PoS Sequencer Pool: Decentralizing an Optimistic Rollup

This was my question as well. Very interesting idea—thank you.

Thank you for the questions and apologies for the delay. These questions are focused on the implementation of the Sequencer Pool. The Metis team is drafting the procedure and implementation for the transition including the answers to these questions; it will be posted in a future blog.

For now, I can help answer any questions related to the architecture of this system.

The proposed architecture resembles more an architecture of a blockchain network trying to anchor data in Ethereum, than an architecture of a rollup. (Don treat this as a dig). If the algorithm of leader election is deterministic and based on publicly verifiable information (ex. something in the L1), it seems to me that the consensus layer will be unnecessary.
Furthermore, adding such a consensus layer is bound to increase transaction costs - more actors need to do more work and need to be incentivized for it.

1 Like

Hey - I think the nodes that compose the sequencer chain can just register their public key on the eth main net. I am not sure why there is a need for separate MPC keys.

The nodes produce the blocks together and sign the blocks that are then posted on ETH main net. The signatures can be aggregated using BLS to make verification less gas expensive