprint so that Qt apps or electron-based one like telegram > were able to open urls on the system, is the issue the fact that the > ubuntu core snap with that functionality didn't ship yet? > > Cheers, > Didier > > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft -------------- next part -------------- An HTML attachment was scrubbed... URL: From didrocks at ubuntu.com Tue Jul 26 13:50:44 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Tue, 26 Jul 2016 15:50:44 +0200 Subject: qt5 and external urls In-Reply-To: <1469538931.5366.0@smtp.gmail.com> References: <4cbe3283-ba8b-1770-801b-d9b809de650d@gmail.com> <5797617A.6010703@ubuntu.com> <1469538931.5366.0@smtp.gmail.com> Message-ID: <57976AB4.90908@ubuntu.com> Le 26/07/2016 15:15, Matthew Aguirre a écrit : > Is there a way that this will work with Java too? > Using the: > Desktop.getDesktop().browse(uri); > > I figured it was a containment thing, but if there is an xdg-open for > some apps, will this extend to Java somehow? I guess that depends on what .browser for getDesktop() is using internally to bind with the system. If you can identify the tool if not xdg-open, we can certainly have a look (I'm not a Java expert) Cheers, Didier From matt.aguirre at gmail.com Tue Jul 26 13:53:16 2016 From: matt.aguirre at gmail.com (Matthew Aguirre) Date: Tue, 26 Jul 2016 09:53:16 -0400 Subject: qt5 and external urls In-Reply-To: <57976AB4.90908@ubuntu.com> References: <4cbe3283-ba8b-1770-801b-d9b809de650d@gmail.com> <5797617A.6010703@ubuntu.com> <1469538931.5366.0@smtp.gmail.com> <57976AB4.90908@ubuntu.com> Message-ID: <1469541196.5366.1@smtp.gmail.com> I get it. I was just assuming that this would not be possible ever. Perhaps I'll look at invoking xdg-open if browse() fails. Java claims it's unavailable on the current platform when run within the snap. -- Matt On Tue, Jul 26, 2016 at 9:50 AM, Didier Roche wrote: > Le 26/07/2016 15:15, Matthew Aguirre a écrit : >> Is there a way that this will work with Java too? >> Using the: >> Desktop.getDesktop().browse(uri); >> >> I figured it was a containment thing, but if there is an xdg-open >> for >> some apps, will this extend to Java somehow? > > I guess that depends on what .browser for getDesktop() is using > internally to bind with the system. If you can identify the tool if > not > xdg-open, we can certainly have a look (I'm not a Java expert) > > Cheers, > Didier -------------- next part -------------- An HTML attachment was scrubbed... URL: From alejandro.vera at gmail.com Tue Jul 26 14:13:45 2016 From: alejandro.vera at gmail.com (Alejandro Vera) Date: Tue, 26 Jul 2016 11:13:45 -0300 Subject: git In-Reply-To: References: <6615E957-1C75-4B89-959E-6AFD45A3924C@canonical.com> Message-ID: Hi guys! Thanks for the answers :D. And yes, an interface for ssh keys would be great. Is it planned? Leo Arias: Yes I know about the repositories, but I wanted to use a stable release. I do not know how to configure the git plugin to use a specific tag (2.9.2 for example). Thats why I use a .tgz Gustavo: My snap works. I haven't tested it a lot since last week beacouse I was waiting for an answer to this email. I was wainting to see if using keys is posible (Where are you from? I am from Chile) On Tue, Jul 26, 2016 at 4:47 AM, Fabio Colella wrote: > Thank you very much > > On 26 July 2016 at 10:41, Zygmunt Krynicki > wrote: > >> >> > Wiadomość napisana przez Fabio Colella w dniu >> 26.07.2016, o godz. 10:26: >> > >> > How can we create new interfaces? I would like to make one to access >> .icons and .theme folders. >> >> Please get started with >> http://www.zygoon.pl/2016/04/anatomy-of-snappy-interface.html >> >> I am working on a new article that dives deeper into the process but this >> should help you get started >> >> Best regards >> ZK > > > > -- > Snapcraft mailing list > Snapcraft at lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- Alejandro Vera http://www.recicleta.cl -------------- next part -------------- An HTML attachment was scrubbed... URL: From pengliu.mail at gmail.com Tue Jul 26 17:35:01 2016 From: pengliu.mail at gmail.com (Peng Liu) Date: Tue, 26 Jul 2016 12:35:01 -0500 Subject: How to sign a snap package? Message-ID: Hi All, Is there any tool we can use to sign a snap package? Thanks. Peng -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Tue Jul 26 17:43:10 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Tue, 26 Jul 2016 19:43:10 +0200 Subject: How to sign a snap =?iso-8859-1?Q?package=3F?= In-Reply-To: Message-ID: <5b228c15-cf17-4c26-bd0e-8c587993986b@ubuntu.com> hi, Am Dienstag, 26. Juli 2016 19:35:01 CEST schrieb Peng Liu : > Hi All, > > Is there any tool we can use to sign a snap package? i think we call that tool the store ;) ciao oli -- Mit Dekko von meinem Ubuntu-Gerät gesendet From me at zygoon.pl Sun Jul 24 05:57:07 2016 From: me at zygoon.pl (Zygmunt Krynicki) Date: Sun, 24 Jul 2016 07:57:07 +0200 Subject: snapd crashes on fedora In-Reply-To: References: Message-ID: <41218916-0B87-47BA-8850-8E9E4FBD1826@zygoon.pl> > Wiadomość napisana przez Adam Kurkiewicz w dniu 24.07.2016, o godz. 00:30: > > Hey guys, I've followed the instructions to install snapd on fedora 23: > > http://snapcraft.io/docs/core/install > > The installation almost worked (I just had to replace a dash with a > hyphen. It's probably a good idea to remove it: it is between the > words "enable" and "now") and I've got a snapd installed on my system. > > However, it would't even start because (as journalctl --unit=snapd told me): > > Jul 23 18:45:24 jordan snapd[10645]: error: fatal: directory > "/var/lib/snapd" must be present Thanks for reporting this. I think this directory is usually included in the packaging (or at least has been earlier). > > I had to go ahead and manually > > sudo mkdir /var/lib/snapd > > which at least allowed me to start the deamon. > > I can three ways to fix it: > > 1) add my workaround to the official instructions > 2) fix snapd to automatically create /var/lib/snapd or > 3) fix zyga/ubuntucore on corp to include this directory (maybe put > some empty file there if one can't create empty directories with corp, > I don't know). Let’s do this ASAP so that the user experience is not impaired. While this is addressed we can look for a more permanent and universal solution. Best regards ZK From peter.chen at lemaker.com Fri Jul 1 01:59:17 2016 From: peter.chen at lemaker.com (peter.chen at lemaker.com) Date: Fri, 01 Jul 2016 01:59:17 -0000 Subject: Automated review Failed Message-ID: <2016070109584975185913@lemaker.com> Hi, All I want to submit gadget for the lemaker guitar to the snap store, but i have some trouble, as shown in the figure below. I'm very sorry, i don't know how to solve, and hope you can give me some suggestions, and the snap package as attached. Best Regards Peter Chen LeMaker Developer Team Making Innovation Easy. Email: support at lemaker.org (Technical Support) product at lemaker.org (Product Distribution) Website: http://www.lemaker.org/ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Catch.jpg Type: image/jpeg Size: 53600 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: lemaker-guitar-gadget_0.1_armhf.snap Type: application/octet-stream Size: 192512 bytes Desc: not available URL: From tinchester at gmail.com Sat Jul 23 00:41:48 2016 From: tinchester at gmail.com (=?UTF-8?Q?Tin_Tvrtkovi=C4=87?=) Date: Sat, 23 Jul 2016 00:41:48 +0000 Subject: Bind to D-Bus from a snap Message-ID: Hello, for a few days now I've been trying to build a snap for Terminix. I'm pretty much done, except when I actually run it, I get: Failed to register: GDBus.Error:org.freedesktop.DBus.Error.AccessDenied: Connection ":1.194" is not allowed to own the service "com.gexperts.Terminix" due to AppArmor policy Is there something in my snapcraft.yml I should be doing differently? -------------- next part -------------- An HTML attachment was scrubbed... URL: From popantrop at gmail.com Sat Jul 23 22:30:10 2016 From: popantrop at gmail.com (Adam Kurkiewicz) Date: Sat, 23 Jul 2016 23:30:10 +0100 Subject: snapd crashes on fedora Message-ID: Hey guys, I've followed the instructions to install snapd on fedora 23: http://snapcraft.io/docs/core/install The installation almost worked (I just had to replace a dash with a hyphen. It's probably a good idea to remove it: it is between the words "enable" and "now") and I've got a snapd installed on my system. However, it would't even start because (as journalctl --unit=snapd told me): Jul 23 18:45:24 jordan snapd[10645]: error: fatal: directory "/var/lib/snapd" must be present I had to go ahead and manually sudo mkdir /var/lib/snapd which at least allowed me to start the deamon. I can three ways to fix it: 1) add my workaround to the official instructions 2) fix snapd to automatically create /var/lib/snapd or 3) fix zyga/ubuntucore on corp to include this directory (maybe put some empty file there if one can't create empty directories with corp, I don't know). Let me know, Adam From peter.chen at lemaker.com Fri Jul 1 01:59:51 2016 From: peter.chen at lemaker.com (peter.chen at lemaker.com) Date: Fri, 01 Jul 2016 01:59:51 -0000 Subject: snap package automated review Failed Message-ID: <2016070109584975185913@lemaker.com> Hi, All I want to submit gadget for the lemaker guitar to the snap store, but i have some trouble, as shown in the figure below. I'm very sorry, i don't know how to solve, and hope you can give me some suggestions, and the snap package as attached. Best Regards Peter Chen LeMaker Developer Team Making Innovation Easy. Email: support at lemaker.org (Technical Support) product at lemaker.org (Product Distribution) Website: http://www.lemaker.org/ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Catch(07-01-09-59-02).jpg Type: image/jpeg Size: 53600 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: lemaker-guitar-gadget_0.1_armhf.snap Type: application/octet-stream Size: 192512 bytes Desc: not available URL: From mail at radeksprta.eu Wed Jul 20 20:51:34 2016 From: mail at radeksprta.eu (Radek Sprta) Date: Wed, 20 Jul 2016 16:51:34 -0400 Subject: Trying to package angband as snap Message-ID: Hey, I tried packaging Angband as as snap. This is what I have as snapcraft.yaml: name: angband version: "4.0.5" summary: Single-player, text-based, dungeon simulation game confinement: devmode apps: angband: command: games/angband plug: [x11, unity7, home, pulseaudio] parts: angband: plugin: autotools configflags: ["--enable-sdl"] source: http://rephial.org/downloads/4.0/angband-4.0.5.tar.gz build-packages: - libsdl1.2-dev - libsdl-ttf2.0-dev - libsdl-mixer1.2-dev - libsdl-image1.2-dev The game gets packaged, but fails to run, because it looks for the files in /share/games/angband instead of $SNAP/share/games/angband I tried packaging it with '--no-install' and 'prefix=$SNAP' options (changing command to angband), but both result in: /snap/angband/100001/command-angband.wrapper: 5: exec: angband: Permission denied The snappy-debug gives suggestions such as: * adjust program to read necessary files from $SNAP, $SNAP_DATA or $SNAP_USER_DATA and * adjust program to execute binaries directly from $SNAP instead of /snap/bin Do I just fail at snapcraft or does the game really need to be patched? Best regards, Radek -------------- next part -------------- An HTML attachment was scrubbed... URL: From jamie at canonical.com Tue Jul 26 18:21:42 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Tue, 26 Jul 2016 13:21:42 -0500 Subject: snap package automated review Failed In-Reply-To: <2016070109584975185913@lemaker.com> References: <2016070109584975185913@lemaker.com> Message-ID: <1469557302.5406.18.camel@canonical.com> On Fri, 2016-07-01 at 09:59 +0800, peter.chen at lemaker.com wrote: > Hi, All > I want to submit gadget for the lemaker guitar to the snap store, but i have > some trouble, as shown in the figure below.  > I'm very sorry, i don't know how to solve, and hope you can give me some > suggestions, and the snap package as attached. > I don't see this up for manual review in the store. Can you request a manual review? Gadget snaps are being redone as is the store approval process for them. Eventually assertions will be in place and specifying a gadget snap won't trigger a manual review. While that is being worked on I've asked in another thread how gadget snaps should be handle in the interim. In the future I suggest asking on IRC in #snappy on Freenode if someone could perform a manual review and you should get a quick response. -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From jamie at canonical.com Tue Jul 26 18:27:33 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Tue, 26 Jul 2016 13:27:33 -0500 Subject: Bind to D-Bus from a snap In-Reply-To: References: Message-ID: <1469557653.5406.22.camel@canonical.com> On Sat, 2016-07-23 at 00:41 +0000, Tin Tvrtković wrote: > Hello, > > for a few days now I've been trying to build a snap for Terminix. I'm > pretty much done, except when I actually run it, I get: > > Failed to register: GDBus.Error:org.freedesktop.DBus.Error.AccessDenied: > Connection ":1.194" is not allowed to own the service > "com.gexperts.Terminix" due to AppArmor policy > > Is there something in my snapcraft.yml I should be doing differently? No. Binding to an arbitrary dbus well-known name is not currently supported in snapd. However there is a POC PR[1] to start the discussion on how to deal with this sort of thing that you might be interested in following. [1]https://github.com/snapcore/snapd/pull/1446 -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From jamie at canonical.com Tue Jul 26 20:42:52 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Tue, 26 Jul 2016 15:42:52 -0500 Subject: Bind to D-Bus from a snap In-Reply-To: References: <1469557653.5406.22.camel@canonical.com> Message-ID: <1469565772.5406.50.camel@canonical.com> On Tue, 2016-07-26 at 20:03 +0000, Tin Tvrtković wrote: > Alrighty, thanks. Guess this will have to wait a little :) > Do keep in mind you can use --devmode in the meantime. :) > On 20:27, Tue, Jul 26, 2016 Jamie Strandboge wrote: > > > > > On Sat, 2016-07-23 at 00:41 +0000, Tin Tvrtković wrote: > > > > > > Hello, > > > > > > for a few days now I've been trying to build a snap for Terminix. I'm > > > pretty much done, except when I actually run it, I get: > > > > > > Failed to register: GDBus.Error:org.freedesktop.DBus.Error.AccessDenied: > > > Connection ":1.194" is not allowed to own the service > > > "com.gexperts.Terminix" due to AppArmor policy > > > > > > Is there something in my snapcraft.yml I should be doing differently? > > No. Binding to an arbitrary dbus well-known name is not currently > > supported in > > snapd. However there is a POC PR[1] to start the discussion on how to deal > > with > > this sort of thing that you might be interested in following. > > > > [1]https://github.com/snapcore/snapd/pull/1446 > > > > -- > > Jamie Strandboge             | http://www.canonical.com > > > > -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From silver.bullet at zoho.com Tue Jul 26 20:47:44 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Tue, 26 Jul 2016 22:47:44 +0200 Subject: How to sign a snap package? In-Reply-To: <5b228c15-cf17-4c26-bd0e-8c587993986b@ubuntu.com> References: <5b228c15-cf17-4c26-bd0e-8c587993986b@ubuntu.com> Message-ID: <20160726224744.216b87e5@utnubu> On Tue, 26 Jul 2016 19:43:10 +0200, Oliver Grawert wrote: >Am Dienstag, 26. Juli 2016 19:35:01 CEST schrieb Peng Liu: >> Is there any tool we can use to sign a snap package? > >i think we call that tool the store ;) Assumed upstream builds snaps, providing them as downloads by an upstream website, for Linux users of all distros, does it mean, that it is required to provide a snap package, a SHA256SUMS and SHA256SUMS.gpg or to use the Ubuntu store? From selene.scriven at canonical.com Wed Jul 27 01:46:51 2016 From: selene.scriven at canonical.com (Selene Scriven) Date: Tue, 26 Jul 2016 19:46:51 -0600 Subject: mpv snap In-Reply-To: <1469036453.20475.9.camel@canonical.com> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> <1469036453.20475.9.camel@canonical.com> Message-ID: <20160727014651.GB23091@mutt.xyzz.org> * Jamie Strandboge wrote: > On classic the interface provides access to the user session's > pulseaudio. On an all snaps system (eg, Ubuntu Core) pulseaudio > doesn't exist on the system and a pulseaudio snap must be > installed. Once installed, the interface is available for other > snaps to connect. Any idea if there are plans to allow ALSA access? Several things don't work (or don't work well) with Pulse running, and I find that a few lines in .asoundrc usually make Pulse unnecessary even for basic desktop purposes. -- Selene From silver.bullet at zoho.com Wed Jul 27 02:06:52 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Wed, 27 Jul 2016 04:06:52 +0200 Subject: mpv snap In-Reply-To: <20160727014651.GB23091@mutt.xyzz.org> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> <1469036453.20475.9.camel@canonical.com> <20160727014651.GB23091@mutt.xyzz.org> Message-ID: <20160727040652.2240a3f9@utnubu> On Tue, 26 Jul 2016 