Meeting Notes
Testing Update
- EEST
- All lingering PRs are merged
- We should wait for Devent 6 to settle before re-filling
- EOFWRAP relies on evmone which is not yet on pectra-5. Will be turned off until evmone is ready (some 7702 edge cases)
- Devnet 6 is a tiny change
- Fuzzing findings
- No new findings
- Fuzzing fails if clients are not on the same pectra devnet level
- Relies on new fields
- Tracing should support new tracing field RIP-7756
- PC zero to container is critical
- Function depth is currently needed
- Clients need to target Getting up to Pectra Devnet 5/6 (whatever the last devnet becomes)
Client & Compiler Update
- Nethermind, Besu, evmone, revm, erigon, all no changes
- Solidity - not much change, merged the base needed and testing.
- Planning for Q1/Q2
- Deciding which optimizations to do
- Vyper
- Not much change
- PyVM EOF support?
- unknown, Mario will ask snake charmers
- Separate from EELS
Spec And Other Issues
-
EOF Metadata Section
- An update after doing a spike for EIP-7834 (metadata section) in
evmoneand EEST. Ref.- Multiple optional sections makes things complicated.
- Very different from other sections, needs new code paths
- No suggestions for improvements, meets it’s goals
- EEST Tests
- could automatically add metadata to standard tests, or a subset
- Probably via a marker
- could automatically add metadata to standard tests, or a subset
- Design wise more optional sections make parsing more complex
instead of a full re-filling.- If we are sure it can be osaka we could make it mandatory
- An update after doing a spike for EIP-7834 (metadata section) in
-
EOFCREATE / TXCREATE potential features and scenarios - an opinion and/or counter proposals.
andCross-chain addresses (deterministic, counterfactual, and nonceless)- Pitor reviewed doc
- Scenario 1b is the preferred solution
- EOFCREATE - sender+salt
- TXCREATE - sender + salt
- Predeploy (fka ISC) to load bootstrap contracts.
- Drop EIP-7720 (EOF create tx)
- Scenario 1b is the preferred solution
- Discussion talked about pre-reployed contracts
- This makes TXCREATE more palitable
- New TXCREATE EIP
- Includes 2 toehold contracts - salt only and salt+=CALLER
- Maybe an address-based nonce-toehold as part of EIP and not ERC
- We need a plan B for not having TXCREATE
- RIP-7740 - RIPs/RIPS/rip-7740.md at master · ethereum/RIPs · GitHub - Preinstall deterministic deployment factories
- Pre-ERC - EOF deterministic deployment contracts ethereum/ERCs@master…shemnon:ERCs:eof/txcreate-factories
- Pitor reviewed doc
-
Start work on Osaka-1 spec changesout of time