Re: Install archos 101xs
Hi, this should actually work. As far as I know, the hardware is pretty similar to the Archos G9 models. If my memory serves me right, the official Archos kernel sources for the 101 XS (aka Gen. 10) also just have a reference to the G9 kernel. Before reading on, please be aware that doing anything described below will void the warranty of your device and may be dangerous. You do this all on your own responsibility and risk. Please also note that there are different XS versions. While the 101 XS is likely to work, others won't work at all as they use an entirely different SoC. In theory, it should be possible to use the same approach as with the Archos G9 versions. The following URL has a step-by-step how-to on how to install it (scroll down to Installation Howto): https://share.basyskom.com/plasma-active/archos_gen9.html However, I am not sure how up-to-date the images linked on that website are. I suggest to try the images that can be downloaded here: http://files.kde.org/plasma/active/4.0/images/archosg9/ http://files.kde.org/plasma/active/3.0/images/archos-g9/armv7hl/ I'd first give the 4.0 version a try and in case that doesn't work try the latest 3.0 version. For the sake of completeness here is the link to the Archos SDE website: http://www.archos.com/de/support/support_tech/updates_dev.html It would be nice if you could keep us updated about your progress. Best regards, Ruediger On 12/01/2013 09:10 AM, Niels Van Laerhoven wrote: Hello I am from Belgium and I use KDE for almost 7 years by now. I've just read about the plasma active project and would like to test it. However, my tablet (Archos 101xs) isn't listed in the wiki. Would it be possible to install it on my Archos? Thanks Niels Van Laerhoven ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Problems while building nexus 7 image for PA4
Hi, I finally found the time to create .ks files and rootfs tarballs: http://ruedigergad.com/2013/09/09/plasma-active-4-for-nexus-7-2012-and-archos-g9/ hth Best regards, Ruediger On 09/09/2013 07:32 PM, Daniel Kreuter wrote: Hi, I tried again, this time with the ks file Ruediger send me this morning on irc, but I get the same error this time only with the libxml2 package. Do I need that devel package on my system? If yes than he isn't able to locate it. Cheers Am 07.09.2013 um 15:14 schrieb Daniel Kreuter daniel.kreute...@gmail.com mailto:daniel.kreute...@gmail.com: Hi, I'm trying to build a new image for my nexus 7 based on the .ks file of Ruediger Gad but I get the following errors: 1. sudo mic create fs plasma-active-armv7hl-nexus7-devel-2013-07-17.ks --pack-to=plasma-active-armv7hl-google-nexus7-pa4-2013-09-07.tar.gz --pkgmgr=zypp --arch=armv7hl 2. mic 0.14 (Mer 0.2011 Mer) 3. Info: Retrieving repo metadata: 4. Info: Retrieving repomd.xml ... DONE 5. Info: Refreshing repository: mer-core ... 6. Info: Refreshing repository: nemo-apps ... 7. Info: Refreshing repository: nemo-hw-nv-tegra3-common ... 8. Info: Refreshing repository: mer-core-debug ... 9. Info: Refreshing repository: mer-updates ... 10. Info: Refreshing repository: kde-mw ... 11. Info: Refreshing repository: nemo-hw-nv-tegra3-nexus7 ... 12. Info: Refreshing repository: kde-apps ... 13. Info: Refreshing repository: nemo-mw ... 14. Info: Refreshing repository: qt-mobility-hack ... 15. Info: Refreshing repository: kde-ux ... 16. Info: Refreshing repository: wonko-nemo-compatibility ... 17. Info: zypp architecture is armv7hl 18. Warning: repo problem: nothing provides qt-components-qt5 needed by qmlgallery-0.1.6-6.4.Nemo.Apps.armv7hl, 19. Warning: repo problem: nothing provides qt-components-qt5 needed by qmlmessages-0.2.1-10.4.Nemo.Apps.armv7hl, 20. Warning: repo problem: nothing provides qt-components-qt5 needed by qmlsettings
Re: PA4
On 07/23/2013 11:57 PM, Maurice de la Ferté wrote: On 23.07.2013 23:27, Maurice de la Ferté wrote: Hi, On 23.07.2013 22:32, Ruediger Gad wrote: It's simply that the nm-applet doesn't receive any mouse events after it was opened. The icon and the applet seem to receive the equivalent of right-click events: in both cases the same pop-up menu appears. I forced the right click even by tinkering with the button mask via some xinput utils. During the tries to fixing this I played with lots of added debugging statements in the involved code. QML-wise, the touch events seem to be not forwarded to the QML part at all. On the C++ side the results are also strange. It seems like only the first key-press is forwarded to the applet and then the rest is just ignored/filtered/masked, whatever. This sounds to me like a might be similar issue which has Marco workaround by the following patch for PA3: https://build.merproject.org/package/view_file?file=use-fake-windows-for-pa3-arm-release.patchpackage=plasma-mobileproject=kde%3Astable%3Auxrev=331c3c1e0bed9d475c544552dc51a951 I do not remember all details anymore, but we had to use a fake window to get touch events back on some PA core features on arm and the patch is still in usage. This is what the plasma-mobile.changes tells about: snip - Add workaround to get buttons on e.g. archos gen9 on some panels resposive again https://bugs.kde.org/show_bug.cgi?id=304875 snap In case this goes in the wrong direction, please ignore. Thanks for the input. I don't know if it is what you intended but, fwiw, I gave it a brief try to remove the use-fake-windows-for-pa3-arm-release patch. The result is that after boot I get a black screen where the workbench should appear. It shows the progress indicator on boot and also shows the volume and brightness indicator but else the screen remains black. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Instant Message client for Plasma Active?
