All Core Devs - Consensus (ACDC) #161, July 24 2025

Protocol Calls & happenings
,
6

ACDC #161 summary

ACDC #161 Action Items

  • fusaka-devnet-3 is live! Continue monitoring (and debugging issues as they arise) to progress towards mainnet.
  • The following EIPs have been CFI’d for Glamsterdam to signal potential CL headliners
    • EIP-7732 (ePBS)
    • EIP-7782 (Six second slots)
    • EIP-7805 (FOCIL)
  • Outcome of next ACDE and the prior list will be used to make a final call on the CL headliner on the next ACDC on 7 August.
  • Community input: please add here Soliciting stakeholder feedback on Glamsterdam headliners if you’d like to signal support for or against one of the above headliner candidates.

ACDC #161 Summary
Fusaka

  • fusaka-devnet-2
    • Running various test scenarios, some syncing issues in clients
  • fusaka-devnet-3
    • Launched ~24 hours ago, will fork to Fusaka soon
    • Looks good from initial assessement
    • Working on MEV workflow implementation
    • Implementing rate limiters on full nodes:
      • 100 Mbps down / 50 Mbps up on 70% of full nodes
      • 30% super nodes with potentially 1 Gbps limit (from 10 Gbps)
      • Previous devnets had no rate limiting
  • Proposal for timelines
    • August for hardening work
    • Release candidates by end of August
    • Target mainnet launch before DevConnect
    • Process requires 30 days between testnet fork and mainnet date
  • Timeline considered ambitious but not impossible

Glamsterdam

  • Series of updates to inform headliner conversation:

Shorter Slot Times Research Update

  • Maria presented preliminary results on slot timings:
    • Metrics measured:
      • Block propagation (excluding timing games)
      • Attestation arrival time (including block propagation, execution, validation)
      • Pure attestation propagation for missed slots
    • Data shows:
      • Block propagation 95th percentile under 1 second for well-connected nodes
      • Full attestation processing 95th percentile of 95th percentile below 4.5 seconds
      • Identified shift in distribution around 1.5s mark where last 10% take longer
    • Next research steps:
      • Collect more data from other relays
      • Analyze blobs and aggregates
      • Investigate locally built blocks and small blocks
      • Research late arrivals to understand contributing factors

Update on ePBS Free Option Problem

  • Christoph and Bruno from Flashbots presented:
    • Problem: Winning builders can invalidate blocks if unhappy with market changes
    • 8-second window gives builders option to fail to send blobs
    • Option value increases with:
      • Market volatility
      • On-chain liquidity
      • Time window length
    • Potential solutions:
      • Shorten free option window
      • Implement slashing penalties
      • “Block-listing” (not preferred as reduces permissionlessness)
      • Separate blobs from transactions with independent fee market
    • Needs further analysis around mitigations and impact

Another ePBS update

Community input

Headliner selection

  • Given that, we discussed how to proceed.
  • Decided to CFI three EIPs with the understanding that we would select one as a headliner:
    • EIP-7732 (ePBS)
    • EIP-7782 (Six second slots)
    • EIP-7805 (FOCIL)
  • This helps narrow scope for next ACDE so a headliner decison can be made there
  • Given this and results of next ACDE, a decision will be made for the CL headliner on the next ACDC
  • One important point of the discussion was how to best capture community input to help guide headliner selection. If you’d like to make a case for one of the above as headliner, please add it here: Soliciting stakeholder feedback on Glamsterdam headliners

Spec Question: SSZ Types for Pureth

1 Like