Posts by Schreibwaise

    unfortunately : no

    i do get:

    Code
    multimedia.vdr-addon$ ls -la
    insgesamt 4376
    drwxrwxr-x 3 lutz lutz    4096 Mär  6 21:20 .
    drwxrwxr-x 3 lutz lutz    4096 Mär  6 17:22 ..
    drwxrwxr-x 2 lutz lutz    4096 Mär  6 17:22 resources
    -rw-rw-r-- 1 lutz lutz 4468118 Mär  6 21:18 service.multimedia.vdr-addon-9.0.107.zip
    lutz@libreelec:/develop/LibreELEC.tv/target/addons/9.0/WeTek_Play_2/arm/service.multimedia.vdr-addon$

    but the zip does not contain "my" plugin

    Look into addons/service/vdr-addon/package.mk . There is a variable PKG_DEPENDS_TARGET where you need to add your plugin. Rebuild the vdr-addon and it should contain your plugin.

    Works like charm ;) - Thx

    Patch #1 for fixing Makefile and lib/Makefile (python-config to be replaced by pkg-config ...) put under

    vdr-plugin-epg2vdr/patches/01_Makefile-python.patch

    Ahh, so the name of the patch file does not need to start with the package name! This kind of information is not documented anywhere AFAIK, so one needs to learn by try and error and investigating other packages.

    Anyway, I really love libreELEC's build system :)

    The error "has no member named 'SetText' ... indicates that it uses Old- Version of tinyxml2 - see here for example.

    I found that under packages/textproc there are two tinyxml's - neither seems suitable to me - what to do?

    These two versions are tinyxml and tinyxml2 and can be considered two distinct packages. But I assume the tinyxml2 is just a too old version for you (it is at 1.0.12 while the newest version available seems to be 6.1.0). See here. You could try to modify the package.mk of tinyxml2 to use the newer version. However I do not know if this might break other parts of the system.

    I am around the python-stuff though - the Makefiles for epg2vdr need to be tweaked to use pkg-config. i recognized that in some directories of packages there are also patches - are those patches applied automatically or how does that work? For the time beeing i would need to patch the epg2vdr-plugin's Makefile to compile .... -

    Yes, those patches get applied automagically when the package is being built. The patch file name needs to obey to some naming scheme, however: It needs to start with the name of the package and needs to end with ".patch". And the patch needs to apply with "-p1". Just have a look at patches of some other package to get the idea.

    Unfortunately I cannot help with your other issues; but I'm following this thread as I'm a VDR user myself. Thanks for working on this!

    I am experiencingin almost daily crash (when I am not watching!!) with Librelec 8.2.2 on Odroid c2.

    I don't know how am I supposed to debug this, specially with a SIGBUS empty !!

    I'm "happy" I'm not the only one seeing this issue. I'm experiencing these crashes since months on five different devices (generic Intel, Raspberry Pi3, vero 4k), with LibreELEC and with OSMC.

    Thread on kodi.tv and crashlog.

    Even see this when I only boot a system and let it on its own.

    Type "df" to find out where your USB stick is mounted. You should see something like:

    /dev/sda1 524008 136328 387680 26% /var/media/LIBREELEC

    Instead of "LIBREELEC" you will see the label of your stick.


    Then remount the drive readonly by giving this command:

    mount -o remount,ro /var/media/LIBREELEC

    Yes, there is a comma between "remount" and "ro". "ro" means "read-only" and is the parameter for the remount option.

    But beware that this solution is not persistent! If you unplug the USB stick and put it back in, it will be mounted writable again and your write protection is gone.

    I'm also new to this but with a wrong dtb.img the device definitely does not boot (tried it myself). So just try it, it is safe as you can always recreate your SD card. For me the already contained (big) dtb.img worked, though. But trying to install to internal always was failing, so I tried several ones of them and finally by using gxl_p212_2g_slowemmc I was able to install current master onto the internal storage of a vero 4k. Future updates via the tarfile are also working now because the updater can read from /proc/device-tree/le-dt-id which dtb.img to use.

    By using keycodes you can map arbitrary keys. You probably are more interested in the symbolic names kodi knows. If I understand correctly you want to use the Back key to enter the home menu when watching live TV. So maybe you might try this in your keymap.xml:

    Code
     <FullscreenVideo>
        <remote>
          <back>ActivateWindow(home)</back>
        </remote>
      </FullscreenVideo>


    If this brings you to the home menu but Live-TV stops, you probably need to change some setting for your skin. With confluence I have "show background 'Now playing' video" enabled.

    Sorry if I used the wrong wording. I'm not a native English speaker, so I probably got it wrong. What I wanted to say by using the term "bogus" is that the error message seems to be wrong since there is no user noticable problem.

    Reproducing is trivial, as I just need to play a recording. Nothing special. And the log file is *huge* since kodi is reporting the error several times per second.

    Anyway, meanwhile I'm starting to understand what's going on: The vdr server is running LE 8.2.3, so the vdr server is still at protocol 10 while the client is already at 12. But the method GetStreamTimes was introduced only with protocol 11, so trying to use this is expected to fail. I'm amazed how well this all is working nevertheless (so far I did not notice any other problem). I'm now using a modified version of the client to prevent my disk being filled with kodi logs. I cannot yet update the server.

    Please accept my apologies for the noise. At least I learned something new and the sound driver for this specific device will get a new module parameter in the kernel, as I was investigating this issue with one of the ALSA developers and we found out that it worked with older kernels only by luck (just choosing another device is not enough as this will lead to broken volume control; the new parameter will allow to get the same behaviour as with kernel 4.11.12).