On 07/14/2013 07:49 PM, Marco Martin wrote: On Fri, Jul 5, 2013 at 7:09 PM, David Talmage talm...@acm.org mailto:talm...@acm.org wrote: Is there an IM client for PA? I don't see one in the integration repo. There is also an active ui for kde telepathy, i don't know if is actively developed. Its developer will sure be happy for an extra hand :) Fwiw, in the following blog post I wrote about my experience with XMPP(Jabber) in PA using some GUI frontends from Nemo: http://ruedigergad.com/2013/03/13/jabberxmpp-with-latest-plasma-active-on-nexus-7/ Please note that the required libs and GUIs are only installed by default on the PA rootfs tarballs for Nexus7 and ArchosG9 that you can get from here: http://files.kde.org/plasma/active/3.0/images/ The latest development snapshots should come with everything needed for using XMPP. Afaik, Mer now also ships with telepathy libs for ICQ. However, it won't be easily usable in PA via the approach described in my blog post yet. Please note that this is highly experimental, primarily a proof of concept, and that the user experience is far from optimal. However, it used to work here. hth Ruediger ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Nexus 7 images
Alright, I found the problem. This is actually quite an old issue. To summarize: there are still issues with QT_DEFAULT_RUNTIME_SYSTEM=opengl Back then we had this setting only in: /etc/profile.d/tegra3-env.sh Because of above mentioned issues we disabled that environment variable there. Apparently, that setting crept in again via: /etc/sysconfig/nemo-mobile-hw Afaik, QT_DEFAULT_RUNTIME_SYSTEM=opengl doesn't cause issues on nemomobile. Hence, nemo uses this setting. As a test I set QT_DEFAULT_RUNTIME_SYSTEM=raster in /etc/sysconfig/nemo-mobile-hw and now everything works as expected. If you want to test the latest image, which should be on the mirrors by now, you can edit that file after un-taring the rootfs. I'll try to create rootfs tarballs with a fix for this issue when I have some time. Cheers, Ruediger On 06/06/2013 10:03 AM, Ruediger Gad wrote: Attached, I send the according log files. hth Ruediger On 06/05/2013 09:22 PM, Matthew Dawson wrote: On June 5, 2013 06:03:15 PM Daniel Kreuter wrote: Hi Matthew, can you post snippets of that log file? Maybe Martin has any idea what the generic messages mean. Unfortunately, I don't have that log file anymore, as I've since upgraded my tablet. If someone has PA working on the N7, the error messages appeared in the general log file in /tmp. Also, I've done some GL work. The error message was being printed from the end of the entire draw in KWin, and was some generic OpenGL error. I was working on digging into what exactly was throwing the error, but ran out of time. If I have some time I'll try working on it again, and see if I can figure out it's source. I've played with OpenGL before (but not KWin), and that error isn't helpful without knowing the call causing it. I think it was some GL_INVALID_*. Also, since you mentioned some crashes, I had hunted down the crash in KDED and last I remember, Powerdevil caused some GUI to display, which caused Qt to try and render something using OpenGL ES, which promptly crashed as some Qt(?) class called Context(?) which wasn't initialized. That is from memory, and I was trying to update my tablet at the time I was hunting that down, so I don't have that stack trace either. I actually wanted to first get image generation for the N7 stable so I could more easily deal with it before tackling those issues. But I never have enough time :). I first wanted to deal with SSU as it was causing my images to be missing debug symbols, important for bug hunting! ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: maliit-active and maliit-plugins do not build in kde:devel:ux
On 06/06/2013 02:41 PM, Marco Martin wrote: On Thursday 06 June 2013, Ruediger Gad wrote: Hi, maliit-active and maliit-plugins are currently not building in kde:devel:ux. Reason is an ambiguity for resolving pkgconfig(maliit-1.0) which is provided by libmaliit-devel and maliit-qt4-devel. I tested building maliit-active with adding PkgBR: -libmaliit-devel and it works. I could quickly fix both builds that way if it is OK for you. +1 maliit-active is fixed. With respect to maliit-plugins: why is this in kde:devel:ux? There is already a maliit-plugins package in nemo:devel:mw -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Nexus 7 images
On 06/06/2013 06:14 PM, Marco Martin wrote: On Thursday 06 June 2013, Daniel Kreuter wrote: work on other things? The benefit of opengl would be a better performance because the rendering would happen on the GPU instead of the CPU. What would be nice to know is why opengl isn't working on pa. for qgraphicsview, not really (that's why screnegraph in qml2 exists) Afaik, there is much work on the plasma/KDE graphics engine anyhow, right? Anyway, it is interesting that the UI comes up with opengl at all. Some months ago it died already in a very early stage. In the log file I sent one error repeated very often: kwinactive(554)/kwin KWin::checkGLError: GL error ( PostPaint ): GL_INVALID_OPERATION Btw. I just uploaded a rootfs tarball that includes the fix: http://files.kde.org/plasma/active/3.0/images/nexus7/plasma-active-armv7hl-google-nexus7-devel-mer-latest-2013-06-06.tar.gz.mirrorlist -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: maliit-active and maliit-plugins do not build in kde:devel:ux
On 06/06/2013 07:40 PM, Ruediger Gad wrote: On 06/06/2013 02:41 PM, Marco Martin wrote: On Thursday 06 June 2013, Ruediger Gad wrote: Hi, maliit-active and maliit-plugins are currently not building in kde:devel:ux. Reason is an ambiguity for resolving pkgconfig(maliit-1.0) which is provided by libmaliit-devel and maliit-qt4-devel. I tested building maliit-active with adding PkgBR: -libmaliit-devel and it works. I could quickly fix both builds that way if it is OK for you. +1 maliit-active is fixed. With respect to maliit-plugins: why is this in kde:devel:ux? There is already a maliit-plugins package in nemo:devel:mw PS: fwiw, I also fixed the maliit-plugins build in kde:devel:ux. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Network Management Applet not working
