Blockchain is an unchangeable (unless you own the only authority as we’ve seen, or you fork it) database that can only grow. It HAS to have multiple, identical instances or the owner can corrupt it. I’m not sure I see any benefit here except for potential security. But we’ve seen blockchains get into the terrabytes pretty quickly which isn’t that feasible for large-scale operations.
This is why Ethereum is moving to statelessness. Where only a rolling state is saved but the diffs are broadcast and accessible for a certain period of time before they are dropped. So any system listening to an Ethereum node can just follow the root and listen for just changes they care about.
How do you ensure you are holding an accurate copy of the blockchain in that case? A single holder of a blockchain means the entire thing is insecure and the holder can change or steal whatever they want.
If a new player wants to host a copy of the etherum blockchain, how do they do that when it’s stateless?
It will still have some state, things like merkle roots will still be verifiable, and any derived/state that isn’t kept from the main chain will have some form of proof that links it to the main chain.
Most actual transactions/normal user stuff is moving to layer 2 chains, which commit a hash or proof of their state to the main chain. So they will still carry their state, but rely on the security of Ethereum main net.
There will still be archive nodes that store everything that’s ever happened, but being a full archive node won’t be required to run a normal node as a validator or as a user. This allows the network to be more decentralized without every node needing 2-4TB of fast nvme to even do anything.
You’re right about that for sure. But a few TB isn’t that big a deal anymore, and if there was a need to deal with it we could come up with something… or just not use it at all. 😎
Blockchain is an unchangeable (unless you own the only authority as we’ve seen, or you fork it) database that can only grow. It HAS to have multiple, identical instances or the owner can corrupt it. I’m not sure I see any benefit here except for potential security. But we’ve seen blockchains get into the terrabytes pretty quickly which isn’t that feasible for large-scale operations.
This is why Ethereum is moving to statelessness. Where only a rolling state is saved but the diffs are broadcast and accessible for a certain period of time before they are dropped. So any system listening to an Ethereum node can just follow the root and listen for just changes they care about.
How do you ensure you are holding an accurate copy of the blockchain in that case? A single holder of a blockchain means the entire thing is insecure and the holder can change or steal whatever they want.
If a new player wants to host a copy of the etherum blockchain, how do they do that when it’s stateless?
It will still have some state, things like merkle roots will still be verifiable, and any derived/state that isn’t kept from the main chain will have some form of proof that links it to the main chain.
Most actual transactions/normal user stuff is moving to layer 2 chains, which commit a hash or proof of their state to the main chain. So they will still carry their state, but rely on the security of Ethereum main net.
There will still be archive nodes that store everything that’s ever happened, but being a full archive node won’t be required to run a normal node as a validator or as a user. This allows the network to be more decentralized without every node needing 2-4TB of fast nvme to even do anything.
You’re right about that for sure. But a few TB isn’t that big a deal anymore, and if there was a need to deal with it we could come up with something… or just not use it at all. 😎