19:46:51 -0600, Selene Scriven wrote: >I find that a few lines in .asoundrc usually make Pulse unnecessary >even for basic desktop purposes. Let alone that jackd is able to provide real-time audio abilities, pulseaudio can't provide. It's too funny, that there's no tablet PC or phone available, based on Linux, that could be seriously used by musicians. It's possible for a musician, to use iPear devices, even with a less good infrastructure, than jackd does provide. From mark at ubuntu.com Wed Jul 27 06:43:43 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Wed, 27 Jul 2016 08:43:43 +0200 Subject: How to sign a snap package? In-Reply-To: <20160726224744.216b87e5@utnubu> References: <5b228c15-cf17-4c26-bd0e-8c587993986b@ubuntu.com> <20160726224744.216b87e5@utnubu> Message-ID: On 26/07/16 22:47, Ralf Mardorf wrote: > On Tue, 26 Jul 2016 19:43:10 +0200, Oliver Grawert wrote: >> Am Dienstag, 26. Juli 2016 19:35:01 CEST schrieb Peng Liu: >>> Is there any tool we can use to sign a snap package? >> i think we call that tool the store ;) > Assumed upstream builds snaps, providing them as downloads by an > upstream website, for Linux users of all distros, does it mean, that it > is required to provide a snap package, a SHA256SUMS and SHA256SUMS.gpg > or to use the Ubuntu store? Yes, both work. If you are publishing snaps on your website then it would be recommended to provide a GPG-signed list of digests as you suggest. Simplistically: * on an https web page * have directory listing your snaps and a sha256sums.txt * which is a list of the snaps, digests, and is gpg clear-text signed If you push your snap to the Ubuntu store, then the store will publish signatures which snapd will use to validate the snap on install and on refresh. In future, you'll be able to GPG sign the snap before you push it to the store, so snapd actually checks that YOU built it, not just that the store claims you uploaded it. Different store implementations (the snap format is independent of the store) will take different approaches, I've just outlined how we're doing it in Ubuntu with snapd, and how you can publish raw signed snaps on your site. Hope that helps! Mark From didrocks at ubuntu.com Wed Jul 27 07:02:13 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Wed, 27 Jul 2016 09:02:13 +0200 Subject: Trying to package angband as snap In-Reply-To: References: Message-ID: Le 20/07/2016 à 22:51, Radek Sprta a écrit : > Hey, > > I tried packaging Angband as as snap. This is what I have as > snapcraft.yaml: > > name: angband > version: "4.0.5" > summary: Single-player, text-based, dungeon simulation game > confinement: devmode > > apps: > angband: > command: games/angband > plug: [x11, unity7, home, pulseaudio] > > parts: > angband: > plugin: autotools > configflags: ["--enable-sdl"] > source: http://rephial.org/downloads/4.0/angband-4.0.5.tar.gz > build-packages: > - libsdl1.2-dev > - libsdl-ttf2.0-dev > - libsdl-mixer1.2-dev > - libsdl-image1.2-dev > > The game gets packaged, but fails to run, because it looks for the > files in /share/games/angband instead of $SNAP/share/games/angband > > I tried packaging it with '--no-install' and 'prefix=$SNAP' options > (changing command to angband), but both result in: > /snap/angband/100001/command-angband.wrapper: 5: exec: angband: > Permission denied > > The snappy-debug gives suggestions such as: > * adjust program to read necessary files from $SNAP, $SNAP_DATA or > $SNAP_USER_DATA > and > * adjust program to execute binaries directly from $SNAP instead of > /snap/bin > > Do I just fail at snapcraft or does the game really need to be patched? Hey Radek, The good news is that you are not alone! It's something we encountered in quite some project (having data dir hardcoded at build time based on prefix) without any relocatable path for it. I did file https://bugs.launchpad.net/ubuntu/+source/snapcraft/+bug/1583250/ at the time (and just added snappy upstream bug to it). Gustavo, do you think we can have some actionable actions on this and some ideas of when a proper solution would be provided? I know we talked to have a syscalls redirect LD_PRELOAD lib. For context, we are seeing quite some people having similar issues with desktop apps on the playpen and not pursuing their snap journey due to this. Meanwhile, something that worked for some projects is to "install-via: prefix" like we did for galculator: https://github.com/ubuntu/snappy-playpen/blob/master/galculator/snapcraft.yaml#L22. Another solution, that you mentioned is to hardcode "--prefix=/snap//current" is an immediate fix, but it means: * the snap depends on being installed at that location, which isn't ensured in the future. * then, you end up mostly with you binaries in /snap//current/snap//current/bin (notice the repetition), which is way you have the exec issue and need to move the content post-build to the correct location via a custom plugin or the "organize" keyword. Cheers, Didier From didrocks at ubuntu.com Wed Jul 27 07:06:42 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Wed, 27 Jul 2016 09:06:42 +0200 Subject: Trying to package angband as snap In-Reply-To: References: Message-ID: <3e60ac84-0929-1b3f-91de-f1d3418dfbda@ubuntu.com> + Gustavo on that one Le 27/07/2016 à 09:02, Didier Roche a écrit : > Le 20/07/2016 à 22:51, Radek Sprta a écrit : >> Hey, >> >> I tried packaging Angband as as snap. This is what I have as >> snapcraft.yaml: >> >> name: angband >> version: "4.0.5" >> summary: Single-player, text-based, dungeon simulation game >> confinement: devmode >> >> apps: >> angband: >> command: games/angband >> plug: [x11, unity7, home, pulseaudio] >> >> parts: >> angband: >> plugin: autotools >> configflags: ["--enable-sdl"] >> source: http://rephial.org/downloads/4.0/angband-4.0.5.tar.gz >> build-packages: >> - libsdl1.2-dev >> - libsdl-ttf2.0-dev >> - libsdl-mixer1.2-dev >> - libsdl-image1.2-dev >> >> The game gets packaged, but fails to run, because it looks for the >> files in /share/games/angband instead of $SNAP/share/games/angband >> >> I tried packaging it with '--no-install' and 'prefix=$SNAP' options >> (changing command to angband), but both result in: >> /snap/angband/100001/command-angband.wrapper: 5: exec: angband: >> Permission denied >> >> The snappy-debug gives suggestions such as: >> * adjust program to read necessary files from $SNAP, $SNAP_DATA or >> $SNAP_USER_DATA >> and >> * adjust program to execute binaries directly from $SNAP instead of >> /snap/bin >> >> Do I just fail at snapcraft or does the game really need to be patched? > Hey Radek, > > The good news is that you are not alone! It's something we encountered > in quite some project (having data dir hardcoded at build time based on > prefix) without any relocatable path for it. I did file > https://bugs.launchpad.net/ubuntu/+source/snapcraft/+bug/1583250/ at the > time (and just added snappy upstream bug to it). > > Gustavo, do you think we can have some actionable actions on this and > some ideas of when a proper solution would be provided? I know we talked > to have a syscalls redirect LD_PRELOAD lib. For context, we are seeing > quite some people having similar issues with desktop apps on the playpen > and not pursuing their snap journey due to this. > > Meanwhile, something that worked for some projects is to "install-via: > prefix" like we did for galculator: > https://github.com/ubuntu/snappy-playpen/blob/master/galculator/snapcraft.yaml#L22. > Another solution, that you mentioned is to hardcode > "--prefix=/snap//current" is an immediate fix, but it means: > * the snap depends on being installed at that location, which isn't > ensured in the future. > * then, you end up mostly with you binaries in > /snap//current/snap//current/bin (notice the > repetition), which is way you have the exec issue and need to move the > content post-build to the correct location via a custom plugin or the > "organize" keyword. > > Cheers, > Didier > > From ogra at ubuntu.com Wed Jul 27 09:29:00 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Wed, 27 Jul 2016 11:29:00 +0200 Subject: How to sign a snap package? In-Reply-To: <20160726224744.216b87e5@utnubu> References: <5b228c15-cf17-4c26-bd0e-8c587993986b@ubuntu.com> <20160726224744.216b87e5@utnubu> Message-ID: <1469611740.4310.11.camel@anubis> hi, Am Dienstag, den 26.07.2016, 22:47 +0200 schrieb Ralf Mardorf: > On Tue, 26 Jul 2016 19:43:10 +0200, Oliver Grawert wrote: > >Am Dienstag, 26. Juli 2016 19:35:01 CEST schrieb Peng Liu: > >> Is there any tool we can use to sign a snap package? > > > >i think we call that tool the store ;) > > Assumed upstream builds snaps, providing them as downloads by an > upstream website, for Linux users of all distros, does it mean, that it > is required to provide a snap package, a SHA256SUMS and SHA256SUMS.gpg > or to use the Ubuntu store? > you need to use "a" store ... doesn't have to be the ubuntu store, i know at the snappy sprint in heidelberg some representatives of the other distros expressed interest to run their own store ... there is no strict requirement how the verification or signing is done by any store, the design of snappy is flexible enough to allow anything they/you want to implement ;) we know how *our* store does it/will do it ... but that doesn't mean anyone else *needs* to use that mechanism or follow us with the implementation. snappy is designed with the freedom of differentiation in mind, it is all a matter of how much you want to invest, if the will is zero you can always use our store, if you have the resources you can totally build all your own infrastructure around the core... the one thing you cant change is the core design though, to keep compatibility among implementations intact and keep snaps installable across all possible targets. ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From ogra at ubuntu.com Wed Jul 27 09:54:21 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Wed, 27 Jul 2016 11:54:21 +0200 Subject: mpv snap In-Reply-To: <20160727014651.GB23091@mutt.xyzz.org> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> <1469036453.20475.9.camel@canonical.com> <20160727014651.GB23091@mutt.xyzz.org> Message-ID: <1469613261.4310.21.camel@anubis> hi, Am Dienstag, den 26.07.2016, 19:46 -0600 schrieb Selene Scriven: > * Jamie Strandboge wrote: > > On classic the interface provides access to the user session's > > pulseaudio. On an all snaps system (eg, Ubuntu Core) pulseaudio > > doesn't exist on the system and a pulseaudio snap must be > > installed. Once installed, the interface is available for other > > snaps to connect. > > Any idea if there are plans to allow ALSA access? Several things > don't work (or don't work well) with Pulse running, and I find > that a few lines in .asoundrc usually make Pulse unnecessary even > for basic desktop purposes. well, you dont really need alsa *access* you can ship the libs and config in your snap (and have full control over your dependencies), what you need is direct device access for that setup ... that way you can also have all realtime love you can imagine without the underlying host system interfering ... so instead of an alsa interface we should have an audio-device interface that you can use more flexible... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From mark at ubuntu.com Wed Jul 27 10:24:27 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Wed, 27 Jul 2016 12:24:27 +0200 Subject: Trying to package angband as snap In-Reply-To: <3e60ac84-0929-1b3f-91de-f1d3418dfbda@ubuntu.com> References: <3e60ac84-0929-1b3f-91de-f1d3418dfbda@ubuntu.com> Message-ID: <5ebb91a9-89d2-0009-7d4b-5357e9bb62b3@ubuntu.com> Would it be useful to be able to hard-map a set of directories for this specific snap to subdirectories in the snap writable directory? Something like this: plug: interface: overlay-directories versioned: [ /var/lib/games/foo, /usr/share/foo ] common: [ /etc/foo ] When that plug is connected, processes in THIS snap see content at that location which is ACTUALLY in $WRITABLE/snapd/overlay-directories/var/lib/games/foo for example. If snaps try to use this mechanism to share content it will fail since these directories are fake mounts just for those processes, other processes would see totally different directories. Would that work, would that be useful? It's sort of a "very specific and controlled overlay-fs" that works by directory rather than as a true overlay. Mark From mark at ubuntu.com Wed Jul 27 10:35:21 2016 From: mark at ubuntu.com (Mark Shuttleworth) Date: Wed, 27 Jul 2016 12:35:21 +0200 Subject: mpv snap In-Reply-To: <1469613261.4310.21.camel@anubis> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> <1469036453.20475.9.camel@canonical.com> <20160727014651.GB23091@mutt.xyzz.org> <1469613261.4310.21.camel@anubis> Message-ID: Both raw device access and alsa access are interesting. ALSA is mediated (multiple snaps can get it) and raw device is exclusive. Mark On 27/07/16 11:54, Oliver Grawert wrote: > hi, > Am Dienstag, den 26.07.2016, 19:46 -0600 schrieb Selene Scriven: >> * Jamie Strandboge wrote: >>> On classic the interface provides access to the user session's >>> pulseaudio. On an all snaps system (eg, Ubuntu Core) pulseaudio >>> doesn't exist on the system and a pulseaudio snap must be >>> installed. Once installed, the interface is available for other >>> snaps to connect. >> Any idea if there are plans to allow ALSA access? Several things >> don't work (or don't work well) with Pulse running, and I find >> that a few lines in .asoundrc usually make Pulse unnecessary even >> for basic desktop purposes. > well, you dont really need alsa *access* you can ship the libs and > config in your snap (and have full control over your dependencies), what > you need is direct device access for that setup ... that way you can > also have all realtime love you can imagine without the underlying host > system interfering ... > > so instead of an alsa interface we should have an audio-device interface > that you can use more flexible... > > ciao > oli > > -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From ogra at ubuntu.com Wed Jul 27 12:23:35 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Wed, 27 Jul 2016 14:23:35 +0200 Subject: naming of our kernel and gadget snaps ... Message-ID: <1469622215.4310.49.camel@anubis> hi, we currently have lots of snaps called "canonical-$device" or "canonical-$device-linux" (first one is gadget, second is kernel) there is a desire of getting rid of the redundant canonical in there (it is owned by the canonical user, that is shown prominently everywhere, so there is no need to have it additionally in the package name) and still be a bit descriptive of the content. == Kernel == our current kernel snaps are largely created by simply using everything the kernel meta package installs from the archive but will be extended by mesa and GL libraries and potentially by hardware related udev rules. since the majority of content comes from the linux meta packages i think we should use these as top level names ... i.e. for the current amd64 kernel it would just be: linux-generic_$packageversion.snap (with the version extracted from the meta .deb at snap creation time) if there are additional pieces inside, they should extend the namespace: linux-generic-nvidia_$packageversion.snap for a generic amd64 kernel snap that ships the nvidia drivers alongside. (as you might know, we cant ship all binary drivers in the same package easily, as they can ship their own libGL/GLES which would cause breakage) i think for the kernel that is relatively easy and sensible since our official kernel snaps simply come from archive packages. == Gadget == the more complicated bit where i'm seeking input and ideas from others are the gadget snaps ... should they have "gadget" in their names, should they just be the device name, etc ? we used to have just "beaglebone" or "raspberry-pi" snaps in the store in the past which was re-defined after some sprint to be "canonical-pi2" or "canonical-pc". i was digging through various sprint notes but can not find the reasoning for the rename. my best proposal would be to go back to the former and have "dragonboard", "raspberry-pi2", "raspberry-pi3", "pc-i386", "pc-amd64" does anyone have any better suggestions ? ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From silver.bullet at zoho.com Wed Jul 27 13:30:04 2016 From: silver.bullet at zoho.com (Ralf Mardorf) Date: Wed, 27 Jul 2016 15:30:04 +0200 Subject: mpv snap In-Reply-To: <1469613261.4310.21.camel@anubis> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> <1469036453.20475.9.camel@canonical.com> <20160727014651.GB23091@mutt.xyzz.org> <1469613261.4310.21.camel@anubis> Message-ID: <20160727153004.6f533128@utnubu> On Wed, 27 Jul 2016 11:54:21 +0200, Oliver Grawert wrote: >well, you dont really need alsa *access* you can ship the libs and >config in your snap (and have full control over your dependencies), >what you need is direct device access for that setup Yesno, there could be the need to mix audio streams, jackd, dmix or what ever else, so direct access to the devices isn't helpful, as soon as two snaps try to access one audio device. IIUC you already made pulseaudio available ;), but pulseaudio doesn't fit to everybody's needs. Jackd might not fit to this mpv direct alsa access example, but dmix does. OTOH I'm not using dmix, so I don't know how it works, perhaps dmix could be considered to quasi be the audio device. Regards, Ralf -- Death of ROXTerm https://sourceforge.net/p/roxterm/discussion/422638/thread/60da6975/ From dgarrod at extremenetworks.com Wed Jul 27 15:50:38 2016 From: dgarrod at extremenetworks.com (David Garrod) Date: Wed, 27 Jul 2016 15:50:38 +0000 Subject: How do I share a namespace between snap commands? Message-ID: I'm very new to snappy so please excuse me for any questions that seem obvious. I have snapcraft.yaml that looks as follows: -------------------------------------------------------------------------------------------------------------------------------------------- % cat snapcraft.yaml name: openswitch-appliance version: 0.3.0-rc0-1 summary: OpenSwitch NOS for the Appliance. description: A standalone version of the OpenSwitch NOS (www.openswitch.net) Appliance targeted at an Oracle VirtualBox running Ubuntu-Core with one management interface and seven "front-panel" network adapters. confinement: devmode apps: start-openswitch: command: usr/sbin/start-openswitch plugs: [ network, network-control, network-bind, network-manager, network-observe, gsettings ] stop-openswitch: command: usr/sbin/stop-openswitch plugs: [ network, network-control, network-bind, network-manager, network-observe, gsettings ] vtysh: command: usr/sbin/start-vtysh shell: command: bin/bash plugs: [ network, network-control, network-bind, network-manager, network-observe, gsettings ] parts: ops-init: plugin: make source: src/ops-init stage-packages: [ binutils, file, strace, bash, iputils-ping, traceroute ] openswitch: plugin: x-openswitch # source: https://github.com/ops-snappy/ops-build.git # source: ../ops-build source: /builds/dgarrod/ops-build platform: appliance # skip-native-build: true skip-native-build: true stage-packages: [ resolvconf ] after: [ ops-init ] -------------------------------------------------------------------------------------------------------------------------------------------- The important point here is that there are multiple commands that are part of the snap. I start our snap using the command: sudo /snap/bin/openswitch-appliance.start-openswitch Inside the SNAP this creates a couple of namespaces (swns and nonet): /sbin/ip netns add swns /sbin/ip netns add nonet The problem is that if I then start up another command in the SNAP, specifically a bash shell using: sudo /snap/bin/openswitch-appliance.bash I can't get at the swns or nonet namespaces from the bash shell that gets started. I get errors like: root at Snappy-Appliance:/home/netop# ip netns exec swns bash RTNETLINK answers: Invalid argument RTNETLINK answers: Invalid argument setting the network namespace "swns" failed: Invalid argument root at Snappy-Appliance:/home/netop# ip netns pids swns RTNETLINK answers: Invalid argument RTNETLINK answers: Invalid argument Why is this? I thought all commands for my "snap" are somehow part of the same "snap", even though they are in different process trees. It apparently isn't the case. How do I get namespaces to play across these commands. Note that currently I'm installing my SNAP --devmode and I'm also connecting all the network related plug/slots. sudo snap connect openswitch-appliance:network-control ubuntu-core:network-control sudo snap connect openswitch-appliance:network-manager ubuntu-core:network-manager sudo snap connect openswitch-appliance:network-observe ubuntu-core:network-observe The only way I've sort of worked around this is to start up a "bash" shell using my snap bash command and then actually starting my subsystem from within this bash shell. So now all processes only went through the one "command". But for a number of reasons this isn't going to work for me. Another way of asking my questions is how do I get all my "apps" in my SNAP to be able to play nicely together? Are they one "snap" or not, even when started separately? Thanks, Dave Garrod ________________________________ DISCLAIMER: This e-mail and any attachments to it may contain confidential and proprietary material and is solely for the use of the intended recipient. Any review, use, disclosure, distribution or copying of this transmittal is prohibited except by or on behalf of the intended recipient. If you have received this transmittal in error, please notify the sender and destroy this e-mail and any attachments and all copies, whether electronic or printed. -------------- next part -------------- An HTML attachment was scrubbed... URL: From dgarrod at extremenetworks.com Wed Jul 27 16:00:54 2016 From: dgarrod at extremenetworks.com (David Garrod) Date: Wed, 27 Jul 2016 16:00:54 +0000 Subject: How do I get a postinst stage properly executed - traceroute will not install correctly Message-ID: I'm trying to get the "traceroute" package to install properly inside my SNAP. To do this I've added traceroute to the stage-packages: line in my snapcraft.yaml. The files to get put into the; /snap/openswitch-appliance/x1/... $SNAP tree and indeed I can then see them from inside the SNAP. root at Snappy-Appliance:/home/netop# find $SNAP -name "traceroute*" /snap/openswitch-appliance/x1/usr/bin/traceroute-nanog /snap/openswitch-appliance/x1/usr/bin/traceroute.db /snap/openswitch-appliance/x1/usr/bin/traceroute6.db /snap/openswitch-appliance/x1/usr/share/doc/traceroute /snap/openswitch-appliance/x1/usr/share/man/man1/traceroute-nanog.1.gz /snap/openswitch-appliance/x1/usr/share/man/man1/traceroute.db.1.gz /snap/openswitch-appliance/x1/usr/share/man/man1/traceroute6.db.1.gz But the problem is that the installation isn't properly finished because the the postinst stage didn't get run. If you look at the postinst stage for the traceroute package you'll see what I've listed below. In other words for this package to work it is meant to build some softlinks to /etc so that "traceroute" is actually made available. How am I meant to do this is the SNAP world? Thanks, Dave Garrod % cat DEBIAN/postinst #!/bin/sh set -e case "${1}" in configure) update-alternatives --install \ /usr/bin/traceroute traceroute /usr/bin/traceroute.db 100 \ --slave /usr/sbin/traceroute traceroute.sbin \ /usr/bin/traceroute.db \ --slave /usr/share/man/man1/traceroute.1.gz traceroute.1.gz \ /usr/share/man/man1/traceroute.db.1.gz for _ALTERNATIVE in traceroute6 lft traceproto do update-alternatives --install \ /usr/bin/${_ALTERNATIVE} ${_ALTERNATIVE} /usr/bin/${_ALTERNATIVE}.db 100 \ --slave /usr/share/man/man1/${_ALTERNATIVE}.1.gz ${_ALTERNATIVE}.1.gz \ /usr/share/man/man1/${_ALTERNATIVE}.db.1.gz done for _ALTERNATIVE in tcptraceroute do update-alternatives --install \ /usr/sbin/${_ALTERNATIVE} ${_ALTERNATIVE} /usr/sbin/${_ALTERNATIVE}.db 100 \ --slave /usr/share/man/man8/${_ALTERNATIVE}.8.gz ${_ALTERNATIVE}.8.gz \ /usr/share/man/man8/${_ALTERNATIVE}.db.8.gz done ;; abort-upgrade|abort-remove|abort-deconfigure) ;; *) echo "postinst called with unknown argument \`${1}'" >&2 exit 1 ;; esac exit 0 ________________________________ DISCLAIMER: This e-mail and any attachments to it may contain confidential and proprietary material and is solely for the use of the intended recipient. Any review, use, disclosure, distribution or copying of this transmittal is prohibited except by or on behalf of the intended recipient. If you have received this transmittal in error, please notify the sender and destroy this e-mail and any attachments and all copies, whether electronic or printed. -------------- next part -------------- An HTML attachment was scrubbed... URL: From jamie at canonical.com Wed Jul 27 17:03:00 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Wed, 27 Jul 2016 12:03:00 -0500 Subject: How do I share a namespace between snap commands? In-Reply-To: References: Message-ID: <1469638980.5406.135.camel@canonical.com> On Wed, 2016-07-27 at 15:50 +0000, David Garrod wrote: ... > > The important point here is that there are multiple commands that are part of > the snap. I start our snap using the command: > > sudo /snap/bin/openswitch-appliance.start-openswitch > > Inside the SNAP this creates a couple of namespaces (swns and nonet): > > /sbin/ip netns add swns > /sbin/ip netns add nonet > > The problem is that if I then start up another command in the SNAP, > specifically a bash shell using: > > sudo /snap/bin/openswitch-appliance.bash > > I can't get at the swns or nonet namespaces from the bash shell that gets > started. I get errors like: > > root at Snappy-Appliance:/home/netop# ip netns exec swns bash > RTNETLINK answers: Invalid argument > RTNETLINK answers: Invalid argument > setting the network namespace "swns" failed: Invalid argument > root at Snappy-Appliance:/home/netop# ip netns pids swns > RTNETLINK answers: Invalid argument > RTNETLINK answers: Invalid argument > > Why is this? I thought all commands for my "snap" are somehow part of the same > "snap", even though they are in different process trees. It apparently isn't > the case. How do I get namespaces to play across these commands. > I'm going to answer only part of this so everyone is on the same page and let the snappy team comment on why this might not be working for you. Snaps can be thought of as running inside a container, but the snappy container differs from other containers like LXD and Docker[1] in that the application runs (primarily) within the global namespace. We actively made this design choice so the application is fully integrated with the rest of the system and to facilitate arbitrary snap connections via carefully designed snappy interfaces. For example, rather than using Linux file, IPC or pid namespaces for isolation, we use MAC (ie, AppArmor). Snappy leverages many different techniques to provide strong isolation though, such as seccomp syscall filtering, bind mounts, device cgroups, private mount namespace for app-specific /tmp, and a devpts new instance. To the point, the design of snappy is precisely so that apps within the same snap can work together (but also interact with other snaps in controlled ways via interfaces). To your specific question, I would expect one app within your snap to be able to create a network namespace for another app in your snap to access. All of the apps in your snap are running within the global network namespace. That said, because of how snapd sets up bind mounts and the private mount namespace for /tmp for each app's runtime in your snap, it sounds like there is a bug and something might be getting in the way there (eg, a bind mount is missing, a bind mount is hiding something from your other app, your snap needs a shared /tmp, etc). This might explain why running all the commands from within your snap's bash shell works when it doesn't across apps. The snappy team can comment on this (though, perhaps filing a bug at https://bug s.launchpad.net/snappy/+filebug with a simple reproducer would be the fastest way to get this resolved). > Note that currently I'm installing my SNAP --devmode and I'm also connecting > all the network related plug/slots. > This is good since we don't yet have an interface that supports setting up network namespaces. Once your snap is working in --devmode, please file a bug adding the snapd-interface tag. I'm not sure if this would be an extension of network-control or a new interface-- we can discuss that at the appropriate time. Jamie [1] Note, for people who want LXD and Docker, snaps are in active development and you can expect them to work wonderfully when delivered as snaps to your classic or all snaps systems :)   -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From jamie at canonical.com Wed Jul 27 17:30:37 2016 From: jamie at canonical.com (Jamie Strandboge) Date: Wed, 27 Jul 2016 12:30:37 -0500 Subject: How do I share a namespace between snap commands? In-Reply-To: <1469638980.5406.135.camel@canonical.com> References: <1469638980.5406.135.camel@canonical.com> Message-ID: <1469640637.5406.149.camel@canonical.com> On Wed, 2016-07-27 at 12:03 -0500, Jamie Strandboge wrote: > On Wed, 2016-07-27 at 15:50 +0000, David Garrod wrote: > ... > > > > > > The important point here is that there are multiple commands that are part > > of > > the snap. I start our snap using the command: > > > > sudo /snap/bin/openswitch-appliance.start-openswitch > > > > Inside the SNAP this creates a couple of namespaces (swns and nonet): > > > > /sbin/ip netns add swns > > /sbin/ip netns add nonet > > > > The problem is that if I then start up another command in the SNAP, > > specifically a bash shell using: > > > > sudo /snap/bin/openswitch-appliance.bash > > To the point, the design of snappy is precisely so that apps within the same > snap can work together (but also interact with other snaps in controlled ways > via interfaces). John Lenton mentioned to me on IRC that this might be that you are trying to call applications from /snap/bin from within your snap, but it isn't clear to me that this is what you are doing. Just in case, all apps within a snap may execute anything in $SNAP, $SNAP_DATA and $SNAP_USER_DATA and these directories are shared among all apps within the same snap. Apps are not allowed to execute anything in /snap/bin (their own or otherwise). If you are trying to run things in /snap/bin, simply adjust your snap to run your executables directly, ideally referencing the various $SNAP* environment variables. (If you are unfamiliar with the $SNAP* environment variables, within your bash app do 'set|grep SNAP' or install hello-world and run 'hello-world.env'). -- Jamie Strandboge | http://www.canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: This is a digitally signed message part URL: From dgarrod at extremenetworks.com Wed Jul 27 18:58:12 2016 From: dgarrod at extremenetworks.com (David Garrod) Date: Wed, 27 Jul 2016 18:58:12 +0000 Subject: How do I share a namespace between snap commands? In-Reply-To: <1469640637.5406.149.camel@canonical.com> References: <1469638980.5406.135.camel@canonical.com> <1469640637.5406.149.camel@canonical.com> Message-ID: Thank you for your initial answers. Re: > John Lenton mentioned to me on IRC that this might be that you are trying to call applications from /snap/bin from within your snap, but it isn't clear to me that this is what you are doing. No. From within the SNAP I don't reference /snap/bin. That is only referenced when I start up one of the snap applications listed in snapcraft.yaml file. I am familiar with the $SNAP* environment variables and directory structure of the SNAPs. I can simplify the problem statement I've found the following: From the base system start up the SNAP. I use the command that just gives me a bash prompt: netop at Snappy-Appliance:~$ sudo /snap/bin/openswitch-appliance.shell [sudo] password for netop: So now I'm in the SNAP and I'm going to create a network namespace and use it: root at Snappy-Appliance:/home/netop# root at Snappy-Appliance:/home/netop# which ip /snap/openswitch-appliance/x1/bin/ip root at Snappy-Appliance:/home/netop# ip netns add xxx root at Snappy-Appliance:/home/netop# root at Snappy-Appliance:/home/netop# root at Snappy-Appliance:/home/netop# ip netns exec xxx bash root at Snappy-Appliance:/home/netop# ip netns list xxx root at Snappy-Appliance:/home/netop# ip netns pids xxx 2650 2655 root at Snappy-Appliance:/home/netop# exit exit root at Snappy-Appliance:/home/netop# ip netns pids xxx root at Snappy-Appliance:/home/netop# ip netns list xxx root at Snappy-Appliance:/home/netop# As you can see from the above it works fine. Now in another window I'm going to fire up the bash command in another instanced of this snap and try and use the xxx network namespace. As you can see the: ip netns list command errors out with: root at Snappy-Appliance:/home/netop# ip netns list RTNETLINK answers: Invalid argument RTNETLINK answers: Invalid argument The 'ip netns exec xxx bash' fails in a similar way. I've also included an strace of the command in case that throws any light on the matter. One thing I do want to note is the following. From the first window where I created the xxx namespace: root at Snappy-Appliance:/home/netop# df -a Filesystem 1K-blocks Used Available Use% Mounted on /dev/sda1 15349744 5357264 9189716 37% / udev 488872 0 488872 0% /dev devpts 0 0 0 - /dev/pts tmpfs 508136 172 507964 1% /dev/shm hugetlbfs 0 0 0 - /dev/hugepages mqueue 0 0 0 - /dev/mqueue tmpfs 101628 10132 91496 10% /run tmpfs 5120 4 5116 1% /run/lock sunrpc 0 0 0 - /run/rpc_pipefs tmpfs 101628 72 101556 1% /run/user/1000 gvfsd-fuse 0 0 0 - /run/user/1000/gvfs sysfs 0 0 0 - /sys securityfs 0 0 0 - /sys/kernel/security tmpfs 508136 0 508136 0% /sys/fs/cgroup cgroup 0 0 0 - /sys/fs/cgroup/systemd cgroup 0 0 0 - /sys/fs/cgroup/hugetlb cgroup 0 0 0 - /sys/fs/cgroup/cpuset cgroup 0 0 0 - /sys/fs/cgroup/cpu,cpuacct cgroup 0 0 0 - /sys/fs/cgroup/pids cgroup 0 0 0 - /sys/fs/cgroup/net_cls,net_prio cgroup 0 0 0 - /sys/fs/cgroup/memory cgroup 0 0 0 - /sys/fs/cgroup/devices