I want to check when will we need cryptography in blockchain field.I list all I know here:
Consensus algorithm
2.Signing transaction (including different kinds on digital signature )
3.Layer2’s verification on Layer1
4.Verifying the cross chain tx
5.Rollup with different digital signature
6.Cross-shard
Multi-party computation for privacy (in some situations ZK-SNARKs are not sufficient, particularly when you want to maintain state that no one can decrypt)
Thanks Vitalik
it seems that the most cryptography in Consensus layer is hard code. Could Ewasm help ?
And could application layer get more flexibility for programming ? For example,zk rollup is hard to interact with general smart contract on Eth so they need to invent new lang like Zinc.
Ewasm could certainly be used to write cryptography more efficiently, though in many cases it’s not needed; rather, what’s needed is for EVM to get better support for big-integer operations, as that’s where the bulk of the cost of a lot of cryptographic operations comes in.
For example,zk rollup is hard to interact with general smart contract on Eth so they need to invent new lang like Zinc.
Yeah, this is a problem. The EVM is not very friendly to living inside of cryptographic systems like ZK-SNARKs (or STARKs, or MPC, or FHE). Ewasm is a little better, but the best would be a VM specifically designed for ZK-friendliness.