Execution Layer Meeting 198

[timbeiko]

Meeting Info

• Oct 10, 2024, 14:00-15:30 UTC
• Stream: https://youtube.com/live/YQwdKE0d8LI
• Zoom: shared on #allcoredevs Discord channel shortly before the call
• Ethereum Protocol Calls Calendar subscription

Agenda

• (Oct 11 Deadline) Pectra Contract audit RFP
• Pectra Implementation updates:
  • devnet-4
    • CL spec release: v1.5.0-alpha.8
    • 7702 eth rpc schema
    • BLS precompile pricing
      • Remove MUL precompiles?
  • devnet-3 updates
  • devnet-4 launch timelines?
  • Public testnet name: MooDeng?
• EIP Discussions
  • EIP-7623 update
  • EIP-7783
• CL Discussions
  • Add EIP: Reduce Slot Time for Lower Peak Bandwidth EIPs#8931
  • https://github.com/ethereum/builder-specs/pull/101/files#r1778224447
  • Staking bandwidth considerations (context)

[ryanberckmans]

Before we add more blobs, imo we must develop a rigorous definition of staking bandwidth requirements to help maximize credible neutrality, especially so we know where the validators can and can’t migrate should a disaster warrant such a migration.

https://ethresear.ch/t/wheres-the-home-staking-bandwidth-research/20507

https://x.com/ryanberckmans/status/1839663502035747119

[lightclient]

EIP-7702 eth rpc schema: ethereum/execution-apis#575

[nerolation]

Quick update on EIP-7623, it's now more elegant to implement. (h/t @tanishqjasoria)

[Giulio2002]

I would like to present ethereum/EIPs#8933 EIP-7783

[chfast]

The gas costs for BLS precompiles still don't make sense. It looks gas costs for MSM were based on the version without subgroup checks and subgroup checks were added later, made the complexity linear but gas costs remained unchanged. More info in https://ethereum-magicians.org/t/eip-2537-bls12-precompile-discussion-thread/4187/81.

[timbeiko]

@lightclient @chfast @Giulio2002, I've added your topics to the agenda 👍

@nerolation could you please share a link to the changes for people to review ahead of the call?

@ryanberckmans I've added your topic to the agenda at the end, given it's very open-ended, somewhat more of a CL topic (although it has EL implications) and that we may tangentially get to a similar discussion when covering EIP-7783. Are there specific questions you'd like to bring up on the call? If so, we can try and prioritize them.

[nerolation]

For EIP-7623, you can check out the execution specs PR here:
ethereum/execution-specs#966.

This addresses some rare edge cases where a transaction could run out of gas, but then receive a refund afterward (due to EIP-7623). If the refund had been applied earlier, the transaction wouldn't have run out of gas in the first place.

[ralexstokes]

worth flagging v1.5.0-alpha.8 should be ready as a stable pectra-devnet-4 CL spec target :)

[chfast]

I propose a simplification to the BLS precompiles (EIP-2537): ethereum/EIPs#8945.

[ralexstokes]

I have a primarily CL-specific question, so I'd schedule this at the end of the call and if we don't have time then we can skip; but if possible I'd like to raise the question of the best structure for the Builder API following the execution layer requests in Pectra (cf. https://github.com/ethereum/builder-specs/pull/101/files#r1778224447)

[abcoathup]

Bikeshedding FYI:
🦛 MooDeng leading the Eth Magicians vote for the public Pectra testnet name. (for signaling purposes only)
https://ethereum-magicians.org/t/naming-the-public-pectra-testnet/21263

[benaadams]

EIP-7782 (slot time) to be discussed with EIP-7783 (gaslimit auto increase)

[Giulio2002]

EIP-7782 (slot time) to be discussed with EIP-7783 (gaslimit auto increase)

This is also probably worth discussing on next week's ACDC as this is mostly a CL change rather than an execution change - although it has execution layer implication

[timbeiko]

@benaadams added right after 7783, and then @ralexstokes's point after that 😄