A while ago, I posted about my plan to build a Lemmy client using the Plebbit protocol.
The response was, honestly, full of hate. I wasn’t expecting praise or anything, but I didn’t think people would react so negatively to the idea of something truly decentralized.
But here I am again. Still believing that Plebbit is the only real self-hosted social media protocol out there.
Let me explain why, in the most direct way I can:
– Plebbit is serverless. – There are no global admins. – It does not rely on any central server. – It can’t be censored or taken down. – It works like BitTorrent, but for social media. – No subreddit can go offline as long as one peer is online.
Every subreddit (called a "subplebbit") is its own world. Mods can ban users, remove posts, or run things how they want. But there’s no “head office.” Nothing above them.
And yes, Plebbit already has support for NSFW subs like /pol and others. It doesn’t need approval from anyone.
I see Plebbit as the Bitcoin of social media. Pure, peer-to-peer. No middlemen. No backdoors. No central kill switch.
It reminds me of what the internet was supposed to be—free, open, uncensorable.
Sadly, most devs I’ve met online don’t really understand peer-to-peer tech deeply. Some barely know cryptography. That’s okay, but it also makes real decentralization hard to appreciate.
If you’ve never read the Plebbit whitepaper,
https://github.com/plebbit/whitepaper
please do. It’s not just another protocol. It’s a whole different way of thinking about social interaction online.
I’m still planning to build that client. I don’t care if the first reactions were negative. I’m not doing this for approval. I’m doing it because I genuinely believe in it. But reviews matter too.
So every user is their own moderator... which just sounds like a ton of extra work vs something like Mastodon where I can pick a server whose moderation practices I agree with, is already decentralized into countless servers and allows the user to spin up their own instance.
Keyword filtration as a moderation technique is woefully ineffective vs trolls who simply find "clever" new ways to harass with intentional misspellings, dogwhistles, etc.
Meanwhile, you're pitching this thing as "uncensorable" which automatically appeals to the worst elements available. Maybe I'm wrong and it'll be the perfect format for internet discussion, but I'm going to have to see that actually happen before I jump on board.
Once we have the labeling services it will be easy and a single click to use someone's else labeling.
Also each community moderate however they see fit, as a community owner you're incentivized to keep their community free of spam and derailing posts etc.
P2P is superior to federation in many ways though
collapsed inline media
I agree, but it's not just keywords, it's community-based labeling services, so you could have 10+ people labeling on a single content-labeling extension. You could also have AI agent sifting through the network and labeling content with minimal human intervention.
I mean... I guess I don't really see the point, but I absolutely encourage you to go ahead. I always enjoy being proven wrong when I'm being negative.