# EPF Cohort 5 - Week 8 Date: 05/08/2024 ### Updates 1. We had the bi-weekly progress update with the team. Ashley is done preparing an initial version of the spec. 2. Started on the libp2p integration for the rated list. Faced some design questions * Should we use a discv5 or libp2p-kad-dht for discovery? there was a general consensus that discv5 is too bulky for a research project. Personally, I don't think the two are different since libp2p uses provider records and discv5 uses ENR records. * How do bootnodes boot? examples within the libp2p project either specify bootnodes for an already existing network or establish a network of bootnodes through `mdns`. Since this is only a one time problem we bootstrapped the bootnodes one by one. * Which transport should we use, UDP-based or TCP-based? QUIC is faster than TCP and can prove to be useful in production ready nodes. We left this decision for an opportunity of experimentation. 3. Read the entire peerDAS [spec](https://github.com/ethereum/consensus-specs/blob/dev/specs/_features/eip7594/das-core.md). > You can find the code [here](https://github.com/hopinheimer/rated-list-research) ### Next Steps 1. Parametrize node degree through the connections manager. 2. Orchestration using one of the those fancy ethpandaops tools. 3. Research on measuring metrics using libp2p, see how probelab does it. 4. Research on how integration tests can be written. Will be useful to test edge cases of rated list. 5. Implement basic funtionality of rated list. ### Personal Notes This week the output of the EPF was more tangible than the other weeks. It feels great and I am excited for the coming weeks.