cgroup 0 0 0 - /sys/fs/cgroup/perf_event cgroup 0 0 0 - /sys/fs/cgroup/blkio cgroup 0 0 0 - /sys/fs/cgroup/freezer pstore 0 0 0 - /sys/fs/pstore debugfs 0 0 0 - /sys/kernel/debug fusectl 0 0 0 - /sys/fs/fuse/connections proc 0 0 0 - /proc systemd-1 0 0 0 - /proc/sys/fs/binfmt_misc /dev/loop1 36352 36352 0 100% /snap/openswitch-appliance/x1 /dev/loop2 128 128 0 100% /snap/hello-world/26 /dev/loop3 128 128 0 100% /snap/hello-world/27 /dev/loop0 66432 66432 0 100% /snap/ubuntu-core/122 VirtualBox_Shared_Folder 500106580 203225352 296881228 41% /media/sf_VirtualBox_Shared_Folder /dev/sr0 56868 56868 0 100% /media/netop/VBOXADDITIONS_5.0.26_108824 auto.direct 1487997952 1150649344 322231296 79% /roch/blizzard_fw2 auto.direct 907138816 701365856 196556960 79% /roch/blizzard_fw /dev/loop0 66432 66432 0 100% /bin /dev/loop0 66432 66432 0 100% /sbin /dev/loop0 66432 66432 0 100% /lib /dev/loop0 66432 66432 0 100% /lib64 /dev/loop0 66432 66432 0 100% /usr /dev/sda1 15349744 5357268 9189712 37% /tmp devpts 0 0 0 - /dev/pts devpts 0 0 0 - /dev/ptmx tmpfs 101628 11416 90212 12% /run/netns nsfs 0 0 0 - /run/netns/xxx Note the "nsfs" mount of /run/netns/xxx. Here is the identical 'df -a' from the 2nd window, i.e. the one that doesn’t work. Note that this DOES NOT have the same line: root at Snappy-Appliance:/home/netop# df -a Filesystem 1K-blocks Used Available Use% Mounted on /dev/sda1 15349744 5357272 9189708 37% / udev 488872 0 488872 0% /dev devpts 0 0 0 - /dev/pts tmpfs 508136 172 507964 1% /dev/shm hugetlbfs 0 0 0 - /dev/hugepages mqueue 0 0 0 - /dev/mqueue tmpfs 101628 11416 90212 12% /run tmpfs 5120 4 5116 1% /run/lock sunrpc 0 0 0 - /run/rpc_pipefs tmpfs 101628 72 101556 1% /run/user/1000 gvfsd-fuse 0 0 0 - /run/user/1000/gvfs sysfs 0 0 0 - /sys securityfs 0 0 0 - /sys/kernel/security tmpfs 508136 0 508136 0% /sys/fs/cgroup cgroup 0 0 0 - /sys/fs/cgroup/systemd cgroup 0 0 0 - /sys/fs/cgroup/hugetlb cgroup 0 0 0 - /sys/fs/cgroup/cpuset cgroup 0 0 0 - /sys/fs/cgroup/cpu,cpuacct cgroup 0 0 0 - /sys/fs/cgroup/pids cgroup 0 0 0 - /sys/fs/cgroup/net_cls,net_prio cgroup 0 0 0 - /sys/fs/cgroup/memory cgroup 0 0 0 - /sys/fs/cgroup/devices cgroup 0 0 0 - /sys/fs/cgroup/perf_event cgroup 0 0 0 - /sys/fs/cgroup/blkio cgroup 0 0 0 - /sys/fs/cgroup/freezer pstore 0 0 0 - /sys/fs/pstore debugfs 0 0 0 - /sys/kernel/debug fusectl 0 0 0 - /sys/fs/fuse/connections proc 0 0 0 - /proc systemd-1 - - - - /proc/sys/fs/binfmt_misc /dev/loop1 36352 36352 0 100% /snap/openswitch-appliance/x1 /dev/loop2 128 128 0 100% /snap/hello-world/26 /dev/loop3 128 128 0 100% /snap/hello-world/27 /dev/loop0 66432 66432 0 100% /snap/ubuntu-core/122 VirtualBox_Shared_Folder 500106580 203225352 296881228 41% /media/sf_VirtualBox_Shared_Folder /dev/sr0 56868 56868 0 100% /media/netop/VBOXADDITIONS_5.0.26_108824 auto.direct - - - - /roch/blizzard_fw2 auto.direct - - - - /roch/blizzard_fw /dev/loop0 66432 66432 0 100% /bin /dev/loop0 66432 66432 0 100% /sbin /dev/loop0 66432 66432 0 100% /lib /dev/loop0 66432 66432 0 100% /lib64 /dev/loop0 66432 66432 0 100% /usr /dev/sda1 15349744 5357272 9189708 37% /tmp devpts 0 0 0 - /dev/pts devpts 0 0 0 - /dev/ptmx binfmt_misc 0 0 0 - /proc/sys/fs/binfmt_misc engstore3:/fw2/blizzard 1487997952 1150649344 322231296 79% /roch/blizzard_fw2 maand-engstore1:/fw1/blizzard 907138816 701365984 196556832 79% /roch/blizzard_fw Below is the long output from the second window referenced above before the mentioning of the 'df -a' commands. So my question is why does the 'ip netns' command not work in the 2nd window? Thanks, Dave netop at Snappy-Appliance:~$ sudo /snap/bin/openswitch-appliance.shell [sudo] password for netop: root at Snappy-Appliance:/home/netop# root at Snappy-Appliance:/home/netop# which ip /snap/openswitch-appliance/x1/bin/ip root at Snappy-Appliance:/home/netop# ip netns list RTNETLINK answers: Invalid argument RTNETLINK answers: Invalid argument xxx root at Snappy-Appliance:/home/netop# ip netns exec xxx bash RTNETLINK answers: Invalid argument setting the network namespace "xxx" failed: Invalid argument root at Snappy-Appliance:/home/netop# ps PID TTY TIME CMD 2695 ? 00:00:00 sudo 2696 ? 00:00:00 bash 2708 ? 00:00:00 ps root at Snappy-Appliance:/home/netop# root at Snappy-Appliance:/home/netop# root at Snappy-Appliance:/home/netop# strace ip netns list execve("/snap/openswitch-appliance/x1/bin/ip", ["ip", "netns", "list"], [/* 32 vars */]) = 0 brk(NULL) = 0x141c000 access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory) mmap(NULL, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f1ff7520000 access("/etc/ld.so.preload", R_OK) = -1 ENOENT (No such file or directory) open("/var/lib/snapd/lib/gl/tls/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/var/lib/snapd/lib/gl/tls/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/var/lib/snapd/lib/gl/tls/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/var/lib/snapd/lib/gl/tls", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/var/lib/snapd/lib/gl/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/var/lib/snapd/lib/gl/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/var/lib/snapd/lib/gl/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/var/lib/snapd/lib/gl", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0 open("tls/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("tls/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/tls/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/tls/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/tls/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/tls", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu", {st_mode=S_IFDIR|0755, st_size=76, ...}) = 0 open("/snap/openswitch-appliance/x1/usr/lib/tls/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib/tls/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/tls/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib/tls", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib", {st_mode=S_IFDIR|0755, st_size=1456, ...}) = 0 open("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/tls/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/tls/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/tls/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/tls", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu", {st_mode=S_IFDIR|0755, st_size=646, ...}) = 0 open("/snap/openswitch-appliance/x1/lib/tls/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib/tls/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/tls/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib/tls", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/x86_64/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib/x86_64", 0x7ffc033f74f0) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/libdl.so.2", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) stat("/snap/openswitch-appliance/x1/lib", {st_mode=S_IFDIR|0755, st_size=146, ...}) = 0 open("/etc/ld.so.cache", O_RDONLY|O_CLOEXEC) = 3 fstat(3, {st_mode=S_IFREG|0644, st_size=96833, ...}) = 0 mmap(NULL, 96833, PROT_READ, MAP_PRIVATE, 3, 0) = 0x7f1ff7508000 close(3) = 0 access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory) open("/lib/x86_64-linux-gnu/libdl.so.2", O_RDONLY|O_CLOEXEC) = 3 read(3, "\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0\240\r\0\0\0\0\0\0"..., 832) = 832 fstat(3, {st_mode=S_IFREG|0644, st_size=14608, ...}) = 0 mmap(NULL, 2109680, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x7f1ff70f9000 mprotect(0x7f1ff70fc000, 2093056, PROT_NONE) = 0 mmap(0x7f1ff72fb000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x2000) = 0x7f1ff72fb000 close(3) = 0 open("/var/lib/snapd/lib/gl/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("tls/x86_64/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("tls/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("x86_64/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/x86_64-linux-gnu/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/usr/lib/x86_64-linux-gnu/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("/snap/openswitch-appliance/x1/lib/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory) open("/lib/x86_64-linux-gnu/libc.so.6", O_RDONLY|O_CLOEXEC) = 3 read(3, "\177ELF\2\1\1\3\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0P\t\2\0\0\0\0\0"..., 832) = 832 fstat(3, {st_mode=S_IFREG|0755, st_size=1864888, ...}) = 0 mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f1ff7507000 mmap(NULL, 3967488, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x7f1ff6d30000 mprotect(0x7f1ff6ef0000, 2093056, PROT_NONE) = 0 mmap(0x7f1ff70ef000, 24576, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x1bf000) = 0x7f1ff70ef000 mmap(0x7f1ff70f5000, 14848, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x7f1ff70f5000 close(3) = 0 mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f1ff7506000 mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f1ff7505000 arch_prctl(ARCH_SET_FS, 0x7f1ff7506700) = 0 mprotect(0x7f1ff70ef000, 16384, PROT_READ) = 0 mprotect(0x7f1ff72fb000, 4096, PROT_READ) = 0 mprotect(0x656000, 4096, PROT_READ) = 0 mprotect(0x7f1ff7522000, 4096, PROT_READ) = 0 munmap(0x7f1ff7508000, 96833) = 0 socket(PF_NETLINK, SOCK_RAW|SOCK_CLOEXEC, NETLINK_ROUTE) = 3 setsockopt(3, SOL_SOCKET, SO_SNDBUF, [32768], 4) = 0 setsockopt(3, SOL_SOCKET, SO_RCVBUF, [1048576], 4) = 0 bind(3, {sa_family=AF_NETLINK, pid=0, groups=00000000}, 12) = 0 getsockname(3, {sa_family=AF_NETLINK, pid=2722, groups=00000000}, [12]) = 0 open("/proc/self/ns/net", O_RDONLY) = 4 sendto(3, "\34\0\0\0Z\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\10\0\3\0\4\0\0\0", 28, 0, NULL, 0) = 28 recvmsg(3, {msg_name(12)={sa_family=AF_NETLINK, pid=0, groups=00000000}, msg_iov(1)=[{"\34\0\0\0X\0\0\0\0\0\0\0\242\n\0\0\0\0\0\0\10\0\1\0\377\377\377\377", 16384}], msg_controllen=0, msg_flags=0}, 0) = 28 close(4) = 0 socket(PF_NETLINK, SOCK_RAW|SOCK_CLOEXEC, NETLINK_ROUTE) = 4 setsockopt(4, SOL_SOCKET, SO_SNDBUF, [32768], 4) = 0 setsockopt(4, SOL_SOCKET, SO_RCVBUF, [1048576], 4) = 0 bind(4, {sa_family=AF_NETLINK, pid=0, groups=00000000}, 12) = 0 getsockname(4, {sa_family=AF_NETLINK, pid=-353713190, groups=00000000}, [12]) = 0 open("/var/run/netns", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 5 fstat(5, {st_mode=S_IFDIR|0755, st_size=60, ...}) = 0 brk(NULL) = 0x141c000 brk(0x1445000) = 0x1445000 getdents(5, /* 3 entries */, 32768) = 72 open("/var/run/netns/xxx", O_RDONLY) = 6 sendmsg(4, {msg_name(12)={sa_family=AF_NETLINK, pid=0, groups=00000000}, msg_iov(1)=[{"\34\0\0\0Z\0\1\0\231\2\231W\0\0\0\0\0\0\0\0\10\0\3\0\6\0\0\0", 28}], msg_controllen=0, msg_flags=0}, 0) = 28 recvmsg(4, {msg_name(12)={sa_family=AF_NETLINK, pid=0, groups=00000000}, msg_iov(1)=[{"0\0\0\0\2\0\0\0\231\2\231W\332\303\352\352\352\377\377\377\34\0\0\0Z\0\1\0\231\2\231W"..., 32768}], msg_controllen=0, msg_flags=0}, 0) = 48 write(2, "RTNETLINK answers: Invalid argum"..., 36RTNETLINK answers: Invalid argument ) = 36 close(6) = 0 getdents(5, /* 0 entries */, 32768) = 0 brk(0x143d000) = 0x143d000 close(5) = 0 open("/var/run/netns", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 5 fstat(5, {st_mode=S_IFDIR|0755, st_size=60, ...}) = 0 getdents(5, /* 3 entries */, 32768) = 72 fstat(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(136, 17), ...}) = 0 open("/var/run/netns/xxx", O_RDONLY) = 6 sendmsg(4, {msg_name(12)={sa_family=AF_NETLINK, pid=0, groups=00000000}, msg_iov(1)=[{"\34\0\0\0Z\0\1\0\232\2\231W\0\0\0\0\0\0\0\0\10\0\3\0\6\0\0\0", 28}], msg_controllen=0, msg_flags=0}, 0) = 28 recvmsg(4, {msg_name(12)={sa_family=AF_NETLINK, pid=0, groups=00000000}, msg_iov(1)=[{"0\0\0\0\2\0\0\0\232\2\231W\332\303\352\352\352\377\377\377\34\0\0\0Z\0\1\0\232\2\231W"..., 32768}], msg_controllen=0, msg_flags=0}, 0) = 48 write(2, "RTNETLINK answers: Invalid argum"..., 36RTNETLINK answers: Invalid argument ) = 36 close(6) = 0 write(1, "xxx\n", 4xxx ) = 4 getdents(5, /* 0 entries */, 32768) = 0 close(5) = 0 exit_group(0) = ? +++ exited with 0 +++ root at Snappy-Appliance:/home/netop# ________________________________ DISCLAIMER: This e-mail and any attachments to it may contain confidential and proprietary material and is solely for the use of the intended recipient. Any review, use, disclosure, distribution or copying of this transmittal is prohibited except by or on behalf of the intended recipient. If you have received this transmittal in error, please notify the sender and destroy this e-mail and any attachments and all copies, whether electronic or printed. From didrocks at ubuntu.com Thu Jul 28 06:14:16 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Thu, 28 Jul 2016 08:14:16 +0200 Subject: How do I get a postinst stage properly executed - traceroute will not install correctly In-Reply-To: References: Message-ID: Le 27/07/2016 à 18:00, David Garrod a écrit : > > I’m trying to get the “traceroute” package to install properly inside > my SNAP. To do this I’ve added > > > > traceroute > > > > to the > > > > stage-packages: > > > > line in my snapcraft.yaml. The files to get put into the; > > /snap/openswitch-appliance/x1/… > > > > $SNAP tree and indeed I can then see them from inside the SNAP. > > > > root at Snappy-Appliance:/home/netop# find $SNAP -name "traceroute*" > > /snap/openswitch-appliance/x1/usr/bin/traceroute-nanog > > /snap/openswitch-appliance/x1/usr/bin/traceroute.db > > /snap/openswitch-appliance/x1/usr/bin/traceroute6.db > > /snap/openswitch-appliance/x1/usr/share/doc/traceroute > > /snap/openswitch-appliance/x1/usr/share/man/man1/traceroute-nanog.1.gz > > /snap/openswitch-appliance/x1/usr/share/man/man1/traceroute.db.1.gz > > /snap/openswitch-appliance/x1/usr/share/man/man1/traceroute6.db.1.gz > > > > But the problem is that the installation isn’t properly finished > because the the postinst stage didn’t get run. If you look at the > postinst stage for the traceroute package you’ll see what I’ve listed > below. > > > > In other words for this package to work it is meant to build some > softlinks to /etc so that “traceroute” is actually made available. > > > > How am I meant to do this is the SNAP world? > Hey Dave, Looking at the postinst, it doesn't seem this is actually that you need to execute any postinst in this specific case. The update-alternative calls here are just to make available a default "traceroute" binary pointing to your "usr/bin/traceroute.db" (same for manual page). You can expose it directly through apps to get the same result: apps: traceroute: command: $SNAP/usr/bin/traceroute.db If the issue is that other commands from this package are using "traceroute" internally, you can ship a symlink yourself that you ensure are in your $PATH snap. Cheers, Didier > > > Thanks, > > > > Dave Garrod > > > > % cat DEBIAN/postinst > > #!