Execution Layer Meeting 203

[timbeiko]

Meeting Info

• Jan 16, 2025, 14:00-15:30 UTC
  • 📆 Scheduling note: ACDE is cancelled on Jan 2 for the holidays. A Pectra interop testing call will be held that day for those around.
• Stream: https://youtube.com/live/uh1hZCE4k0w
• Zoom: shared on #allcoredevs Discord channel shortly before the call
• Ethereum Protocol Calls Calendar subscription

Agenda

• Pectra updates
  • devnet-5 readinesss
  • Update EIP-7623: Clarify the gas refunds handling EIPs#9227
  • Update EIP-7840: Add BaseFeeUpdateFraction EIPs#9240
  • Blob base fee at fork boundary
  • Update EIP-2537: Clarify the gast cost calculation formula for MSM EIPs#9245
  • EIP-7702 delegation introspection
  • Audit updates
  • Testnet releases + scheduling
• Holesky gas limit (context)
• RPC Standardization
• Hardware & bandwidth requirements
• EIP Editors' Workshop ethcatherders/EIPIP#372

[abcoathup]

Eth Magicians wiki post: https://ethereum-magicians.org/t/all-core-devs-execution-acde-203-january-16-2025/22346

[fjl]

I would like to discuss RPC standardization on this meeting. Specifically, I would like to resolve the following:

• The role of rpc-compat hive tests. How do clients perceive the value of these tests? We maintain them as a separate artefact in the execution-apis repo so that clients can pull them into their unit tests. But nobody does that, probably. They also run on hive, but there are a lot of minor client issues that prevent them from being green. Ideally we'd reach a point where all clients pass all tests.
• Spec considerations: we have some stuff in the spec for backwards-compatibility, like v signature value on transactions. It's hard to remove because the RPC client space is so open-ended, so have opted for keeping this around. I know some client-specific fields exist, such as author on blocks for Nethermind. Should these be added to the spec as well?
• How to deal with extension proposals: there are a few open issues on the execution-apis repo where people proposed new methods. Sometimes these proposals would require extensive client changes. What's the process for landing an RPC extension?

Update: I made a document about the state of the RPC spec process, and some open questions: https://notes.ethereum.org/@fjl/rpc-standards-2025-01

[chfast]

EIP-7623 spec clarification. Requires updating a lot of tests.

[Sophia-Gold]

I'd like to ask client teams to immediately set the default gas limit on Holesky to 36M and once Pectra is on Holesky increase to 60M.

The idea is for Holesky to always have a higher gas limit than mainnet so we can spot bugs earlier and have an idea what increases are safe.

[chfast]

Potential EIP-2537 (BLS precompiles) update: ethereum/EIPs#9245.

[ralexstokes]

we should discuss clarifying behavior of the blob base fee computation at the fork boundary (and may need a clarification to EIP-7691 to make this explicit)

[Jrachman]

Given all of the recent discussion about EXTCODE*’s behavior with 7702, @dave-carroll7 and I put together a technical writeup into code introspection, its relationship with EOF, and a simple path forward: https://hackmd.io/@otim/H1Q7yCHDyl

This seems to be the final sticking point for 7702 and hope that we can reach closure at ACD!

[poojaranjan]

EIP Editors' Workshop

• Date & Time: Jan 17th, 2025 at 16:00 UTC
• Agenda & Livestream

[kevaundray]

I would like to set a call for finalization on the hardware requirements for the next ACD and tentatively start a discussion on 25/25 for bandwidth given focil and mev-boost and 50/50 given focil but self-building

[Jrachman]

Appreciate the conversation today! As requested, here is a PR that reflects our proposed changes to the written spec: ethereum/EIPs#9248

OP: #1227 (comment)

[akashkshirsagar31]

Podcast (audio only) - https://open.spotify.com/episode/6WFi88O6o6VKgYMD19JNRq?si=sbVRElyCQCm2dpF3rMiCGw

[timbeiko]

Closed in favor of #1253