Philosophical arguments to decentralize everything are great, but literally nobody outside of this industry cares. The key to success, ironically, for decentralization, is for everyone to care. The first step is to ship a product people actually want to use that has tangible benefits: right now, every philosophical site like this is dead on arrival because it requires ground swell that doesn't exist without a compelling product.
I would say the key to decentralization success is making decentralized services that users don't even know are decentralized. As long as using a decentralized service requires more effort than centralized services it will only be enthusiasts using the products.
In my opinion, even that is not sufficient. In most cases, decentralization will only succeed if its architecture gives a significant advantage over the alternative centralized one.
Let’s say you make a social network that people don’t need to know is decentralized. It’s easy to use! Cool, well anyone who has tried to launch a centralized social network will tell you that getting market share is very difficult. A good UX is necessary but not sufficient... you need to outcompete the centralized incumbents. If the decentralized design helps (maybe people like censorship-resistance), then you could succeed. If it doesn’t — and perhaps makes development more difficult — then you are much less likely to gain adoption.
Often you can sort of moderate it yourself based on what you choose to see. Anyone can host a racist site on the internet, but people aren't really turned off from the entire internet by it. Similarly, if there were a racist subreddit, right now that feels weird because reddit is actively hosting it and choosing to allow it. But if there were some kind of decentralized reddit and people made a racist sub on that, I don't think people would view it the same way.
As for scammers, imagine something like twitter. If I follow my friends and celebrities I like, scammers will be relegated to the comments section of their tweets (I wouldn't see their actual tweets because I wouldn't follow a scammer). Then you could use some kind of web of trust system to filter the comments - perhaps I would only request comments for people I'm following, or people followed by people I'm following, etc.
You compare a pull to a push there. Sure on a service where you pull the information, you wouldn't pull what you don't want. Internet is a good example of that, your example of twitter-like is also pull.
It fail when it's a push instead. That would be anything that have suggestion, any public conversation, like a forum, or twitter, even private conversation in some case, like an email service, etc...
For sure you won't pull what you don't want, that's by definition, but you moderation is needed for the push system, where you don't control everything that goes to you.
Most service are push too, we want things to be suggested, we don't always know exactly what we want, we don't know even if it exist.
Ah, the fallacy of the killfile and transitive reputation. Sorry, but that doesn’t work and has been shown over and over again to not work for large scale communities.
This is crux of the problem with their take-off. A few weeks ago I set up a personal Matrix homeserver and when my non-technical partner asked me what I was working on I didn't even know where to start. She is not computer-illiterate by any means, but being outside the industry, a concept such as federation is completely foreign.
Even trying to explain federation to my work colleagues who are in the industry is difficult, because it is such a fundamental shift from the normal web infrastructure.
They do, but what I think they might not understand is that there is a protocol behind it all that at least theoretically allows for anyone to run their own server without a third party, which I think is the aspect of e-mail that you're trying to call attention to, no?
Part of the point that I'm trying to make is that e-mail as it stands today - with the majority of messages going through the servers of a single corporation - isn't all that decentralized in practice.
When you mention e-mail to the layperson without additional explanation, what they have in mind is a fairly centralized system.
To be fair, Matrix as it stands today isn't very decentralized either. The dev team reports that their big matrix.org server has something like 40k monthly users. And while there are a ton of tiny little personal homeservers out there, all the action is really happening on matrix.org.
This struggle is real! I, too, still get surprised when trying to explain to my work colleagues - who are in the tech industry also. Even when i can get some modicum of understanding from them, they rarely see the point of it all...they're almost totally brainwashed - i see no better term for it than this.
I think you and the parent post both have captured necessary, but not sufficient, attributes.
The problem is that before any of that matters, assuming your goal is mass adoption beyond narrow niches of the already interested, you need to convince people they should care. This is fundamentally marketing and sales, which are the sorts of efforts many people trying to build these things are actively disdainful of. Moreover, who is the messenger to public? The very same decentralization that makes these sorts of solutions interesting actively works against the sort of marketing and sales activities that would get the non-initiated engaged.