/bin/sh > > > > set -e > > > > case "${1}" in > > configure) > > update-alternatives --install \ > > /usr/bin/traceroute traceroute > /usr/bin/traceroute.db 100 \ > > --slave /usr/sbin/traceroute traceroute.sbin \ > > /usr/bin/traceroute.db \ > > --slave /usr/share/man/man1/traceroute.1.gz > traceroute.1.gz \ > > /usr/share/man/man1/traceroute.db.1.gz > > > > for _ALTERNATIVE in traceroute6 lft traceproto > > do > > update-alternatives --install \ > > /usr/bin/${_ALTERNATIVE} > ${_ALTERNATIVE} /usr/bin/${_ALTERNATIVE}.db 100 \ > > --slave > /usr/share/man/man1/${_ALTERNATIVE}.1.gz ${_ALTERNATIVE}.1.gz \ > > > /usr/share/man/man1/${_ALTERNATIVE}.db.1.gz > > done > > > > for _ALTERNATIVE in tcptraceroute > > do > > update-alternatives --install \ > > /usr/sbin/${_ALTERNATIVE} > ${_ALTERNATIVE} /usr/sbin/${_ALTERNATIVE}.db 100 \ > > --slave > /usr/share/man/man8/${_ALTERNATIVE}.8.gz ${_ALTERNATIVE}.8.gz \ > > > /usr/share/man/man8/${_ALTERNATIVE}.db.8.gz > > done > > ;; > > > > abort-upgrade|abort-remove|abort-deconfigure) > > > > ;; > > > > *) > > echo "postinst called with unknown argument \`${1}'" >&2 > > exit 1 > > ;; > > esac > > > > > > > > exit 0 > > > > > ------------------------------------------------------------------------ > > DISCLAIMER: > This e-mail and any attachments to it may contain confidential and > proprietary material and is solely for the use of the intended > recipient. Any review, use, disclosure, distribution or copying of > this transmittal is prohibited except by or on behalf of the intended > recipient. If you have received this transmittal in error, please > notify the sender and destroy this e-mail and any attachments and all > copies, whether electronic or printed. > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From selene.scriven at canonical.com Thu Jul 28 07:11:42 2016 From: selene.scriven at canonical.com (Selene Scriven) Date: Thu, 28 Jul 2016 01:11:42 -0600 Subject: mpv snap In-Reply-To: <20160727153004.6f533128@utnubu> References: <0e885aa2-3802-ff0e-9d39-46087b6e4461@canonical.com> <1469036453.20475.9.camel@canonical.com> <20160727014651.GB23091@mutt.xyzz.org> <1469613261.4310.21.camel@anubis> <20160727153004.6f533128@utnubu> Message-ID: <20160728071142.GD23091@mutt.xyzz.org> * Ralf Mardorf wrote: > On Wed, 27 Jul 2016 11:54:21 +0200, Oliver Grawert wrote: > >well, you dont really need alsa *access* you can ship the libs and > >config in your snap (and have full control over your dependencies), > >what you need is direct device access for that setup > > Yesno, there could be the need to mix audio streams, jackd, dmix or > what ever else, so direct access to the devices isn't helpful, as soon > as two snaps try to access one audio device. IIUC you already made > pulseaudio available ;), but pulseaudio doesn't fit to everybody's > needs. Jackd might not fit to this mpv direct alsa access example, but > dmix does. OTOH I'm not using dmix, so I don't know how it works, > perhaps dmix could be considered to quasi be the audio device. Yes, I was thinking probably dmix, since it should be able to give multiple snaps simultaneous audio. Of course, for a music-oriented device it'd be nice to have a jackd service instead, and many general-purpose devices get along well with pulseaudio. But for my personal use I've found alsa with dmix to be the most efficient and reliable option most of the time. -- Selene From woodrow.shen at canonical.com Thu Jul 28 09:50:54 2016 From: woodrow.shen at canonical.com (Woodrow Shen) Date: Thu, 28 Jul 2016 17:50:54 +0800 Subject: no hardware description in Gadget snap for building an image Message-ID: Hi all, It seems the latest revision 8 has a lack of hardware section for gadget snap. I think someone is working on it now, but hope we can make sure the snap uploaded to edge channel can work at least. What I saw the content in revision 8 is == name: canonical-pc version: 16.04-0.2 summary: AMD64 generic package description: | This package contains a simple OEM snappy package for system configuration type: gadget Thanks, -- Snapcraft mailing list Snapcraft at lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft -------------- next part -------------- An HTML attachment was scrubbed... URL: From dietmar.winkler at dwe.no Thu Jul 28 13:52:01 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Thu, 28 Jul 2016 15:52:01 +0200 Subject: Problems with PREFIX when building a snap Message-ID: Hi, I'm currently trying to buld my first snap. With the help of the #snappy irc channel I could finally create a snapp package but it seems to be empty and non of the binary are actually contained. Running `snapcraft` on https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml generates this structure: https://gist.github.com/dietmarw/422354f1906bcb2babacd0168349ea55 Doing `sudo snapcraft install` then gives https://gist.github.com/dietmarw/7622ba197f64b4b89dbc79b60b86d348 One thing that threw snapcraft earlier was the definition of "IPEPREFIX" . But that is needed by the Makefile of the project. Setting it to something that is writable by the user running snapcraft was the only way to get it working. I've ended up using IPEPREFIX=$SNAPCRAFT_STAGE (as mentioned in https://developer.ubuntu.com/en/snappy/build-apps/snapcraft-advanced-features/) but still don't know if this is the correct "installation" directory as required by the Makefile. The snap is supposed to just compile ipe just like described here: https://github.com/otfried/ipe-wiki/wiki/Downloading,%20Compiling,%20and%20Installing%20Ipe#debian-ubuntu-mint which works without problems. I'm a but at a loss of what it is I'm doing wrong. Also snapcraft is not really reporting any problems or what is missing :-( Anyone got an idea what is missing? -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From dgarrod at extremenetworks.com Thu Jul 28 16:32:24 2016 From: dgarrod at extremenetworks.com (David Garrod) Date: Thu, 28 Jul 2016 16:32:24 +0000 Subject: How do I get a postinst stage properly executed - traceroute will not install correctly In-Reply-To: References: Message-ID: Thanks for your reply. Re: Ø Looking at the postinst, it doesn't seem this is actually that you need to execute any postinst in this specific case. The update-alternative calls here are just to make available a default "traceroute" binary pointing to your "usr/bin/traceroute.db" (same for manual page). You can expose it directly through apps to get the same result: apps: traceroute: command: $SNAP/usr/bin/traceroute.db If the issue is that other commands from this package are using "traceroute" internally, you can ship a symlink yourself that you ensure are in your $PATH snap. Yes I know I can do some special messing around to try and fix up the installation but I don't think that is a good idea. How can I get the package to install fully and correctly inside the SNAP, i.e. properly execute the "postinst" commands. If I install the package outside the SNAP using 'dpkg install' or 'apt-get install' it properly installs on a UBUNTU system. What I'm looking for is the mechanism to get the package installed inside the SNAP. The current behavior just seems wrong to me. Are there any plans to fix it? Thanks, Dave ________________________________ DISCLAIMER: This e-mail and any attachments to it may contain confidential and proprietary material and is solely for the use of the intended recipient. Any review, use, disclosure, distribution or copying of this transmittal is prohibited except by or on behalf of the intended recipient. If you have received this transmittal in error, please notify the sender and destroy this e-mail and any attachments and all copies, whether electronic or printed. -------------- next part -------------- An HTML attachment was scrubbed... URL: From james.jesudason at canonical.com Thu Jul 28 16:34:07 2016 From: james.jesudason at canonical.com (James Jesudason) Date: Thu, 28 Jul 2016 17:34:07 +0100 Subject: Problems with PREFIX when building a snap In-Reply-To: References: Message-ID: If you refer to http://snapcraft.io/create it mentions the different directories and their uses: "parts/ contains one subdirectory for each part: gnu-bash and gnu-hello. Each part's subdirectory has a src/ directory (to where the code is pulled), build/ where the build happens, and install/ where make install is processed. stage/ is where all the content from parts//install is copied and consolidated in a single directory. prime/ is the actual snap content. The stage/ directory is copied there. Binaries are stripped, wrappers are generated by Snapcraft, and the snap metadata files are also generated there. The final .snap file is just a squashfs archive of this prime/ directory. You can use and debug the snap's content from here" So you need to get the "make" process to deliver the binaries to the "parts//install" directory. In your case, this can be done via: - IPEPREFIX=../../../install (there's probably a more elegant way of getting path, but that seems to work ). Then, you need to make sure that the command to launch the app is correct. In your case, this looks like: apps: ipe: command: bin/ipe With those changes, I was able to get a ipe_7.2.5_amd64.snap file built. You can then install it using: sudo snap install ipe_7.2.5_amd64.snap On 28 July 2016 at 14:52, Dietmar Winkler wrote: > Hi, > > I'm currently trying to buld my first snap. With the help of the > #snappy irc channel I could finally create a snapp package but it > seems to be empty and non of the binary are actually contained. > > Running `snapcraft` on > https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml > > generates this structure: > https://gist.github.com/dietmarw/422354f1906bcb2babacd0168349ea55 > > Doing `sudo snapcraft install` then gives > https://gist.github.com/dietmarw/7622ba197f64b4b89dbc79b60b86d348 > > One thing that threw snapcraft earlier was the definition of > "IPEPREFIX" . But that is needed by the Makefile of the project. > Setting it to something that is writable by the user running snapcraft > was the only way to get it working. I've ended up using > > IPEPREFIX=$SNAPCRAFT_STAGE > > (as mentioned in > > https://developer.ubuntu.com/en/snappy/build-apps/snapcraft-advanced-features/ > ) > > but still don't know if this is the correct "installation" directory > as required by the Makefile. > > The snap is supposed to just compile ipe just like described here: > > https://github.com/otfried/ipe-wiki/wiki/Downloading,%20Compiling,%20and%20Installing%20Ipe#debian-ubuntu-mint > > which works without problems. > > I'm a but at a loss of what it is I'm doing wrong. Also snapcraft is > not really reporting any problems or what is missing :-( > > Anyone got an idea what is missing? > > -- > /Dietmar/ > > Secure email communication: > https://encrypt.to/dietmar.winkler at dwe.no > Public OpenPGP key: 0x235E6689 > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dietmar.winkler at dwe.no Thu Jul 28 17:36:41 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Thu, 28 Jul 2016 19:36:41 +0200 Subject: Problems with PREFIX when building a snap In-Reply-To: References: Message-ID: Thanks James, you are right the build works now. But I hope you don't mind me asking (since I really want to understand it) how you ended up at that IPEPREFIX path. The source is in: ~/tmp/snap/ipe/parts/ipe/src For the build that gets copied over to ~/tmp/snap/ipe/parts/ipe/build make gets executed in the subdir "src", which means now we are in ~/tmp/snap/ipe/parts/ipe/build/src Now I try to travers 3 levels up as defined by - IPEPREFIX=../../../install and I end up in ~/tmp/snap/ipe/parts/ but here is only the "ipe" dir the install is one level below: ~/tmp/snap/ipe/parts[master]$ tree -L 2 . └── ipe ├── build ├── install ├── src ├── state └── ubuntu So how come - IPEPREFIX=../../../install works but - IPEPREFIX=../../install would not (which I would find logical)? I'm rather puzzled. One other question, is there somewhere a list of available "plugs". I guessed I would need "x11" but only found this reference in some examples. I did not manage to find any specific documentation on this. Thanks, On 28 July 2016 at 18:34, James Jesudason wrote: > If you refer to http://snapcraft.io/create it mentions the different > directories and their uses: > > "parts/ contains one subdirectory for each part: gnu-bash and gnu-hello. > Each part's subdirectory has a src/ directory (to where the code is pulled), > build/ where the build happens, and install/ where make install is > processed. > > stage/ is where all the content from parts//install is copied and > consolidated in a single directory. > > prime/ is the actual snap content. The stage/ directory is copied there. > Binaries are stripped, wrappers are generated by Snapcraft, and the snap > metadata files are also generated there. The final .snap file is just a > squashfs archive of this prime/ directory. You can use and debug the snap's > content from here" > > > So you need to get the "make" process to deliver the binaries to the > "parts//install" directory. In your case, this can be done via: > > - IPEPREFIX=../../../install > > (there's probably a more elegant way of getting path, but that seems to > work). > > Then, you need to make sure that the command to launch the app is correct. > In your case, this looks like: > > apps: > ipe: > command: bin/ipe > > With those changes, I was able to get a ipe_7.2.5_amd64.snap file built. You > can then install it using: > > sudo snap install ipe_7.2.5_amd64.snap > > > > On 28 July 2016 at 14:52, Dietmar Winkler wrote: >> >> Hi, >> >> I'm currently trying to buld my first snap. With the help of the >> #snappy irc channel I could finally create a snapp package but it >> seems to be empty and non of the binary are actually contained. >> >> Running `snapcraft` on >> https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml >> >> generates this structure: >> https://gist.github.com/dietmarw/422354f1906bcb2babacd0168349ea55 >> >> Doing `sudo snapcraft install` then gives >> https://gist.github.com/dietmarw/7622ba197f64b4b89dbc79b60b86d348 >> >> One thing that threw snapcraft earlier was the definition of >> "IPEPREFIX" . But that is needed by the Makefile of the project. >> Setting it to something that is writable by the user running snapcraft >> was the only way to get it working. I've ended up using >> >> IPEPREFIX=$SNAPCRAFT_STAGE >> >> (as mentioned in >> >> https://developer.ubuntu.com/en/snappy/build-apps/snapcraft-advanced-features/) >> >> but still don't know if this is the correct "installation" directory >> as required by the Makefile. >> >> The snap is supposed to just compile ipe just like described here: >> >> https://github.com/otfried/ipe-wiki/wiki/Downloading,%20Compiling,%20and%20Installing%20Ipe#debian-ubuntu-mint >> >> which works without problems. >> >> I'm a but at a loss of what it is I'm doing wrong. Also snapcraft is >> not really reporting any problems or what is missing :-( >> >> Anyone got an idea what is missing? >> >> -- >> /Dietmar/ >> >> Secure email communication: >> https://encrypt.to/dietmar.winkler at dwe.no >> Public OpenPGP key: 0x235E6689 >> >> -- >> Snapcraft mailing list >> Snapcraft at lists.snapcraft.io >> Modify settings or unsubscribe at: >> https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From dietmar.winkler at dwe.no Thu Jul 28 18:01:00 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Thu, 28 Jul 2016 20:01:00 +0200 Subject: Multiple app per part generate "dot" executables Message-ID: Hi, I have one part that when compiled provides several binaries. See : https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L50 The problem is that the only "clean" bin binary is ipe all the other apps get installed as: $ ls /snap/bin/ipe* /snap/bin/ipe /snap/bin/ipe.ipeextract /snap/bin/ipe.ipescript /snap/bin/ipe.ipe6upgrade /snap/bin/ipe.iperender /snap/bin/ipe.ipetoipe So how can I avoid having to call the apps with "ipe." in front? -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From michael.vogt at canonical.com Thu Jul 28 16:49:04 2016 From: michael.vogt at canonical.com (Michael Vogt) Date: Thu, 28 Jul 2016 18:49:04 +0200 Subject: no hardware description in Gadget snap for building an image In-Reply-To: References: Message-ID: <20160728164904.GS2824@bod> On Thu, Jul 28, 2016 at 05:50:54PM +0800, Woodrow Shen wrote: > Hi all, Hi Woodrow, thanks for your mail! > It seems the latest revision 8 has a lack of hardware section for gadget > snap. > I think someone is working on it now, but hope we can make sure the snap > uploaded to edge channel can work at least. You are correct, we are currently working on a new and improved format to describe the bootloader for a device. This work is currently in progress and because the old format stops working we removed it. We publish new gadget snaps with the new format soon. ubuntu-image will be the tool to consume them then. Cheers, Michael From dietmar.winkler at dwe.no Thu Jul 28 18:05:21 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Thu, 28 Jul 2016 20:05:21 +0200 Subject: Getting QT apps running. Message-ID: Hi, when trying to run this "ipe" qt app I got the error message: $ ipe This application failed to start because it could not find or load the Qt platform plugin "xcb". I found out that I need to include the libqt5gui5 package as stage package (just like in the hello-world-desktop example, But even though I added it to my yaml file I still get the same error message. https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L44 (I'm actually not sure if I really need the "after: [qt5conf]" part.) So what else am I missing? -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From kyle.fazzari at canonical.com Thu Jul 28 18:21:36 2016 From: kyle.fazzari at canonical.com (Kyle Fazzari) Date: Thu, 28 Jul 2016 11:21:36 -0700 Subject: Getting QT apps running. In-Reply-To: References: Message-ID: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> On 07/28/2016 11:05 AM, Dietmar Winkler wrote: > Hi, > > when trying to run this "ipe" qt app I got the error message: > $ ipe > This application failed to start because it could not find or load the > Qt platform plugin "xcb". > > I found out that I need to include the libqt5gui5 package as stage > package (just like in the hello-world-desktop example, But even though > I added it to my yaml file I still get the same error message. > > https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L44 > (I'm actually not sure if I really need the "after: [qt5conf]" part.) > > So what else am I missing? The problem is that your application isn't looking in the right place for the platform plugins. That's where the qt5conf part comes in handy-- it includes a launcher script to set that environment for you. You need to use the launcher from the qt5conf part in order to launch the app. apps: ipe: command: bin/ipe Needs to be: apps: ipe: command: qt5-launch bin/ipe And so on. -- Kyle Fazzari (kyrofa) Software Engineer Canonical Ltd. kyle at canonical.