--- tags: eth2devs description: Notes from the regular Eth2 implementers call image: https://benjaminion.xyz/f/favicon-96x96.png --- # Eth2 Implementers’ Call #50 - 2020-10-15 [Quick contemporaneous notes by Ben Edgington] Agenda: https://github.com/ethereum/eth2.0-pm/issues/187 Livestream: https://youtu.be/L4Dvlgxku1g ## Testing and Release updates [v1.0rc0 spec](https://github.com/ethereum/eth2.0-specs/releases/tag/v1.0.0-rc.0) is out (release candidate). Primarily mainnet config things. **Please run the test vectors**. v1.0 spec release is fairly imminent. Changes: discv5.1 wire format; BLS v04 spec that disallows the zero private key. Fuzz testing: Teku has been added to the structural differenial fuzzer. JNI interface with Teku/instantiating the JVM makes things quite slow, but this is unavoidable. Custom fuzzing engine coming soon. ## Zinken Good work on Zinken. The late Prysm release caused issues for the block explorer, but otherwise all was well. ## Medalla Medalla testnet is not v1.0 conformant. We also expect most people to turn off their nodes when mainnet comes. Therefore, we are considering a new long-running v1.0 testnet/devnet without such wide participation. There have been suggestions to upgrade Medalla to v1.0 spec. [Preston] Even though we might see a big drop in participation, we will still be able to gather syncing data etc and Medalla can still serve a purpose. As of today, participation is low, but worth running for a few weeks to see the inactivity leak in action and otherwise stress-test clients. The BLS v04 upgrade is quite simple; the discv5.1 may be more complex. [Terence] Maintaining two branches is complicated, so we should either upgrade Medalla or drop it. [Afri] Good chance to test the upgrade process. [Proto] Re discv5.1: we can have nodes that run both v5 and v5.1 and bridge discovery. [Danny] Alternatively all clients could switch over at a fixed point in time. General feeling is that supporting multiple network configurations is a pain. No consensus emerging - **Danny to open an issue to continue the conversation with a view to resolving next week**. ## Client updates **Lodestar** Participated on Zinken with four validators. There was some incompatibility between the validators and beacon node, but the beacon node is online and keeping up :tada: Will fix up the validator interaction, and should be able to participate in future. Have improved validator duties handling. **Nimbus** Audit is in final stages. Now in review and fix stage. Plan to join public attacknets. Are up to v1.0 of the spec. Plan to release binaries for Linux, Mac and Windows. Renamed repo `nimbus-eth2`. Implemented weak subjectivity checkpoint start. Question: how do others handle the history of Eth2 deposits in this scenario? Working on interop tests for gossipsub v1.1. Also reducing the database size and memory usage. **Teku** Enabled moderately strict weak subjectivity handling. Dealing with various issues from the audit. The Teku validators can now optionally run in a separate process. We have implemented enough of the standard API to support this and are working through the remainder. Memory usage reductions (0.5GB down). Gossipsub v1.1 ready to be merged. **Prysm** New release that cleans up lots of deprecated flags. Working on slash protection interchange. Discv5.1. BLS v04. Peer scoring. Aligning with v1.0rc. Released WebUI - please give feedback! **Lighthouse** Good Zinken launch. New breaking release with a lot of changes. Includeds WSP checkpoint, uPnP support. Passing v1.0 spec tests. Progress on WebUI. [EIP 3030](https://github.com/ethereum/EIPs/pull/3030/) proposal for standard interface for remote signing. ## Research Updates **Vitalik** Data availability sampling. Been thinking about simplifying Phase 1: having data on chain but nothing else. Want to reduce the level of dependence on committees in order to make the protocol more robust. Working on how to formulate this. [Discussion on shard transitions, and Kate commitments...] **TXRX** [Mikhail] Team is losing the Tx (the Texans) part, but work continues. Pyspec transpiler work is ongoing; looking at transpiling into Rust for a pure functional representation. Working on withdrawals to the Eth1 shard. Eth1/Eth2 transition work ongoing, expect updates soon. **Leo (BSC)** Looking into Prometheus metrics for four clients. Have created a [spreadsheet](https://docs.google.com/spreadsheets/d/1obCGque-BMHwMlX_SuFj_jCr6MiEGxNIoyYbOYYGaaQ/edit?usp=drivesdk) - would like advice as to whether these metrics are equivalent between clients. Would be interested in creating some standard naming around metrics. ## Networking Discovery v5.1 is the big thing. Client implementations are near complete. ## Spec discussion Danny is collecting some complex scenarios to add to the test vectors. ## Open discussion Nothing. * * * # Chat highlights From Leo BSC to Everyone: 03:38 PM https://docs.google.com/spreadsheets/d/1obCGque-BMHwMlX_SuFj_jCr6MiEGxNIoyYbOYYGaaQ/edit?usp=drivesdk