blockchain

Monad VS MegaETH-Bankless Episode and Debate


Modified:

Overview

Recently a bankless spotify epsode about high performant evm blockchain is popular, let's discorver:

Spotify Episode

Twitter Thread

Takeaways

General

  1. High performance can be achieved in various ways: single-thread optimization / parallel computing; with different expressions.
  2. Monad's monolithic approach vs. MegaETH's modularity.
  3. Ethereum settlement + 3rd party data availability (DA) works.
  4. Crypto should be fun; limited edition strategies for peripheral products.

Debate on Full Nodes

https://x.com/keoneHD/status/1826488867152605320

In an interview, Keone@Monad and Lei@MegaETH had a disagreement on the definition of a full node. The core issue is whether trusting the sequencer when syncing the state from the sequencer during full node startup constitutes a trust assumption. Keone later expanded on this debate on Twitter.

  • Keone Hon@Monad In the context of an exchange or merchant scenario, under megaeth architecture, how long should the transaction confirmation period be when using the so-called full node?

  • Vitalik@Ethereum The main focus is not on the definition of a full node. There are two levels of confirmation for transactions based on Ethereum L2:

    1. Pre-confirmation at the sequencer level, using a sequencer slashing mechanism to prevent malicious behavior.
    2. L1 confirmation for a deeper level of security.

  • Keone Hon@Monad In a fraud-proof system, should a transaction be re-executed to ensure it is not being cheated?

  • Vitalik@Ethereum Introducing stakesure: https://arxiv.org/pdf/2401.05797 will provide instant feedback (either success or compensation). ZK and validity proofs will solve this issue.

  • Keone Hon@Monad Replaying transactions is obviously very costly, and megaeth will not implement zk validity proofs for now. I still believe that in real production environments, being able to run a full node (to execute all transactions) is very important for the payment process.

Monad

  1. Monad DB / Monad BFT / parallel (async) execution; 100-200 validators distributed globally.
  2. The biggest execution bottleneck is state access.
  3. Monad emphasizes decentralization, characterized by low-cost hardware / validator count / full node count.
  4. Monad DB is a core innovation, with Merkle tree optimizations.
  5. Monad uses async execution; decoupled from consensus.
  6. Monad has community-driven design mascots.
  7. Currently there is a devnet; mainnet will be fully open-source before launch.
  8. Mainnet is expected early next year.
monad-mascot

MegaETH

  1. MegaETH is an ETH Layer 2: sequencer / full node / verifier node.
  2. MegaETH uses fraud proofs / eigen data availability (DA).
  3. MegaETH's verifier uses a stateless client.
  4. MegaETH is based on Ethereum for decentralization.
  5. Emphasizes low latency: e.g., real-time games.
  6. Incubation program: Mega Mafia.
  7. MegaETH uses Bunny as its mascot.
  8. Mainnet is expected early next year.
megaeth-mascot

EVM

  1. It has a 256-bit width issue, but fundamentally it is sound, with strong ecosystem and network effects, and it is evolving.
  2. Vitalik mentioned 3 regrets, including precompiled contracts?
相关文章推荐