Ethereum Core Devs Meeting 133 Agenda

[timbeiko]

Meeting Info

• March 4, 2022, 14:00 UTC
  • Note: Daylight Savings time will start in Americas before the next ACD (Mar 18), but the call is "fixed" at 14:00 UTC - double check your local time!
• Duration: 90 minutes
• Youtube Stream: https://youtu.be/ZKPElqIfteU
• Zoom: To be shared on #allcoredevs Discord server shortly before the call

Agenda

• Merge Updates
  • Kiln 🔥🧱 Updates
    • Public launch/documentation for clients
• Shanghai Planning #450
  • EIP-4863: Beacon chain push withdrawals
  • EIP-4844: Shard Blob Transactions
  • EIP 3651: Warm COINBASE
• EIP discussions:
  • EIP-3978: Reprice reverted SSTORE/CREATE/SELFDESTRUCT/LOGX operations to 100 gas via gas refund mechanism due to state non-modification
• Announcements:
  • EL Merge Client Release Naming Breakout Room #491

[k06a]

I'd like to discuss EIP-3978: Reprice reverted SSTORE/CREATE/SELFDESTRUCT/LOGX operations to 100 gas via gas refund mechanism due to state non-modification. I know some of you wouldn't like the idea one more time to reprice opcodes, but I believe this concrete case is important from multiple pov: reverted transactions would cost much less, and reverted subcalls will be significantly cheaper and in the end, this repricing sounds fair.

Right now after the bad swap, it is cheaper to revert swap subcall and transfer tokens back to the user (to have a gas refund) instead of reverting whole transaction - this is ridiculous.

[timbeiko]

Quite the EIP name, @k06a! Added to the agenda 😄

[ralexstokes]

i'd like to discuss EIP-4863 for inclusion in Shanghai, which specifies beacon chain push withdrawals: ethereum/EIPs#4863

after some feedback starting from the "pull" style on the last call, we seem to have more weak consensus in this direction

[protolambda]

I'd like to discuss EIP-4844 for inclusion in Shanghai, which specifies blob transactions: https://eips.ethereum.org/EIPS/eip-4844

During ETHDenver we have built (experimental) execution and consensus layer prototypes, which we can also discuss.

Execution layer implementation: protolambda/go-ethereum#1 (by EIP authors)

Consensus layer implementation:

• prysmaticlabs/prysm@kiln...blobs by Prysmatic Labs
• sigp/lighthouse@unstable...realbigsean:mini-danksharding by Sigma Prime (single man team during hackathon, may be incomplete)

Some technical feedback I would specifically like:

• EIP-2718 defined typed tx encoding for within the execution payload, and we extend this with network encoding now that contains additional contents. In go-ethereum this meant we had to define "wrap data" (next to regular inner tx) and additional encoding/decoding. Can we extend EIP-2718 / define a network-layer counterpart?
• The "wrap data" greatly simplifies distribution of the blob through the tx pool, storage in local tx pool journal, and RPC usage, since the blob tx without is not valid. Retrieving blob contents lazily in places like the tx pool might be more optimal though (pull full blob data only once). What is the best method to relay this tx wrapping data?
• This tx type uses SSZ encoding in the execution layer for the first time (and so does the withdrawal transactions proposal), any technical arguments against SSZ here?
• Feedback how we can adopt KZG commitments (critical to forward compatibility of this tx type).

[timbeiko]

Added @ralexstokes @protolambda 😄

[wjmelements]

I plan to join to discuss EIP 3651: Warm COINBASE for inclusion in Shanghai

[djrtwo]

accompanying doc for EIP 4863
https://notes.ethereum.org/@djrtwo/B1Urzhzgq

[timbeiko]

Closed in favor of #492