Note: This issue is to gain feedback on if there are any emergency topics to discuss. If no compelling topics are proposed in this issue, we would propose to cancel the call in favor of a breakout room (TBD at ACD).
Update: Call not cancelled.
Meeting Time: Monday, March 16, 2026 at 14:00 UTC (60 minutes)
*Summary by @Nerolation [copied from Eth R&D Discord]
Should we reuse the Message IDs from snap v1 for snap v2, or should we keep counting? another question I’d add: Should we make sure that snap v2 is rolled out across clients with Glamsterdam to avoid some clients not supporting it (thus not being able to support nodes that already use it during sync)? Would be great to get a temp check if client think this is feasible for glamsterdam. The efficiency improvements and simplification are quite significant, thus we should actively consider it.
Summary by @potuz [copied from Eth R&D Discord]
The team discussed meeting structure and updates on various development topics including BlobNodeZero implementation, devnet progress, and benchmarking flag implementations across different clients. The group addressed issues with EVM fuzzing scenarios and EPBS DevNet Zero functionality across different clients, with some experiencing problems while others worked correctly. The conversation ended with discussions about accelerating the Standards Finalization Initiative process for EIPs to better align with network upgrades, though concerns were raised about potential interdependencies between EIPs.
The team discussed the structure of the meeting, which included usual updates and addressing immediate blockers, followed by an experiment with breakout rooms for Block Level Access List and ePBS topics. Barnabas provided updates on BlobNodeZero, mentioning some issues and the finalization of the spec for partial size implementation. The rollout plan involves deploying on testnets, starting with a few validators on Sepolia and Goerli, with further investigation into clustering effects before enabling it by default in client logic.
The team discussed progress on devnet 2, where Stefan reported 100% participation and Erigon fixed proposal issues. Marc provided an update on parallel execution, noting it ran stably for 12 hours before breaking, and Alexi is investigating parallel state loading. The group agreed to continue general updates in the main call while moving detailed discussions to breakout rooms. Emma confirmed Reth has implemented all three benchmarking flags on their devnet 2 and 3 branches, and Andrew from Erigon committed to checking and implementing the necessary flags if not already done.
Stefan reported issues with EVM fuzzing scenarios breaking in devnet 3, though most clients handle normal transactions well. Dragan mentioned finding existing spec tests that need fixing. Barnabas provided updates on EPBS DevNet Zero, noting that Prysm, Lighthouse, and Lodestar are working correctly, while Nimbus and Teku have issues, and exits do not work yet despite deposits being functional. The team discussed plans for Alpha 3 release and potential updates to PTC before the next event.
Spencer proposed accelerating the SFI (Standards Finalization Initiative) process for EIPs to better align with network upgrades, particularly EIP 7723. He noted that Steel only merges EIPs into their main branch once they’re SFI’d, but they’ve been experiencing conflicts when creating devnet branches due to EIPs not being available early enough. Parithosh expressed concerns about SFI’ing EIPs too early, particularly those with interdependencies with other EIPs, suggesting a middle path where EIPs without foreseeable interactions could be SFI’d, but this decision should be made by ACD rather than ACDT.
The team discussed EIPs ready for SFI, with Spencer proposing to present a filtered list on Thursday for ACD’s approval. Toni suggested starting with EIPs from devnet 2, while Barnabas advised against SFI for EIPs that are not fully finalized, particularly mentioning AT37 which uses hard-coded versions. The group agreed to focus on EIPs with minimal interactions and adequate testing before SFI. The conversation ended with participants moving to breakout rooms for further discussion.
The meeting appeared to be a brief closing session with Parithosh thanking everyone for attending. No specific decisions, action items, or substantive discussions were captured in this transcript segment.
5f$2q!iZ)5f$2q!iZ)5f$2q!iZ)YouTube Stream Links: