NaibofTabr

Nice save, and a fantastic PSA.

Also I’m a big fan of sleeping on a problem as a path to a solution. I’m not sure how exactly that skill develops, but it’s definitely something that I’ve done a few times over the years.

well… fuck.

OK, so you want to install Linux on an old laptop after wiping the hard drive.

Sure, you would have root control for that OS. Make sure that there is no BIOS/UEFI password set before you start, or that you have the password.

Do not use the root account for regular use, especially if you will be connecting this laptop to the Internet. Log in as a normal user account and escalate your privilege as needed.

I also highly recommend gparted for editing the hard drive partitions before you install the OS. Don’t install goarted, just run it live from a USB drive.

Do you know which Linux distribution you want to install?

er, what exactly do you mean by “from the scratch”? and what exactly constitutes “my own system”? are you planning on designing your own logic circuits, manufacturing your own chips, and then writing your own firmware to run them?

Perfect explanation.

Thank you, I try. It’s always tricky to keep network infrastructure explanations concise and readable - the Internet is such a complicated mess.

People like paying for convenience.

Well, I would simplify that to people like convenience. Infrastructure of any type is basically someone else solving convenience problems for you. People don’t really like paying, but they will if it’s the most convenient option.

Syncthing is doing this for you for free, I assume mostly because the developers wanted the infrastructure to work that way and didn’t want it to be dependent on DNS, and decided to make it available to users at large. It’s very convenient, but it also obscures a lot of the technical side of network services which can make learning harder.

This kind of thing shows why tech giants are giants and why selfhosted is a niche.

There’s also always the “why reinvent the wheel?” question, and consider that the guy who is selling wheels works on making wheels as a full-time occupation and has been doing so long enough to build a business on it, whereas you are a hobbyist. There are things that guy knows about wheelmaking that would take you ten years to learn, and he also has a properly equipped workshop for it - you have some YouTube videos, your garage and a handful of tools from Harbor Freight.

Sometimes there is good reason to do so (e.g. privacy from cloud service data gathering) but this is a real balancing act between cost (time and money, both up-front and long-term), risk (privacy exposure, data loss, failure tolerance), and convenience. If you’re going to do something yourself, you should have a specific answer to the question, and probably do a little cost-benefit checking.

But if I’m reading the materials correctly, I’ll need to set up a domain and pay some upfront costs to make my library accessible outside my home.

Why is that?

So when your mobile device is on the public internet it can’t reach directly into your private home network. The IP addresses of the servers on your private network are not routable outside of it, so your mobile device can’t talk to them directly. From the perspective of the public internet, the only piece of your private network that is visible is your ISP gateway device.

When you try to reach your Syncthing service from the public internet, none of the routers know where your private Syncthing instance is or how to reach it. To solve this, the Syncthing developers provide discovery servers on the public internet which contain the directions for the Syncthing app on your device to find your Syncthing service on your private network (assuming you have registered your Syncthing server with the discovery service).

This is a whole level of network infrastructure that is just being done for you to make using Syncthing more convenient. It saves you from having to deal with the details of network routing across network boundaries.

Funkwhale does not provide an equivalent service. To reach your Funkwhale service on your private network from the public internet you have to solve the cross-boundary routing problem for yourself. The most reliable way to do this is to use the DNS infrastructure that already exists on the public internet, which means getting a domain name and linking it to your ISP gateway address.

If your ISP gateway had a static address you could skip this and configure whatever app accesses your Funkwhale service to always point to your ISP gateway address, but residential IP addresses are typically dynamic, so you can’t rely on it being the same long-term. Setting up DynamicDNS solves this problem by updating a DNS record any time your ISP gateway address changes.

There are several DynDNS providers listed at the bottom of that last article, some of which provide domain names. Some of them are free services (like afraid.org) but those typically have some strings attached (afraid.org requires you to log in regularly to confirm that your address is still active, otherwise it will be disabled).

Oh, yes that does change the meaning.

Hey man, the tech literate people were saying this:

It’s the VCs and the marketing people who were pushing cloud services as the next big thing.

https://dayssince.itwasdns.net/

They should be powered on if you want to retain data on them long-term. The controller should automatically check physical integrity and disable bad sections as needed.

I’m not sure if just connecting them to power would be enough for the controller to run error correction, or if they need to be connected to a computer. That might be model specific.

What server OS are you using? Are you already using some SSDs for cache drives?

Any backup is better than no backup, but SSDs are really not a good choice for long-term cold storage. You’ll probably get tired of manually plugging them in to check integrity and update the backups pretty fast.

That’s ok we’ll just refactor it with AI.

Well shit, yeah, that “MUST be accepted and parsed” is pretty explicit. That sucks. What is even the point of revising standards? How the fuck do we ever get rid of some of these bad ideas?

#18 seems really bad, like no-one-has-ever-sanity-checked-this bad.

Yeah I feel like the correct answer for anything obsoleted by a more recent RFC should be “Invalid”.

Installing an OS will always be a hurdle. Most people don’t want to spend that much time thinking about how their computer works, they just want to turn it on and have it work. For more people to use Linux, it will have to be preinstalled.

After that, it needs to be stable. If the audio stops working, most people don’t think “maybe I need to roll back my driver” or “maybe ALSA has muted my output channel for some reason”, they just think “my computer is broken”. These kind of problems have to go away, or at least be reduced to <1% of users.

Also, very few people are going to have any patience for any kind of difficulty related to “oh you have to add a different repository to your package manager to play common media formats” or w/e (e.g. AUR or Ubuntu Multiverse &etc). Normal people spend exactly 0 time considering what codecs they might need to install to listen to some music, or where they might need to get those codecs from, or whether those codecs are open or proprietary or freeware or whatever.

both AD and GPO are fucking incredible pieces of software.

AD is really the only way to manage an organization with thousands of endpoints and users.

I have some hope that someone in the EU will develop a competing product now that they’re pushing to get away from Microsoft, but it doesn’t exist yet.

We asked 100+ AI models to write code.

The Results: AI-generated Code

no shit son

That Works

OK this part is surprising, probably headline-worthy

But Isn’t Safe

Surprising literally no one with any sense.

Realistically no organization has so many endpoints that they need IPv6 on their internal networks. There’s no reason to deal with more complicated addressing schemes except on the public Internet. Only the border devices should be using IPv6.

Hopefully if an organization has remote endpoints which are connecting to the internal network over the Internet, they are doing that through a VPN and can still just be assigned IPv4 addresses on dedicated VLANs when they connect.

“We have investigated ourselves and found no problem.”