Well, most people struggle to grasp the concept of digital private property; they understand what it means to own a pair of shoes, but they don't get what it means in digital because they see a screen and, to them, what they're looking at is just there, on the screen in their hand, not coming from someone else's servers.

This simple phenomenon makes it easy to set up certain business models and, besides, the shift towards cloud+mobile, essentially modern mainframes + dumb terminals, serves perfectly to deny the average person "the power of the desktop" and, with it, personal ownership.

You need to define what being a sysadmin means. Imagine a distro like Umbrel, Start9, Univention, etc essentially a distro that serves a web app with a series of "pre-configured" services by default. Imagine it's not based on obsolete distros like Debian, but on declarative distros like NixOS, where devs don't fork but simply maintain the config with every update, giving users something that practically never breaks between releases.

Being a sysadmin in this case means knowing how to deploy a distro on a home server, basically a generic desktop computer recycled for this purpose, following a fairly trivial wizard, and using the services. If something goes wrong, there's the community to ask for a hand maybe with Nostr/WoT as comms and social score backbone + economic model behind.

The skills you need cut across every discipline; it's IT knowledge that even a greengrocer needs to have today.

The problem is that today's devs don't understand this; it's not their world because they haven't experienced real FLOSS. They were just born inside some giant company, using its services, without knowing anything else. This is why there's been a push to eliminate or marginalise Ops, and today, with LLMs, they're trying to marginalise devs, just as doctors are being marginalised while glorifying nurses, as "paramedics" etc. In other words, anyone with PERSONAL ability and PERSONAL culture is being crushed, because through these they build their PRIVATE property and aren't just a number, a stereotypical Ford-model worker to be managed on an assembly line, but a capable individual who stands on their own two feet.

You have every right, then, to complain about the state of complication rather than the complexity of modern software, but you also have the need to know just enough to be sovereign in society in your time. Otherwise, well, you can only be a slave, and even if that doesn't sound pleasant described so bluntly, it's necessary to digest it and choose.

Very true, but how many people actually have a home server? Because that’s the prerequisite we should all have, just as we should all have a domain name, IPv6 with a global per host, and a basic IT literacy that spans every discipline, yet is lacking in almost every discipline, often including PhDs in CS and CE.

Nowadays we write everything on computers, except perhaps for a few notes, but most people don't know how to typeset a decent document, they don't know how to generate a graph, do basic maths, or draw in a simple CAD or freehand style. In other words, most are incapable of communicating their thoughts using the tools of their time at the minimum level necessary to be defined as an "intellectual" or a "cultured person".

On the Bitcoin side, there's also a software problem: we need a single, trivial-to-deploy application that does everything, modular if you like, but unified and well-documented. Not Bitcoin Core and LND and Alby Hub and Zeus... we need to be able to say, "Want your own bank? Right, go install ... | cargo build ... | uv install ... and you've already got what you need; nix develop if you want, or check your distro's package manager, you don't need anything else. If you want, put NGINX in front of it, but it's not necessary, you're up and running with just this app. Go to [::1]:12345 and you've got the web UI with your on-chain addresses, you create a Lightning channel on the fly, manage its liquidity, and you've also got the documentation, links to mobile Lightning wallets, logs, etc. You also find a textual dump of the config to makes anything easy reproducible. You don't need anything other than to study this single, cross-platform, well-documented project". That’s how you spread it to the masses; as it stands today, it's beyond the level of effort even most nerds are willing to put in, so people end up on CExs and living on someone else's services.

There's Umbrel, Start9, etc which is good, but it's still not enough and it's lagging behind because the people behind these distros are stuck in a bygone era without realising it; they don't offer a solid enough foundation for the end user.

A chopping board, cut at a sharp angle to get a useful shape and a good final surface, starting with a height of about 7cm using a chainsaw. Then, place it on a workbench and secure it by screwing it into the side. Set up two flat guides at the necessary height, and place a long enough board across them with a router mounted on top; use this to plane the rock-hard wood to a "mirror" finish. Once one side is done, flip it over and do the second. Use a chisel or the router itself to create a drip groove that ends in a drain at one end of the "ellipse".

