Posts by yubby

    Oh queso, if someone else can confirm the 'issue' that I believe now exists with the Chrome ".deb" file download (executed via "curl" in the "chrome-downloader" script) in "/storage/.kodi/addons/browser.chrome/bin" installed by the "Chrome" addon..

    Apparently, Google has removed the "dl.google.com" directory paths and contents, forcing a 'web browser app' method of obtaining the installation files, no longer allowing specific selection of a version (or knowledge of what version is to be downloaded).

    Such that, I created a work-around in which I manually downloaded the 'latest' DEB linux version (via their website), copying it to a 'Configs' sub-directory on my LE unit (which I keep for all the 'manual' stuff I have to install) and modifying the "chrome-download" script to copy it into the 'temp' area where it expects the 'curl downloaded' file to be placed...

    It 'faked it out' quite well, and now have version 125.0.6422.141 installed...

    I haven't found any information on a 'new' direct-download site, which may mean that all updates will have to be performed 'manually' in the future...

    (Using LE 12.0 generic legacy X86_64)


    > diff chrome-downloader.orig chrome-downloader
    12c12,14
    < CHROME_FILE="google-chrome-stable_109.0.5414.74-1_amd64.deb"
    ---
    > ##CHROME_FILE="google-chrome-stable_125.0.6422.141_amd64.deb"
    > CHROME_FILE="google-chrome-stable_current_amd64.deb"
    37c39
    < echo "Downloading Chrome"
    ---
    > ##echo "Downloading Chrome"
    40,50c42,59
    ---
    > ##rm -f ${CONTROL_FILE} ${DATA_FILE}
    > ##(
    > ##  curl -# -O -C - https://dl.google.com/linux/chrome/deb/pool/main/g/google-chrome-stable/${CHROME_FILE} 2>${DATA_FILE}
    > ##  touch ${CONTROL_FILE}
    > ##) | \
    > ##  while [ : ]; do
    > ##    [ -f ${DATA_FILE} ] && prog="$(tr '\r' '\n' < ${DATA_FILE} | tail -n 1 | sed -r 's/^[# ]+/#/;s/^[^0-9]*//g')" || prog=
    > ##    kodi-send --action="Notification(Downloading Chrome,\"${prog:-0.0%}\",3000,${ICON})" >/dev/null
    > ##    [ -f ${CONTROL_FILE} ] && break
    > ##    sleep 4
    > ##  done
    >
    > ##rm -f ${CONTROL_FILE} ${DATA_FILE}
    >
    >
    > ## Copy manually-downloaded chrome image to 'temp' directory:
    > cp -p ~/Configs/addons/${CHROME_FILE}  $ADDON_DIR/tmp_download/
    >

    LE also supports C2 which has been dropped by CE; again with the upstream codebase which works better on older hardware (vdec drivers work and were written for older hardware which is one reason why things don't work so well on newer hardware)

    But, when I check the 'downloads' page, I see that 'all' of the AMlogic releases are still only up to version 9.0.2 (which I 'have' used on the C2 in the past...).

    I would like to jump on this thread, to inform of the same situation, when playing local 'music files' (MP3's on the internal disk drive), which does not appear to be seen as 'activity', and the LE (X86 PC) unit goes through it's 'sleep/shutdown' steps at the configured time settings (I have 15/30/45 for screensaver, display turnoff, shutdown/sleep)

    If I catch it in time (just after the TV goes black), and press any button on my remote control, it quickly re-awakens and continues playing the current track...

    I wish I had read 'this thread' last night, when I had updated a friend's Libre system to the 10.0.1 version.

    Chrome would not launch if 'custom audio device' was selected (regardless of output device, though I was specifically attempting to use the DisplayPort output 0,3, from the same vendor-list of devices encountered by ppkstat).

    I will test 'locally', though now I have to wait until he returns from his trip to test the new 'syntax' on his...

    So, is the next 'update' for the Odroid 'C2' and 'C4' still targeted for the next major release ('10.x') ?

    Though I 'would' prefer to only deal with '1' distribution... I had wanted the latest version available (to match my x86 installs) which CE offered...

    And, they looked like Mentiads 'to me'...

    Quote

    1-627c88a26a9f1d3bd50e67ea5924065e67834864.png

    chewitt

    LE 9.2 does not exist for Amlogic devices. We got fed-up with working in an awful vendor kernel and changed our focus to development of mainline kernel support. LE10 (Kod 19) is planned to have support for C2 again.

    Or, has the Odroid C2 (and the upcoming Odroid 'C4') development and releases been 'delegated' to the CoreELEC group ?

    I've been testing 'their' 9.2.2 release without issue (though they don't have 'your' repository(ies) )...

    Just an FYI (1 year later). I had begun swapping-out the units that used the Nvidia-based HDMI connections, replacing with mini-PC's that have integrated DisplayPort 1.2 connections, feeding off of the embedded Intel GPU, usually the HD4000 (which I then used DP to HDMI cables).

    The issues with the 4K TV's is no longer an issue, which may indicate that there is an issue with specific graphic chipsets, age and compatibility, etc...

    Or not...

    I purchased 'this', which within Windows 10, linked at 433mbps, and was configured and connected to my 'ac' router without issue (and streams full 4K videos without 'drops').

    Alfa AWUS036ACS ($15 - $20 on that 'bay' site, and comes with a USB extension cable, so you can relocate the antenna)

    Note: using LE 9.0.2 for Odroid


    BUT, I just downloaded the 'nightly' dev image, and tested it (briefly), discounting everything I just said...

    MCE remote 'button pushes' aren't being 'seen' via CEC, though the limited TV remote-control's are (and I'm aware that the MCE keyboard configuration will now have to be implemented differently, to see the 'text' keys).

    Thanks for that info. I'm currently setting up '3' Odroid-C2 units, and ran into the 'duplicate button response' (which turns out to have been a 'triple' since I also plugged in a USB-MCE-IR-Receiver).

    Didn't know there was such 'flexibility' in the configuration, and am glad I was able to disable the 'TV off' functionality, as it grew tiresome to have to keep turning it back on...

    And, was concerned that the 'built-in' IR receiver on the Odroid card wouldn't be 'visible' and hence 'unresponsive' to button pushes, but with CEC forwarding the commands via the TV's IR receiver, this makes the placement of the Odroid inconsequential (except for WiFi receptivity issues).

    ALSO, currently using LE 9.0.2 which some have stated the 9.x versions broke at least some of the CEC functions, which 'I' haven't encountered... (yet ?)

    I believe it's going to be limited to wifi USB dongles that match what's in the kernel driver directories:

    At the linked gh page
    Note: Kernel 4.2 has been released which contains a driver for mt7601u, this repository is now deprecated.

    so I guess it should just work

    Apparently 'not' for the Odroid version: