Odaily Planet Daily News: Vitalik.eth expressed on the X platform that there has been an increasing discussion about the ongoing role of L2 in the Ethereum ecosystem recently. As the progress of L2 entering phase 2 is far slower than expected, and L1 itself is expanding, it is anticipated that the Gas limit will significantly increase by 2026, which means that the original vision of L2 as Ethereum's 'branded sharding' is no longer applicable and new paths need to be found.

Vitalik Buterin pointed out that L1 no longer needs L2 as branded sharding, and that L2 cannot or does not want to meet the attributes required for true branded sharding. He suggested that L2 should identify value-added points beyond 'scaling', such as privacy, efficiency for specific applications, extreme levels of scaling, non-financial application design, ultra-low latency, and built-in oracles. When dealing with ETH or other Ethereum assets, L2 should at least reach phase 1 and support maximum interoperability with Ethereum.

In addition, Vitalik Buterin stated that over the past few months he has become more convinced of the value of native Rollup precompiles, especially after having the ZK-EVM proofs required for scaling L1. This precompile will make EVM verification without a security council feasible. He believes that research should be conducted on how to design this precompile so that it can verify the EVM part when L2 contains 'EVM and other content.' This will make achieving secure, robust, trustless interoperability with Ethereum easier and enable synchronous composability.