The Gnosis Core Devs Call is a pivotal part of the Gnosis ecosystem, fostering idea sharing, collaboration, and continuous innovation within the community.

Representatives from various teams, including Erigon, Gateway, Nethermind, Geth, Gnosis DevOps, Gnosis Core Devs, Gnosis DevRels, Gnosis Comms, and community contributors. The diverse attendance ensures a thorough and insightful discussion from multiple angles.

For those unable to attend, the full recording can be found on YouTube.

Client Team Updates

EL

• Nethermind:
  Working on Shutter
  No more xdai(_archive) configs in the new Nethermind release (1.21.0)
  Potential issues in Avado

• Erigon:
  The header download functionality is acting up a bit from time to time
  Somnath is working on it
  Transaction pool should work on Gnosis but hasn’t gone through extensive testing
  Same for validating

• Geth:
  Was able to sync 1.5m blocks
  Geth is now running out of memory on the old branch
  Did a rebase and is cleaning up some code, but that new branch isn’t syncing

Chain Infra

• Gateway
  Is running 100% of the RPC traffic of the Gnosis RPC

Dencun

• EIP-4788

• We believe that the HISTORY_BUFFER_LENGTH variable doesn’t matter that must

• Having 1.14 days or 0.47 days give roughly the same UX

• We’re going to keep that number and thus deploy the exact same contracts

• Gnosis: https://gnosis.blockscout.com/address/0x000F3df6D732807Ef1319fB7B8bB8522d0Beac02

• Chiado: https://gnosis-chiado.blockscout.com/address/0x000F3df6D732807Ef1319fB7B8bB8522d0Beac02

EIP-4844

• Big block experiment

• The network suffers from big blocks (we spammed it super cheaply for a 10 - 20 minutes)

• The head participation went down quite a bit (close to 20%)

• Min gas price will be 1 gwei to be on the same scale as normal transactions (1 gwei minimal priority fee)

• Nethermind wrote a tool to spam blobs and it broke a lot of clients

• These might just have been early implementation bugs

• EIP-7514

• https://hackmd.io/jWVHqjmrQ1aXtpt6iTNTdw?view

• An attacker could get a 67% share in the network in 2.5 days because the entry queue scaled way too quickly

Hive Tests

• Merged the dencun tests from upstream

• Will work with Egor to catch up as fast as possible

Devnet

• We have a Devnet with Nethermind, Teku and Lodestar

• Lighthouse doesn’t work for custom configs anymore, Carlos is in communication with them

• There are some issues, probably related to peering

• Also deployed the blob scanner and spammer

‍