Ethereum’s move from proof-of-work to proof-of-stake is seen as an extremely important step, as developers view it as a key enabler for several further development goals. However, merging and future partitioning upgrades come with some tradeoffs.
Even before the launch in 2015, Ethereum developers intended to replace the Proof-of-Work (PoW) consensus mechanism with an alternative mechanism: Proof-of-Stake (PoS). While it was considered too technically risky to launch the network with anything other than PoW, the eventual move to PoS was a key development goal for Ethereum developers and a long-awaited milestone on their roadmap.
The development roadmap itself was created before the launch of Ethereum and already existed when the network was still in the testnet phase. Many changes described in the evolutionary roadmap have already been implemented, but the move to PoS, one of the most difficult and costly changes, has not yet been completed and is again delayed.
Previous Proof of Stake Delays
The reasons for the delays are varied and complex, but they can more or less be attributed to the fact that the secure implementation of PoS has proven to be much more difficult than developers initially thought. Many prototypes were proposed and evaluated, but problems continued to arise, requiring several years of bug fixes and new designs to be developed.
We can even observe the time lags of migration, often referred to as melting, using blockchain data. Knowing that implementing PoS is always at hand, the Ethereum protocol comes with a hard-coded exponential increase in mining difficulty. This mechanism is known as the “difficulty bomb” and is intended to make difficulty and mining income no longer match. Thus, miners are forced to abandon the PoW chain and the alternative PoS chain remains the only viable chain.
In the image above, three different “detonations” of the difficulty bomb can be seen. They are visible both as exponential increases in block times and rapid discrepancies between hash rate and difficulty. However, since the PoS was not operational at the time of the detonation (or imminent detonation), the Ethereum devs picked up the hard bomb five times.
Although there is no specific date for the merger at the time of writing (and the merger has just been pushed back from H1 2022 to H2 2022). But there are signs that warrant cautious optimism that the PoS launch may in fact be imminent this time around.
Future sharding is also based on merging
Besides PoS (merge), the second important part of the next phase of Ethereum is the introduction of sharding. Sharding is a blockchain protocol scaling technique in which the protocol increases its throughput by dividing the blockchain into multiple blockchains (shards). This allows individual computers to choose which of the many blockchains they want to work on. Sharing can increase the overall throughput of the protocol without increasing the computational effort of the individual computers working there. In other words, Ethereum will be able to process a lot more information while relying on relatively infrequent users to provide distributed computing power using ordinary consumer computers.
In the computing world, this is called the horizontal scaling technique. Horizontal scaling is characterized by increasing throughput/performance by adding more individual computers to a network. The alternative is vertical scaling, where scaling is only achieved by increasing the throughput/power of each computer on the network.
In the world of blockchain protocols, increasing block size or increasing block frequency (decreasing the target time between blocks) are examples of vertical scaling. They require all participating computers to be very powerful (which is expensive). Conversely, sharding enables additional throughput/capacity, i.e. the ability to process many more transactions and smart contracts per second at a much lower cost by adding more participants to the network. It is assumed that they will all take care of separate fragments.
Proof of stake reintroduces the need for trust
Stopping PoW mining comes with significant trade-offs which, in turn, lead to a drastic reduction in power consumption, which also means a massive reduction in carbon emissions in our current global mining system. electricity production. In summary, with the adoption of PoS, Ethereum will lose or completely abandon censorship resistance, trust minimization, and decentralization. Additionally, the attack surface will increase significantly due to increased code complexity.
PoS reintroduces the requirement to trust other network participants when connecting to the network. The reason for this is that the stake is an internal quantity of the blockchain network. This means that you cannot know who has what stake unless you know which blockchain is the right one. This means that before a user can verify that the blockchain in front of them is working correctly, they must first trust someone to tell them what the blockchain is. This is a problem when a new user or an old user has a choice of multiple conflicting blockchains presented by a malicious actor. Since creating a PoS blockchain costs nothing, dishonest participants can mass-create fake records and present them to strangers.
With proof of work, on the other hand, the proof (computing power) is outside the system. This means that when a new user or old user encounters two conflicting blockchains, they can easily check which blockchain is correct (the one with the most accumulated processing power is correct by definition) based on accumulated work or power. treatment. In a PoS system, the only solution is to introduce checkpoints. This in turn requires being able to rely on other participants to tell you which blockchain was the right one at various times in the past. PoS therefore creates the need to trust other network participants in several new ways, which it must balance against its benefits.
Resistance to censorship and centralization
PoS also has the disadvantage of reducing censorship resistance. Censorship resistance, in this context, means the ability of the network to resist the actions of a network participant who attempts to prevent some or all transactions from being included in the transaction record. The only effective way to achieve this is to control more than 51% of block producers – miners in a PoW system, stakers in a PoS system. A company that controls the majority of block producers can simply refuse to commit any or all transactions to the blockchain, thereby censoring one or all parties.
In a PoW system, miners need to consume an external resource and also need external capital (hardware). This can be achieved without the majority miner knowing, which means there is a mechanism by which a censor can lose their place as majority miner.
In a PoS system, there is no such fallback according to the rules of the protocol. Once an entity has acquired a majority stake in the system, it will steadily increase its share of the total stake, and nothing can force it to sell any part of its stake. This means that it is impossible to disturb their position. The only way out is to resort to a “hard fork” within the framework of a social consensus, which means nothing more than centralized management by a select committee – which is by definition the opposite. of decentralization.
Sharing further reduces decentralization
The trade-off between scaling a blockchain network and keeping throughput low is that the network as a whole looks more like a client-server network than a peer-to-peer network, thus losing important benefits of decentralization. Why is that? In short: to be a full peer in a blockchain network – that is, someone who participates in the network without having to trust another network participant – a user must be able to fully verify each network event.
With multiple blockchains (or a single large blockchain) to verify, the compute and bandwidth resources needed for a full peer increase dramatically, leaving fewer and fewer users able to afford the privilege of be full peers. This leads to a reintroduction of trust. Because all users who are not able to verify all shards (or a single large blockchain) must trust other users to tell them the truth about what is happening on other shards (or the large blockchain that they can no longer verify themselves).
A high degree of decentralization is a coveted but traditionally hard-to-define feature of peer-to-peer networks. The reason for this is that a network with as many peers as possible is unstoppable due to the large number of participants, all of which must be stopped for the network to be completely cleared. Thus, the more expensive it is to be a full-fledged peer, the less decentralized a network will be in general, resulting in less security in exchange for higher scalability; also known as the blockchain trilemma.