All Core Devs - Consensus (ACDC) #176, Mar 19, 2026

Protocol Calls & happenings
,
1

Agenda

March 5, 2026 meeting canceled in favor of async version on Eth R&D discord.

March 5 Agenda:

Meeting Time: Thursday, March 19, 2026 at 14:00 UTC (90 minutes)

GitHub Issue

1 Like
3

Video, transcript & chatlog

News coverage

Resources

March 5 async

4

Will use this thread to drop some notes on the 5 March agenda. As stated above, we are handling this week’s agenda with async work on Discord. Check this thread Discord or here for discussions.

Notes for 5 March

agenda: All Core Devs - Consensus (ACDC) #176, Mar 19, 2026 · Issue #1941 · ethereum/pm · GitHub

Glamsterdam

epbs-devnet-0

There was a bit of a rocky start but the devnet looks reasonably stable now.
Curious if anyone wants to do any specific testing/analysis on this devnet before moving to devnet-1.

epbs-devnet-1

With the first devnet live, we should identify what we need to
include before we are ready for the next devnet. My understanding is that there are some
parts of the alpha.2 spec we have deferred.

glamsterdam-devnet-0

This also needs to sync with the BAL devnets, but we should identify if we need an epbs-devnet-2
beyond devnet-1 and what it should contain before we are ready to merge CL and EL for the first Glamsterdam devnet.

variable PTC deadline

One such item we had under discussion was the need for a varible PTC deadline: Variable PTC deadline by fradamt · Pull Request #4843 · ethereum/consensus-specs · GitHub.
After some discussions, I’d propose we keep it simple and not target this feature for Glamsterdam.
It is nice but adds more complexity on top of what is already a big set of changes.
AIUI this feature supports a potential gas limit increase on top of what we get without it, but I think
we still have room to raise the limit with other Glamsterdam changes so should be ok without this.
A quantitative analysis of how much more we could raise the gas limit with this feature would be a good starting place
to support its inclusion.

SSZ in the Engine API

giulio2002 wanted to bring up two PRs that would add SSZ support to the Engine API:

barnabasbusa also introduced this PR feat: add ssz to engine api by barnabasbusa · Pull Request #764 · ethereum/execution-apis · GitHub with similar effect.

There’s active discussion around “why SSZ” here: Discord

One path forward is to align on having this feature in the first place (my read is that most lean yes), and then getting one version of a spec in place. Check the discord thread for more info.

JWT secret standardization

bomanaps wanted to bring up the standardization of how we handle the --jwt-secret flag across client implementations. My understanding is that this was in a good enough place but if you’d like to chime in on the PR I expect they will try to push it forward. engine: default location for reading/writing jwt secrets by lightclient · Pull Request #297 · ethereum/execution-apis · GitHub

Hegotá

With the selection of FOCIL as the CL Hegotá headliner, our current process implies we need to block on selecting the EL Hegotá headliner before proceeding with further fork scoping. The conversation still seems to be ongoing, so I’ve updated the scoping process post to extend headliner selection until 26 March which gives two more ACDE calls for headliner selection there. From there, we would take the next ~month for non-headliner EIP proposals. If ACDE decides on what to do with the EL Hegotá headliner earlier than 26 March, we can adjust timelines accordingly.

1 Like
5

Meeting Summary:

The team discussed implementation challenges with ePBS on DevNet Zero, focusing on issues with chain reorganization and block roots, with particular concern about changing the 4788 specification. The group reviewed various technical proposals including PTC implementation, payload re-organization, and a variable payload deadline, while also discussing the SSZ Engine API and potential improvements. The team agreed to target mid-April for the Glamsterdam DevNet Zero implementation and planned to make key decisions about PTC and other specifications at the following week’s meeting.

