Ethereum co-founder Vitalik Buterin shared thoughts about a proposed technical change in Ethereum’s consensus design.
The discussion focuses on replacing Casper FFG with a new finality gadget called Minimmit. The idea is to simplify block finality while balancing network safety and recovery behavior.
Under the current theory of Casper FFG, validators will follow these steps to verify a block. They will sign to justify the block, and after that, they will sign again to finalize the block.
However, this is a robust system, and it requires a higher level of collaboration between validators. In the Minimmit model, validators will only sign once.
The tradeoff is the margin of security. Casper FFG has a goal of being about 33% fault-tolerant, so it remains secure even when about a third of the active stakers are behaving dishonestly.
Minimant reduces this minimum to about 17%. But Buterin says that this still has actual economic disincentives for attackers.
Also Read: Ethereum Signals Bullish Divergence, Targets $2,050 Amid Monthly Loss Risk
The main idea is that, in the case of Minimmit, it is focused on stopping censorship attacks, not on reversing finality. In other words, a finality reversal attack is one that is easy to detect and has obvious proof of the crime. Such attacks could lead to a large loss of staked assets.
This type of attack is worse because it is harder to defend against. According to Buterin, who is part of Ethereum, when an attacker has over half of the staking power and cannot easily finalize blocks, people will create minority soft forks.
This will ensure that the network remains active even when there is pressure. The idea is to make it so that uncertain events will produce multiple chains instead of quickly finalizing data that might be false.
The other concern is related to bugs in the software. If the client software contains a bug, then finalization too early may cause the blockchain to be locked in the wrong state.
Minimmit makes it harder for accidental finalization by increasing the percentage of clients needed, i.e., 83%, as opposed to 67% in older assumptions.
The idea is that finality should, in fact, be final. If the network is able to pause finalization in uncertain conditions, the developers and validators will have time to act.
Buterin also said that if attackers are going to spend billions of dollars to disrupt the network, they would probably think of other ways to do it, other than the staking system.
Also Read: Ethereum Targets Next Growth Phase With Powerful Scaling and ZK-EVM Roadmap
