FOCIL Breakout #34, May 19, 2026

Protocol Calls & happenings
1

Agenda

  • Svalbard interop updates
  • Development updates
  • Spec updates
  • Testing updates
  • and more

Meeting Time: Tuesday, May 19, 2026 at 13:00 UTC (60 minutes)

GitHub Issue

2

Meeting Summary:

The Fossil Breakout 34 meeting focused on recapping the Svalbard interop progress, client development updates, and backend testing status. The team discussed achieving local devnet between Lystore and Bezu, onboarding HitRex, and reaching consensus on spec changes including non-finality enforcement and frame transaction compatibility. Client development updates included Mehdi reporting that Teku published a new docker image with GLS changes and is working on EPBS fixes, while Ethan noted Lighthouse is working but experiencing some orphaning and reorging issues. The team decided to target basing the first Fossil devnet (DevNet 0) on GLAM DevNet 4, with plans to launch pending GLAM’s stability over the next week or two.

Click to expand detailed summary

Justin and Jihoon discussed a new Discord channel called EIP, which Jihoon questioned the necessity of. They agreed to keep the channel but considered potentially renaming and moving the existing Inclusion List channel to it, similar to how they handled the Payload Builders channel. The conversation ended with uncertainty about whether they should also move the BLS and EPBS channels.

The team discussed progress from the Svalbard interop, including the establishment of a local devnet between Lowestore and Bezu, onboarding HitRex, and agreement on spec changes regarding non-finality and frame transaction compatibility. Client development updates were shared, with Mehdi reporting progress on Teku’s Docker image and fixes for EPBS issues, while Eitan mentioned Lighthouse’s working implementation but noted some block orphaning issues. The team discussed plans for updating to the 1.7 alpha 8 spec, with Jihoon suggesting they should target the latest release and be flexible about rebasing once confident about launching the first devnet.

The team discussed potential functional differences in the fossil architecture and agreed to investigate further. Jihoon committed to looking into FCO implementation and testing between CL clients. The group reviewed backend testing status, noting that EL testing was complete with both Bazel and HitRex passing tests, while CL interop was working with Besu and AetherX in local devnet environments. Justin received approval to remove the outdated fossil testing checklist from the repository and replace it with an automatically generated version based on existing tests.

The team discussed launching public .NET and decided to base Hagara DevNet 0 on GLAM DevNet 4, with plans to launch in 2-4 weeks if GLAM DevNet 4 stabilizes. They agreed to continue discussions about Fossil-specific items in this current meeting format rather than moving to ACDT immediately, as more focus time is needed for Amsterdam issues. Fabio was tasked with updating his prototype to be based on GLAM DevNet 4, and the team set an end-of-May target for the launch.

Next Steps:

  • Mehdi: Finalize and publish a new Docker image for Teku with the latest EPBS bug fixes merged.
  • Eitan: Debug and resolve the block orphaning and payload reorg issues observed in Lighthouse’s Fossil implementation.
  • Jihoon: Investigate the differences between the current Fossil spec and the 1.7 alpha 8 spec (including FCU V4 changes) and share findings with the team.
  • Jihoon: Review FCU versioning changes and determine compatibility with the current Fossil spec, then share results after the call.
  • Justin: Remove the outdated Fossil testing checklist from the repo and work toward an auto-generated checklist based on existing tests.
  • Mehdi: Look into merging/renaming the Inclusion List Discord channel into the new Fossil/EIP channel to preserve chat history.
  • Fabio: Update the Besu Fossil prototype to be based on Glamsterdam DevNet 4 in preparation for Hagora DevNet 0.
  • Jihoon: Monitor Glamsterdam DevNet 4 stability over the next one to two weeks and, if stable, coordinate the launch of Hagora DevNet 0 (targeting approximately two to four weeks out).

Recording Access:

3

YouTube recording available: https://youtu.be/D_PJR0IFQKI

4

Hegotá devnet-0 comes into view.

Svalbard Interop Recap

  • A Lodestar <> Besu local devnet was achieved during the Svalbard interop. ethrex was onboarded shortly afterward.

  • For the next couple of months, the priority is stabilization. This includes finding spec bugs, surfacing unknown spec issues and expanding test coverage.

  • During the Svalbard FOCIL breakout, people agreed to disable inclusion list enforcement during non-finality. Under non-finality, nodes are disconnecting from each other, EL is syncing and CL is processing attestations from different chains, so enforcing ILs would worsen the chain’s liveness rather than help it.

  • If frame transaction ships with Hegotá, inclusion list validation logic will change. An omitted IL transaction will be checked at a fixed index of the payload, not at the end of the payload. The motivation is to avoid builders re-executing frame transactions repeatedly. With a 60M gas limit and 10K gas per frame transaction per IL, a builder could otherwise execute up to ~100M gas in the worst case. Note that this change will only apply if frame transaction is included in Hegotá.

  • Beyond these, no major spec changes are expected, though minor changes may surface during devnets.

Development Updates

  • Teku: Mehdi published a Docker image based on a branch that has all Gloas changes merged. An interop with Lodestar worked.

  • Lighthouse: Eitan has a working implementation but is investigating an issue where one to two blocks per epoch are getting orphaned.

  • Besu: Fabio confirmed Besu is close to rebasing on top of glamsterdam-devnet-4.

  • ethrex: Iván confirmed ethrex passes all Hive tests.

Testing Updates

  • Felix from STEEL is helping with EL testing. EL has basic test vectors and hive tests passing on both Besu and ethrex.

  • There is no proper CL testing yet. CL interop currently works against either Besu or ethrex on a local devnet.

Hegotá Devnet-0 Scope

  • hegotá-devnet-0 is tentatively based on glamsterdam-devnet-4. If glamsterdam-devnet-4 stabilizes within a week or two, the target is to launch hegotá-devnet-0 about a week later.

Q: Is the FOCIL breakout ready to merge into ACDT?
A: It remains useful to discuss FOCIL-specific topics for now. ACDT is still busy with Glamsterdam. This issue can be revisited in a month or two, depending on Glamsterdam progress.

1 Like