Posts by chewitt

    Google claims the Vorke Z6 plus is the same as some Tanix box which had board pictures showing QCA9377 as the SDIO module. SDIO WiFi should probe and load automagically regardless of device-tree content, so not sure why it doesn't, but BT is serial UART based and needs the correct content in device-tree to probe and load drivers. VFD also needs content in device-tree and some detective work.

    If it's a Gbit Ethernet box the Q200 device-tree should work (not Q201 which is for 10/100 boxes). I'd guess Ethernet not-working correctly is the reason for adding the USB Ethernet adaptor?

    I'll think about creating something, but start with logs from Q200 dtb please. If you have the original dts or dtb file from Android that would also be useful to have.

    For other boxes; start individual threads else I easily lose track of things.

    Code
    WP2:~ # find / -name brightness
    /sys/devices/platform/leds/leds/wetek-play:wifi-status/brightness
    /sys/devices/platform/leds/leds/wetek-play:ethernet-status/brightness
    /sys/devices/platform/leds/leds/blue:power/brightness

    Let's say the blue:power LED is the annoying one:

    Code
    echo 0 > /sys/devices/platform/leds/leds/blue:power/brightness

    Should make it turn off.. and we can script this to run after startup:

    Code
    echo "(sleep 10 && echo 0 > /sys/devices/platform/leds/leds/blue:power/brightness)&" > /storage/.config/autostart.sh

    Should result in the LED being turned off 10 seconds after boot starts. You will obviously need to modify the /path/to/brightness for whatever LEDs are on the board. NB: this only works if the LEDs are GPIO controllable; some LEDs are hardwired.

    You can power the board off from flirc, but not power on as the USB device is not awake/listening once off. It may be possible to configure a power on event from the IR sensor, but that normally depends on the low-level boot firmware and the device being suspended rather than powered off (so the sensor can receive the power-on code). jernej would need to comment on that possibility.

    The Cubox keymap is probably for an older Linux kernel and will need to be transcribed into the 'toml' format used now. There is a section in the wiki on creating custom keymaps. It needs to be updated to reference the toml format but the files you need to create and their locations are still correctly described. There are lots of existing keymaps in /usr/lib/udev/rc_keymaps/ to crib the toml format from.

    Or the whole point of flirc is that you can program it to respond to any remote. If you install the config app on a PC it should take no more than 2 mins to map all the core Kodi functions to remote buttons.

    There is a prometheus agent add-on for LE users that want to monitor stats on their HTPC (everyone else prefers to just watch movies). If you are trying to develop an application; LE is the wrong codebase to work from as our distro packaging will be rather restrictive.

    Re-reading the original screenshot: You ran emmctool without actually having the AMLGX file to write on the /storage partition so it's errored and written no data to the device. Hence boot has still worked. If the box boots; that's where advice stops because I have no interest in wading into the shitfest of atetmpting support for internal installs on hardware I don't have. I've deleted the GT1 image from my test share.

    I deleted the other random post that bumped a thread from 2019. Install the System-Tools add-on and you have 7-zip as an alternative to normal zip.

    And you have a low-spec ARM board with an older Mali 450 GPU. It uses exactly the same mesa/lima graphics as LE10 so I would not expect any significant change to GUI behaviour. IMHO scrolling and navigation performance is more dependent on the remote device being used than anything else.

    VNC is VNC .. and works with the legacy Pi codebase in LE 9.2.8. However VNC stopped working on RPi and all platforms using V4L2/GBM from LE10 onwards. It's an inherent problem with how zero-copy graphics pipelines work and a "fix" will be non-trivial to engineer; and we're not aware of anyone working on something.