Ethereum upgrade travel map: a review of upgrade history and future planning

Author: Panzhi Xiong, research director smell chain

Details of the 8 upgrade history and future plans of Ethereum, and the impact of each upgrade.

Ethereum has undergone eight network upgrades in history. When we try to map the height of each forked block of history to a linear timeline, we will find that the frequency of the Ethereum network upgrade is gradually decreasing, and it has not even been upgraded throughout the year. This article does not attempt to explore why the frequency of Ethereum upgrades is decreasing, but the most intuitive guess may be because the Ethereum Foundation needs to focus some of its engineers on the development of 2.0, or because the platform mature It is not necessary to perform high-frequency forks. After all, this is very troublesome for the nodes running the network, and the client needs to be upgraded.

The latest hard fork upgrade was completed around 8 o'clock on December 8th, Beijing time. The upgrade code was " Istanbul " and 6 EIP proposals were activated to upgrade the network. Ethereum co-founder Vitalik Buterin said that after this upgrade, ordinary transactions per second (TPS) will increase by about 5% to 10%, and for Layer 2 technology Rollup, it can increase about 4 times.

As we all know, the entire project of Ethereum is gradually promoted in four stages, divided into: Frontier, Homestead, Metropolis and Serenity . The first three stages can be called Ethereum 1.0, while Tranquility can be called Ethereum 2.0. Since 2.0 is a very large project that spans several years, 1.0 using the proof of work mechanism and 2.0 upgraded to proof of equity are likely to coexist and evolve separately over a long period of time.

The next hard fork upgrade is codenamed " Berlin ". It is planned to introduce ProgPoW mining algorithm, which will weaken the influence of the dedicated mining equipment ASIC on the network. Since 1.X and 2.0 may evolve in parallel for several years, the fairness of mining algorithms is still a topic worth exploring for 1.X networks. Of course, with the gradual completion of 2.0, the "difficulty bomb" originally implanted in the network will continue to play a role. When Ethereum enters the "Ice Age" and it is difficult for 1.X miners to generate new blocks, users will have to switch to The "quietness" of 2.0.

Of course, the core technology of 2.0 is the sharding technology , which can greatly improve the efficiency of the distributed system of Ethereum. However, some other important technologies are also arranged in 2.0, which can further improve network performance, such as eWASM. So this kind of technology may also test some features on the 1.X chain before 2.0 is fully launched.

This article has sorted out the roadmap page in the official Ethereum document, and tried to show the impact of each upgrade in history. Of course, since the most important significance of most upgrades is to improve the basic technical capabilities or performance of the network, the content of this part will not be expanded in detail.

Some things to watch:

  1. Ethereum did not implant a "difficulty bomb" when it launched on the mainnet;
  2. After the difficulty bomb was implanted, two hard forks to delay the difficulty bomb have been performed, and the third hard fork to delay the difficulty bomb has occurred as soon as this month;
  3. Ethereum has not performed a network upgrade in 2018;
  4. After Istanbul, the upgrade codes of Ethereum will be named in the order of Devcon;
  5. The future development cycle will be centered on EIP, and only when EIP is mature enough will it consider merging into the upgrade.

Frontier: Enter the first stage

On July 30, 2015, this is the date when the Ethereum genesis block was generated and the date when the first phase of Ethereum started.

Ethereum "frontier" is a decentralized software platform

Frontier Thawing: implanting difficulty bombs

On September 7, 2015, after this upgrade (that is, a block height of 200,000), a difficulty adjustment mechanism will be introduced, also known as a "difficulty bomb". The difficulty of computing power in the network will be experienced in about a year. Exponential growth, so it is guaranteed that Ethereum will once again undergo a hard fork within the next year.

The purpose of introducing the difficulty bomb is to provide the network with a mechanism to migrate from PoW to PoS. When the computing power is too high and miners cannot dig out any block, it will be the best time for the network to convert to PoS.

Homestead: Entering the second stage, there is finally a graphical interface wallet

March 14, 2016 (π day), this is the first hard fork of Ethereum and the date of the second phase of the roadmap. The most important feature in this version is the optimization of smart contracts and the introduction of new code for the smart contract language Solidity. In addition, a desktop wallet Mist has been released in this version, and users can use Mist to hold assets or use smart contracts. The Mist project was subsequently announced for termination in early 2019.

Ethereum "home" is a blockchain application platform

DAO Fork: Birth of ETC and ETH

This is why Ethereum has two networks. In 2016, a decentralized autonomous organization project called The DAO raised $ 150 million through token issuance. In June, The DAO's contract was used by hackers, and ETH worth tens of millions of dollars was stolen by unknown hackers. There are still many details about the subsequent development of this incident. In short, most community members believe that a hard fork should be carried out to recover stolen funds from hackers. So on July 20, 2016, a hard fork called "DAO Fork" was performed. Many community members do not recognize this behavior. They believe that decentralized blockchain platforms should not perform any human intervention, and transactions that have occurred cannot be cancelled. The divisive communities produced two networks after the hard fork, Ethereum Classic, where hackers still have funds, and Ethereum, where stolen funds were returned to investors.