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 836 bytes Desc: OpenPGP digital signature URL: From dietmar.winkler at hit.no Thu Jul 28 18:34:27 2016 From: dietmar.winkler at hit.no (Dietmar Winkler) Date: Thu, 28 Jul 2016 20:34:27 +0200 Subject: Getting QT apps running. In-Reply-To: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> Message-ID: Hi, thanks. That let me start the app ... almost. I'm now getting: $ ipe Qt: Session management error: None of the authentication protocols specified are supported QGtkStyle could not resolve GTK. Make sure you have installed the proper libraries. [string "require "main""]:1: module 'main' not found: no field package.preload['main'] no file '../../../install/share/ipe/7.2.5/lua/main.lua' no file '/usr/local/lib/lua/5.2/main.so' no file '/usr/lib/x86_64-linux-gnu/lua/5.2/main.so' no file '/usr/lib/lua/5.2/main.so' no file '/usr/local/lib/lua/5.2/loadall.so' no file './main.so' stack traceback: [C]: in function 'require' [string "require "main""]:1: in main chunk On 28 July 2016 at 20:21, Kyle Fazzari wrote: > On 07/28/2016 11:05 AM, Dietmar Winkler wrote: >> Hi, >> >> when trying to run this "ipe" qt app I got the error message: >> $ ipe >> This application failed to start because it could not find or load the >> Qt platform plugin "xcb". >> >> I found out that I need to include the libqt5gui5 package as stage >> package (just like in the hello-world-desktop example, But even though >> I added it to my yaml file I still get the same error message. >> >> https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L44 >> (I'm actually not sure if I really need the "after: [qt5conf]" part.) >> >> So what else am I missing? > > The problem is that your application isn't looking in the right place > for the platform plugins. That's where the qt5conf part comes in handy-- > it includes a launcher script to set that environment for you. > > You need to use the launcher from the qt5conf part in order to launch > the app. > > apps: > ipe: > command: bin/ipe > > Needs to be: > > apps: > ipe: > command: qt5-launch bin/ipe > > And so on. > > -- > Kyle Fazzari (kyrofa) > Software Engineer > Canonical Ltd. > kyle at canonical.com > > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft > -- Dietmar Winkler, Assistant professor University College of Southeast Norway Faculty of Technology Campus: Porsgrunn http://me.dwe.no Kjølnes ring 56, 3918 Porsgrunn, Norway From mhall119 at ubuntu.com Thu Jul 28 19:18:08 2016 From: mhall119 at ubuntu.com (Michael Hall) Date: Thu, 28 Jul 2016 15:18:08 -0400 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> Message-ID: <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> "qt5conf" has been replaced be the newer "desktop/qt5" part and "qt5-launch" replaced by "desktop-launch" in the command. These new desktop launchers have fixed a lot of the common problems that the original qt5conf and gtk-equiqalents ran into. Michael Hall mhall119 at ubuntu.com On 07/28/2016 02:34 PM, Dietmar Winkler wrote: > Hi, > > thanks. That let me start the app ... almost. I'm now getting: > > $ ipe > Qt: Session management error: None of the authentication protocols > specified are supported > QGtkStyle could not resolve GTK. Make sure you have installed the > proper libraries. > [string "require "main""]:1: module 'main' not found: > no field package.preload['main'] > no file '../../../install/share/ipe/7.2.5/lua/main.lua' > no file '/usr/local/lib/lua/5.2/main.so' > no file '/usr/lib/x86_64-linux-gnu/lua/5.2/main.so' > no file '/usr/lib/lua/5.2/main.so' > no file '/usr/local/lib/lua/5.2/loadall.so' > no file './main.so' > stack traceback: > [C]: in function 'require' > [string "require "main""]:1: in main chunk > > > > On 28 July 2016 at 20:21, Kyle Fazzari wrote: >> On 07/28/2016 11:05 AM, Dietmar Winkler wrote: >>> Hi, >>> >>> when trying to run this "ipe" qt app I got the error message: >>> $ ipe >>> This application failed to start because it could not find or load the >>> Qt platform plugin "xcb". >>> >>> I found out that I need to include the libqt5gui5 package as stage >>> package (just like in the hello-world-desktop example, But even though >>> I added it to my yaml file I still get the same error message. >>> >>> https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L44 >>> (I'm actually not sure if I really need the "after: [qt5conf]" part.) >>> >>> So what else am I missing? >> >> The problem is that your application isn't looking in the right place >> for the platform plugins. That's where the qt5conf part comes in handy-- >> it includes a launcher script to set that environment for you. >> >> You need to use the launcher from the qt5conf part in order to launch >> the app. >> >> apps: >> ipe: >> command: bin/ipe >> >> Needs to be: >> >> apps: >> ipe: >> command: qt5-launch bin/ipe >> >> And so on. >> >> -- >> Kyle Fazzari (kyrofa) >> Software Engineer >> Canonical Ltd. >> kyle at canonical.com >> >> >> -- >> Snapcraft mailing list >> Snapcraft at lists.snapcraft.io >> Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft >> > > > From didrocks at ubuntu.com Fri Jul 29 06:15:22 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Fri, 29 Jul 2016 08:15:22 +0200 Subject: How do I get a postinst stage properly executed - traceroute will not install correctly In-Reply-To: References: Message-ID: Le 28/07/2016 à 18:32, David Garrod a écrit : > Thanks for your reply. > > > > Re: > > > > Ø Looking at the postinst, it doesn't seem this is actually that you > need to execute any postinst in this specific case. The > update-alternative calls here are just to make available a default > "traceroute" binary pointing to your "usr/bin/traceroute.db" (same for > manual page). You can expose it directly through apps to get the same > result: > apps: > traceroute: > command: $SNAP/usr/bin/traceroute.db > > If the issue is that other commands from this package are using > "traceroute" internally, you can ship a symlink yourself that you ensure > are in your $PATH snap. > > Yes I know I can do some special messing around to try and fix up the > installation but I don’t think that is a good idea. How can I get the > package to install fully and correctly inside the SNAP, i.e. properly > execute the “postinst” commands. > > > > If I install the package outside the SNAP using ‘dpkg install’ or > ‘apt-get install’ it properly installs on a UBUNTU system. What I’m > looking for is the mechanism to get the package installed inside the SNAP. > > > > The current behavior just seems wrong to me. Are there any plans to fix it? I'll let the snapcraft team answering this, but I don't think this is even doable. Postinst scripts executes as root and do system level actions. This goes against what we are trying to do with snapd, as this void the transactional update/rollback mechanism (what does those scripts do?) and security garantees as those scripts have access to the whole system, while running as root. Cheers, Didier From didrocks at ubuntu.com Fri Jul 29 06:19:08 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Fri, 29 Jul 2016 08:19:08 +0200 Subject: Multiple app per part generate "dot" executables In-Reply-To: References: Message-ID: <03e7a387-bd9b-b229-3819-a25aa1c1f545@ubuntu.com> Le 28/07/2016 à 20:01, Dietmar Winkler a écrit : > Hi, > > I have one part that when compiled provides several binaries. > > See : https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L50 > > The problem is that the only "clean" bin binary is ipe all the other > apps get installed as: > > $ ls /snap/bin/ipe* > /snap/bin/ipe /snap/bin/ipe.ipeextract /snap/bin/ipe.ipescript > /snap/bin/ipe.ipe6upgrade /snap/bin/ipe.iperender /snap/bin/ipe.ipetoipe > > So how can I avoid having to call the apps with "ipe." in front? > Hey, This is the core part of the snapd design: commands are namespaced (with a default command shorten if app_name == snap_name). That way, you can avoid clash and conflicts when multiple snaps are installed on the system as they can't claim the same binary names, compared to a traditional packaging format. -------------- next part -------------- An HTML attachment was scrubbed... URL: From dietmar.winkler at hit.no Fri Jul 29 06:33:38 2016 From: dietmar.winkler at hit.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 08:33:38 +0200 Subject: Multiple app per part generate "dot" executables In-Reply-To: <03e7a387-bd9b-b229-3819-a25aa1c1f545@ubuntu.com> References: <03e7a387-bd9b-b229-3819-a25aa1c1f545@ubuntu.com> Message-ID: Hi, so is there any way around that. I don't want to have to multiply the snaps for each subapplication. From a usesage perspective this pretty much would be a show stopper to use snap over classical packages when suddenly the same applications have to be called differently because they have been installed as a snap and not as a deb. On 29 July 2016 at 08:19, Didier Roche wrote: > Le 28/07/2016 à 20:01, Dietmar Winkler a écrit : > > Hi, > > I have one part that when compiled provides several binaries. > > See : https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L50 > > The problem is that the only "clean" bin binary is ipe all the other > apps get installed as: > > $ ls /snap/bin/ipe* > /snap/bin/ipe /snap/bin/ipe.ipeextract /snap/bin/ipe.ipescript > /snap/bin/ipe.ipe6upgrade /snap/bin/ipe.iperender /snap/bin/ipe.ipetoipe > > So how can I avoid having to call the apps with "ipe." in front? > > Hey, > > This is the core part of the snapd design: commands are namespaced (with > a default command shorten if app_name == snap_name). > > That way, you can avoid clash and conflicts when multiple snaps are > installed on the system as they can't claim the same binary names, > compared to a traditional packaging format. > > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -- Dietmar Winkler, Assistant professor University College of Southeast Norway Faculty of Technology Campus: Porsgrunn http://me.dwe.no Kjølnes ring 56, 3918 Porsgrunn, Norway From timo.jyrinki at canonical.com Fri Jul 29 07:14:12 2016 From: timo.jyrinki at canonical.com (Timo Jyrinki) Date: Fri, 29 Jul 2016 10:14:12 +0300 Subject: Getting QT apps running. In-Reply-To: <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> Message-ID: Changing qt5conf to desktop/qt5 and qt5-launch to desktop-launch does not help in ipe's case it seems, even though it gets rid of a few warnings. The problem is with the IPE_PREFIX that ipe requires to be set, and I'm also unable to find a combination that would allow ipe to build (../../../install works) and ipe to find its main.lua file from the correct directory. It seemed useful to remove the "bin/" prefix though after the changes. -Timo (snapcraft newbie, testing because Qt mentioned) On Thu, Jul 28, 2016 at 10:18 PM, Michael Hall wrote: > "qt5conf" has been replaced be the newer "desktop/qt5" part and > "qt5-launch" replaced by "desktop-launch" in the command. These new > desktop launchers have fixed a lot of the common problems that the > original qt5conf and gtk-equiqalents ran into. > > > Michael Hall > mhall119 at ubuntu.com > > On 07/28/2016 02:34 PM, Dietmar Winkler wrote: >> Hi, >> >> thanks. That let me start the app ... almost. I'm now getting: >> >> $ ipe >> Qt: Session management error: None of the authentication protocols >> specified are supported >> QGtkStyle could not resolve GTK. Make sure you have installed the >> proper libraries. >> [string "require "main""]:1: module 'main' not found: >> no field package.preload['main'] >> no file '../../../install/share/ipe/7.2.5/lua/main.lua' >> no file '/usr/local/lib/lua/5.2/main.so' >> no file '/usr/lib/x86_64-linux-gnu/lua/5.2/main.so' >> no file '/usr/lib/lua/5.2/main.so' >> no file '/usr/local/lib/lua/5.2/loadall.so' >> no file './main.so' >> stack traceback: >> [C]: in function 'require' >> [string "require "main""]:1: in main chunk >> >> >> >> On 28 July 2016 at 20:21, Kyle Fazzari wrote: >>> On 07/28/2016 11:05 AM, Dietmar Winkler wrote: >>>> Hi, >>>> >>>> when trying to run this "ipe" qt app I got the error message: >>>> $ ipe >>>> This application failed to start because it could not find or load the >>>> Qt platform plugin "xcb". >>>> >>>> I found out that I need to include the libqt5gui5 package as stage >>>> package (just like in the hello-world-desktop example, But even though >>>> I added it to my yaml file I still get the same error message. >>>> >>>> https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L44 >>>> (I'm actually not sure if I really need the "after: [qt5conf]" part.) >>>> >>>> So what else am I missing? >>> >>> The problem is that your application isn't looking in the right place >>> for the platform plugins. That's where the qt5conf part comes in handy-- >>> it includes a launcher script to set that environment for you. >>> >>> You need to use the launcher from the qt5conf part in order to launch >>> the app. >>> >>> apps: >>> ipe: >>> command: bin/ipe >>> >>> Needs to be: >>> >>> apps: >>> ipe: >>> command: qt5-launch bin/ipe >>> >>> And so on. >>> >>> -- >>> Kyle Fazzari (kyrofa) >>> Software Engineer >>> Canonical Ltd. >>> kyle at canonical.com >>> >>> >>> -- >>> Snapcraft mailing list >>> Snapcraft at lists.snapcraft.io >>> Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft >>> >> >> >> > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft From james.jesudason at canonical.com Fri Jul 29 08:26:45 2016 From: james.jesudason at canonical.com (James Jesudason) Date: Fri, 29 Jul 2016 09:26:45 +0100 Subject: Problems with PREFIX when building a snap In-Reply-To: References: Message-ID: On 28 July 2016 at 18:36, Dietmar Winkler wrote: > Thanks James, > > you are right the build works now. But I hope you don't mind me asking > (since I really want to understand it) how you ended up at that > IPEPREFIX path. > > The source is in: > > ~/tmp/snap/ipe/parts/ipe/src > > For the build that gets copied over to > > ~/tmp/snap/ipe/parts/ipe/build > > make gets executed in the subdir "src", which means now we are in > > ~/tmp/snap/ipe/parts/ipe/build/src > > Now I try to travers 3 levels up as defined by > > - IPEPREFIX=../../../install > > and I end up in > > ~/tmp/snap/ipe/parts/ > > but here is only the "ipe" dir the install is one level below: > > ~/tmp/snap/ipe/parts[master]$ tree -L 2 > . > └── ipe > ├── build > ├── install > ├── src > ├── state > └── ubuntu > > So how come > > - IPEPREFIX=../../../install > > works but > > - IPEPREFIX=../../install > > would not (which I would find logical)? > > ​It's more to do with the build script than with snapcraft. Since there are multiple directories called "install" it's hard to see what's happening. If you try something like "../xxxx" instead, then it's easier to spot what the script is doing. ​ ​When the script was running, I spotted th​at it said: make[2]: Entering directory '/parts/ipe/build/src/ipe' ​So, that's where the extra level comes from.​ > > One other question, is there somewhere a list of available "plugs". I > guessed I would need "x11" but only found this reference in some > examples. I did not manage to find any specific documentation on this. > > ​The plugs are documented with snapd rather than snapcraft. There's a list here: https://developer.ubuntu.com/en/snappy/guides/interfaces/ But the definitive list is the source code (sometimes the tests there can be helpful): https://github.com/snapcore/snapd/tree/master/interfaces/builtin​ ​The "playpen" examples can be helpful as well: https://github.com/ubuntu/snappy-playpen ​ -------------- next part -------------- An HTML attachment was scrubbed... URL: From dietmar.winkler at hit.no Fri Jul 29 09:19:17 2016 From: dietmar.winkler at hit.