On 06/02/2013 11:01 PM, Ruediger Gad wrote: On 06/01/2013 02:25 PM, Ruediger Gad wrote: On 05/31/2013 12:38 PM, Ruediger Gad wrote: On 05/30/2013 07:18 PM, Marco Martin wrote: On Thursday 30 May 2013, Ruediger Gad wrote: Marco also suggested to create a small sample app to try different combinations of windows flags and how they affect the behavior. Unfortunately, I do not have the time to write such an app right now but I found something that may be useful: http://qt-project.org/doc/qt-4.8/widgets-windowflags.html This app seems to be pretty much what Marco suggested? Do you have any ideas where window flags may be set for applets in the top bar or the corresponding popups? Generally, do you have ideas on how to solve the issue? another hack that came to mind, is patching popupapplet.cpp, and making all popup windows fullscreen (maximized, maybe even without alpha channel,extra kwin atoms and so forth) for small screens it could even look it makes sense, so not feel as such and hack Thanks for the hints and ideas Marco. :) As you already mentioned in the other thread, the problem is really weird indeed. I'll try to summarize what I tested so far: My first thought was that events get not forwarded properly to the popup, so I placed debug output in the event handler methods in dialog.cpp and popupapplet.cpp. With this debug output I could not find any difference between the non-working networkmanagement applet and the working battery applet. The next try was to see whether the bounding box etc. are properly set. Again, everything seems fine here as well; no apparent difference between the two applets. Another try I did some days ago was to simply put a big MouseArea in the top item of the NMPopup.qml code to see if, generally, mouse events reach the widget. Here, I couldn't get any events at all. I even set the z property to something insanely high like 1001. So, to summarize, it seems like the events get lost somewhere in the stack. In dialog.cpp and popupapplet.cpp it looked like that they are there (I was looking for mousePressEvent). But then they seem to get lost somewhere after that. Alright, I think I'll throw in the towel for now. For some reason dialog.cpp receives mouse press events but the actual NMPopup.qml doesn't. In the meantime I also tried various settings like setAcceptTouchEvents or grabMouse but with no success. My last debug output attempts can be found here: https://build.merproject.org/package/show?package=NetworkManager-kdeproject=home%3Awonko%3Apa-networkmanagement-tests https://build.merproject.org/project/show?project=home%3Awonko%3Abranches%3Akde%3Adevel%3Aux Be aware that this is quite a mess. Even the patches are named stupidly. As workaround I'll use nmcli for now. Connecting to a WPA wlan via nmcli is actually pretty straight forward (as mer user): nmcli wifi connect SSID password PASSWORD I hope someone else has more success on this one... Just to correct myself, I was missing a small but important 'd': nmcli d wifi connect SSID password PASSWORD jfyi This problem exists on the Nexus 7 as well. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Nexus 7 images
Apparently, the download didn't finish properly yesterday. Seems like it failed just after I left the computer, of course ;) I deleted the file and re-upload it right now. On 06/05/2013 12:27 AM, Ruediger Gad wrote: Hi, uploading a new devel image for Nexus7 right now (plasma-active-armv7hl-google-nexus7-devel-mer-latest-2013-06-04.*). Unfortunately, this has quite a number of issues and is pretty much unusable. On boot you will discover some segfaults. You can ignore these (hopefully) as the UI will come up after some time. Unfortunately, there will be the next problem. The UI freezes instantly under some circumstances (e.g., when trying to pull down the top bar). You should(tm) be able to switch between activities, which shows that it is generally working, but you will encounter lots of situations in which the UI just freezes. I uploaded the .ks file from which the image was created as well. Right now, the most stable variant for arm seems to be the Archos G9 version. I just uploaded a new image for these today as well (plasma-active-armv7hl-archos-gen9-pvr-devel-mer-latest-2013-06-01.*). Only big issue there is that the network manager applet does not accept input. This can be worked around via nmcli on the command line. Upload eta for the Nexus 7 image is ~30 minutes. BR, Ruediger On 06/04/2013 08:54 PM, Jan Grulich wrote: Hi, I don't suppose it's easily possible to build my own images with Pa for Nexus 7. So my question is, could someone make some newer images for Nexus 7 with the latest PA (it should be PA 4 RC1), because I would like to test plasma-nm on some real hardware and not only in Virtualbox. The latest image is two month old and didn't work quite well if I remember correctly. Thanks a lot :) Cheers, Jan -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Nexus 7 images
Just to clarify: when I speak about devel, I mean the devel projects or branch at build.merproject.org, e.g.: https://build.merproject.org/project/show?project=kde%3Adevel%3Aux etc. When I speak about testing, I mean the corresponding testing versions, e.g., https://build.merproject.org/project/show?project=kde%3Atesting%3Aux etc. Afaik, the devel version at build.merproject.org runs a (more or less?) continous integration against the master of the accompanying git repositories. So it should, more or less, correspond to git master. With respect to building images: I use the following command line from mer sdk to build the nexus7 image: PATH=$PATH:/sbin sudo mic create fs plasma-active-armv7hl-google-nexus7-devel-mer-latest-2013-06-04.ks --pack-to=plasma-active-armv7hl-google-nexus7-devel-mer-latest-2013-06-04.tar.gz --pkgmgr=zypp --arch=armv7hl Unfortunately, I am short of time right now. The mer wiki should provide more help. On 06/05/2013 10:36 AM, Daniel Kreuter wrote: When you talk of devel I always think of the devel branch of plasma active, and master the master branch which should always be in a releasable state, which might not always be the case. But good that you pointed out that what you mean with devel means it's based on the master branch. Can you point me to some guides/tutorials on how to build these images? Then I can try to fin out where these problems ome from and try to fix them if I can, so I don't need to wait for your great work and help you a bit. So long On Jun 5, 2013 9:38 AM, Ruediger Gad r@gmx.de mailto:r@gmx.de wrote: What do you mean with master? Afaik, devel is based on master from the git repositories (Please correct me if I am mistaken.). Nonetheless, iirc, the same issue also occurs with testing. On 06/05/2013 09:15 AM, Daniel Kreuter wrote: Would it be possible to provide a pa image based on master? Or are the problems you mentioned the sam on master and devel? On Jun 5, 2013 12:27 AM, Ruediger Gad r@gmx.de mailto:r@gmx.de mailto:r@gmx.de mailto:r@gmx.de wrote: Hi, uploading a new devel image for Nexus7 right now (plasma-active-armv7hl-google-nexus7-devel-mer-latest-__2013-__06-04.*). Unfortunately, this has quite a number of issues and is pretty much unusable. On boot you will discover some segfaults. You can ignore these (hopefully) as the UI will come up after some time. Unfortunately, there will be the next problem. The UI freezes instantly under some circumstances (e.g., when trying to pull down the top bar). You should(tm) be able to switch between activities, which shows that it is generally working, but you will encounter lots of situations in which the UI just freezes. I uploaded the .ks file from which the image was created as well. Right now, the most stable variant for arm seems to be the Archos G9 version. I just uploaded a new image for these today as well (plasma-active-armv7hl-archos-gen9-pvr-devel-mer-latest-2013-06-01.*). Only big issue there is that the network manager applet does not accept input. This can be worked around via nmcli on the command line. Upload eta for the Nexus 7 image is ~30 minutes. BR, Ruediger On 06/04/2013 08:54 PM, Jan Grulich wrote: Hi, I don't suppose it's easily possible to build my own images with Pa for Nexus 7. So my question is, could someone make some newer images for Nexus 7 with the latest PA (it should be PA 4 RC1), because I would like to test plasma-nm on some real hardware and not only in Virtualbox. The latest image is two month old and didn't work quite well if I remember correctly. Thanks a lot :) Cheers, Jan -- http://ruedigergad.com _ Active mailing list Active@kde.org mailto:Active@kde.org mailto:Active@kde.org mailto:Active@kde.org https://mail.kde.org/mailman/__listinfo/active https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Nexus 7 images
