I haven’t finished setting it up yet, so I can’t offer an opinion yet, but the way I’ve gone is a doorbell with electric lock control- mine does 2 locks, and you can set the 2nd lock to be a doorbell instead. The ‘doorbell’ output is a contact closure, which I just wired to a standard doorbell chime. Network down means I still have a doorbell.

That’s a big bet on the RAM modules. If LPCAMM2 doesn’t take off, one of the most likely upgrade paths is going to be a lot more expensive.

This might be a time when an ignorant point of view helps, like mine. I don’t know anything about Wayland- Is it like Zululand but where the Way tribe lives? Anyway, my naive assumption is OP was wondering if you can run it on a server that does not have a display plugged in.

I think avoiding containers is the way I’m going to go on my next attempt. I’ll still have to put it in an lxc or a VM on my proxmox, but all in one will hopefully reduce some problems. The sonarr/radarr split was what I was referring to with the above or below an hour comment.

I hate how fragmented they are. I’ve given up on various guides out there for ‘setting up the arr stack’ because of getting bogged down in since miniature detail that, IMHO, shouldn’t even be a thing. I get that hosting seperate services has advantages. But the disadvantage of giving up on the whole thing because you have to sort out networking and file permission issues between the service that downloads video files over an hour long and the service that downloads video files under an hour outweighs those advantages.

You could also just shoo them outside. I’d prefer to not have fly guts all over my hands, and I don’t really have the need to kill every fly I see.

I find another part of this story pretty wild- The US ‘rebuild’ was done over dry land, then a canal was dug underneath.

In my world, very much not software development, the two roles would essentially be the same thing. What’s the difference for you?