Fragmented, uncoordinated, and probably just word-of-mouth messaging against well coordinated, well budgeted, and financially motivated opposing interests seem to me to be the bigger hurdles to overcome than assembling the right software or protocols.
Agreed. Consumers don't care about about implementation details, they just want better experiences. I think this is what most decentralized services miss the mark on. There's an interesting new protocol called Handshake (https://handshake.org) that's trying to decentralize DNS. The main benefit is that it can improve security by shifting the root of trust in DNS from a certificate authority based system to a distributed network. If you think about traditional DNS, 99% of people using it (everyone on the internet) don't even know that it exists. For Handshake, I think it's possible that millions of people will use it and get the security benefits without knowing it's decentralized.
My favourite piece of irony is that one of the first political allies of decentralisation has historically been the Roman Chaotic Church, one of the most hierarchical organisations on earth. During the first industrial revolution it outlined the concept of Distributism[1] in the encyclical (mailinglist) Rerum novarum as a response to the growing social upheaval resulting from the increasing automation of labor and the centralisation of wealth in the hands of the few because of the economics of scale, and the thread this posed to the church. This publication has had a profound impact on society and was a direct influence on the general principle of European Law called Subsidiarity[2] that states that problems are best solved at a local level and only need super-state interference if the local level proofs insufficient. It makes me very skeptical about the direction the EU (and it member states) has taken with its increasing attention to data sovereignty and the host of projects and initiatives it funds in this space: blockchains, edge-computing, digital-twins, privacy-preserving-artificial-intelligence and what-not. The really successful decentralised projects where grass-root affairs, things like torrent, wikipedia and bitcoin, and stand in sharp contrast to the astro-turfing of recent ears where big and powerful institutions have usurped the idea if decentralisation to serve an agenda and to get elected. The Tor network is a counter example because it was funded my the US military, but I'd argue that investigative journalist are better served with a VPN hosted in a jurisdiction with bad diplomatic ties to their respective country of residence (some political awareness is always going to be faster and more convenient than crypto).
That's not hard to conceive, mesh networks. It's unbelievable I can't share a picture or video privately, phone to phone, at high speed and convenience, without being at the mercy of a fast and possibly expensive internet connection for both parties to send bytes around half the world to be snooped upon by one or several dystopian outfits. And then back again. Of course first I also have to negotiate joint usage of the same remote service.
So here is a very compelling use-case for people who don't care. Sadly it conflicts with the interests of the device maker duopoly.
Add to this that decentralisation introduces engineering challenges that centralisation doesn't and it's abundantly clear why fee decentralised projects work.
For my money, I've always believed the right way to make the internet communal again is about ownership, not protocols. I'd much rather be a member of a cooperative messaging app than struggle through the technical issues of a decentralised one. Even if it meant Wikipedia style begging once a year. But that's just me.
the Safe Network is the only one I've seen that would give the average person real motivation to use it
- earning money for sharing resources (drive space, cpu cores) with the network
- 1 time cost for storing a file (possibly eternally) vs the current recurring subscription
- only 1 password needed to access the network vs the current way of every site needing a separate password
bear in mind its not even finished so maybe these things won't happen in the end
I feel like home NATs were one of the biggest developments that curbed decentralization and led to the dominance of a "producer vs. consumer" internet. STUN/TURN and IPv6 are big enablers of a move back to the roots of the internet, right?
NAT isn't the problem. Most NAT punching isn't terribly hard after a lightweight rendezvous. And VPSs with public IPs have been available for $20/mo for the past decade and a half. The problem is the user experience of p2p software, due to less funding to polish it up and market it hard, due to its fundamental goal of eliminating middlemen.
> NAT punching isn't terribly hard after a lightweight rendezvous.
Has that been true for very long? It seems like nowadays it is trivial, but a decade ago it was extremely common for users to have to do manual port forwarding to receive connections.
