Posts by chewitt

The hardware problem on my board was dry solder joints on the UART pins causing noise. I resoldered the pins and I have no boot issues with upstream u-boot. Someone else forced a UART line to positive or something? (forget what, but it's commented somewhere) and that solved their problem. WeTek did a reasonable job of design and an above average (for 2015/16 standards) job of supporting devices; but ultimately they're just another Amlogic board manufactured by a Chinese ODM manufacturer (Videostrong is on the schematics). We're now 7-8 years on from the manufacturing date and some boards are showing some age related issues, which is also not unusual.

In the absence of a debug log ..

NB: LSPCON is unlikely in an nVidia ION box .. it is not a NUC. It is much older than NUCs and their LSPCON silliness.

There's a warning message about unsupported ABI versions between Xorg and the nVidia driver, and EGL fails to initialise. Those are prob. related. Not sure what to suggest other than downgrading to old releases until you find something that works.

On your end you don't really need the compatibility so use EXT4 on the drive and perhaps Transmit to shift files. SFTP is slower than NFS/SMB but the client will queue files and transfer things reliably. On the parental end .. exFAT is really the only option for plug/play with macOS unless you add third-party drivers to the OS.

The best investment(s) I made for reliable playback were some Synology NAS devices in the network. Simple to copy media to after ripping etc. and the LE device becomes a simple/dumb playback client accessing content from an SMB share and with nothing dangling/attached to the RPi(4). On the parental end I have a similar arrangement where the NAS also acts as the media device, but also a local TimeMachine destination so I don't rely upon elderly minds remembering to make backup copies on a USB device (they are good at hoarding data but backup habits evade them). Once in a while we copy new data from removable drives to the parental NAS where there's disk redundancy and also remote sync of specific dirs back to the NAS at mine to effect some off-site backup of important files.

That's a little beyond my knowledge, but FWIW i've never seen those timeouts on other G12B devices using the kernel (and u-boot) sources that LE has been using for ages. I'd suggest you post the Q to the linux-amlogic kernel mailing list.

video=HDMI-A-1:1920x1080M@60

^ You can append that to boot params in cmdline.txt and it will force initial resolution to 1920x1080 .. or perhaps 1280x720M@60 if it's an older set without FullHD resolutions.

I'd use EXT4 for Linux filesystems and simply use the network to transfer files; thus avoiding the dependency on exFAT or NTFS for Windows compatibility. There's nothing wrong with Samba/SMB for the server. We don't enable/provide NFS. You can always use an SFTP (SCP) client to transfer files over SSH.

LE is not a company. We have no products. There is no Engineering, no Project Management, no Product Management, no Marketing, .. and no Sales. We do have a support forum. You are welcome to use it (as-is) .. or not. Your choice.

It was related to reserved memory regions.

See: https://patchwork.kernel.org/project/linux-…[email protected]/

Interesting. Thanks for the tip..

LE11 uses Python 3.11.12. What are the add-ons?

LibreELEC is $free open-source software not a PRODUCT from a company, and no user from any timezone has a right to instant or rapid support from the unpaid volunteers who give up their free time to work on the distro and staff this forum.

The legacy images do not contain vendor u-boot, they are designed to boot from SD card (with vendor u-boot on eMMC to initiate boot) or to replace the Android image on eMMC (again working with the original vendor u-boot on eMMC). If no u-boot on eMMC .. no boot.

The vendor u-boot code "works" but is horrible. Upstream u-boot also "works" and is cleaner to work with for packaging. Your Hub has a hardware problem. If you fix that (in hardware) the upstream u-boot works fine.

Read the sections on scrapers and how to use directories/filenames to get things to scrape accurately. They apply to offline media and the local filesystem scraper as much as online scrapers. If you store media in the right structure and filenames it all "just works" .. and if you stick with some homebrew structure that doesn't follow the rules the results are probably doesn't work.

Old LE version will be using the NTFS-3G userspace filesystem driver and current LE version will be using a new in-kernel NTFS filesystem driver. That will explain differences in how things are mounted and perhaps different error messages. It doesn't explain how things got corrupted in the first place; but with USB media PSU's and or the simple fact of NTFS + removable media =

Quote from quosek

dial tcp: lookup registry-1.docker.io on 192.168.1.1:53: no such host.

^ that's a DNS resolution failure. You might need to investigate Docker comms.

Quote from ejwjohn

So i am struggling to understand why someone has gone to lengths to make this rule?? it's not in any standard that I am aware of, and it's easy to do on other devices so slightly puzzled.

You are welcome to complain to the upstream maintainers of ConnMan who authored/coded the connection manager whose features and capabilities we expose via the dbus agent in the Settings add-on. TL/DR; we didn't code anything or go to any lengths.

Unless you're planning to submit upstream patches to change worklows; the workaround will be to create a systemd service that runs before kodi.target and after network-online.target, that applies your preferred DNS server config to the active ethernet service using a connmanctl command. You'll find some useful prior art for something similar in other forum posts from users trying to change service (routing) order for VPN connections (also done with connmanctl commands).

NB: initial post approval applies to everyone and while your posts might not be visible to all users, they will be visible to mods/staff who are most likely to be the people replying. It is done solely for anti-spam reasons only, and is effective, and the algorithm that auto-manages it normally lifts restrictions quickly. If you're deeply offended by it or ragingly impatient, you're probably in the wrong forum.