Meeting Time: Monday, October 27, 2025 at 14:00 UTC (60 minutes)
Mario Vega facilitated the call
Barnabas
Other updates
Minhyuk Kim
Barnabas
Marius van der Wijden
Mario Vega
Kamil
Raxhvl
Stefan
Justin T.
Lukasz
Keri
Marius
Dan provided a summary
execution-spec-tests will now contain the test fixtures, and test fixture releases, but no code.PS: This is a quick note, please suggest any changes here.
The meeting covered updates and discussions about various testnets, including the Sepolia and Hoodi forks, as well as the deprecation of the Holesky testnet, with Barnabas providing specific timing details for upcoming events. The team reviewed performance testing efforts, including improvements in stability and gas limit discussions, while also addressing concerns about test case documentation and devnet updates. The conversation ended with discussions about ePBS releases and consensus specifications, including updates on test cases and storage-related issues, along with an announcement about the “weld” project and contributor guidelines.
The meeting focused on updates and discussions related to the Sepolia and Hoodi forks, as well as the deprecation of the Holesky testnet. Barnabas confirmed that BPO2 was scheduled to happen in 9 hours, followed by the Hoodi Fork, and noted that BPO1 on Hoodi would be reviewed a week later. The team discussed the status of the devnet, which was expected to remain online until midnight, and emphasized the need for participants to update their notes and stay online during the transition. Minhyuk raised concerns about the removal of Holesky, highlighting its utility for testing upcoming forks, and Barnabas explained the challenges of coordinating Hoodi’s larger user base. The conversation ended with a reminder for clients to confirm and update their gas limits to 60 million, as it was the default setting for most clients.
The team discussed performance testing and profiling efforts, with Kamil reporting on stateful and compute tests, including improvements in stability and the creation of a custom Docker file for detailed tracing. They also addressed gas limit increases, with Marius and others discussing EIPs related to gas pricing and precompiles, while Kamil noted that recent benchmarking showed improvements but required further evaluation. The conversation ended with a brief mention of Block Access List updates and a reminder about Rahul’s presentation on that topic.
Rahul presented updates on a new release with 20 test cases for Coinbase withdrawals and edge cases, and highlighted a bug that escaped detection by all three clients during early testing. He proposed maintaining a formal Markdown document for test cases to improve collaboration and suggested linking it to EIPs, which sparked a discussion on whether to keep test cases within EIPs or in an external document. The team also discussed updates on devnets, with Stefan confirming they would be ready by the end of the day, and Mario noted that Holesky has officially stopped finalizing.
The team discussed updates on ePBS, including a new Consensus-specs release with additional ePBS tests for COAS and a rejected proposed structure change to attestation. They noted that a missing Gossip condition check was identified for fixing. The next ePBS breakout call is scheduled for next Friday, and clients were encouraged to attend. Łukasz raised concerns about the need to update ETH simulation tests to reflect the reverted error code change and discussed potential issues with transaction type resolution logic. Keri suggested opening issues in the Execution APIs repo to address these concerns. Mario mentioned a discussion about EIP 7610 regarding storage dangling in contracts, which was deemed underspecified.
The team discussed an edge case involving contract deployments to accounts with no code but existing storage, which was previously addressed by EIP-684. They debated whether to leave the storage as-is or implement a self-destruct approach, with Gary expressing concerns about the semantic implications of leaving storage unchanged. The group agreed to continue the discussion offline and revisit it at the upcoming ACDE meeting. Additionally, danceratopz provided an update on the “weld” project, announcing that Osaka contributors can now start PRing to the default execution-specs branch, while Amsterdam contributors should hold off until the fork is rebased.
3cR@^ky!)3cR@^ky!)YouTube Stream Links: