TechFlow reports, July 05, according to Vitalik Buterin (@VitalikButerin) post, Ethereum researchers recently convened a meeting in Berlin, updating the protocol's long-term development roadmap (strawmap.org). Vitalik pointed out that "Lean Ethereum" is not a single upgrade, but a series of improvements to be implemented in phases over the next three to four years, with importance comparable to "the Merge", covering the reconstruction of almost every core module of the protocol, main contents include:
- Validation Mechanism: Introduce recursive STARKs, replacing the existing direct re-execution method, becoming a first-class core component of the protocol
- Quantum Security: Priority significantly elevated, all quantum-vulnerable components will be replaced, quantum-safe Blob design is underway
- Consensus Layer: Decouple availability and finality, achieving one to two round finality, with better security and lower latency
- State Layer: Existing dynamic state remains unchanged, but new scalable state types will be added (such as UTXO storage, ring buffers, etc.), expected by 2030 Ethereum will possess 2TB dynamic state + 100TB new state, after migration of ERC20, NFT and other applications, over 10x Gas fee reduction can be achieved
- Privacy: Upgraded from an additional feature to a first-class objective, integrated throughout designs such as Mempool, state tree, etc.
- VM: In addition to EVM, leanISA or RISC-V will be introduced, the long-term goal is for the protocol layer to only directly recognize the underlying VM, with EVM relegated to a high-level language compilation layer
- Scaling: Gas limit, Blob quantity, and block time will be adjusted multiple times over the next approximately five years, the Glasterdam upgrade is expected to bring a significant Gas limit increase
Vitalik stated that H-star (Hegota) may be the last "pre-Lean" style fork, starting from I-star, subsequent upgrades will fully embody the Lean philosophy.



