Consensus-layer Call 146

[ralexstokes]

Consensus-layer Call 146

prev: call 145

Meeting Date/Time: Thursday 2024/11/28 at 14:00 UTC
Meeting Duration: 1.5 hours
stream

1. Electra
   • mekong updates (deposit processing bug)
   • devnet-5 spec
     • EIP-7685 updates for empty requests: engine: exclude empty requests in requests list execution-apis#599 should be merged
     • finalize EIP-7742: include? Update EIP-7742: update the required EL headers and the gas fee mechanism EIPs#9047
   • Include? eip7251: Do not change creds type on consolidation consensus-specs#4020
     • if so, target devnet-5 or later?
   • Blob throughput increase in Pectra
     • Move to (target, max) of (6,9): https://eips.ethereum.org/EIPS/eip-7691
       • samcm update: https://ethresear.ch/t/block-arrivals-home-stakers-bumping-the-blob-count/21096
     • For any given adjustment, do we want to include EIP-7623?
2. PeerDAS / Blob scaling
3. Research, spec, etc.
   • EIP-7805: FOCIL
   • Consensus-layer Call 146 #1200 (comment)
4. Open discussion/Closing remarks
   • Discuss potential ACD improvements

[nerolation]

I want to quickly summarize the ongoing discussion about increasing the blob throughput in Pectra and hope we can discuss this further.

There are different proposals:

Blob increase EIP, increasing target to 6 and max to 9:
ethereum/EIPs#9045

Calldata cost increase EIP:
https://eips.ethereum.org/EIPS/eip-7623

And with touching the blobs, we might want to adjust the base fee mechanism:
https://ethresear.ch/t/on-blob-markets-base-fee-adjustments-and-optimizations/21024

Personally, I find the blob increase the most important, followed by limiting the EL payload size through 7623. The base fee adjustments are more on the nice-to-have side and nothing can really go wrong if we don't ship them. Still, it would be cleaner to do them directly with the blob increase, open to debate.

[mkalinin]

I’d like to gather feedback on the following change to the consolidation processing:

• eip7251: Do not change creds type on consolidation consensus-specs#4020

[g11tech]

bringing up the following PR to take a call on to include this change in eip 7742 or to go ahead without any changes for pectra
(as discussed in devcon pectra meeting)

ethereum/EIPs#9047

cc: @adietrichs @bkellerman @nerolation @barnabemonnot

[soispoke]

I'd like to introduce EIP-7805 to present FOCIL and its core properties

[timbeiko]

If time permits, I'd like to discuss potential ACD improvements we brainstormed at the pre-devcon R&D workshop: https://ethereum-magicians.org/t/allcoredevs-network-upgrade-ethmagicians-process-improvements/20157/51

Not urgent, and can be moved to next week's ACDE if the agenda is too busy.

[samcm]

I'd like to briefly present https://ethresear.ch/t/block-arrivals-home-stakers-bumping-the-blob-count/21096 with regards to EIP7691 if the agenda/time permits 🙏

[siladu]

I won't be able to attend the call but I would like check in on where clients are at with using the graffiti to help with execution client diversity analysis. This was the reason @ethDreamer introduced engine_getClientVersionV1 https://github.com/ethereum/execution-apis/blob/main/src/engine/identification.md

There was a talk at Devcon by Migalabs https://www.youtube.com/watch?v=HIg1bscGqOE
featuring some new client diversity analysis via the graffiti using tools such as https://github.com/migalabs/goteth and
https://clientcode.supermajority.info/ but it's not as useful until every client implements the API and graffiti changes.
Pectra is a good milestone to aim for since all clients will be forced to update.

As far as I know the following issues are outstanding and progressing them would help with this effort:
status-im/nimbus-eth2#6668
prysmaticlabs/prysm#13558
erigontech/erigon#12581

Does anyone know of any blockers to progressing this on either CL or EL side?

[fjl]

I see my PR ethereum/execution-apis#599 is on the agenda.

It seems that again the reason for this change has come into question. Unfortunately, I cannot participate in the call today, and thus can't be there to defend it. I have left some more comments on the PR explaining my position: empty requests do not lead to any difference in the block, I am only proposing to leave them out on the engine API as a convention.

[cortze]

Following @samcm 's upper comment, from ProbeLab we would also like to make an update on the bandwidth measurement study we've been doing over the last week from multiple geographical regions.

[akashkshirsagar31]

Podcast (audio only) - https://open.spotify.com/episode/5FF4NnBSvPfLxoRh5t02FA?si=ps4UAMuBSDK32zSIVcBNFg

[abcoathup]

@soispoke can you share the EIP7805 slides you presented?

[abcoathup]

Eth Magicians: https://ethereum-magicians.org/t/all-core-devs-consensus-acdc-146-november-28-2024/21793

[soispoke]

@soispoke can you share the EIP7805 slides you presented?

ok course, here's the link: https://docs.google.com/presentation/d/1BYZQOmNkv-nLe25NmPOowhBLPVa76kDDibR_nR0-RGU/edit?usp=sharing