Is there a way yet to in-place upgrade or is it still only “flash a new SD”?

I use Jellyfin as a backend for my Kodi boxes (I have 3, and JF keeps them in sync). I used to have a YouTube plugin, but YT broke that this year.

Personally, I use Kodi for that. It works very well with minimal keyboard and no mouse (though it can handle both), so much so that I’ve run it for years using only an IR remote.

Really? Such as?

Slackware 1.2, because it came on a CD in the back of a fat paperback manual I got at Barnes and Noble. It was only later that I learned what a distro is.

Currently on Fedora with a Frankenstein desktop of my own concoction.

This is how my partner and I do our notifications. We’ve got “him”, “her”, and “us”, depending on who needs the notification. Whenever either of us gets a new device, I add it to either of our groups and then works.

Skipping forward/back between scenes mostly. It’s either that or the time skip, which works, but is more work and less accurate.

It’s more because they provide an ONVIF interface or an RTSP stream that makes them self-hosting darlings. Them being Chinese white-labels and cheap is mainly a side-bonus.

What are your recommendations if not them?

Wait, you object to their feely-distributable firmware updates? Seriously? Without those, your CPU is vulnerable to exploits and known hacks.

Really? Which ones?

You mean besides Fedora?

From what I’ve seen from running it the last year or so, yes, most Z2M releases add/change a large number of things. I use the Docker container, and I backup my mapped data directory between releases, but I have had no release related issues. Sometimes new items or features appear in Homeassistant, but it has always worked for me.

I agree that it is awesome to use Kodi as a Jellyfin front end. I use some CoreElec ARM boxes that directly support all the codecs I use. My problem is that the Jellyfin plugin only updates content at startup, so I have to restart Kodi before it picks up changes made since the last restart. I’ve asked around and either no one else is having the issue or they didn’t chime in.

As far as I can tell, it has always used the dotNet 6 framework.

Odd, I only have to reboot mine for updates. Other than that it seems fine running on a Linux VM with 2GB RAM, after the initial setup.

And it uses the dotNet runtime 6 so I’m unclear on what roadmap you refer to.

‘dd’ works, but I prefer ‘shred’. It does a DoD multi-pass shred by default, so I usually use ‘shred -vn1z /dev/(drive)’. That gives output, does a one-pass random write followed by one-pass zero of the disk. More than that just wastes time, and this kinda thing takes hours on large spinners. I also use ‘smartmontools’ to run SMART tests against my drives regularly to check their health.

6th gen works, 8th gen and up works better.

As long as you have enough RAM, you won’t get much more speed. 4GB should be enough. A minimal Linux install plus Jellyfin takes less than 16GB on disk, and anything is fast enough.

Fanless Intel runs a little hot for my taste, but it’s your build. I’ve run tiny/mini/micro systems that were virtually silent but still had a CPU fan to help move heat out.

I’m not using nginx, though good to note the web socket. Mine is all local-net. The plugins auto-detect my server at its correct hostname:port at first install.

I am, yes. It came pre-installed. I even have retention days set to unlimited. The startup sync seems to work, but I hate restarting it daily. The live sync is what doesn’t seem to be working for me.

That is so odd. I’m doing exactly that right now. I’ve got Finamp playing music in the background as I write this.