1/8 Web3 is broken. That’s what they say. We don’t think so, but we believe Web3 apps are still very fragile, which puts us very far from the truly decentralized blockchain economy. Still, we know where to move and what must be done to make it really decentralized 🧵
2/8 Yes, we can't fix everything, but we know how to fix the RPC layer. There are several main issues, the most crucial of which is that domain name censorship remains a crucial point of failure as the provider is a regulated entity.
3/8 And yes, front-end hosting could be distributed but it just would increase the time required for an attack. The hosting provider would prefer to terminate services for a customer upon a request from an official rather than investigate the issue.
4/8 An RPC node indeed would be less censorship-vulnerable if each dApp maintained its own #RPC, yet even in that case a node could go down, get censored, or become compromised. That’s not the answer.
5/8 We cannot rely on something that is unreliable by design. The blockchain protocol seems to be the most sustainable entity, leaning on a widely distributed network of professional players: validators and miners.
6/8 We firmly believe the RPC layer should be a cluster of an excessive number of geographically distributed RPC nodes, pretty much like the L1 PoS validators consensus concept. And who can implement it if not validators? So, there we go.