Greetings everyone, and welcome to the weekly Gnosis Core Devs Call. Just a quick reminder that this meeting takes place every Wednesday.
Participants: Erigon, Gateway, Nethermind, Geth, Gnosis DevOps, Gnosis Core Devs, Gnosis DevRels, Gnosis Comms team and the contributors.
April 19, 2023
Client Team Updates
- Nethermind: No updates.
- Erigon: No updates.
- Geth: Was able to import the state (simulate snap sync)Is now starting the node in full sync, but it’s not connecting / downloading blocks. Once this is done, Guillaume should have a working prototype
- Prysm: No updates.
- Nimbus: No updates.
Gateway: No peers found on Devnet (probably just the different genesis).
Wrong configuration was shared. There’s now a mismatch for outside nodes. There’s no Erigon node running at this point. 80% attestation rate (all nodes ran by Nethermind) Shapella went through
Lion: does Nethermind have metrics for the network?
I.e. withdrawal amounts per block etc. Apparently not.
Might potentially be added. Most credentials are still 0x00. Meaning that no withdrawals are currently happening.
Defines the number of validators that can join or exit the network in a specific time frame. On Gnosis that number is a bit high, meaning that it would theoretically be possible to get a big stake in the network very quickly. This would require a consensus client code change. Technically a change on the deposit contract could also limit inflow. But that break a lot of assumptions. It doesn’t change anything in the case of outflows.Attacking the light client bridge is exclusively an inflow problem
Shutterized Beacon Chain: is there progress?
We’ve got more precise specs on a PR
Lion: how specific do you need the spec to be? It seems like EL clients have more lax requirements for specs, is that true? On CL everything is specced quite precisely.
Hive: Will start writing Hive tests by the end of this week
Improving the genesis loading for Erigon and Nethermind, but hasn’t been tested on Erigon yet.
Lion: can we agree on which conditions tests would be considered good enough to move to testnet / mainnet?
Ruben thinks it would be enough if we can just run the contract specific parts (balances etc)