Second Layer Scaling - Week 16, 2022

[bk0code]

Second Layer Scaling - Week 16, 2022

Assembly Group

• We compiled a list of topics to be discussed during the next research summit.
• We worked on a sketch of the governance in the Assembly ecosystem. Now we have two researchers and one student working on this topic.
• Two team members met in Dublin for a mini-summit. They discussed many topics, including the security of the root committee selection, modeling the token distribution, and academic papers we will work on in the future.
• Two external researchers joined our weekly meetings. We hope to introduce them to our work and get them involved in the future.

ZK Task Force

• Experimenting with gnark (go) and bellman (rust) libraries
• Progressing with the zkEVM part of the report
• Finalizing the zk primer on zk Applications and proof systems evolution
• Focusing on Pinocchio, Groth16, PlonK arithmetization, Plookup, Schwartz-Zippel lemma and Kate commitments
• Progressing with the family tree
• Reviewing Stateless DLTs & light clients implementation, added Plumo to our focus

Data Availability

This week we have discussed the network model of Assembly and the roles of each entity there.
Some questions to be discussed:

• Can everyone be an access node? Is there any stake locked to become an access node?
• Are the same attacks applicable for ZK rollup as for Validium?
• What happens if a fisherman doesn't have enough information to produce a fraud-proof? In the classical DA setting, light clients simply don't accept the block under consideration
• If roles of identities are known this may be problematic. Who knows who is a very important question.
• DA can be used when fishermen try to get information from validators/access nodes
• How many honest validators (in total and in each committee)?

Some questions that might need to be further addressed:

• Which setting has to be considered: optimistic or zk?
• Are there access nodes with a partial view? I.e. access light nodes
• Who writes to the SC chain (SCC)? E.g. participants using the atomic composability in the SCC - do they need a full view on all involved SCCs?
• Who would perform the DA check and for whom?
• On chain summaries (scaling solutions) are NOT equivalent to having the data available. Particularly in an optimistic setting. But is it enough?