See my other comment on this thread. Basically I have a shared mount point for the two containers and TubeSync writes video metadata to NFO files.
loganb
TubeSync has an option to write metadata to NFO files. Then you just tell Jellyfin to not run any scrapper and just use said NFO files. It's not perfect but it gets you a title and description for the video.
I use TubeSync to do the downloading and then have Jellyfin as a frontend player. Seems to work pretty good for me and was pretty quick to stand up in docker.
Or when you hear "I just have a quick question" and you instantly know its time to get comfy as its gonna be a long ass phone call.
I've been using fedora on a small intel 6th gen or newer mini pc. I then cook up some custom launch scripts that cause JMP to run at login. I use cockpit and a CMK agent for remote monitoring and management.
I got sick of the lack certificate management on Android TV and how much you need to do to make it reasonably private.
If you are on the latest mesa drivers (hence fedora over a more LTS release), and you install Jellfin Media Player via flatpak, everything should just work with hardware decoding.
But you know those repairs will outlive the rest of the pants.
You can self-host the kiwix server in docker and grab .zim files for whatever wiki you want to host. Wikipedia is one of those files.
I can also vouch that Android Auto works in a work profile.
That's most likely the problem. In my experience, nearly all tor exit nodes are flagged as such and captchas are nearly impossible to "pass" when using such an exit node. I would try using a free VPN to test. Try protonvpn without an account and see if you can get past the captcha.
Are you using a VPN? It might be that changing your exit IP might help. I've noticed captchas get harder to pass if your on a VPN that has a lot of traffic trying to pass captchas. Probably DDoS protection.
It's OK I was literally OMW to be that guy.
This message was approved by your local democracy officer.