Ethereum Transition to ETH2 Gets a Formal Improvement Proposal

Ethereum developers have long been planning to transition the blockchain to a more energy-efficient, secure, and scalable network.

They now have an official Ethereum Improvement Proposal to work with.

ConsenSys researcher Mikhail Kalinin has created a pull request for EIP-3675, which would upgrade the blockchain's consensus mechanism to proof of stake, a system that allows people to validate transactions and create new blocks based on how much ETH they contribute. The current proof-of-work system instead relies on "miners," who compete to be the first to validate transactions.

The proposal, which sets the stage for "the merge" to Ethereum 2.0, will be discussed at Friday's core developers meeting. 

But, contrary to some speculation, that doesn't mean ETH2 is necessarily happening this year. Tim Beiko, an Ethereum Foundation dev who coordinates upgrades, told Decrypt that a 2021 merge is unlikely in his opinion. "Everything would have to go perfectly," he said.

The intent of Ethereum 2.0 is to allow more—and speedier—transactions on the network. The blockchain has often struggled with congestion, as it's not able to process any more than about 15 transactions per second. With decentralized finance (DeFi) applications—which replace financial intermediaries with blockchain-based code—tapping into Ethereum's network, every buy and sell order adds strain. As a result, transaction fees exploded this year, sometimes reaching triple-digit dollar amounts for single-digit transactions.

Ethereum's method for achieving a more functional network is to get rid of miners, the people who run software that process every transaction on the blockchain and create new blocks. Instead of miners eating up electricity in an effort to solve cryptographic problems and get new blocks, "validators" who lock up at least 32 ETH in the new proof-of-stake system will be selected using a pseudorandom algorithm. 

Danny Ryan, who's been coordinating ETH2 efforts, told Decrypt EIP-3675 is just one of many items to check off the list. "We are ensuring that core specs are in place so that post London and Altair, the engineering process can move swiftly," he said.

London and Altair refer to the next two upgrades to the current proof-of-work network. London, scheduled for around August 4, features the long-awaited EIP-1559, which will reduce the supply of ETH. Altair, slated for sometime before October, tweaks the Ethereum 2.0 network, which is currently running but not fully functional.

Whether or not the draft of EIP-3675 quickly gets finalized, developers are running up against a hard deadline. That's because of what's known as the Ethereum "difficulty bomb," code integrated into blockchain in 2015 that increases the difficulty level of mining new blocks. The intent was to put some pressure on developers to get ETH2 done quickly.

EIP-3554, a code change that will be implemented with the London upgrade, gives devs some breathing room, pushing the bomb's detonation back from this summer to this December. If the merge isn't ready by then, devs will still have to upgrade the network. 

And it likely won't be. Beiko wrote last week that being able to merge in December would happen only "in the most optimistic of scenarios."

As Ryan put it, there's still "plenty to do."