Hi, uploading a new devel image for Nexus7 right now (plasma-active-armv7hl-google-nexus7-devel-mer-latest-2013-06-04.*). Unfortunately, this has quite a number of issues and is pretty much unusable. On boot you will discover some segfaults. You can ignore these (hopefully) as the UI will come up after some time. Unfortunately, there will be the next problem. The UI freezes instantly under some circumstances (e.g., when trying to pull down the top bar). You should(tm) be able to switch between activities, which shows that it is generally working, but you will encounter lots of situations in which the UI just freezes. I uploaded the .ks file from which the image was created as well. Right now, the most stable variant for arm seems to be the Archos G9 version. I just uploaded a new image for these today as well (plasma-active-armv7hl-archos-gen9-pvr-devel-mer-latest-2013-06-01.*). Only big issue there is that the network manager applet does not accept input. This can be worked around via nmcli on the command line. Upload eta for the Nexus 7 image is ~30 minutes. BR, Ruediger On 06/04/2013 08:54 PM, Jan Grulich wrote: Hi, I don't suppose it's easily possible to build my own images with Pa for Nexus 7. So my question is, could someone make some newer images for Nexus 7 with the latest PA (it should be PA 4 RC1), because I would like to test plasma-nm on some real hardware and not only in Virtualbox. The latest image is two month old and didn't work quite well if I remember correctly. Thanks a lot :) Cheers, Jan -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Network Management Applet not working
On 05/31/2013 12:38 PM, Ruediger Gad wrote: On 05/30/2013 07:18 PM, Marco Martin wrote: On Thursday 30 May 2013, Ruediger Gad wrote: Marco also suggested to create a small sample app to try different combinations of windows flags and how they affect the behavior. Unfortunately, I do not have the time to write such an app right now but I found something that may be useful: http://qt-project.org/doc/qt-4.8/widgets-windowflags.html This app seems to be pretty much what Marco suggested? Do you have any ideas where window flags may be set for applets in the top bar or the corresponding popups? Generally, do you have ideas on how to solve the issue? another hack that came to mind, is patching popupapplet.cpp, and making all popup windows fullscreen (maximized, maybe even without alpha channel,extra kwin atoms and so forth) for small screens it could even look it makes sense, so not feel as such and hack Thanks for the hints and ideas Marco. :) As you already mentioned in the other thread, the problem is really weird indeed. I'll try to summarize what I tested so far: My first thought was that events get not forwarded properly to the popup, so I placed debug output in the event handler methods in dialog.cpp and popupapplet.cpp. With this debug output I could not find any difference between the non-working networkmanagement applet and the working battery applet. The next try was to see whether the bounding box etc. are properly set. Again, everything seems fine here as well; no apparent difference between the two applets. Another try I did some days ago was to simply put a big MouseArea in the top item of the NMPopup.qml code to see if, generally, mouse events reach the widget. Here, I couldn't get any events at all. I even set the z property to something insanely high like 1001. So, to summarize, it seems like the events get lost somewhere in the stack. In dialog.cpp and popupapplet.cpp it looked like that they are there (I was looking for mousePressEvent). But then they seem to get lost somewhere after that. Alright, I think I'll throw in the towel for now. For some reason dialog.cpp receives mouse press events but the actual NMPopup.qml doesn't. In the meantime I also tried various settings like setAcceptTouchEvents or grabMouse but with no success. My last debug output attempts can be found here: https://build.merproject.org/package/show?package=NetworkManager-kdeproject=home%3Awonko%3Apa-networkmanagement-tests https://build.merproject.org/project/show?project=home%3Awonko%3Abranches%3Akde%3Adevel%3Aux Be aware that this is quite a mess. Even the patches are named stupidly. As workaround I'll use nmcli for now. Connecting to a WPA wlan via nmcli is actually pretty straight forward (as mer user): nmcli wifi connect SSID password PASSWORD I hope someone else has more success on this one... ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Network Management Applet not working
On 05/30/2013 07:18 PM, Marco Martin wrote: On Thursday 30 May 2013, Ruediger Gad wrote: Marco also suggested to create a small sample app to try different combinations of windows flags and how they affect the behavior. Unfortunately, I do not have the time to write such an app right now but I found something that may be useful: http://qt-project.org/doc/qt-4.8/widgets-windowflags.html This app seems to be pretty much what Marco suggested? Do you have any ideas where window flags may be set for applets in the top bar or the corresponding popups? Generally, do you have ideas on how to solve the issue? another hack that came to mind, is patching popupapplet.cpp, and making all popup windows fullscreen (maximized, maybe even without alpha channel,extra kwin atoms and so forth) for small screens it could even look it makes sense, so not feel as such and hack Thanks for the hints and ideas Marco. :) As you already mentioned in the other thread, the problem is really weird indeed. I'll try to summarize what I tested so far: My first thought was that events get not forwarded properly to the popup, so I placed debug output in the event handler methods in dialog.cpp and popupapplet.cpp. With this debug output I could not find any difference between the non-working networkmanagement applet and the working battery applet. The next try was to see whether the bounding box etc. are properly set. Again, everything seems fine here as well; no apparent difference between the two applets. Another try I did some days ago was to simply put a big MouseArea in the top item of the NMPopup.qml code to see if, generally, mouse events reach the widget. Here, I couldn't get any events at all. I even set the z property to something insanely high like 1001. So, to summarize, it seems like the events get lost somewhere in the stack. In dialog.cpp and popupapplet.cpp it looked like that they are there (I was looking for mousePressEvent). But then they seem to get lost somewhere after that. -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: PA in Virtualbox