Tangerine Whistle: Mitigating DoS Attacks

The Ethereum network has been attacked by individuals or teams since September 18, 2016. The network is blocked by a large number of outstanding transactions, similar to a denial of service attack (DoS). So the Ethereum Foundation plans to conduct two hard forks to completely resolve such attacks. The first hard fork will focus on addressing the most urgent network health issues, which are low-priced opcodes. The second hard fork will focus on the issue of secondary priority. For example, an attacker using a large number of empty accounts for flood attacks may increase the size of the unnecessary blockchain ledger.

The first hard fork on October 18, 2016, increased the cost of the Ethereum Virtual Machine (EVM) opcode, which means that the cost of performing such calculations on the chain will be improve. This type of opcode is the basis for preventing spam attacks. This upgrade is to adjust these to mitigate the impact of such attacks.

Spurious Dragon: Complete attack

After a lapse of one month, the second hard fork was carried out on November 22, 2016, hoping to completely resolve a series of attacks that began in September. Four proposals will be jointly addressed, namely: increase protection against heavy attacks, increase the cost of EXP opcodes, clean up the state tree, and adjust the maximum byte limit for smart contracts.

Byzantium: Entering the third stage & delaying the difficulty bomb for the first time

On October 16, 2017, this hard fork included 9 improvement proposals. In addition to updates related to the underlying code such as opcodes and smart contracts, the “difficulty bomb” will be postponed to one and a half years later, and the block reward will be reduced from 5 ETH to 3 ETH. Before removing the difficulty bomb, the block generation time was close to 30 seconds.

Territory of the Byzantine Empire during the Paleolio Dynasty, around 1265 AD

Constantinople: Double Hard Fork & Delayed Difficulty Bomb Again

After more than a year of silence, the Ethereum Foundation will perform the last step of the "Metropolis" on February 28, 2019, which is a hard fork called "Constantinburg", which contains 6 improvements . Interestingly, a hard fork called "Petersburg" will be carried out simultaneously in this upgrade, removing a previous improvement proposal (EIP-1283) in Constantinople.

In this 5 major updates, in addition to technology-related adjustments, the difficulty bomb is postponed again for 12 months, and the block reward is reduced from 3 ETH to 2 ETH.

Ottoman Constantinople, 1572

Istanbul: Performance Optimization

Ethereum 2.0 is planned to start the first phase (phase 0) in 2020. Before 2.0 is fully launched, most users and developers may mainly use Ethereum 1.X, so subsequent 1.X updates are also very important. . The Istanbul Hard Fork was activated on December 8th, and contains 6 improvements, performance improvements and opcodes for optimized parts. Ethereum co-founder Vitalik Buterin said that after this upgrade, ordinary transactions per second (TPS) will increase by about 5% to 10%, and for Layer 2 technology Rollup, it can increase about 4 times.

Panoramic view of Istanbul from the junction of the Bosphorus and the Marmara Sea

Muir Glacier: Delaying the difficulty bomb again

Users and developers continue to find that the block generation interval of Ethereum has been slightly improved recently, which will cause the TPS of the Ethereum network to decrease. The Ethereum developers discussed and someone proposed a hard fork code-named "Murr Glacier" at block height 9,200,000 to remove the difficulty bomb, which is expected to happen around December 31, 2019. According to the developer's data in the discussion group, the block generation time of Ethereum will continue to increase until the difficulty bomb is removed, and may reach 25 to 30 seconds by January 6.

Muir Glacier by Frank La Roche, Photography Records circa 1897

Berlin: Begins named after Devcon's city

Starting from this version, the upgrade code will follow the order of Devcon at the Ethereum Developers Conference, and the first Devcon 0 was held in Berlin. In fact, many improvements were planned for "Istanbul" before, but for various reasons, the improvement proposals that did not have enough time to be placed in Istanbul will be moved to "Berlin". There are 8 EIPs that have been temporarily accepted.

First (Devcon 0) in Berlin with Vitalik Buterin and Gavin Wood in the middle

London, Shanghai

Currently in the official documentation, the two upgrade codes are mentioned, and no other information has been discussed. However, the official mentioned that the future development cycle will be centered on EIP. Only when EIP is mature enough will it consider merging into the upgrade. And will give priority to "on time" and "small scale", rather than "deferred" and "large scale".

The third (Devcon 2) in Shanghai