There are some experimental models made specifically for use with Home Assistant, for example home-llm.

Even though they are tiny 1-3B I’ve found them to work much better than even 14B general purpose models. Obviously they suck for general purpose questions just by their size alone.

That being said they’re still LLMs. I like to keep the “prefer handling commands locally” option turned on and only use the LLM as a fallback.

I’ve had this exact same gripe and can thankfully report that running EarlyOOM has fixed this for me.

Why not set up backups for the Proxmox VM and be done with it?

Also makes it easy to add offsite backups via the Proxmox Backup Server in the future.

This person had the same issue and they’ve just logged out and in again

Not an expert but these systems are fairly self-contained and robust. A few things that can be checked easily is that the fan spins, the radiator is free of debris and some compressors might have a sight glass for the oil level.

Any other checks regarding performance of the system, leaks and refrigerant level require you to perform a full refrigerant discharge and recharge. That takes special equipment and some time so no one in their right mind would do that for free, unless they can then force/guide you into some kind of upsell situation.

Larger systems might have some kind of oil filter/catch-can that you might be able to check easily but I’m not too sure on that.

After all heat pumps are just plain old A/C units with a reversible cycle.

Out of curiosity I’ve let it rate Low<-Tech Magazine, a website run on an ARM SBC powered exclusively with off-grid solar power, and that only achieves 87% / A.

Link to results

Typst

You can use their online web-editor (similar to OverLeaf for LaTeX) or download the open-source engine and run it locally (there are extensions available for many text editors).

Compared to LaTeX I find it much more comfortable to work with. It comes with sane, modern defaults and doesn’t need any plugins just to generate a (localized) bibliography or include links.

Since Typst is very young compared to LaTeX I’m sure that there are numerous docs / workflows that can’t be reproduced at the moment but if you don’t need some special feature I’d recommend giving it a shot.

Not a monetary one, no.

* (there might exist some business power tariffs that coincidentally benefit from this but nothing you’d use at home)

The development of Piper is being driven by the Home Assistant Project. That probably makes it one of the larger OSS TTS projects. Hope may not be lost yet ;)

I started out with WireGuard. As you said its a little finicky to get the config to work but after that it was great.

As long as it was just my devices this was fine and simple but as soon as you expand this service to family members or friends (including not-so-technical people) it gets too annoying to manually deal with the configs.

And that’s where Tailscale / Headscale comes in to save the day because now your workload as the admin is reduced to pointing their apps to the right server and having them enter their username and password.

Getting the configs to work with my personal devices was already a little finicky but doing that for not-so-technical family members was starting to be a bit too much work for me.

I’m hoping that Headscale will cut that down to pointing their app at the server and having them enter their username and password.

Was running Wireguard and am now in the process of changing over to Tailscale (Headscale).

It uses Wireguard for the actual connections but manages all the wireguard configs for you.

The “add to home screen” button turns into an “install” button when Firefox detects that the website is a progressive web app (PWA). Other browsers do the same.

The difference is that a PWA can define a custom icon and name for the “app” button on your home screen and that it can use some clever caching making many PWAs offline capable (meaning you don’t need an internet connection to open the web page).

I understand the reluctance to press “install” but in the case of PWAs the install size is tiny and fully contained in Firefox and you get the added benefit of faster startup / loading times due to caching.

Nextcloud is just a web service. How he or anyone can access it is not determined by nextcloud but by the routers, firewalls, vpns and potentially reverse proxies that are routing the traffic to nextcloud.

With the proper configuration of all traffic handling services it will not be possible to access anything other than the intended endpoint i.e. nextcloud.

Within nextcloud any user can only access their own files plus anything that is explicitly shared to them.