Hi Jan, that's a known issue in the current devel branch. There is also a possible fix. It's just not in the repository yet. BR, Ruediger On 05/29/2013 12:31 PM, Jan Grulich wrote: Hi, I'm trying to install PA in Virtualbox from Basyskom images. The latest Mer testing image worked without any problems but I want to try the latest PA which, as I think, is in the Mer development images. I'm able to install this image, but it ends up with the black screen and I can't do anything. It's quite odd, because I see PA boot animation during start or lock screen, when I leave it inactive for a while. Does anyone have the same problem? Or any idea how to solve it? Thanks. Cheers, Jan -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Bugs(?)/Questions in current PA devel testing
On 05/28/2013 11:01 AM, Marco Martin wrote: On Monday 27 May 2013 21:28:02 Maurice de la Ferté wrote: i don't have at the moment hardware that presents that problem, but one thing that could be tried is to write a small test app with a window with different possible sizes, different possible window flags combination, to see if one in particular reproduces the problem reliably Maybe someone should take a look to that patch request which is dangling for some weeks: https://build.merproject.org/request/show/293 patch seems fine, seems to be about first initialization rther that the accepting input problem I understood Maurice in the way that we could at least use nmcli as a temporary workaround for devs to get wlan? -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: PA in Virtualbox
You can get a built version of plasma-mobile with that patch applied from one of my repositories at build.merproject.org: https://build.merproject.org/project/show?project=home%3Awonko%3Apa-devel-contour-hack Essentially, you just need to copy /usr/bin/contour and /usr/share/autostart/contour.desktop from the plasma-mobile rpm to your installation. But you can of course also install the rpm the usual way. On 05/29/2013 01:34 PM, Jan Grulich wrote: Dne 29.5.2013 12:56, Ruediger Gad napsal(a): Hi Jan, that's a known issue in the current devel branch. There is also a possible fix. It's just not in the repository yet. Is it possible to try this fix? BR, Ruediger On 05/29/2013 12:31 PM, Jan Grulich wrote: Hi, I'm trying to install PA in Virtualbox from Basyskom images. The latest Mer testing image worked without any problems but I want to try the latest PA which, as I think, is in the Mer development images. I'm able to install this image, but it ends up with the black screen and I can't do anything. It's quite odd, because I see PA boot animation during start or lock screen, when I leave it inactive for a while. Does anyone have the same problem? Or any idea how to solve it? Thanks. Cheers, Jan Thanks for the answer, at least I know why it doesn't work. -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Bugs(?)/Questions in current PA devel testing
Sorry, am really short of time right now. On 05/27/2013 07:21 PM, Marco Martin wrote: On Monday 27 May 2013 15:49:41 Ruediger Gad wrote: As you can see in the patch there were also some bugs in the code itself. I will try to submit that patch into git if I find the time. cool, would be appreciated if you can push it :) Will do, but cannot promise when. Unfortunately, there is still the issue that the network manager applet does not accept input. Do you have any hints on how to identify or even solve the problem, or where the problem could be located? unfortunately i really don't know, and is a pretty important issue :/ Agreed, I also think it's very important. Maybe Maurice suggestion can help to workaround the issue? At least for developers it would be easier to get wlan access this way, it maybe even makes debugging this issue a little easier. but seems quite low in the stack.. what i think it could be from is from the touchscreen driver, depending from either the x11 windows flags, or the window size (like windows smaller than a certain size not taking inut? seems utterly weird, but i'm a bit out of ideas) the weird thing is that happens only with those popus, not any app afaik? so depends definitely from one of those two things Another weird observation: in older images the network manager pop-up used to work. It stopped working at some time for no apparent reason. Also, iirc the power pop-up used to not to work (i.e. not accept input) in older versions but it works now. i don't have at the moment hardware that presents that problem, but one thing that could be tried is to write a small test app with a window with different possible sizes, different possible window flags combination, to see if one in particular reproduces the problem reliably -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: PA 4 Rootfs Tarballs for Archos G9
Hi Robin, On 05/23/2013 07:40 AM, robin wrote: hi ruediger, I would very much like to try them out. I am new to plasma active, so please excuse if these are stupid questions: when do I use fb and when pvr / what are the abbreviations? This describes which graphics driver is used. fb is a pure open source driver while pvr uses a closed source driver. The difference is that pvr has 3d acceleration and that video playback works at least a little bit with this one. is your branch pa4 or is it pa3 as the path mentions pa3? It uses the current testing branch so it is based on the upcoming pa4. Please correct me if I am mistaken. in case it is a pa4 preview, will it be easy to upgrade to the final pa4, eg via zypper ref zypper dup? Yes, upgrading should be easily possible this way. best regards and many thanks for keeping the G9 in active development You're welcome. It's nice to see Archos G9 is still being used with pa. :) Just in case, this guide also applies to the new images: https://share.basyskom.com/plasma-active/archos_gen9.html BR, Ruediger robin ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Bugs(?)/Questions in current PA devel testing
I have some more findings and will try to briefly explain these in the hope that they are useful. On 05/26/2013 01:06 AM, Aaron J. Seigo wrote: On Sunday, May 26, 2013 00:50:37 Ruediger Gad wrote: Another thing I found with respect to the devel image is that it is missing /usr/bin/contour which is present in the testing version. I assume contour is responsible for painting the main parts of the UI? No, plasma-device is. contourd should not be required. I just gave it a try with a fresh image based on the devel branch. With the stock image the ui does not come up. Then I manually copied /usr/bin/contour and /usr/share/autostart/contour.desktop from the plasma-mobile rpm of the testing branch. With the copied files in place the ui comes up as it should. ps also shows a running contour process. With respect to the changes in plasma-mobile that lead to contour not being built: There is a new automatism that dynamically determines at build-time whether contourd should be build or not. There seem to be some errors related to this, at least on arm. Attached, I send you a diff with from my attempt to let it build again. Please note that those changes are very hackish and are only intended as experiments. I am not even sure if QtNetworkManager vs. NetworkManagerQt is really a mistake. But I hope that it is suited to give an impression of what may go wrong. Also note that even though the build progressed more with these changes it still failed. I hope you consider this as useful. BR Ruediger ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com diff --git a/CMakeLists.txt b/CMakeLists.txt index b0d15e9..893524e 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -15,7 +15,7 @@ find_package(Nepomuk REQUIRED) find_package(Soprano) find_package(PkgConfig) -pkg_check_modules(QTNETWORKMANAGER QtNetworkManager) +pkg_check_modules(QTNETWORKMANAGER NetworkManagerQt) macro_log_feature( Soprano_FOUND Soprano Semantic Desktop Storing http://soprano.sourceforge.net; diff --git a/contourd/CMakeLists.txt b/contourd/CMakeLists.txt index 9fbeb3e..30486f0 100644 --- a/contourd/CMakeLists.txt +++ b/contourd/CMakeLists.txt @@ -12,6 +12,7 @@ include_directories( ${QT_INCLUDES} ${KDE4_INCLUDES} ${NEPOMUK_INCLUDE_DIR} + /usr/include/NetworkManager ${SOPRANO_INCLUDE_DIR} # QtMobility doesn't work for us anyway: @@ -102,7 +103,7 @@ set(recommendationmanager_SRC set(ADDITIONAL_LINK_LIBRARIES ${ADDITIONAL_LINK_LIBRARIES} -QtNetworkManager +NetworkManagerQt ) diff --git a/contourd/location/network-engines/solid/SolidNetworkNotifier.cpp b/contourd/location/network-engines/solid/SolidNetworkNotifier.cpp index 759042f..4989b15 100644 --- a/contourd/location/network-engines/solid/SolidNetworkNotifier.cpp +++ b/contourd/location/network-engines/solid/SolidNetworkNotifier.cpp @@ -23,8 +23,8 @@ #include solid/control/wirelessnetworkinterface.h #include solid/control/wirelessaccesspoint.h -#include QtNetworkManager/manager.h -#include QtNetworkManager/wirelessdevice.h +#include NetworkManagerQt/manager.h +#include NetworkManagerQt/wirelessdevice.h #include QHash smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Bugs(?)/Questions in current PA devel testing
Sorry, the log file was truncated. Attached I send the full log, including a shutdown, that had been triggered via the lock screen. On 05/25/2013 12:38 PM, Ruediger Gad wrote: Hi, sorry if this had been asked before. As you probably know, I created new PA (devel testing) rootfs tarballs for Archos G9. There are right now issues with both devel and testing. In testing, the network manager applet does not accept input. In devel, the UI starts, as in, X comes up, the startup sound is played and the volume setting bar can be shown. But except that the screen stays completely black. The lock-screen works and can be triggered via the power button. Suspend/resume works as well. Interestingly, when trying to debug this issue I noticed that the usual startactive log files are not in /tmp. I could get the log output by changing the /usr/bin/startactive script to redirect to /home/mer instead of /tmp. Is /tmp wiped or re-mounted at some time during boot? Attached, I send the corresponding .log and .env files for reference. Are these known bugs? Should bug reports be filed for these ones or do you have hints on how to further debug or even fix these issues? BR Ruediger ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com XDG_VTNR=1 XDG_SESSION_ID=c1 QT_DEFAULT_RUNTIME_SYSTEM=meego OPTIONS=-background none AG_SERVICE_TYPES=/usr/share/accounts/service-types USER=mer QML_IMPORT_PATH=/usr/lib/kde4/imports/ AG_SERVICES=/usr/share/accounts/services PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin QT_IM_MODULE=Maliit PWD=/home/mer LANG=C GSETTINGS_BACKEND=gconf QT_GRAPHICSSYSTEM=raster NOTIFY_SOCKET=@/org/freedesktop/systemd1/notify SHLVL=1 XDG_SEAT=seat0 HOME=/home/mer XCURSOR_THEME=plasmamobilemouse AG_PROVIDERS=/usr/share/accounts/providers LOGNAME=mer DBUS_SESSION_BUS_ADDRESS=unix:path=/run/user/1000/dbus/user_bus_socket XDG_RUNTIME_DIR=/run/user/1000 DISPLAY=:0 M_DECORATED=0 G_BROKEN_FILENAMES=1 _=/usr/bin/env QDBusConnection: session D-Bus connection created before QCoreApplication. Application may misbehave. --- StartActive - StartActive: Setting environment variables: StartActive: KDEDIR = /usr StartActive: KDEDIRS = /usr StartActive: KDE_FULL_SESSION = true StartActive: KDE_SESSION_VERSION = 4 StartActive: LD_LIBRARY_PATH = /usr/lib:/usr/lib/ StartActive: PATH = /usr/bin:/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games StartActive: QML_IMPORT_PATH = /usr/lib/kde4/imports StartActive: QT_IM_MODULE = Maliit StartActive: QT_PLUGIN_PATH = /usr/lib/kde4/plugins StartActive: XCURSOR_PATH = /usr/share/icons::~/.icons:/usr/share/icons:/usr/share/pixmaps:/usr/X11R6/lib/X11/icons --- StartActive - StartActive: Initializing DBus StartActive: DBus already running at: unix:path=/run/user/1000/dbus/user_bus_socket StartActive: Loading modules: active StartActive: these are the currently running modules: QSet() StartActive: starting the following modules: QSet(lnusertemp) StartActive: starting module lnusertemp ProcessStarter: lnusertemp nothing to exec - meta-module SPLASH THEME: /usr/share/kde4/apps/ksplash/Themes/ActiveAir file:///usr/share/kde4/apps/ksplash/Themes/ActiveAir/main.qml:45: ReferenceError: Can't find variable: stage1BottomLeftVine ProcessStarter: lnusertemp process finished StartActive: module started lnusertemp 4 ( 673 ms ) StartActive: 0 - QSet() StartActive: 1 - QSet(contour, setup-kde-skel, activitymanager, kdeinit, prestart) StartActive: 2 - QSet(ksmserver, nepomuk) StartActive: 3 - QSet(active, plasma-device) StartActive: Modules that depend on the current one: QSet(prestart) StartActive: these are the currently running modules: QSet() StartActive: starting the following modules: QSet(prestart) StartActive: starting module prestart ProcessStarter: prestart nothing to exec - meta-module ProcessStarter: prestart process finished StartActive: module started prestart 4 ( 677 ms ) StartActive: 0 - QSet() StartActive: 1 - QSet(contour, setup-kde-skel, activitymanager, kdeinit) StartActive: 2 - QSet(ksmserver, nepomuk) StartActive: 3 - QSet(active, plasma-device) StartActive: Modules that depend on the current one: QSet(setup-kde-skel) StartActive: these are the currently running modules: QSet() StartActive: starting the following modules: QSet(setup-kde-skel) StartActive: starting module setup-kde-skel ProcessStarter: setup-kde-skel will wait for the process to finish. ProcessStarter: setup-kde-skel exec process setup-kde-skel wait for /home/mer/.kde/share/config/plasma-device-appletsrc already exists, will not overwrite ProcessStarter: setup-kde-skel process finished StartActive: module started setup-kde-skel 4 ( 815 ms ) StartActive: 0 - QSet() StartActive: 1
Re: Bugs(?)/Questions in current PA devel testing
On 05/25/2013 10:55 PM, Marco Martin wrote: On Saturday 25 May 2013 12:38:05 Ruediger Gad wrote: Interestingly, when trying to debug this issue I noticed that the usual startactive log files are not in /tmp. I could get the log output by changing the /usr/bin/startactive script to redirect to /home/mer instead of /tmp. Is /tmp wiped or re-mounted at some time during boot? not that i know of Apparently, /tmp is now on a tmpfs in memory. This explains why there are no files left after shutdown. iirc this is new? Another thing I found with respect to the devel image is that it is missing /usr/bin/contour which is present in the testing version. I assume contour is responsible for painting the main parts of the UI? This would explain why a black screen is shown but some parts (like lock screen or volume settings) of the UI work. Are these known bugs? Should bug reports be filed for these ones or do you have hints on how to further debug or even fix these issues? i heard some times about windows not accepting input and happening only on arm, unfortunately i have no idea why :/ -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
