A place for developers to talk about building on Farcaster.

/dev

How does Farcaster handle censorship and spam? I watched the YouTube overview, seems like every node is a full sync "archive node" (with 1yr expiry window). Muting happens clientside - What about abusive or illegal content? Can my Hub opt out of storing data from specific authors? Do spammers live rent free on my disk?

Full-stack digital product consultant @ BoundCorp.net

Not rent free. they have to pay 5$ per year I guess.

Training myself to become a public market investor .

Who do they pay $5/yr to? Surely it doesn't go to the other Hub operators. Is there a centralized entity collecting that?

For now goes to protocol. 
Detailed discussed on it here

System is still evolving. But from what I understand. 
$5/year per user will be fixed at the protocol/HUB level . Not an option for each hub to make on their own. Messages from users who's fees not paid will be rejected by Hubs.

Oh yikes, this $5/yr *is* the anti-spam mechanism? And they expect each operator to do their own anti-spam on like the honor system or something? No way to ban or censor bad actors, no way to evict Hubs that allow abuse? I gotta say this looks like a huge oversight, I don't think the incentives line up at all here.

For now goes to protocol. 
Detailed discussed on it here 

https://github.com/farcasterxyz/protocol/discussions/98

Then don’t run a Hub?

How does this work with Ethereum nodes?

But this means if someone is posting illegal content, I can't modify my Hub to exclude it. Client side filtering isn't sufficient when you already have so many custom clients in the ecosystem. The Hubs still bear the burden of spam storage here and every Hub has to republish it or else they are as you say invalid.

It goes to the protocol contract. Hubs are configured to follow that contract for storage. If you modify your hub you’ll have invalid data for the CRDT and other Hubs will eventually stop peering with you.