basicaly we would like to finalise the part about missing non-engine methods that are exposed on the engine port and finalize that
would also discuss the direction of top-up sync and would it be fitted into an eip and proposed for inclusion
Meeting Time: Friday, July 24, 2026 at 13:00 UTC (60 minutes)
The Assisted Engine API Breakout Number 3 meeting focused on testing updates and discussions about the top-up sync functionality. Tamaghna reported that a testing PR was in review and planned to add new SSZ models to existing execution fixtures for verification. The team discussed how to implement top-up sync, with Dustin suggesting it should be part of the overall Benton API package functionality. They reviewed which non-Engine methods should be exposed on the Engine port, with Dustin noting that Nimbus currently uses chain ID and syncing. A significant discussion emerged about the HTTP/2 requirement, with Dustin questioning the empirical evidence for head-of-line blocking issues in the Engine API context. FLCL suggested conducting benchmarks to compare HTTP1.1 and HTTP2 performance. The team also discussed the value of top-up sync based on a real-world case where Nimbus EL successfully participated in a network despite not getting peers due to Teku providing blocks in the correct order. The conversation ended with a discussion about potentially moving the call time to better accommodate participants in Asia and Europe, with Justin suggesting a 4-6 hour time change.
Tamaghna led a meeting about the Assisted Engine API and provided an update on testing, mentioning a PR currently in review and plans to add a new SSZ model to existing execution fixtures. The team discussed the implementation of top-up sync, considering whether it should be proposed as an EIP or added as a requested engine endpoint. No specific decisions were made during the meeting, and further input on top-up sync implementation was requested from the team.
The team discussed integrating SSE REST Engine API functionality into the Benton API package, with Dustin suggesting it should follow the same approach as other updates. Tamaghna proposed starting work on adding the EIP and requested input from other clients regarding the engine endpoint and non-Engine methods to be exposed on the Engine port. The group identified a minimal list of common methods including eth_syncing, eth_getBlockByNumber, eth_getBlockByHash, and eth_chainId for potential inclusion.
Dustin raised a question about the HTTP 2 requirement and head-of-line blocking in the context of a serial API, asking for empirical confirmation of its significance. Tamaghna suggested gathering additional views from other clients, including Lighthouse, and proposed carrying the discussion forward to the next call if needed. The conversation included brief exchanges about JWT, SSZ paths, and specific Ethereum-related functions, but no definitive answers were provided during the meeting.
The team discussed the implementation of HTTP/2 support for the Engine API, with Dustin expressing concerns about the value proposition without quantitative benchmarks. FLCL suggested setting up benchmarks to measure the performance difference between HTTP/2 and HTTP/1.1, and the group agreed to start a conversation about benchmarking in the SSZ channel. Tamaghna proposed reviewing Maurice’s list as a next step for the agenda.
The team discussed whether ELs (Execution Layers) should handle state or blocks for syncing, with Dustin highlighting a specific case where a Nimbus EL bug prevented peer connection in DevNet 7, but Teku successfully worked around this issue by feeding blocks in the correct order. Dustin noted this demonstrated the value of the EL as library approach, even with pre-existing separate clients and execution layers. Tamaghna agreed that this real-world example showed the concept’s usefulness and supported the decision to implement it.
The team discussed moving their call time to better accommodate participants in Asia and Europe. Justin suggested moving the call 4-6 hours earlier, and Tamaghna agreed to check with Marius about the timing and gauge general preferences from the team. Tamaghna committed to following up offline about the new time slot.
fkjPLR#4)fkjPLR#4)fkjPLR#4)YouTube recording available: https://youtu.be/U1dJJL1aV1U