Simplifying Vitalik's "Endgame"
This is a breakdown of the article that Vitalik Buterin, Founder of Ethereum, published recently regarding blockchain design.
Today Vitalik Buterin released an article titled ‘Endgame’. It’s an interesting read and basically talks about the future of blockchain design and how many paths lead to the same destination. However, as you may know Vitalik is an absolute gigabrain who speaks in highly technical terms that most of us may not understand. Since the contents of the article are extremely important, I have made an attempt to simplify his article. Hopefully without missing the point.
The article starts with a discussion about “Big Block Chains”. The type of chain he is referring to is one that compromises on decentralization by increasing the block size. To further simplify it, increasing block size means the amount of transactions that can be included in a single block is very high, often times it is so high that the demand to have a transaction included in the block may never be more than the supply. Hence, the gas fees will never be prohibitively expensive. This type of design compromises on decentralization because the hardware requirements to become a miner and produce blocks is very expensive. Therefore, the mining process is dominated by a few players which means they have major power over the network. This removes the trustless and censorship-resistant elements of blockchains. As a random guy on twitter once said “A blockchain without decentralization is just a very shitty database”.
So Vitalik proposes an alternative design for these “Big Block Chains” which goes as follows:
1. Second tier staking & dividing the blocks into buckets – He proposes dividing one big block into 100 buckets or so essentially making multiple blocks within the main block. These buckets will run like a proof-of-stake system where validators can stake their tokens and be randomly assigned to validate specific buckets making the barrier to entry for a new validator low. The main block will only be mined if all the buckets are signed off on by 2/3 validators. But how do you keep the validators in check.
2. Using rollup technology but in a different scenario – Use fraud proofs (optimistic rollups) or ZK SNARKS (ZK rollups) to ensure that every bucket within the main block is correctly validated. This adds an additional layer of security over the random assigning of stakers.
3. Introducing Data Availability Sampling (DAS) – This allows more users to cheaply verify the validity of the blocks that have been published.
4. Add secondary transaction channels – this is the process of allowing the Stakers/bucket validators to make a list of transactions to be included in the next main block.
Essentially, he is proposing that rather than decentralizing the mining process, you can decentralize the validation process within the blocks so that even though the mining process is centralized, they cannot easily collude and attack the entire network. As he states “the worst they can do is all go offline entirely”. ‘they’ is referring to the miners. This means that the network will temporarily have no miners and no transactions will be going through until the community pools resources to establish a new honest miner/node.
Second Proposal
He then moves on to another proposal where a rollup chain uses shard chains to yield a similar result as above. He depicts a scenario where there is one rollup chain that is very successful in its node/miner implementation essentially allowing thousands of transactions per second. So if this highly performant rollup chain can use the increased data-availability created by the shard chain infrastructure then you may end up with a slightly similar scenario as above. Since it is one highly performant rollup chain, the block producing side is highly centralized but the validation process is decentralized as it is spread amongst the shard chains and uses rollup technology.
Third Proposal
This one is a lot more technical than the previous ones. He talks about a multi-rollup future. Basically, each rollup chain cannot hold anywhere near the amount of activity that takes place on the main chain, so you are likely to have a future where you have multiple rollup chains which are built on top of the base layer which provides data-availability and security. This is a cosmos-like architecture but replace the zones with rollup chains and the hubs with base layer. So in this scenario, you theoretically have everything. Decentralization of the mining process, decentralization of the validation process, censorship resistance, and a trustless environment.
The decentralization of mining aspect however has the potential to diminish over time. It is unlikely, but still possible. This happens because of something Vitalik calls “Cross-domain MEV”. This is a very vast and complex topic but for the purpose of this article I think the main takeaway is that it essentially means miners can work together across the different rollup chains through many different complex strategies. The incentive to do so is high as they will control the block producing market of a highly efficient network. So what can be done to combat this. Vitalik proposes the following:
· Implementing a mechanism of auctioning off block production – It makes sure that when there are centralization tendencies, the entirety of the block production process isn’t taken up by the elite
· Rollups implement a ‘Censorship-resistant Bypass’ – Basically, if the network does end up becoming highly centralized and if those in control try to censor it, then there should be a mechanism to bypass them which is already in-built.
So if you noticed, the third proposed design also leads to the same result. Block production centralized, block validation decentralized, and censorship resistance is maintained.
Vitaliks concluding remarks
Towards the end of the article, Vitalik basically points out that there are many different design choices that can be taken but they all end up with a similar result down the line. He says, that if this future is inevitable then block producing will become a centralized and highly specialized market where only a few players can be successful. Despite all the attempts made to maintain decentralization of the block producing process, there are too many bottlenecks that prevent its success, most notably scalability or the inevitable centralization in the future.
He then talks about how Ethereum is pretty well positioned to embrace this future. If you take a look at their rollup centric roadmap, then you may get a clearer picture of how they are well positioned. Then to end the article he says that there are blockchains that exist today which come under the banner of ‘big block chains’ and even though they are highly centralized now, they have a clear path to achieve some form of decentralization, trustlessness, and censorship resistance. In a multi-chain future, Vitalik sees the world having multiple chains that are at a high level very similar in design.
I hope you enjoyed this article. For more content please subscribe to the substack and follow the medium (ITS FREE)
Follow me on Twitter @LeftsideEmiri
Thank you for reading.