Consensus-layer Call 135

[ralexstokes]

Consensus-layer Call 135

prev: call 134

Meeting Date/Time: Thursday 2024/6/13 at 14:00 UTC
Meeting Duration: 1.5 hours
stream

0. Announcement
   • migration of ethereum-package kurtosis module
   • Consensus-layer Call 135 #1069 (comment)
1. Electra
   • devnet-1
     • specs release v1.5.0-alpha.3
     • open question: EIP-7549: Append new committee_bits field to end of Attestation consensus-specs#3768
     • implementation progress?
   • peerDAS
     • PEERDAS_ACTIVATION_EPOCH recap
     • inclusion in Pectra Meta-EIP
     • proposal to uncouple CL and EL around blob count: EIP-7742: Uncouple blob limits per block across CL and EL consensus-specs#3800
     • how to raise the blob count?, EIP-7623
     • testnet thoughts? keep parallel R&D for now?
     • any other issues/PRs worth discussing?
   • SSZ update
2. Research, spec, etc
   • F-star naming: Fulu, Felis, ?
3. Open discussion/Closing remarks

[barnabasbusa]

Kurtosis team plans to open source their entire project. (https://x.com/KurtosisTech/status/1800258729289355763)
There is currently a plan to migrate the ethereum-package from their repo to the ethpandaops organization on 13th of June.

From their announcement channel:

We will be transferring the Ethereum Package to ethPandaOps this Thursday 2024-06-14

From then on to access the Kurtosis package you will have to do

kurtosis run github.com/ethpandaops/ethereum-package
...

This is a pretty big breaking change, as it will break literally everyone's workflow until they don't change the url. Would like to discuss what this will mean for us/client teams in the coming days/weeks.

[hwwhww]

I'd nudge on getting the finality of the F-star naming again this week.

The ChatGPT-generated proposals: https://hackmd.io/oVQCHuOdT-q2KN9YrD-r7w

In the last call, iirc "Fulu" and "Felis" received a lot of support in the zoom chat. It seems people like shorter names. I'd add that "Foramen" is cool because of the Ramen x Osaka theme (🍜 x🏯)!

p.s. I think Felis looks like a typo of Felix (@fjl 😂)

[fjl]

I'm totally up for having a fork named after me.

[timbeiko]

Draft PR to include PeerDAS in the Pectra Meta EIP once we've clarified how it will be reflected in the CL specs.

[etan-status]

SSZ https://stabilitynow.box

• EIP-7495: SSZ StableContainer
  • Remerkleable implementation ready for merge
• EIP-7688: Forward compatible consensus data structures
  • Teku has joined the EIP-7688 devnet (with Lodestar / Nimbus)
  • Grandine continued on implementation
  • Prysm / Lighthouse still in early stage (we should realistically decide on go/nogo for Pectra until end of Jun, 1 more ACDC)
• EIP-6493: SSZ Transaction Signature Scheme
  • It should be possible for this EIP to coexist with Verkle in the same fork. The others (7495/7688) we should consider for Petra regardless of decision on 6493.
  • Ethereumjs has started implementing a prototype
  • Besu has listed EIP-6493 as a protocol fellowship project
  • Open tradeoff question:
    • Either, CL has to know about the Transaction StableContainer structure (but not the various Profile), in order to know how to merkleize transactions and compute hash_tree_root
    • Or, CL continues processing ByteList of opaque, serialized transactions. But in this case, we need two lists, one containing the serialized transactions (as today), and an additional one containing the transaction roots. Essentially leading to double the hashes on the EL side to compute both roots.
    • Possible to decide later, it's low effort to change, but if someone has a strong opinion, better to know early.
• 🦒 synergizes well with SSZ tree structures growing from the earth.
  • Nimbus received initial ANSI art

[hwwhww]

I'd like to confirm if it's okay to merge ethereum/consensus-specs#3768 in the call.

[timbeiko]

Review request for the PFI/CFI/SFI EIP: ethereum/EIPs#8662

We don't have to discuss on the call, but client team reviews of the proposal would be appreciated 😄

Biggest open question: do we want CFI to have an explicit "gate", i.e. 1 (+?) client team thinks an EIP should be in an upgrade?

[abcoathup]

I'd nudge on getting the finality of the F-star naming again this week.

@hwwhww I dug my F-star name discussion post out of drafts on Eth Magicians. (I had been waiting for Pectra scope to be finalized).
It is a wiki post so feel free to edit.
https://ethereum-magicians.org/t/f-star-name-for-consensus-layer-upgrade-after-electra/20285

[gorondan]

Review request for the PFI/CFI/SFI EIP: ethereum/EIPs#8662

We don't have to discuss on the call, but client team reviews of the proposal would be appreciated 😄

Biggest open question: do we want CFI to have an explicit "gate", i.e. 1 (+?) client team thinks an EIP should be in an upgrade?

ref. explicit "gate" : The current version of ethereum/EIPs#8662 (based on @abcoathup 's EthMag feedback on the way for the Upgrade Meta EIP author to avoid spam proposals) seems fairer than adding an explicit gate.

[ralexstokes]

closing in lieu of #1084