Maybe this trend will reverse eventually but I don't really see the signs yet. The Cryptocurrency crowd keeps shouting "decentralization" but they still fail to create applications that can compete with the centralized alternatives in terms of usability, performance and cost. There have been many attempts at making decentralized social networks but they failed to gain mainstream adoption. IPFS works pretty well but again, hardly anybody uses it.

I'm all for decentralization but there's no denying that there seems to be a path of least resistance towards centralized solutions. They're easier to develop, easier to maintain, easier to upgrade and often easier to use.

So for me decentralization is the objective, but unfortunately it's not "clearly the end game".

Even most of the very technically competent people I know are gradually moving toward central services. I'm part of a co-op of people with collocated servers. We started in 2000. We haven't had a new member in years, and we are gradually losing them. I'm at the point where I should replace my server, and I'm having a hard time coming up with reasons to justify the large capital expense and significant time cost versus moving it to somebody's cloud. And that's not even considering the benefits of moving to hosted services. Not worrying about spam, email deliverability, security patches, et cetera, ad nauseam.

I think part of the problem is that the "decentralize!" crowd is willing to put up with a lot of practical inconveniences as long as something conforms to their ideological desires. Their ideology may be perfectly correct, but until it has practical consequences, most consumers won't shift. So they're going to need to come up with competitive services that are better than the existing ones. Better not just to them, but to regular users.

Regarding your note on the path of least resistance leading to centralized solutions---Glowing Bear/WeeChat is definitely more work to set up than just signing up for Slack. You need a machine that runs WeeChat and get a TLS certificate so that the browser will let you connect securely. That definitely limits it to a somewhat nerdy demographic, even among the HN readers ;)

(Personally, I’m ofc biased, as I’m the dev of https://quasseldroid.info/).

But I think for these solutions to gain more mainstream appeal, we’d need to make the setup much simpler, and work on ideally making it a single-click solution for an organization to set this up for their members. And maybe even provide hosted services (similar to IRCCloud) for the many users that would rather pay than run their own servers.

Now I suspect we'll start seeing clones of facebook / slack etc but with a decentralized backend while offering all the features users care about seamlessly. This might take a while but it'll eventually come.

Even on the technical side, most tech products have convenience as a selling point to being more productive/effective/agile, etc.

Regarding torrents, many game clients will use torrents for their downloads, the user simply doesn't see and deal with the torrenting.

Don't confuse self-hosted with decentralised, not the same :)

Again, this is looking at things backwards IMO. You seem to imply that there's a slow momentum from a centralized web to a decentralized one when in fact there's been a rather fast momentum in exactly the opposite direction over the past decades.

To me what you're saying sounds like "horses are about to become a very common way of moving goods". Maybe you're right but merely looking at the trend it's clearly not going that way at all.

>The backend is basically figured out at this point and we need to focus a bit more on UX.

You'll have to tell me more specifically what you have in mind there because that sounds very optimistic to me. We've had decentralized "backends" for as long as we've had the internet. The web is mostly decentralized by design. Even DNS is distributed across plenty of authorities for the various TLDs (even if each of them is effectively centralized and not anybody can become an authority).

Email is decentralized. BitTorrent is decentralized. IRC is decentralized. We're collectively moving away from these technologies, not towards them. I'm personally still a heavy user of all three of these things but it definitely feels niche now (email obviously isn't but self-hosted email is).

>Regarding torrents, many game clients will use torrents for their downloads, the user simply doesn't see and deal with the torrenting.

Which is pretty much irrelevant in this conversation then. It's about the technology people use to share content with each other, not about how Blizzard chooses to update your WoW client. It's a locked-down, vendor-approved way of distributing software from a centralized authority.

>Don't confuse self-hosted with decentralised, not the same :)

It's not the same but it's related. In general if something is truly decentralized then it becomes self-hostable otherwise it's more distributed than decentralized. Anybody can host their Bitcoin node, their Bittorent peer or their email server. I can't host a Facebook node.

The distributed system does not stop to work then, but the user might risk punishment for using it, which might be even worse than not being able to use it.

The fact that you can't decentralize legislation and physical governance is not insignificant. You can't block the influence of preexisting powerful actors. Those factors do have the power to destroy the decentralization movement, and most likely will.

Maybe those "powerful actors" could prevent decentralization from becoming mainstream, but they can't kill it. Consider marijuana, for example. Use has been demonized for decades by the US and its allies. But that didn't stop an appreciable fraction of the US population from using (or at least, trying). And now it's becoming legal in more and more states.

For Internet decentralization, the driving factors will almost certainly be porn, gambling and prostitution. To the extent that they're driven off the clearnet, demand for them will fuel growth of alternatives. Freedom of expression is essential, of course, but it will be just a side benefit.

The more decentralization is suppressed by "powerful actors", the more it will be dominated by other "powerful actors". That is, by organized crime.

It will never happen, at least not on a scale that will affect significant web traffic.

There is another model, not decentralised but a practical middle-ground: the WordPress model.

Consider the following: Wordpress is an example of a profitable open source app that can easily be installed on countless shared hosting platforms or on a VPS. It's easy to switch hosting providers when you want to (and to take your data with you). It's popularity means that one-click installs are widespread.

Unfortunately, there is no common standard for software installation on the server side, and this lack of an easy installation process for everyone else severely limits self-hosting websites and apps.

Many developers think deploying a server-side web app is a non-issue, or they erroneously think that installing Cloudron/providing Docker instances/typing command line instructions are all "easy". Have you seen the server deployment instructions for "web friendly" languages like Ruby and Python? It's ludicrously complicated. And still developers seen nothing wrong in such install procedures. It's so frustrating.

I wish there was some momentum or traction in making server-side web app installation as universally simple as a one-click Wordpress install. It would also unlock countless opportunities for developers to reach more users or customers. But maybe some developers secretly prefer the complexity? It certainly makes selling a SaaS solution much more attractive over the stupidly complicated self-hosting option.

Only techies care about decentralization. Most people would rather follow a Twitter feed rather than an RSS feed. Most would prefer a mega forum like Reddit rather than multiple, standalone forums with separate accounts. There are also network effects that give centralized platforms more of a competitive advantage.

I keep hearing people talk about the need for decentralized social media, but nobody knows how to make it an attractive, viable option for the masses...especially when such a solution wouldn't be as profitable (or as frictionless) as Facebook, Twitter, etc.

This is the future of a centralized web with products, services and companies that people salaries to develop products and services.

The decentralized internet utopia is dead. Stick a fork in it. It is a fringe populated by the modern equivalent of the grey beards of 1990s.

What this all really does in the end is create a technical caste system and obfuscation of ownership. People fortunate enough to have access and be up to speed on the latest technology (or hire people who are) will reap the rewards of decentralized systems which still belong to authoritarian actors, yet it will be extremely difficult to prove that ownership--especially to laymen. It will always have a nice hazy deniability, and it will be almost impossible to hold anyone accountable for their actions, or prevent or even identify exploitation.

This is absolutely the future of the web. Perpetual and stealthy non-neutral manipulation by the technically advanced and financially powerful is here to stay. I think that as engineers our tunnel vision and intellectual hubris have given us a false sense of security as we developed this hideous system, because we thought it was some kind of purely digital realm where we have real control and real comprehension of what we are doing. But nothing is purely digital. Everything is built on top of the real, analogue world where strong actors have already divvied up and taken ownership of everything.