# EPF Cohort 5 - Week 5 Date: 15/07/2024 ### Updates 1. Jumped ship on the ARM optimization project because I don't think it aligns better than the rated list DHT project idea. Also there were hiccups in getting related resources for the project. 2. Fully immersed myself into kademlia DHT, every nuance of it. IMO the original paper is very ambiguous since it overuses the word "closest". * Especially had a hard time understanding the terminating conditions of the lookup and found out that others did too. https://github.com/libp2p/go-libp2p-kad-dht/issues/290 * Check my notes for interesting insights of my research both on kademlia DHT and libp2p. 3. Dived deep into the implementation of discv5 in go-ethereum. The logic is not broken down into clean segments and IMO geth could benefit with some documentation or design patterns. For people in a hurry * [injecting UDP transport in the lookup logic](https://github.com/ethereum/go-ethereum/blob/71210b0630e4f8dd2e7bcc7c39424fb2382e4f00/p2p/discover/v5_udp.go#L317) - This piece of code in this link shows how the UDP RPC request is passed as function pointer into the abstracted lookup logic (see below link) * [use of alpha parameter](https://github.com/ethereum/go-ethereum/blob/71210b0630e4f8dd2e7bcc7c39424fb2382e4f00/p2p/discover/lookup.go#L121) * There was no evidence for multiple path lookups as suggested in the spec(for resilience) and topic related RPCs were also missing (maybe requires a deeper dive) > Most of the resources of my research are maintained under the notes directory of cohort-five repo, [here](https://github.com/eth-protocol-fellows/cohort-five/blob/main/notes/ChiragMahaveerParmar.md) is link. Hope you find something useful there. ### Next Steps 1. Dive deeper into peerDAS and its DHT requirements 2. Get some communication going(hopinheimer, mario and dankrad) to get clarity on the deliverables of the rated list DHT project. 3. Read more of probelab's analysis reports. The reports are a semblance of peace in gossip-y and chaotic p2p world. ### Personal Notes I have finally jumped over the fence and selected rated list DHT as my project this week. Even though I had been researching on p2p throughout, this is a last minute decision. Will need some time to set stage, hence have pushed proposal to next week.