I would be worried about the battery, as many others pointed out too. I had a somewhat similar project with old MacBooks. I thought taking the battery off would make its fan to spin at 100%, yet turned out it does not. Right now one of my servers is MacBook Pro from 2011, and it kicks. It has no battery installed, and it looks like a pretty good server to me. It starts automatically when there’s a power loss, which is ideal for a server. Can totally recommend.
Not in perpetuity though
Ahaha, I came here to write exactly what the other guys wrote, but they outdid me massively. Bravo, folks, made my day! I couldn’t say it better myself.
Unfortunately, where I am, prices aren’t that generous, but I have an unusual interest in very old hardware, so most of my servers can be bought for like $20 or less. Most of them are junk hardware nobody needs. They’re very useful for many cases, especially if we talk self-hosting / personal servers. I bet you can easily serve a small community with a $20 computer. Recently, I started being very picky about power consumption (we have massive blackouts here in Ukraine, due to the Nazi terrorist state, Russia, shelling our civilian infrastructure; but not just that: power efficient hardware is better for the planet), but one thing that isn’t obvious: you can keep the hardware off when unused. It’s trivial to turn it on, or wake up, depending on the use case.
So, if anyone new reading this, I urge you to try this with any spare computer you can find. You may be surprised how good it may work with Linux (server or even desktop).
I may need to do a proper write up on that one, as it’s quite a long story actually. The main thing is, though: it’s hard to believe how actually good modern hardware is. You can do a lot with a Raspberry Pi, not to say Mac mini. Over the decades, I accumulated some old hardware which serves till it would break. And then I’d replace it with similar hardware, I guess. E.g. I have a media server, which is a simple lighttpd web server running on an obsolete Intel Atom computer. It’s slow, but it does its job at serving files. I have old MacBook Pro hosting my Immich library.
All that is rather a self-hosted territory, but I’m pretty confident I can host a production grade stuff off my own hardware. I have been using the cheapest options at DO/Hetzner, so I definitely can host off a maxed out Mac mini or similar.
All that lives at home, or in the office (I am allowed to have some hardware there). I think the main restriction is the gigabit internet… but hey! That’s plenty too!
If I’d ever come to some serious level of hosting, yes, I think I’d co-locate at some data centre. I see very little need in cloud services, as all I need can be done with the bare metal. And it’s so much fun! Feels like magic, seeing an old laptop you’d throw out otherwise, serving as what would be a full-blown server just two decades ago.
Hetzner was good for me, but I went all-in on bare metal, with no looking back.
My bet is that the tech bros hang in the same bar, and they’re just doing that for fun, and they don’t really care about the plebes not getting it.
Glad I left Digital Ocean like almost a decade ago, yet it still feels I’ve been with them for too long.
deleted by creator
Why would it be? It’s like Windows is OK, and Linux is not. But actually, you can do the very same thing, interface wise.
There are plenty of printers that just work, it’s been a long time I’ve seen a printer that doesn’t work. Do you know that sudo can be passwordless? It’s a completely made up problem, probably with the lack of any understanding of the system.
What’s specifically wrong with Linux for a grandmother? Or does she has some weird edge case with some very important niche Windows software? I’d call modern Gnome being even better than modern macOS in terms of simplicity. Telling Windows is better than macOS and Linux for a grandmother is weird. Unless the grandmother is a heavy Windows user who knows all the hotkeys and menus. If she barely uses browser with not much else, I don’t even know.
So, Linux didn’t help your Stockholm’s syndrome yet.
North American English, innit?
If you don’t setup or activate exit node, no traffic is routed through any of your nodes. All you have is the access to the nodes. Which is what you need. I tested exit nodes only recently, they’re very easy to setup as well, but I found no practical need for my use case.
I think installing and logging in should be trivial remotely. Like hey mum, install this app, and log in (trivial with Google or Apple accounts). The rest is on you. Just test the waters yourself first, you’ll get the idea, it’s pretty straightforward. Even if it’s not what you’re looking for, you’d have more information and skills to move to the next thing.
Yes, it’s exactly like this.
Hey, but in general Fedora is very good and I can recommend it. This annoying update behaviour bears some meaning for some reason, so I can tolerate that.
It’s not that the system forces you to update. It’s you who is in control. Most times updates take minutes, with version upgrades taking like half an hour (twice a year).
I’d suggest you to investigate either Tailscale or similar solutions. I’m using Tailscale, and it’s really easy to set up. It can automatically connect to the VPN when you access their resource, and the internet works as well. So technically, they can be connected all the time. That’s much safer than the alternative of just opening a port, and dealing with things like CGNAT.
The alternatives to Tailscale I know about are Headscale (which you need self-hosting), Netbird, WireGuard. At least, but there are more.
And search for tunnels as well. You could utilise Cloudflare Tunnel, but I wouldn’t go that way.
I’d suggest testing waters with Tailscale as it’s the easiest, and tweak from there. They have a YouTube channel which helps at starting, I found it just recently. (I use them for a year or two now.)
But will it install those updates during the reboot? Or will it update just to apply the updates?
As the Gnome version (non-mutable, meaning not Silverblue) does not update a thing, even a browser. It updates during the reboot, by rebooting into some special state, akin to macOS (and perhaps Windows, haven’t been using one for decades) then it reboots again.
Since the latest update of Fedora (50? I lost the count at this point) it does the reboot even when you ask to shutdown the computer after updates.
I haven’t checked whether dnf update updates things right away (like browser), mostly because that’s a shared family computer, which I don’t touch most of the time. And I try to use it as I use macOS (as a normie would use), just to check out whether it’s a viable system yet. (It mostly is, but there are weird bugs that’s easy to resolve only when you’re knowledgeable of Linux.)
But it does not do the extra reboot and install updates thing. That’s why I’m puzzled why is it there in the first place?
I wrote you a long comment from my mobile, and it looks like it was lost.
Can you elaborate on that topic? I use ydotool for this, I’m on Sway.