PA 4 Rootfs Tarballs for Archos G9
Hi, I uploaded fresh rootfs tarballs of the plasma active testing branch for Archos G9 to: http://files.kde.org/plasma/active/3.0/images/archos-g9/armv7hl/ There are tarballs for both fb and pvr. BR Ruediger PS: Thanks a lot to Marco for the heads-up with respect to nemo-mobile-session. :) -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: repos and kickstarts for pa4
Hi, last night, I gave it a quick try to create images for Nexus7. Unfortunately, both versions, devel and testing didn't run well out of the box. I cannot promise anything as I am very busy with other stuff these days but I'll try to look into this again tonight. At least for the testing version the UI came up. Unfortunately, it froze pretty soon afterwards. I'll keep you updated. Cheers, Ruediger On 05/21/2013 11:11 AM, Daniel Kreuter wrote: I would linke Tod hell Festung, if so eine (maybe Ruediger?) would provide an image for the nexus 7 device. So long Am 20.05.2013 21:46 schrieb Marco Martin notm...@gmail.com mailto:notm...@gmail.com: On Mon, May 20, 2013 at 9:20 PM, Hillel Lubman shtetl...@gmail.com mailto:shtetl...@gmail.com wrote: May be Plasma Active can set up a donations page so anyone interested would help key developers to be equipped with important devices (like key x86 and ARM ones). This can even out the effort since right now, ARM feels like a lower priority and bugs there are discovered slower. The problem is not really that, I don't think more devices to current developers could help much. To scale in a manageable fashion, more devices is completely secondary to have more people involved and willing to maintain adaptations. Cheers, Maco Martin ___ Active mailing list Active@kde.org mailto:Active@kde.org https://mail.kde.org/mailman/listinfo/active ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
New Rootfs Tarball for Nexus 7
Hi, I just uploaded a new Plasma Active rootfs tarball for Nexus 7: http://files.kde.org/plasma/active/3.0/images/nexus7/plasma-active-nexus7-devel-mer-latest-2013-03-28.tar.gz.mirrorlist This is the first one that uses the new repositories at build.merproject.org. Please be aware that I didn't have the time to test the new rootfs tarball yet. So, it might not even boot properly. Regards, Ruediger -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Re: Archos G9 devel image
Hi, attached I send two .ks files: one for creating a testing and one for creating a devel rootfs tarball. Please note that these .ks files will generate armv7hl installations instead of armv7l as can be downloaded from basyskom. Both use the pvr driver. I briefly tested the generated tarballs: The testing one seems to work ok. It booted fine and I could make a suspend/resume cycle without problems. The devel one boots and the GUI comes up. However, here the GUI got somewhat stuck. Nothing responds except actions being triggered via the hardware buttons (volume, lock screen). I don't know if this may be caused by my device battery being nearly empty (At least I saw some message at bootup that may have caused the GUI to lock up.). I will test that later after letting the device charge a little. I generated the rootfs tarballs as follows: PATH=$PATH:/sbin sudo mic create fs plasma-active-archos-gen9-pvr-mer-devel-2013-02-23.ks --pack-to=plasma-active-archos-gen9-pvr-mer-devel-2013-02-23.tar.gz --pkgmgr=yum --arch=armv7hl Up to now, the bits for the armv7hl hardware adaptation are only hosted in my private repositories on cobs@meego: https://build.pub.meego.com/project/show?project=home%3Awonko%3Aarchos-g9-pa-adaptation https://build.pub.meego.com/project/show?project=home%3Awonko%3Aarchos-g9-pa-adaptation%3Apvr If there is the interest I (or someone else) could push this stuff into a little bit more official place. Would creating own projects for an Archos G9 armv7hl adaptation make sense? Imho integrating the changes with the existing Archos G9 armv7l adaptation would be pretty complicated as we would need to distinguish between soft and hard float for building some packages in a single common project. I could also upload the images somewhere if you like to test those yourself. Regards, Ruediger On 02/22/2013 09:10 AM, Hillel Lubman wrote: On 22/02/13 03:05, Yannick Kiekens wrote: As work-around to get latest devel stuff running on your device, just add the following repo to latest testing based archos gen9 image (try the pvr one): zypper ar -f http://repo.pub.meego.com/Project:/KDE:/Devel/CE_UX_PlasmaActive_armv7l/ develop This gave an error while doing the upgrade of kdebase (something about a directory that can't be deleted) I'll have more time to look at it this weekend To bad to hear about the support problems, it is nice hacking a tablet with vendor support. greatings Yannick Kiekens There was a similar issue with switching Nexus 7 build from testing to devel. See here in the comments: http://aseigo.blogspot.com/2013/01/plasma-active-stable-devel-and-bleeding.html may be it'll be helpful. Regards, Hillel Lubman. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com # -*-mic2-options-*- -f loop --save-kernel -*-mic2-options-*- # # Do not Edit! Generated by: # kickstarter.py # lang en_US.UTF-8 keyboard us timezone --utc Europe/Berlin rootpw mer desktop --autologinuser=mer user --name mer --groups audio,video --password mer repo --name=ce-utils --baseurl=http://repo.pub.meego.com/Mer:/Tools:/Testing/Mer_Core_armv7hl/ --save repo --name=mer-core --baseurl=http://releases.merproject.org/releases/latest/builds/armv7hl/packages/ --save repo --name=mer-core-debuginfo --baseurl=http://releases.merproject.org/releases/latest/builds/armv7hl/debug/ --save repo --name=mer-shared --baseurl=http://repo.pub.meego.com/CE:/MW:/Shared/Mer_Core_armv7hl/ --save #repo --name=mer-shared --baseurl=http://repo.merproject.org/obs/nemo:/devel:/mw/latest_armv7hl/ --save repo --name=mer-extras --baseurl=http://repo.pub.meego.com/Project:/KDE:/Mer_Extras/CE_MW_Shared_Mer_Core_armv7hl/ --save repo --name=plasma --baseurl=http://repo.pub.meego.com/Project:/KDE:/Devel/CE_UX_PlasmaActive_armv7hl/ --save repo --name=adaptation-archos-gen9 --baseurl=http://repo.pub.meego.com/home:/wonko:/archos-g9-pa-adaptation/Project_KDE_Devel_CE_UX_PlasmaActive_armv7hl/ --save repo --name=adaptation-archos-gen9-pvr --baseurl=http://repo.pub.meego.com/home:/wonko:/archos-g9-pa-adaptation:/pvr/Project_KDE_Devel_CE_UX_PlasmaActive_armv7hl/ --save %packages # ce_tools repository # # mer-core repository @Mer Core Utils # connman-test diffutils openssh-clients vim-enhanced tar @Mer Core # Mer Core defines following packages (06 dec 2011) # basesystem bash boardname coreutils deltarpm e2fsprogs file filesystem fontpackages-filesystem # kbd lsb-release meego-release nss pam passwd prelink procps readline rootfiles rpm setup # shadow-utils shared-mime-info systemd-sysv time udev usbutils util-linux xdg-user-dirs zypper @Mer Connectivity # Mer Connectivity defines following packages (06 dec 2011) # bluez connman crda iproute iputils net-tools ofono wireless-tools wpa_supplicant #remove connman
Re: Archos G9 devel image
PS: After letting it charge a little the devel image also recovered. Some part of the GUI (kwin or something) had crashed. So, the devel version, in general, works as well. :) On 02/23/2013 11:36 AM, Ruediger Gad wrote: Hi, attached I send two .ks files: one for creating a testing and one for creating a devel rootfs tarball. Please note that these .ks files will generate armv7hl installations instead of armv7l as can be downloaded from basyskom. Both use the pvr driver. I briefly tested the generated tarballs: The testing one seems to work ok. It booted fine and I could make a suspend/resume cycle without problems. The devel one boots and the GUI comes up. However, here the GUI got somewhat stuck. Nothing responds except actions being triggered via the hardware buttons (volume, lock screen). I don't know if this may be caused by my device battery being nearly empty (At least I saw some message at bootup that may have caused the GUI to lock up.). I will test that later after letting the device charge a little. I generated the rootfs tarballs as follows: PATH=$PATH:/sbin sudo mic create fs plasma-active-archos-gen9-pvr-mer-devel-2013-02-23.ks --pack-to=plasma-active-archos-gen9-pvr-mer-devel-2013-02-23.tar.gz --pkgmgr=yum --arch=armv7hl Up to now, the bits for the armv7hl hardware adaptation are only hosted in my private repositories on cobs@meego: https://build.pub.meego.com/project/show?project=home%3Awonko%3Aarchos-g9-pa-adaptation https://build.pub.meego.com/project/show?project=home%3Awonko%3Aarchos-g9-pa-adaptation%3Apvr If there is the interest I (or someone else) could push this stuff into a little bit more official place. Would creating own projects for an Archos G9 armv7hl adaptation make sense? Imho integrating the changes with the existing Archos G9 armv7l adaptation would be pretty complicated as we would need to distinguish between soft and hard float for building some packages in a single common project. I could also upload the images somewhere if you like to test those yourself. Regards, Ruediger On 02/22/2013 09:10 AM, Hillel Lubman wrote: On 22/02/13 03:05, Yannick Kiekens wrote: As work-around to get latest devel stuff running on your device, just add the following repo to latest testing based archos gen9 image (try the pvr one): zypper ar -f http://repo.pub.meego.com/Project:/KDE:/Devel/CE_UX_PlasmaActive_armv7l/ develop This gave an error while doing the upgrade of kdebase (something about a directory that can't be deleted) I'll have more time to look at it this weekend To bad to hear about the support problems, it is nice hacking a tablet with vendor support. greatings Yannick Kiekens There was a similar issue with switching Nexus 7 build from testing to devel. See here in the comments: http://aseigo.blogspot.com/2013/01/plasma-active-stable-devel-and-bleeding.html may be it'll be helpful. Regards, Hillel Lubman. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
[Active] [Bug 310508] Repaint issue in Task Switcher
https://bugs.kde.org/show_bug.cgi?id=310508 Ruediger Gad r@gmx.de changed: What|Removed |Added CC||r@gmx.de --- Comment #4 from Ruediger Gad r@gmx.de --- Jfyi, I filed a submit request for a patch to solve this issue for the current trunk testing of Plasma Active on MeeGo cobs: https://build.pub.meego.com/request/show/7761 -- You are receiving this mail because: You are on the CC list for the bug. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
Archos G9 pvr with susepend/resume
Hi, there was an issue in the PlasmaActive Archos G9 pvr version that prevented proper suspend/resume. More particular, the problem was that the display wasn't coming back on resume and stayed black. Now, there is a workaround to fix this: https://build.pub.meego.com/package/show?package=x11-resume-hackproject=Project%3AKDE%3AMer_Extras%3AAdaptation%3AArchos-gen9%3Apvr Thanks to this workaround suspend/resume had been enabled again. This change should have made it already in the latest weekly image. Unfortunately, I couldn't test the image yet. The version in MeeGo cobs (from which the image was built) already works properly. As one consequence, the config settings that disabled the suspend/resume functionality in the pvr version had been reverted. Regards, Ruediger -- http://ruedigergad.com smime.p7s Description: S/MIME Cryptographic Signature ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
[Bug 307691] Image Viewer does not open image from local directory (using DirModel)
https://bugs.kde.org/show_bug.cgi?id=307691 Ruediger Gad r@gmx.de changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED Latest Commit||http://commits.kde.org/plas ||ma-mobile/47a6624960f38d49f ||d98949f9d0647ed9ee61044 --- Comment #4 from Ruediger Gad r@gmx.de --- Git commit 47a6624960f38d49fd98949f9d0647ed9ee61044 by Ruediger Gad. Committed on 15/10/2012 at 23:11. Pushed by ruedigergad into branch 'master'. Add via QML callable get method to DirModel. The get method returns a QVariantMap. Currently this maps contains the keys url and mimeType, but it can be easily extended. This obsoletes the previous hack. M +1-8 applications/filebrowser/imageviewerpart/contents/ui/ViewerPage.qml M +11 -5components/dirmodel/dirmodel.cpp M +2-1components/dirmodel/dirmodel.h http://commits.kde.org/plasma-mobile/47a6624960f38d49fd98949f9d0647ed9ee61044 -- You are receiving this mail because: You are the assignee for the bug. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active
[Bug 307691] New: Image Viewer does not open image from local directory (using DirModel)
https://bugs.kde.org/show_bug.cgi?id=307691 Bug ID: 307691 Severity: normal Version: unspecified Priority: NOR Assignee: active@kde.org Summary: Image Viewer does not open image from local directory (using DirModel) Classification: Unclassified OS: Linux Reporter: r@gmx.de Hardware: Other Status: UNCONFIRMED Component: General Product: Active On PlasmaActive (image: basyskom-plasma-active-archos-gen9-omapfb-tablet-mer-testing-120921-1829.tar.bz) it is not possible to open images from a local directory (e.g. mounted sd-card) via the built-in image viewer. Icons for the images are displayed. However, clicking on an image yields in a black screen that shows loading The image bar at the bottom works. When running the image viewer from the command line the error is as follows: file:///usr/share/kde4/apps/plasma/packages/org.kde.active.imageviewerpart/contents/ui/ViewerPage.qml:218: TypeError: Result of expression 'fileBrowserRoot.model.get' [undefined] is not a function. I assume the reason is that when navigating a local filesystem a DirModel (more precisely KDirModel wrapped via DirModel in plasma-mobile) is used for fileBrowserRoot.model. However, there is no get method that can be invoked via QML. Hence, the above error is occurs. Reproducible: Always Steps to Reproduce: 1. Open image viewer 2. Try to open an image from a local file and not the gallery. Actual Results: Black screen showing loading... is displayed. Expected Results: The selected image should be displayed. -- You are receiving this mail because: You are the assignee for the bug. ___ Active mailing list Active@kde.org https://mail.kde.org/mailman/listinfo/active