AMA #14 · 50% staking cap actually conservative

EF Protocol AMA #14 · 2025-08-29

Justin Drake defended the cap, saying the deposit share keeps rising

It covered a wide range, starting with the most upvoted question on the issuance curve and going through zkEVM real-time proving and RISC-V, ePBS and 6-second slots in Glamsterdam, and Foundation organizational operations. The Foundation side acknowledged that the share of staked ETH will keep rising under the current curve, and answered that separating provers does not harm censorship resistance and that L1 will not jump into the ultra-low-latency race.

Why it matters

Reducing issuance and capping staking directly change ETH holders' yields and the structure of the LST market, and researchers see the transition to 3SF as a chance to revisit them. If zkEVM proofs become mandatory around 2027 as Justin Drake envisions, validator hardware requirements will fall but a new role, the prover, will emerge. The halt of Verkle trees and the shelving of 6-second slots effectively set the scope of the next several upgrades.

Discussion

Issuance curve and staking cap

The most upvoted questions were how to reach consensus on the long-term issuance curve, and whether every rational holder ends up staking under the current curve. The Foundation side mostly answered yes, and the side effects of the 50% cap proposal Justin Drake raised on a podcast and the risks of ETH treasury companies (DAT) were also covered.

zkEVM real-time proving and RISC-V

Questions covered whether censorship resistance holds if block building and proof generation are left to a few operators, how RISC-V should be introduced, and whether the native rollup EXECUTE precompile must re-execute. The Foundation side answered that provers carry low risk because only one honest prover is needed, and drew the line that RISC-V adoption will be gradual and replacing the EVM is not currently being researched.

Glamsterdam: ePBS, 6-second slots, headliner selection

Questions followed on whether ePBS (EIP-7732) encourages builders to miss slots on purpose, why 6-second slots (EIP-7782) were left out, and whether it is appropriate that the proposal with more development progress effectively becomes the headliner. The Foundation side said mitigations exist, but admitted the process needs improvement to separate feature selection from solution selection.

Gas limit, blobs, and state growth

Questions covered governance of the gas limit set by staker votes, the blob limit after Fusaka, the direction of Verkle trees, and whether things can hold up without state rent. The Foundation side said the gas limit should in fact be raised faster, and said for the first time officially that Verkle trees have been halted.

Foundation organization and client teams

Questions covered whether Geth should be moved out of the Foundation, whether the Foundation should have its own consensus client, how client teams make a living, and whether transparency has improved. Geth developers and researchers had differing views, and several recently created teams and processes were introduced.

Wallet standards and privacy

Questions covered the state of EIP-7702 adoption, smart account interoperability standards, contract state privacy, and the progress of the Foundation's privacy wallet Kohaku. The Foundation side answered that ERC-5792 is the key to resolving fragmentation between wallets and dapps, and that state encryption still has no answer.

Glossary

EIPs covered

Forkcast record · ← Core Dev Calls