As for VPSs, that's basically a non-starter in terms of user experience, similar to the issue you raise with p2p software. For the audience of HN, NAT etc. are a non-issue, but for real democratization to occur, the process needs to be as simple as "install app, run app, share link provided by app to friends".
NAT punching is establishing communication without forwarding a port. Think like making a TCP connection with a connect() at each end (no listen()). The ability has been there the whole time, but I'm sure libraries have gotten better - again because of development effort.
The point about VPSs is that extra public IPs have been easily available. If simply being able to receive incoming connections was sufficient, the process of setting up a VPS would have been easily automated. In reality, server-based protocols such as HTTP/DNS are insufficient for decentralization, and developing replacements takes work.
1. A name points to a single authority for a resource.
2. Even worse, the single authority is bound to a single (logical) authoritative server.
Using pubkeys for names, like Freenet or IPNS, solves (2) but not (1). Which does allow content to be wholly retrieved from peers (a massive step up!), but still relies on a single authority. So I would think that well-known authorities will still crop up, centralize, and abuse market inefficiency. Although much easier to fork, so more pressure to behave (imagine being able to say hey use my faceboot alternative at http://newsite.com, and have all the content automatically there, without the hostile software!).
I've been thinking about this for quite some time, and ideally (1) would be solved as well. The only system I've seen in the wild that attempts to distribute authority itself is Camlistore's "claims". But the idea being that a "far" link should truly be changeable by the end user, rather than to a singular authority.
You probably meant this given the mention of cloud hosted public IP addresses, but note that UDP hole punching doesn't work for all Internet connections. It's more of a way to reduce the amount of traffic that needs hairpinned vs a general purpose solution.
I think a bigger hurdle is that consumer service ISPs have restrictions on self-hosting. If it wasn't that way then I can see more self-hosting being a norm especially with some consumer routers playing as host at least for some devices attached (external HDDs particularly).
Not all, but a lot of home routers will block any incoming connections. Even if you have an IPv6 address, you'll need manual configuration or something like upnp to make it work. And with the current systems we have available... it's probably better that way.
True. Although those are not great solutions. Stun may simply fail with weird routers, or because there's lots of new connections and ports are not guessed correctly. Turn requires tunneling your traffic through a third party. Both still need you to actively reach outside and advertise your availability via some openly reachable node.
I like the idea, want to participate but am NOT signing up for yet another stupid newsletter. Please publish an RSS feed (I see that it's Github Pages ... perhaps your site generator can do it?)!
Someone noted that the forums are Google Groups and the site, blog etc. are hosted on Github. Shouldn't dog-fooding inherently be part of this strategy?
Inoreader recently added support for subscribing to email newsletters with an email address it gives you and it acts like they're an RSS feed and combines it into all your other subscriptions: https://blog.inoreader.com/2020/02/declutter-your-inbox-subs...
Might not be enough to switch for if you're already happy with what you use for RSS, but I think it's a great feature.
I didn’t know there were Jekyll templates with rss. I’d send a pr, but Im not into a project trying to promote decentralization that adds a donate button before an RSS feed.
I've been to some events here in London organized by redecentralize.org, they were fantastic. I'm quite active in that sphere, mostly SSB for me, so going there was like being a fish returning to the sea. I've learned so much and was so energized with all the conversations and demos. If you're in London or nearby, I advise to keep an eye for the next event and don't miss it. If you end up going and don't know anyone, look for me, I'm usually with a adesive with the same nick you see here. I'll introduce you to people and show you cool tech.
I was reading their list of technology (https://github.com/redecentralize/alternative-internet) and I think book looks really good, it's solid time-tested technology and I can't believe I'd never seen it in the 'alternative-internet' scene.
Scaling systems that have to run out of the browser by default is tough, but we're doing this on $0 in server costs, in javascript, with millions of users.
I'm disappointed to see a 'decentralization' platform promoting a product which promotes Pocket. And this in its top 5 recommendations, even. I understand the problem of purity spiraling vs pragmatism, but at least they should have a warning "this product promotes services which are antagonistic to decentralization."
