By The Numbers: The Byzantium Hard Fork

As of publication, we are more than 14,000 blocks deep into the Byzantium hard fork, which went live on the Ethereum mainnet on October 15, 2017. As announced, the block reward has adjusted down from 5 to 3 Ether.

The real concern is that a significant percentage of nodes have not upgraded their clients to versions that are compatible with Byzantium. Although 59.1 percent of nodes running Geth have upgraded to v1.7.2 or later, just 38.4 percent of nodes running Parity have upgraded to v1.7.6 or later.

While Geth v1.7.1 is compatible with the Byzantium hard fork, it might be subject to a Denial of Service (DoS) vulnerability, which could cause the nodes to drop offline.

Just three days after the Byzantium hard fork, it’s too soon to tell if other Ethereum forks could gain any traction. Prior to the hard fork, the price of Ether was hovering around $340, perhaps in anticipation of a bull run. As of publication, the price of Ether is now $305, more a reflection of market uncertainty than any indictment of Byzantium.

Shortly before the fork, Parity Technologies, led by Gavin Wood, expressed concern. In a quality assurance technique for software called “fuzzing,” bugs were reportedly being discovered in Byzantium shortly before its release.

Parity proposed delaying the fork until a period of two weeks had passed with no discovered bugs, a proposal that was rejected by other developers including Hudson Jameson.

One of the most notable developments is that average block time has fallen substantially – due to an adjustment to the difficulty formula. In the early part of October, the average block time was approximately 29 seconds. Post-Byzantium, as of October 17, 2017, the average block time is approximately 14 seconds. According to Etherscan, on Tuesday, the Ethereum network also achieved an all-time high of 505,611 transactions.

The reduction in block fullness corresponds to the reduced block time. As blocks are mined more quickly, the blocks themselves are smaller (fewer transactions per block). See the chart below which illustrates the reduction in the average block size.

Undoubtedly, there will be more developmental challenges ahead, and many questions remain about Proof-of-Stake. But, at least in the early going, part 1 of Metropolis endures.