On May 17th, 2018, Token Summit III was held in New York City. The event featured an array of speakers and presenters discussing the current state of the token-based economy and its future. However, one presenter stole the show with the announcement of a new blockchain advancement.
Cornell professor Emin Gun Sirer announced new consensus protocols for blockchain networks that involve the combination of, what Sirer called “classic consensus” and “Nakamoto consensus” for blockchain security.
Sirer said, “They rely on randomness, and they rely on random interactions, and yet they ensure after the interactions everyone has decided the same thing.”
The new model was released in a whitepaper entitled, Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Cryptocurrencies, penned under a pseudonymous group going by the name of “Team Rocket” after the beloved Pokémon characters. The paper outlines three protocols: Snowflake, Snowball, and Avalanche. The whitepaper describes aspects of the new model relying on random sampling to determine consensus:
“Inspired by gossip algorithms, this new loved ones gains its basic safety via a deliberately metastable mechanism. Specifically, the system operates by frequently sampling the community at random and steering the proper nodes toward the same result.”
As previously stated, this model is a result of the combination of two already existing models. Nakamoto consensus requires miners to agree on transactions within blocks before they can be added to the blockchain, while classical consensus relies on a model of a two-thirds plus one majority for consensus.
Along with being a computer science professor at Cornell, Sirer is co-director at The Initiative For CryptoCurrencies & Contracts and also has several patents to his name, including a method for implementing a virtual machine on distributed service components.
Although the paper is believed to describe a “simple, yet powerful” consensus model, not everyone was impressed with the new methodology, especially Ethereum developer Vlad Zamfir who has been hard at work on a the proof-of-stake consensus model known as Casper. Zamfir took to Twitter to voice his displeasure of the announcement, calling the model “the worst of both worlds, because “it’s not asynchronously safe and it’s probabilistic,” which is not the way in which a blockchain consensus model should operate, in his opinion. According to Zamfir, “We never get to take a probabilistic design of the community for granted.”
Only time will tell if Snowflake, Snowball, and Avalanche will be effective models of blockchain consensus for the industry.
Dan is a freelance cryptocurrency and blockchain content writer. He has written content for startups, ICOs, financial planners, venture capital firms, and more. Previously he founded an e-commerce company that grew to $1 million in revenue and profitability in less than 3 years. Dan has a degree in Economics and Finance from Bentley University.