Cut as shown in the photo, there isn't much else they can do. If you want, you can use wood carving discs (grinder discs, made like wood rasps for carving, dangerous to use, so be careful) to shape other things, for example, a stylised fruit bowl to be decorated later with a pyrography pen or chisel, but this requires much more skill than the chopping board and the practical daily use isn't great.

If there's still a long piece left and you like polenta, you can make a classic Italian "polenta stick", literally a stick about 80-90cm long, with a diameter that's comfortable to hold and a wider, flattened end. But carving it from a log requires a certain level of skill; you can even do it without a lathe by starting the cut with a chainsaw and then setting up a "DIY lathe" using a drill and a router. It's a long, frustrating job for a result that doesn't make much sense anymore now that we have self-stirring pots for polenta, but it can be done, just like you can make large kitchen spoons.

Nowadays, most people need a programmer, but big companies, just as they get rid of doctors to make way for less competent and therefore more docile paramedics, are looking for code monkeys. In the same way, sysadmins have largely been phased out in an attempt to have people who are easier to boss around. The result is that we no longer have anything new, and the number of constant problems is skyrocketing.

Being technically decentralised doesn't make it practically so. The Web is technically a hypertext, running on a partially interconnected mesh network, yet nowadays the bulk of traffic flows between a handful of giant hubs, to the point where "marginal" social networks stay that way simply due to a lack of critical mass, and not having an account on some giant's servers is a communication problem for many. We have, and consider normal, major communication systems that only talk to themselves. XMPP is decentralised by design, yet when it was popular, Google was the main player and its changes were adopted even if they were unsuitable for most, simply because they were needed to interoperate with it. When Google abandoned it, the users vanished and XMPP essentially died, having become irrelevant.

To put it another way, yes, Nostr is decentralised by design, but this peculiar design makes it practically centralised, and it is, or rather will be in the future if it succeeds, a problem. Just see Primal as an example.

I casually discover it :)

Not in a rush, I don't know how to contribute but I curiously observe hoping for a good progress because Nostr IMO have much potential, but lack some aspects to really succeed and loosing a potential success for just some issues well... It's a waste...

The problem I see is that messages aren't "evenly spread across the network" but are concentrated on a few relays, and it even happens that some replies only reach certain relays. The result is centralisation on the most popular relays.

The solution I see is the classic DHT, basically algorithmically spreading content across every node in the network, which will have storage and bandwidth quotas in its parameters. This leaves the administrator free to reject certain content or keep some in full, but fundamentally every message is split in chunks and some chunks are automatically spread across nodes. Historical examples include Usenet on one hand, as a decentralised paradigm, and eMule/KAD or ZeroNet for the distributed one.

The solution I see for further improvement is to not have "Nostr client only", but client+relay in a single package, with potential support for Tor or ZeroNet or something similar for those behind a NAT. Those with an exposed host can provide a way to punch through holes in NAT for those who don't, but basically every client is also a relay and blossom server.

I hope that's clearer now, and I don't see any bullshit or rudeness in what I'm writing. Meanwhile, as a fairly new user curious about Nostr and from the old-school *nix background, I see a tense community that's not very interested in the rest of the world, which isn't a good thing for achieving success.

I'm desktop (nearly) only... When on the go I'll try but it's a rare use. Most of the time (being full-remote worker) I'm on my desktop and since on NixOS I have only Gossip (broken since months) I tend to use web clients.

But I'd really dream of a wisp web built-in the MOAR :)

Idem, significa che anche su Nostr non siamo così marginali :)

Si, io in genere a chi scrive in inglese rispondo in inglese semplicemente perché non so qualche sia la madrelingua della controparte e ad oggi avendo vinto loro l'ultima guerra mondiale l'inglese è la lingua franca come prima lo era il francese...