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 11:19:17 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> Message-ID: So I decided to hack the source of the ipe package in order to get rid of the path issue possibly causing other problems. So the updated snapcraft.yaml pulls in "my version" that does not need a IPEPREFIX any more. https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml But as Timo, I found a couple of other problems. I first tried to only use the x11 and home plug but then launching ipe simply does nothing and sits there until I kill the command. I then tested the unity7 and home plug. Now I'm getting: --8<--- $ ipe ln: failed to create symbolic link '/home/dietmarw/snap/ipe/x17/.themes/themes': Read-only file system (process:8847): Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale. Gtk-Message: Failed to load module "overlay-scrollbar" Gtk-Message: Failed to load module "gtk-vector-screenshot" Gtk-Message: Failed to load module "gail" Gtk-Message: Failed to load module "atk-bridge" Gtk-Message: Failed to load module "unity-gtk-module" Gtk-Message: Failed to load module "canberra-gtk-module" Qt: Session management error: None of the authentication protocols specified are supported [string "require "main""]:1: module 'main' not found: no field package.preload['main'] no file '/share/ipe/7.2.5/lua/main.lua' no file '/usr/local/lib/lua/5.2/main.so' no file '/usr/lib/x86_64-linux-gnu/lua/5.2/main.so' no file '/usr/lib/lua/5.2/main.so' no file '/usr/local/lib/lua/5.2/loadall.so' no file './main.so' stack traceback: [C]: in function 'require' [string "require "main""]:1: in main chunk ---8<---- The main.lua file is actually present $ tree /snap/ipe/current/share/ipe/7.2.5/lua/ /snap/ipe/current/share/ipe/7.2.5/lua/ ├── actions.lua ├── editpath.lua ├── main.lua ├── model.lua ├── mouse.lua ├── prefs.lua ├── properties.lua ├── shortcuts.lua └── tools.lua On 29 July 2016 at 09:14, Timo Jyrinki wrote: > Changing qt5conf to desktop/qt5 and qt5-launch to desktop-launch does > not help in ipe's case it seems, even though it gets rid of a few > warnings. > > The problem is with the IPE_PREFIX that ipe requires to be set, and > I'm also unable to find a combination that would allow ipe to build > (../../../install works) and ipe to find its main.lua file from the > correct directory. > > It seemed useful to remove the "bin/" prefix though after the changes. > > -Timo (snapcraft newbie, testing because Qt mentioned) > > On Thu, Jul 28, 2016 at 10:18 PM, Michael Hall wrote: >> "qt5conf" has been replaced be the newer "desktop/qt5" part and >> "qt5-launch" replaced by "desktop-launch" in the command. These new >> desktop launchers have fixed a lot of the common problems that the >> original qt5conf and gtk-equiqalents ran into. >> >> >> Michael Hall >> mhall119 at ubuntu.com >> >> On 07/28/2016 02:34 PM, Dietmar Winkler wrote: >>> Hi, >>> >>> thanks. That let me start the app ... almost. I'm now getting: >>> >>> $ ipe >>> Qt: Session management error: None of the authentication protocols >>> specified are supported >>> QGtkStyle could not resolve GTK. Make sure you have installed the >>> proper libraries. >>> [string "require "main""]:1: module 'main' not found: >>> no field package.preload['main'] >>> no file '../../../install/share/ipe/7.2.5/lua/main.lua' >>> no file '/usr/local/lib/lua/5.2/main.so' >>> no file '/usr/lib/x86_64-linux-gnu/lua/5.2/main.so' >>> no file '/usr/lib/lua/5.2/main.so' >>> no file '/usr/local/lib/lua/5.2/loadall.so' >>> no file './main.so' >>> stack traceback: >>> [C]: in function 'require' >>> [string "require "main""]:1: in main chunk >>> >>> >>> >>> On 28 July 2016 at 20:21, Kyle Fazzari wrote: >>>> On 07/28/2016 11:05 AM, Dietmar Winkler wrote: >>>>> Hi, >>>>> >>>>> when trying to run this "ipe" qt app I got the error message: >>>>> $ ipe >>>>> This application failed to start because it could not find or load the >>>>> Qt platform plugin "xcb". >>>>> >>>>> I found out that I need to include the libqt5gui5 package as stage >>>>> package (just like in the hello-world-desktop example, But even though >>>>> I added it to my yaml file I still get the same error message. >>>>> >>>>> https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L44 >>>>> (I'm actually not sure if I really need the "after: [qt5conf]" part.) >>>>> >>>>> So what else am I missing? >>>> >>>> The problem is that your application isn't looking in the right place >>>> for the platform plugins. That's where the qt5conf part comes in handy-- >>>> it includes a launcher script to set that environment for you. >>>> >>>> You need to use the launcher from the qt5conf part in order to launch >>>> the app. >>>> >>>> apps: >>>> ipe: >>>> command: bin/ipe >>>> >>>> Needs to be: >>>> >>>> apps: >>>> ipe: >>>> command: qt5-launch bin/ipe >>>> >>>> And so on. >>>> >>>> -- >>>> Kyle Fazzari (kyrofa) >>>> Software Engineer >>>> Canonical Ltd. >>>> kyle at canonical.com >>>> >>>> >>>> -- >>>> Snapcraft mailing list >>>> Snapcraft at lists.snapcraft.io >>>> Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft >>>> >>> >>> >>> >> >> -- >> Snapcraft mailing list >> Snapcraft at lists.snapcraft.io >> Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft -- Dietmar Winkler, Assistant professor University College of Southeast Norway Faculty of Technology Campus: Porsgrunn http://me.dwe.no Kjølnes ring 56, 3918 Porsgrunn, Norway From dietmar.winkler at dwe.no Fri Jul 29 09:24:26 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 11:24:26 +0200 Subject: Problems with PREFIX when building a snap In-Reply-To: References: Message-ID: Hi James, OK now I understand. Still not all resolved but that can be handled in the other thread. https://lists.ubuntu.com/archives/snapcraft/2016-July/000561.html Thanks again, On 29 July 2016 at 10:26, James Jesudason wrote: > > > On 28 July 2016 at 18:36, Dietmar Winkler wrote: >> >> Thanks James, >> >> you are right the build works now. But I hope you don't mind me asking >> (since I really want to understand it) how you ended up at that >> IPEPREFIX path. >> >> The source is in: >> >> ~/tmp/snap/ipe/parts/ipe/src >> >> For the build that gets copied over to >> >> ~/tmp/snap/ipe/parts/ipe/build >> >> make gets executed in the subdir "src", which means now we are in >> >> ~/tmp/snap/ipe/parts/ipe/build/src >> >> Now I try to travers 3 levels up as defined by >> >> - IPEPREFIX=../../../install >> >> and I end up in >> >> ~/tmp/snap/ipe/parts/ >> >> but here is only the "ipe" dir the install is one level below: >> >> ~/tmp/snap/ipe/parts[master]$ tree -L 2 >> . >> └── ipe >> ├── build >> ├── install >> ├── src >> ├── state >> └── ubuntu >> >> So how come >> >> - IPEPREFIX=../../../install >> >> works but >> >> - IPEPREFIX=../../install >> >> would not (which I would find logical)? >> > It's more to do with the build script than with snapcraft. Since there are > multiple directories called "install" it's hard to see what's happening. If > you try something like "../xxxx" instead, then it's easier to spot what the > script is doing. > > When the script was running, I spotted that it said: > > make[2]: Entering directory '/parts/ipe/build/src/ipe' > > So, that's where the extra level comes from. > > >> >> >> One other question, is there somewhere a list of available "plugs". I >> guessed I would need "x11" but only found this reference in some >> examples. I did not manage to find any specific documentation on this. >> > > The plugs are documented with snapd rather than snapcraft. There's a list > here: > https://developer.ubuntu.com/en/snappy/guides/interfaces/ > > But the definitive list is the source code (sometimes the tests there can be > helpful): > https://github.com/snapcore/snapd/tree/master/interfaces/builtin > > The "playpen" examples can be helpful as well: > https://github.com/ubuntu/snappy-playpen > -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From ogra at ubuntu.com Fri Jul 29 09:27:18 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 29 Jul 2016 11:27:18 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> Message-ID: <1469784438.4310.58.camel@anubis> hi, Am Freitag, den 29.07.2016, 11:19 +0200 schrieb Dietmar Winkler: ... > specified are supported > [string "require "main""]:1: module 'main' not found: > no field package.preload['main'] > no file '/share/ipe/7.2.5/lua/main.lua' > no file '/usr/local/lib/lua/5.2/main.so' > no file '/usr/lib/x86_64-linux-gnu/lua/5.2/main.so' > no file '/usr/lib/lua/5.2/main.so' > no file '/usr/local/lib/lua/5.2/loadall.so' > no file './main.so' > stack traceback: > [C]: in function 'require' > [string "require "main""]:1: in main chunk > > ---8<---- > > The main.lua file is actually present > > $ tree /snap/ipe/current/share/ipe/7.2.5/lua/ > /snap/ipe/current/share/ipe/7.2.5/lua/ > ├── actions.lua > ├── editpath.lua > ├── main.lua > ├── model.lua > ├── mouse.lua > ├── prefs.lua > ├── properties.lua > ├── shortcuts.lua > └── tools.lua > any reason to not create another startup wrapper that sets up a proper environmant for you ? http://ipe.otfried.org/manual/manual_40.html talks about IPELUAPATH ... i'd just creaate a small shell script that you call from the desktop-launcher to add additional environment variables ... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From dietmar.winkler at dwe.no Fri Jul 29 09:32:41 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 11:32:41 +0200 Subject: Getting QT apps running. In-Reply-To: <1469784438.4310.58.camel@anubis> References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> Message-ID: Hi, I'm not sure if this will help. According to the output it _is_ already looking at the right place. > no file '/share/ipe/7.2.5/lua/main.lua' but $ tree /snap/ipe/current/share/ipe/7.2.5/lua/ /snap/ipe/current/share/ipe/7.2.5/lua/ ├── actions.lua ├── editpath.lua ├── main.lua ... -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From seb128 at ubuntu.com Fri Jul 29 09:39:46 2016 From: seb128 at ubuntu.com (Sebastien Bacher) Date: Fri, 29 Jul 2016 11:39:46 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> Message-ID: Le 29/07/2016 à 11:32, Dietmar Winkler a écrit : >> no file '/share/ipe/7.2.5/lua/main.lua' > but > > $ tree /snap/ipe/current/share/ipe/7.2.5/lua/ > /snap/ipe/current/share/ipe/7.2.5/lua/ It's look in /share but the actual directory is /snap/ipe/current/share, You might be able to workaround it with some environment variable but you might also be down to https://bugs.launchpad.net/ubuntu/+source/snapcraft/+bug/1583250 "upstream use of build-time defined DATADIR incompatible with snaps relocation" There is work ongoing to resolve those situation with smart syscall redirections but until that land you might be able to workaround using "/snap/ipe/current/share" as a build prefix and reorganize files in your prime if needed... Cheers, Sebastien Bacher -------------- next part -------------- An HTML attachment was scrubbed... URL: From ogra at ubuntu.com Fri Jul 29 09:40:04 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 29 Jul 2016 11:40:04 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> Message-ID: <1469785204.4310.64.camel@anubis> Am Freitag, den 29.07.2016, 11:32 +0200 schrieb Dietmar Winkler: > Hi, > > I'm not sure if this will help. According to the output it _is_ > already looking at the right place. > > > no file '/share/ipe/7.2.5/lua/main.lua' > > but > > $ tree /snap/ipe/current/share/ipe/7.2.5/lua/ > /snap/ipe/current/share/ipe/7.2.5/lua/ > ├── actions.lua > ├── editpath.lua > ├── main.lua > ... > well, it looks in /share, not in /snap .... try this: --- #! /bin/sh export IPELUAPATH=$SNAP/share/ipe/7.2.5/lua exec $SNAP/bin/ipe --- save it as wrapper.sh (make it executable), add a copy plugin part to your snapcraft.yaml that copies it in the toplevel of your snap and change: https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L52 to read: command: desktop-launch wrapper.sh if needed also set the other variables to their proper paths ... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From dietmar.winkler at dwe.no Fri Jul 29 09:43:04 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 11:43:04 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> Message-ID: On 29 July 2016 at 11:39, Sebastien Bacher wrote: > It's look in /share but the actual directory is /snap/ipe/current/share, Ah yes of course. I was assuming that the snap only knows about what's inside $SNAP and not the root file system. Hence I did not think those were to different locations. -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From gustavo.niemeyer at canonical.com Fri Jul 29 10:02:55 2016 From: gustavo.niemeyer at canonical.com (Gustavo Niemeyer) Date: Fri, 29 Jul 2016 12:02:55 +0200 Subject: no hardware description in Gadget snap for building an image In-Reply-To: <20160728164904.GS2824@bod> References: <20160728164904.GS2824@bod> Message-ID: For the record, please note this change was announced in this list a couple of months ago, and that we don't actually have official Ubuntu Core 16 images yet. Polishing the format is being done while finalizing those images. On Jul 28, 2016 8:03 PM, "Michael Vogt" wrote: > On Thu, Jul 28, 2016 at 05:50:54PM +0800, Woodrow Shen wrote: > > Hi all, > Hi Woodrow, > > thanks for your mail! > > > It seems the latest revision 8 has a lack of hardware section for gadget > > snap. > > I think someone is working on it now, but hope we can make sure the snap > > uploaded to edge channel can work at least. > > You are correct, we are currently working on a new and improved format > to describe the bootloader for a device. This work is currently in > progress and because the old format stops working we removed it. We > publish new gadget snaps with the new format soon. ubuntu-image will > be the tool to consume them then. > > Cheers, > Michael > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/snapcraft > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dietmar.winkler at dwe.no Fri Jul 29 10:45:38 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 12:45:38 +0200 Subject: Getting QT apps running. In-Reply-To: <1469785204.4310.64.camel@anubis> References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> Message-ID: Thanks Oli, I've implemented it as you suggested and got one step further: $ ipe ln: failed to create symbolic link '/home/dietmarw/snap/ipe/x22/.themes/themes': Read-only file system (process:31604): Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale. Gtk-Message: Failed to load module "overlay-scrollbar" Gtk-Message: Failed to load module "gtk-vector-screenshot" Gtk-Message: Failed to load module "gail" Gtk-Message: Failed to load module "atk-bridge" Gtk-Message: Failed to load module "unity-gtk-module" Gtk-Message: Failed to load module "canberra-gtk-module" Qt: Session management error: None of the authentication protocols specified are supported error loading module 'main' from file '/snap/ipe/x22/share/ipe/7.2.5/lua': cannot read /snap/ipe/x22/share/ipe/7.2.5/lua: Is a directory stack traceback: [C]: in ? [C]: in function 'require' [string "require "main""]:1: in main chunk but now I'm stuck again. On 29 July 2016 at 11:40, Oliver Grawert wrote: > Am Freitag, den 29.07.2016, 11:32 +0200 schrieb Dietmar Winkler: >> Hi, >> >> I'm not sure if this will help. According to the output it _is_ >> already looking at the right place. >> >> > no file '/share/ipe/7.2.5/lua/main.lua' >> >> but >> >> $ tree /snap/ipe/current/share/ipe/7.2.5/lua/ >> /snap/ipe/current/share/ipe/7.2.5/lua/ >> ├── actions.lua >> ├── editpath.lua >> ├── main.lua >> ... >> > > well, it looks in /share, not in /snap .... > > try this: > --- > #! /bin/sh > > export IPELUAPATH=$SNAP/share/ipe/7.2.5/lua > > exec $SNAP/bin/ipe > --- > > save it as wrapper.sh (make it executable), add a copy plugin part to > your snapcraft.yaml that copies it in the toplevel of your snap > and change: > https://github.com/dietmarw/snaps/blob/master/ipe/snapcraft.yaml#L52 > to read: > > command: desktop-launch wrapper.sh > > if needed also set the other variables to their proper paths ... > > ciao > oli -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From ogra at ubuntu.com Fri Jul 29 11:08:25 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 29 Jul 2016 13:08:25 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> Message-ID: <1469790505.13113.2.camel@ubuntu.com> hi, On Fr, 2016-07-29 at 12:45 +0200, Dietmar Winkler wrote: > Thanks Oli, > > I've implemented it as you suggested and got one step further: > > ... > '/snap/ipe/x22/share/ipe/7.2.5/lua': > cannot read /snap/ipe/x22/share/ipe/7.2.5/lua: Is a directory >  wow, so googling a bit [1] seems to suggest that IPELUAPATH actually wants patterns o_O (how weird) ...  try something like: export IPELUAPATH=$SNAP/share/ipe/7.2.5/lua/*.lua or: export IPELUAPATH=$SNAP/share/ipe/7.2.5/lua/?.lua ciao oli [1] http://webcache.googleusercontent.com/search?q=cache:hge_N7NYCM4J:c omments.gmane.org/gmane.comp.graphics.ipe.general/808&num=1&client=ubun tu&hs=bEn&hl=de&gl=de&strip=1&vwsrc=0 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From dietmar.winkler at dwe.no Fri Jul 29 11:40:29 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 13:40:29 +0200 Subject: Getting QT apps running. In-Reply-To: <1469790505.13113.2.camel@ubuntu.com> References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> <1469790505.13113.2.camel@ubuntu.com> Message-ID: Hi, > export IPELUAPATH=$SNAP/share/ipe/7.2.5/lua/?.lua That was it. The app now launches. Still loads of menus and icons missing in the UI but I guess this is down to: Gtk-Message: Failed to load module "overlay-scrollbar" Gtk-Message: Failed to load module "gtk-vector-screenshot" Gtk-Message: Failed to load module "gail" Gtk-Message: Failed to load module "atk-bridge" Gtk-Message: Failed to load module "unity-gtk-module" Gtk-Message: Failed to load module "canberra-gtk-module" Qt: Session management error: None of the authentication protocols specified are supported during start up. And of course the theme link that can not be generated: ln: failed to create symbolic link '/home/dietmarw/snap/ipe/x4/.themes/themes': Read-only file system So I guess currently there is not much I can do about the gtk modules not getting loaded. Other then wait that some day perhaps snappy will support this. -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From ogra at ubuntu.com Fri Jul 29 11:58:21 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 29 Jul 2016 13:58:21 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> <1469790505.13113.2.camel@ubuntu.com> Message-ID: <1469793501.13113.21.camel@ubuntu.com> hi, On Fr, 2016-07-29 at 13:40 +0200, Dietmar Winkler wrote: > Hi, > > > > > export IPELUAPATH=$SNAP/share/ipe/7.2.5/lua/?.lua > > That was it. The app now launches. Still loads of menus and icons > missing in the UI but I guess this is down to: > > Gtk-Message: Failed to load module "overlay-scrollbar" > Gtk-Message: Failed to load module "gtk-vector-screenshot" > Gtk-Message: Failed to load module "gail" > Gtk-Message: Failed to load module "atk-bridge" > Gtk-Message: Failed to load module "unity-gtk-module" > Gtk-Message: Failed to load module "canberra-gtk-module" > Qt: Session management error: None of the authentication protocols > specified are supported > > during start up. And of course the theme link that can not be > generated: > > ln: failed to create symbolic link > '/home/dietmarw/snap/ipe/x4/.themes/themes': Read-only file system > > So I guess currently there is not much I can do about the gtk modules > not getting loaded. Other then wait that some day perhaps snappy will > support this. i doubt it is related to the Gtk messages given your app is Qt (none of the above modules are essential)  ...  