I wonder if you could compile solidity into an intermediate representation, which can be further compiled to:

1. EVM bytecode
2. Zk-friendly VM

Half baked idea that requires associating the ZK instructions with the deployed contracts.

You could run a sidecar along with main chain that’s zk-friendly. 

Idk how but maybe you could easily generate ZKPs of a tx’s state update from just:

1. State root
2. Tx hash
3. Signature

—-

The proof should be done over the zk friendly sidecar so it can be much quicker. Then maybe we can have commit/reveal block building?

Working on building zkTLS infra/products. A ruthless empiricist and meme connoisseur.

http://opacity.network

Appreciate the simplification haha

It would be“zk-friendly EVM equivalent” or some shit.

A lot of the zk EVM state proofs do the shit off of the state directly. Not ZK friendly so slow.

The idea is what if you do it at the compiler level? Can you more efficiently maintain a zk friendly version of the canonical chain on the side

This is what we are doing at vyper with venom IR- it's theoretically friendly for zk backends