It's all good in theory but in practice "trustless" simply means that trust relationships are obscured. If you want to go trustless you go in the woods and live like hunter-gatherer. In a complex society there is always someone who is running things. The only choice you have is between someone who is close to you (that autocrat who actually might be your friend) and someone who is far away (the developer who is taking VC money and doesn't give a shit about your problems)
Technological tricks that increase your power as a user (key based identity, data portability) can work in both kinds of systems.
Can you be a little more specific? Like I said, I wasn’t using it while it “failed”, so I don’t have much experience here.
you still need to have some form of collaborative filtering / moderation.
But the advantage of those sorts of systems is that you can have many more types of moderation and collaborative filtering that aren’t “a single autocrat who owns or rents the hardware has absolute control”. Such as Aether’s democracy; or Scuttlebutt’s “subjective moderation”/public blocking; or maybe a hierarchy where the person at the top (say, the founder) delegates control to moderators to run subforums but he doesn’t have direct control over those subforums, and those moderators can delegate further and so on ad infinitum, like DNS; or a normal political system but with protection against certain abuses of power, such as shadowbanning. Sure, you could try and implement something like those systems in a standard server-focused model, but the single autocrat with the hardware can break the rules any time they like.
@Hyolobrika@strypey it was a rhetorical question: the answer is "they failed".
If we know that they failed, what makes you think that emulating their p2p approach is a good idea?
Fully p2p and "trustless" systems makes zero sense when designing social systems. It is costly to operate and you *still* need to have some form of collaborative filtering / moderation.
The idea is not to have "one community for everyone" but a simple onboarding guide for those wishing to leave Reddit. Let them come to the places that they are "used to" at first, and once we got rid of Reddit we can start optimizing the new system.
@strypey@raphael If you’re going to have one community for everyone, there should be some way of allowing multiple moderation styles to co-exist within it, like Aether used to have.
@raphael > for extra points, maybe publish them to a Lemmy community?
It would be good if we could have one TIL community across the whole fediverse. Lemmy immediately seemed less useful to me when I learned that there can be a totally disconnected TIL community on every Lemmy instance. This makes it difficult, if not impossible, for a critical mass of people to converge on a community, as they can on a subReddit.
@Hyolobrika@strypey@raphael I just don't think these ideas are new. Most political ideas don't work. Ideas that work generally lead to one of the two outcomes I previously described: many small autocrats or one big organization that controls everything
Whether the person is my friend or not, all other things being equal, it’s better for him to have less power than more. And even if he is my friend, he might not be other users’.
Technological tricks that increase your power as a user (key based identity, data portability) can work in both kinds of systems.
Yes, but can radically new (for the internet) political systems such as those I described?
@Hyolobrika@strypey@raphael As far as I know Aether and Freenet are single implementations and fully depend on their developers. Only a few people use them so we don't even know how they scale.
IPFS is owned by Protocol Labs, and there are no independent implementations. IPFS will disappear along with Protocol Labs. Also, after all those years it is still too slow for the web and useless without gateways. Torrents is the only system in your list that is truly decentralized. However, they are only successful because trackers exist
@Hyolobrika@strypey@raphael Trustless is not bad, it is a meaningless buzzword. Even if you use a perfect peer to peer system you need to trust people who maintain it. Even if they are benevolent, they can remove some feature that is very important for you or abandon the project, and this is just as bad as admin who is banning you from the server
That can change, and that's a trust relationship that exists regardless of how centralised or trustless a protocol is. I still don't see why it's bad for a protocol to be trustless.
@Hyolobrika@strypey@raphael It is not that simple because perfect p2p doesn't exist and in federations you have more choice. But if you compare these models on a high level then yes p2p is better.
But you have to trust the developers somewhat anyway. The difference is that with a centralised or federated protocol you have to trust developers plus server admins whereas with a P2P (or to some degree otherwise key-based) protocol you just have to trust developers. Put that way it's obvious which is better.
@Hyolobrika Yes, I'm happy to see this. But those who depend on IPFS should be worried, because it doesn't have a strong community of volunteer contributors and may not survive
@silverpill I agree with your larger point though. Pure P2P is a pipe dream. After about 25 years of watching with interest, I've never seen one that worked on even the scale of the fediverse, and lasted more than 10 years.
@silverpill > ["exit to community"] basically means "we're cutting expenses, now you're on your own"
Not necessarily, and that's not how this announcement reads to me. It's better for a protocol and its reference implementations to be under independent governance. So I'm giving them the benefit of the doubt that this is more like Jabber -> XMPP Foundation or New Vector -> Matrix Foundation than it is like Sandstorm or ScuttleButt. We'll see.