have you tried setting the other variables described in http://ipe.otfried.org/manual/manual_40.html looks to me like IPEICONDIR might give you icons for example ...when set to the right location ...  i would just try to play with these vars a bit, i bet you can sort most of your issues when they point to the right places ... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From mabnhdev at gmail.com Fri Jul 29 12:07:01 2016 From: mabnhdev at gmail.com (MikeB) Date: Fri, 29 Jul 2016 08:07:01 -0400 Subject: Snap daemonization ordering and manual interface connections Message-ID: On 7/14/2016 18:25 Mark Shuttleworth wrote: > Great topic for the sprint in Heidelberg indeed! > > We could easily prompt the user appropriately, and design for CLI and > GUI experiences shaped by the snap publisher. > > Mark Was any decision or progress made on these issues in Heidelberg? I'm really blocked by failure problems while daemonizing applications before the manual connections are made. Any suggested workarounds? Thanks, Mike -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.lenton at canonical.com Fri Jul 29 12:19:58 2016 From: john.lenton at canonical.com (John Lenton) Date: Fri, 29 Jul 2016 13:19:58 +0100 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> Message-ID: On 29 July 2016 at 11:45, Dietmar Winkler wrote: > ln: failed to create symbolic link > '/home/dietmarw/snap/ipe/x22/.themes/themes': Read-only file system that seems strange. What is read-only in that path? From dietmar.winkler at dwe.no Fri Jul 29 12:21:32 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 14:21:32 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> Message-ID: John, On 29 July 2016 at 14:19, John Lenton wrote: > On 29 July 2016 at 11:45, Dietmar Winkler wrote: >> ln: failed to create symbolic link >> '/home/dietmarw/snap/ipe/x22/.themes/themes': Read-only file system > > that seems strange. What is read-only in that path? my take here is that it's because it's in a dot directory. But then again its generated by snap itself. So this should not happen. -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From ogra at ubuntu.com Fri Jul 29 12:22:28 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 29 Jul 2016 14:22:28 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> Message-ID: <1469794948.13113.22.camel@ubuntu.com> hi, On Fr, 2016-07-29 at 13:19 +0100, John Lenton wrote: > On 29 July 2016 at 11:45, Dietmar Winkler > wrote: > > > > ln: failed to create symbolic link > > '/home/dietmarw/snap/ipe/x22/.themes/themes': Read-only file system > > that seems strange. What is read-only in that path? there is a known bug in the desktop launcher that tries to move the link around on upgrades as i understand, dider was working on a fix ... it is actually harmless ... ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From didrocks at ubuntu.com Fri Jul 29 12:22:56 2016 From: didrocks at ubuntu.com (Didier Roche) Date: Fri, 29 Jul 2016 14:22:56 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> Message-ID: <45e748dd-5f19-96d8-3b58-c9ab80d60247@ubuntu.com> Le 29/07/2016 à 14:19, John Lenton a écrit : > On 29 July 2016 at 11:45, Dietmar Winkler wrote: >> ln: failed to create symbolic link >> '/home/dietmarw/snap/ipe/x22/.themes/themes': Read-only file system > that seems strange. What is read-only in that path? > That was this bug from the destkop-launcher: https://github.com/ubuntu/snapcraft-desktop-helpers/issues/4. I fixed it yesterday, so any new build pulling it should get it fixed. It doesn't have any visible consequence to the user (apart if we updated the platform themes between upgrade). Didier From dietmar.winkler at dwe.no Fri Jul 29 12:53:53 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 14:53:53 +0200 Subject: Getting QT apps running. In-Reply-To: <1469793501.13113.21.camel@ubuntu.com> References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> <1469790505.13113.2.camel@ubuntu.com> <1469793501.13113.21.camel@ubuntu.com> Message-ID: Seting the rests of the paths helped. I now have icons. The menu is still missing which reminded me that there currently is an issue with appmenu-qt5 which when naively compiled causes some keyboard shortcuts not to work. https://github.com/otfried/ipe-wiki/wiki/FAQ#ipe-doesnt-work-correctly-on-ubuntu I was wondering if it is possible to generate a variant of desktop/qt5 without appmenu-qt5 staged. I've found the definition here: https://wiki.ubuntu.com/Snappy/Parts but would still need the source to point to. -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From joe.talbott at canonical.com Fri Jul 29 13:34:26 2016 From: joe.talbott at canonical.com (Joe Talbott) Date: Fri, 29 Jul 2016 09:34:26 -0400 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> <1469790505.13113.2.camel@ubuntu.com> <1469793501.13113.21.camel@ubuntu.com> Message-ID: Note the remote parts are now at https://wiki.ubuntu.com/snapcraft/parts Cheers, Joe On Fri, Jul 29, 2016 at 8:53 AM, Dietmar Winkler wrote: > Seting the rests of the paths helped. I now have icons. The menu is > still missing which reminded me that there currently is an issue with > appmenu-qt5 which when naively compiled causes some keyboard shortcuts > not to work. > https://github.com/otfried/ipe-wiki/wiki/FAQ#ipe-doesnt-work-correctly-on-ubuntu > > I was wondering if it is possible to generate a variant of desktop/qt5 > without appmenu-qt5 staged. > > I've found the definition here: https://wiki.ubuntu.com/Snappy/Parts > but would still need the source to point to. > > -- > /Dietmar/ > > Secure email communication: > https://encrypt.to/dietmar.winkler at dwe.no > Public OpenPGP key: 0x235E6689 > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft From dietmar.winkler at dwe.no Fri Jul 29 14:01:52 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 16:01:52 +0200 Subject: Getting QT apps running. In-Reply-To: References: <83f185c8-1300-a503-f55d-9aa3f052c07a@canonical.com> <82d6e45f-eeab-03fb-df36-31adc7cf79b4@ubuntu.com> <1469784438.4310.58.camel@anubis> <1469785204.4310.64.camel@anubis> <1469790505.13113.2.camel@ubuntu.com> <1469793501.13113.21.camel@ubuntu.com> Message-ID: Thanks for the pointer. Reading through the documentation of desktop-helper again I noticed that I might actually try the x11 plug instead of unity7 since it is the appmenu that causes problems in the first place. But now ipe would not launch at all. Also I don't get any error messages (apart from the Gtk warnings). Are there any debug logs I could search for to see what is getting stuck? On 29 July 2016 at 15:34, Joe Talbott wrote: > Note the remote parts are now at https://wiki.ubuntu.com/snapcraft/parts > > Cheers, > Joe > > On Fri, Jul 29, 2016 at 8:53 AM, Dietmar Winkler wrote: >> Seting the rests of the paths helped. I now have icons. The menu is >> still missing which reminded me that there currently is an issue with >> appmenu-qt5 which when naively compiled causes some keyboard shortcuts >> not to work. >> https://github.com/otfried/ipe-wiki/wiki/FAQ#ipe-doesnt-work-correctly-on-ubuntu >> >> I was wondering if it is possible to generate a variant of desktop/qt5 >> without appmenu-qt5 staged. >> >> I've found the definition here: https://wiki.ubuntu.com/Snappy/Parts >> but would still need the source to point to. >> >> -- >> /Dietmar/ >> >> Secure email communication: >> https://encrypt.to/dietmar.winkler at dwe.no >> Public OpenPGP key: 0x235E6689 >> >> -- >> Snapcraft mailing list >> Snapcraft at lists.snapcraft.io >> Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From dietmar.winkler at dwe.no Fri Jul 29 14:55:05 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 16:55:05 +0200 Subject: Getting pdflatex to work inside a spap packge Message-ID: Hi, I'm currently trying to package my first snap package [1]. That application depends on pdflatex which I included as stage package stage-packages: - texlive-latex-base pdflatex can be executed but when an application from inside the snap starts it it runs into trouble since pdflatex wants to access /etc/texmf/web2c/texmf.cnf which it cant. It seems that the path "/etc/texmf/web2c/texmf.cnf" is actually hard coded in the pdflatex application. I tried things like setting the TEXMFPATH but that does not have an influence. I was wondering if there is a solution for this without rewriting the source of pdflatex. The more I'm working with snap packages the more I'm getting the feeling that it has one fundamental flaw: containerising the application but at the same time not automatically translating paths to inside the container. This means that one has to tweak the original source application in order to get it to work as a snap package. But I'd say that the application should be totally oblivious to how it is deployed. -- /Dietmar/ [1] https://github.com/dietmarw/snaps/tree/master/ipe Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From ogra at ubuntu.com Fri Jul 29 15:02:21 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 29 Jul 2016 17:02:21 +0200 Subject: Getting pdflatex to work inside a spap packge In-Reply-To: References: Message-ID: <1469804541.13113.36.camel@ubuntu.com> hi, On Fr, 2016-07-29 at 16:55 +0200, Dietmar Winkler wrote: >  > It seems that the path "/etc/texmf/web2c/texmf.cnf" is actually hard > coded in the pdflatex application. I tried things like setting the > TEXMFPATH but that does not have an influence. > try TEXMFCONFIG for that one ...   and google indicates that there is no TEXMFPATH but a TEXMFHOME instead ... [1] might be helpful ciao oli [1] http://tex.stackexchange.com/questions/30494/how-to-make-latex-see- local-texmf-tree -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From dietmar.winkler at dwe.no Fri Jul 29 15:10:07 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 17:10:07 +0200 Subject: Getting pdflatex to work inside a spap packge In-Reply-To: <1469804541.13113.36.camel@ubuntu.com> References: <1469804541.13113.36.camel@ubuntu.com> Message-ID: Hi, > try TEXMFCONFIG for that one ... > > and google indicates that there is no TEXMFPATH but a TEXMFHOME instead > ... [1] might be helpful Sorry I should have said I've tried all those already (also TEXMFHOME). As I said that path is hard coded. To be more precise in /usr/sbin/update-texmf: TXMF=/etc/texmf TXMF_CNF=$TXMF/web2c/texmf.cnf :-( -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From ogra at ubuntu.com Fri Jul 29 15:32:48 2016 From: ogra at ubuntu.com (Oliver Grawert) Date: Fri, 29 Jul 2016 17:32:48 +0200 Subject: Getting pdflatex to work inside a spap packge In-Reply-To: References: <1469804541.13113.36.camel@ubuntu.com> Message-ID: <1469806368.13113.38.camel@ubuntu.com> hi, On Fr, 2016-07-29 at 17:10 +0200, Dietmar Winkler wrote: > Hi, > > > > > > try TEXMFCONFIG for that one ... > > > > and google indicates that there is no TEXMFPATH but a TEXMFHOME > > instead > > ... [1] might be helpful > > Sorry I should have said I've tried all those already (also > TEXMFHOME). > As I said that path is hard coded. To be more precise in > /usr/sbin/update-texmf: > > TXMF=/etc/texmf > TXMF_CNF=$TXMF/web2c/texmf.cnf > you could just replace these paths ...  (when does that command run ? at runtime or at install time ?)  ciao oli -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: This is a digitally signed message part URL: From dietmar.winkler at dwe.no Fri Jul 29 15:47:55 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 17:47:55 +0200 Subject: Getting pdflatex to work inside a spap packge In-Reply-To: <1469806368.13113.38.camel@ubuntu.com> References: <1469804541.13113.36.camel@ubuntu.com> <1469806368.13113.38.camel@ubuntu.com> Message-ID: Hi, On 29 July 2016 at 17:32, Oliver Grawert wrote: > you could just replace these paths ... Well then I would to manually change the code in the stage (?) area which would be not at all an optimal workflow. > (when does that command run ? at runtime or at install time ?) It happens when the ipe application triggers pdflatex. Working example: After having build and installed the ipe snap from https://raw.githubusercontent.com/dietmarw/snaps/master/ipe/ you can run ipe.ipetoipe -pdf test.ipe on https://raw.githubusercontent.com/dietmarw/snaps/master/ipe/test/test.ipe and should get: $ ipe.ipetoipe -pdf test.ipe Document test.ipe has 1 pages (1 views) pdflatex: /etc/texmf/web2c/texmf.cnf: Permission denied There was an error trying to run Pdflatex. -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From dietmar.winkler at dwe.no Fri Jul 29 16:06:40 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 18:06:40 +0200 Subject: Getting pdflatex to work inside a spap packge In-Reply-To: References: <1469804541.13113.36.camel@ubuntu.com> <1469806368.13113.38.camel@ubuntu.com> Message-ID: Having manually replaced the location in /usr/bin/update-texmf did not change anything. So it is most likely embedded in the compiled binary of pdflatex. :-( On 29 July 2016 at 17:47, Dietmar Winkler wrote: > Hi, > > On 29 July 2016 at 17:32, Oliver Grawert wrote: >> you could just replace these paths ... > > Well then I would to manually change the code in the stage (?) area > which would be not at all an optimal workflow. > >> (when does that command run ? at runtime or at install time ?) > > It happens when the ipe application triggers pdflatex. > > Working example: > After having build and installed the ipe snap from > > https://raw.githubusercontent.com/dietmarw/snaps/master/ipe/ > > you can run > > ipe.ipetoipe -pdf test.ipe > > on https://raw.githubusercontent.com/dietmarw/snaps/master/ipe/test/test.ipe > > and should get: > $ ipe.ipetoipe -pdf test.ipe > Document test.ipe has 1 pages (1 views) > pdflatex: /etc/texmf/web2c/texmf.cnf: Permission denied > There was an error trying to run Pdflatex. > > -- > /Dietmar/ > > Secure email communication: > https://encrypt.to/dietmar.winkler at dwe.no > Public OpenPGP key: 0x235E6689 -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From dietmar.winkler at dwe.no Fri Jul 29 16:07:21 2016 From: dietmar.winkler at dwe.no (Dietmar Winkler) Date: Fri, 29 Jul 2016 18:07:21 +0200 Subject: Getting pdflatex to work inside a spap packge In-Reply-To: References: <1469804541.13113.36.camel@ubuntu.com> <1469806368.13113.38.camel@ubuntu.com> Message-ID: Sorry I meant of course "/usr/sbin/update-texmf" -- /Dietmar/ Secure email communication: https://encrypt.to/dietmar.winkler at dwe.no Public OpenPGP key: 0x235E6689 From luke.williams at canonical.com Fri Jul 29 17:03:58 2016 From: luke.williams at canonical.com (Luke Williams) Date: Fri, 29 Jul 2016 10:03:58 -0700 Subject: no hardware description in Gadget snap for building an image In-Reply-To: References: <20160728164904.GS2824@bod> Message-ID: <8789389C-DEB2-4B97-B6E2-BB1E6479721B@canonical.com> So in the interim, how do we build custom images with custom kernels? Should we use the image from mvo and then apply our kernel snaps via ’snap install’ for testing? Thanks, Luke Williams - Technical Partner Manager, Network Switches/Ubuntu-Core email: luke.williams at canonical.com http://www.canonical.com/ | http://www.ubuntu.com > On Jul 29, 2016, at 3:02 AM, Gustavo Niemeyer wrote: > > For the record, please note this change was announced in this list a couple of months ago, and that we don't actually have official Ubuntu Core 16 images yet. > > Polishing the format is being done while finalizing those images. > > > On Jul 28, 2016 8:03 PM, "Michael Vogt" > wrote: > On Thu, Jul 28, 2016 at 05:50:54PM +0800, Woodrow Shen wrote: > > Hi all, > Hi Woodrow, > > thanks for your mail! > > > It seems the latest revision 8 has a lack of hardware section for gadget > > snap. > > I think someone is working on it now, but hope we can make sure the snap > > uploaded to edge channel can work at least. > > You are correct, we are currently working on a new and improved format > to describe the bootloader for a device. This work is currently in > progress and because the old format stops working we removed it. We > publish new gadget snaps with the new format soon. ubuntu-image will > be the tool to consume them then. > > Cheers, > Michael > > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft > -- > Snapcraft mailing list > Snapcraft at lists.snapcraft.io > Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/snapcraft -------------- next part -------------- An HTML attachment was scrubbed... URL: