Ethereum after the Merge: Rollups, PoW Forks, and Scaling Doubts

Ethereum reinvents itself among rollups, PoW forks, and doubts

Ethereum changed its consensus engine and still hasn't solved what hurts users most: fees. The Merge arrived, proof of work was shut down, and the network switched to staking validation, but the cost of moving a token barely budged. The original plan was more ambitious: turning Ethereum 1 —a Layer 1 chain with mining— into something resembling a Layer 2, adding the beacon chain, sharding, new virtual machines, and proof of stake. The calendar shifted so many times that the roadmap became a recurring joke within the ecosystem.

What is Ethereum 2 and why was the Merge delayed?

Ethereum 2 is a phased reconstruction: the beacon chain provides consensus, sharding distributes the load, and rollups execute transactions off the main chain before returning them. On paper, Layer 1 stops being where everything happens and becomes a coordinator. The problem is that each phase was announced with a date, and every date was moved.

The final test before the mainnet was Goerli/Prater: Bellatrix on August 4 and the Merge between September 6 and 12. The definitive jump was set for September 15, in a week packed with macro references: the CPI data on the 13th and the Federal Reserve meeting on the 20th and 21st. The coincidence fueled all sorts of readings about whales releasing ethereum to rescue the Nasdaq, and whether that calendar responded to something beyond engineering.

Among developers, the feeling remained of 'changing the engine while flying the plane.' Vitalik Buterin simply wished luck to those defending a proof-of-work fork. Little enthusiasm and considerable haste to close the matter.

Layer 2s that already work without waiting for Ethereum

Polygon (MATIC) is the most advanced solution and is not an independent network: it is an Ethereum 1 sidechain that offers today what Ethereum 2 promises for tomorrow. Behind it are Optimism, an optimistic rollup, Arbitrum, zkSync, and a growing list. Ethereum itself releases its code so rollups proliferate: the more, the better for scaling.

The strategy map is divided into three. Sidechains are the old way. Plasmas and channels never worked and do not support smart contracts. Rollups, optimistic or zero-knowledge, are the new path, and several are already in production.

The most repeated argument against these alternatives is not technical, it is market-driven. A contract deployed on a network with few people only aspires to those few people. Without interoperability, this current holds, any brilliant chain is eclipsed by Ethereum, which is 'the Microsoft of crypto': the place where you must be to promote yourself, not necessarily where you grow.

What remains of Ethereum's PoW forks?

After the Merge, three proof-of-work consensus successor chains theoretically remained: Ethereum Classic (ETC), Ethereum PoW (ETHW), and Ethereum Fair (ETHF). The advantage over previous forks is that applications already exist on Ethereum, and copying them is a cut-and-paste exercise.

The question is whether that is enough. Many platforms announced they would migrate to proof of stake and not support the fork, and there is talk of freezing smart contracts, leaving associated tokens worthless. This is compounded by the risk of spoofing: fake stablecoins from the PoW fork are a perfect lure.

Technical pedagogy also has a backlog. Obtaining fork tokens requires connecting to the new network via RPC and doing so from a wallet with no funds, because theft attempts are frequent. Spoiler: not everyone bothered to explain it.

MEV and the four entities producing blocks

The next front is MEV. Whoever produces a block can see a pending transaction, jump ahead, and keep the profit. The response has been Flashbots, which hide transactions from producers, but with an uncomfortable side effect: the network centralizes. In one specific week, 80% of new blocks came from four entities, and Flashbots comply with OFAC sanctions.

The proposal to break this bottleneck is to move block creation to a separate chain. This is the step Buterin's roadmap calls the Scourge. Meanwhile, ConsenSys and 18 other firms signed up for Egypt's COP27 with a climate platform. Image washing or conviction, the gesture exists.

Shapella: staking is released and price rises

Shanghai, with the Shandong testnet as a rehearsal, arrived in 2023 with the most anticipated proposal: EIP 4895, allowing withdrawal of ether deposited in the beacon chain. Logic suggested a wave of sales. The opposite occurred: the price went to $2,100 per unit, and the dominant reading was one of confidence.

The side effect almost no one highlighted is decentralization. If validation can be done on a Linux computer, without power requirements, the entry barrier disappears, and the node ceases to be the domain of large exchanges and centralized services. Some argue the opposite: that in proof of stake, the network has no intrinsic value and the price could fall below $100. A year after the consensus change, the balance was a downtrend and growing distance from Bitcoin.

Not all institutional actors accompanied. Paxful announced the withdrawal of ether from its market on December 22, with a message from its CEO appealing to 'economic apartheid' and Bitcoin's mission. The argument is debatable; the decision, a fact.

What to watch from now on

The prediction, with reservations: if Ethereum wants to be the settlement layer for everything, its problem is no longer technical, it is distribution. Rollups will grow because Ethereum needs them, not because the user notices the improvement. And if block creation remains in four hands and Layer 2s number in the dozens, decentralization will remain a slogan. That said, no one has yet found the chain that takes its place.