Click to expand detailed summary

The team discussed several key issues related to the ePBS (Execution Payload Blockroot Specification) implementation on DevNet Zero. Barnabas reported that multiple clients (Prysm, Lighthouse, Lodestar) were experiencing issues with chain reorganization, particularly with empty blocks. Potuz outlined three main technical issues that need resolution, including the BTC/PTC acquisition problem and concerns about block roots when execution payloads are missing. The team debated how to handle these issues, with Dima from Lido expressing strong concerns about changing the 4788 specification, arguing it would require significant effort to adapt their existing permissionless proofs system. The group agreed to conduct further analysis on potential solutions, particularly around using parent block roots as a mitigation strategy, and to engage with affected protocols like Rocket Pool and Eigenlayer to understand the impact of any changes.

The team discussed implementing changes related to PTC (Proposer Time Commitment) and payload re-organization in the blockchain. Potuz explained the benefits of keeping PTC in the block to ensure proposer views are respected, while Stokes raised concerns about the complexity of the current proposals. The group reviewed several PRs (pull requests) for implementation, with Potuz expressing a preference for simplicity and minimal state impact. They agreed to further analyze the options and make a decision by the next ACDT meeting on Monday, acknowledging that one of these changes would require a new devnet. The discussion also touched on the need to address variable PTC deadlines in future devnets.

The team discussed implementing a variable payload deadline independent of the PTC deadline, with Toni sharing concerns about how compression could affect block sizing and suggesting clients should make the decision. Mark presented data showing potential gains of 300-500 milliseconds for execution with a variable deadline, though the group agreed more thought was needed on the implementation approach. The team aligned on targeting mid-April for Glamsterdam DevNet Zero, which would involve merging CL and EL work streams including both BALs and PBS, with a decision on PTC to be made the following Monday.

The team discussed two main topics: SFI decisions and the SSZ Engine API. Regarding SFI decisions, they agreed to postpone the discussion to the following week’s ACDE call, as it was determined to be more appropriate for the EL side rather than the current CL call. On the SSZ Engine API topic, Barnabas proposed an opt-in implementation that would allow CLs to support it immediately while ELs could opt in later, without requiring a fork. Marius suggested using this opportunity to rethink the engine API entirely, including versioning and struct definitions, though this would delay implementation.

The team discussed potential improvements to the Engine API, with a focus on implementing SSZ (Simple Serializable JSON) as a first step. They agreed on an incremental approach, starting with SSZ for blob serialization to improve performance, particularly regarding zero-copy operations. The group also reviewed EIP-8148, which proposes adding a custom sweep threshold feature for OX02 validators to increase adoption rates. Dima presented the EIP, explaining how it would allow validators to set custom thresholds between 33 ETH and 2048 ETH, providing more flexibility than the current binary choice between these values. The team noted that while this feature was considered for Pectra but not implemented due to concerns about complexity and timeline, it could be valuable for the Hegata release.

Next Steps:

  • Prysm team (Terence): Merge PR to fix missing blocks issue within the next 2 days
  • Lighthouse team: Investigate and fix exit processing issue on DevNet Zero
  • Client developers: Review and provide opinions on PTC issue PRs (especially PR 5020) by Monday ACDT meeting
  • stokes: Track the syncing semantics and re-orgs issue (create or find existing issue to document this)
  • stokes: Follow up with Rocket Pool regarding 4788 issue impact on their architecture
  • stokes and team: Conduct impact analysis on 4788 changes with Eigenlayer and other affected protocols
  • All teams: Work towards Glamsterdam DevNet Zero by mid-April (before Devconnect)
  • Barnabas: Review and refine PR 764 for SSZ Engine API support
  • Giulio: Refine SSZ Engine API PRs across multiple clients (estimated one week of work)
  • EL teams: Finalize EL headliner for Pecora on next week’s ACD call
  • stokes: Open non-headliner proposal window for Pecora after next week’s ACD call (approximately one month duration)
  • stokes: Double-check E3 search post for formal Pecora timeline tracking

Recording Access:

6

YouTube Stream Links: