__hetz

Closest thing I have to a formal education in IT is taking keyboarding, QB, then VB in high school in the early 00s. The only thing I came away with from that was the basic concepts; variables, integers, strings, loops, functions, recursion, etc. I remember nearly nothing about VB. Self taught some Python after that, went to Linux and started learning Bash, CLI tooling, editing confs, working on small portions of languages I didn’t know like Haskell (Xmonad) and eLisp (emacs), running containers… Long enough ago I remember manually setting up Apache and MySQL a few times, editing whatever init system I was using to ensure they started, etc.

Time and cost are big determining factors. Sometimes I’d like to know more but never enough to become indebted to educational bills for something I’m not making a career out of. Same reason I can install, remove, rekey and pick various locks, and have installed some access control stuff in the past under an employer, but I’m not shelling out to become licensed Locksmith or Security Systems Technician myself. Tools are comparatively inexpensive, and I have many from past employment, and YouTube is still free and archiveable… add in a bit of “dorking” to find manuals and whatnot and self-teaching just makes more sense for me. Same reasons I’m entering my shade tree mechanic arc these days.

I can say that TubeArchivist is essentially single-user. Multiple accounts on a single TA instance will share the same underlying library with a ./library_dir/channel_id/video_id.ext directory structure. You could run an instance per user for individual TA libraries but, last I looked, the Jellyfin plugin doesn’t support multiple TA instances so it still wouldn’t be able to get the metadata from any instance(s) other than the one it’s configured to. It was a requested feature, on the plugin’s GitHub, but I’m not sure if it’s actively being implemented. The only workaround I can think of for now would be using a shared TA library and then manually tagging the videos in JF and using parental controls to hide them. Billy won’t see vids tagged #SusieYT, Susie has #BillyYT added to her blacklist, etc.

This has been my solution. Subscribe to a playlist or an entire channel, set up the schedule, and you’re done. If you’re a supporter of the project I believe it has real-time monitoring for new uploads.

I will note that, somewhat recently, YouTube changed the URL format for playlists and this has broken subscribing/downloading playlists that use the new format. I haven’t checked today but there’s a patch that was still waiting to be merged. Unless there’s a new release incorporating it, you’ll have to apply it yourself or feed the URLs in manually for now.

Unfortunately I can’t help in that regard. I keep everything local/unexposed so my solution for them was just running Jellyfin at their place. I was already rsyncing some stuff to a NAS I set up for them (and vice versa), as off-site backup. Since the files were already there it made the most sense to just give them their own instance.

I don’t have the link(s) on hand but there’s a Tizen build of Jellyfin for Samsung TVs. It runs rather slow on my old tube so I wouldn’t recommend it outside of a last resort. It’s actually smoother for me to just open the app on the TV and then remote control it from a browser/app on another device (my Steam Deck is my homelab universal remote). But you can use the Tizen dev tools or a simpler docker container to push it to the TV.

For my folks I got a cheap Walmart brand Android box (Onn 4k Plus). I installed Jellyfin from the app store then black hole’d the thing because I’m wary of cheap Android apps and their history of supply chain attacks. It’s much more responsive and also leaves me with the option of installing additional stuff like Smart Tubes, Retro Arch and whatnot.