The Avalanche platform
How a team of computer science PhDs is translating an “anonymous” whitepaper into a real-life platform.
Decentralized systems are becoming a common concept today. Decentralized applications will become a commodity in the future. But the non technical enthusiasts struggle to see the remaining obstacles to conquer a (much) larger user base. Actually even some technical enthusiasts get sometimes overwhelmed by their passion and forget there are still many questions to answer.
And as experts tackle the issues yet to be resolved, many frameworks appear in the world of open distributed systems. One of the most adopted is the tryptic Scalability – Decentralization – Security. But even in this tryptic experts do not agree on definitions : How to define Scalabilty? Is it TPS? Finality? What is decentralization? Is it the number of users? Participating nodes? What is security? Is it industrial grade functional languages? Good verification frameworks?
With the emergence of many Decentralized Finance projects, it becomes clear that a network should be able to handle a high number of transactions (TPS), to quickly validate the transactions (Finality) and to scale participation to a high number of nodes. In addition to these practical considerations, a network should be as green as possible in order to gain long term adoption.
To make things easier to grasp, let’s attempt to start with the objective of an open distributed system or platform (be it a blockchain or a DAG): A number of machines (or nodes) working together to maintain a network and a ledger. The assumptions in a permissionless system are that anyone can participate in the network, and that these machines do not trust each other. But there needs to be trust in the result of their collaboration. So a collaboration mechanism is imperative to ensure this global trust. There are important concepts to verify this output :
- A consensus protocol : That is a protocol (Algorithm, architecture and properties) allowing the machines to agree on the state of the network and ledger at each instant.
- Anti-attacks mechanisms : That is the characteristics of the network minimizing the risks of “double-spend” or any other hackers’ attack.
For example, here are some concepts that could be misunderstood – Taken from this excellent series of articles on consensus protocols :
Proof-of-Work (PoW) is not a Consensus Protocol. It is a strategic element that can be added to certain types of Consensus Protocols, which otherwise would be unreliable or even fully dysfunctional, in order to make them work by slowing them down, and reducing their throughput. PoW can also function as a defence mechanism against Sybil attacks. For example, in Bitcoin, PoW is used in Nakamoto Consensus Protocol, which is based on the rule of the Longest Chain.
Proof-of-Stake (PoS) is not a Consensus Protocol. Just like in the case of PoW, PoS is a method for reducing the number of nodes that are allowed to partake in adding, or voting for, the new block on the blockchain, or another kind of record or transaction onto the network-common archive of any type. PoS can also reduce the risk of Sybil attacks.
Now let’s go back to the history of consensus protocols :
- In the 70s : Research began to find ways to solve the consensus problem in shared networks.
- In the 80s & 90s : We saw the emergence of some well known protocols : Quorum-Based; Paxos/Raÿt, PBFT. Led by the work of Leslie Lamport and Barbara Liskov, two turing award winners.
- In 2008 : Nakamoto’s breakthrough protocol, based on the longest chain and proof of work.
- In 2018 : The Snow family paper was dropped by “team rocket” on IPFS, based on epidemic protocols and gossip networks. The team was in contact of prof. Emin Gun Sirer from Cornell University. Notwithstanding the debate on the identity of the writers of the paper, it seems it offers a brand new way of achieving consensus.
Then, the Avalanche platform started taking form :
- In 2019 : The team revealed Avalabs, an entity dedicated to developping a platform based on the protocol.
- In 2020 : Avalabs successfully launched a testnet with more than 1000 validators, and closed an initial distribution of the tokens with more than 100 countries represented.
It is worth noting that the core team has years of contributions in the open distributed systems ecosystem :
- Emin Gun Sirer released a proof of work paper in 2002 before the Nakamoto paper came out. He was involved in bitcoin and Ethereum in their early days, and has advised a number of established projects.
- Ted Yin developped the Hotstuff protocol, which is the one tested with Facebook’s Libra.
- Kevin Sekniqi published research papers on stablecoins frameworks and encryption security.
- John Wu has a good knowledge of the digital asset management world, and has been involved in the finance sphere since 1992.
Interestingly, the team tried to target only one business case at the begining of their project, which was DeFi. Their ambition is to build the internet of finance, which could disrupt an industry as a whole. And they are surrounded by great minds to assist and onboard as many users and participants as possible. To make this grand vision come true.
Avalanche is an open-source platform for launching highly decentralized applications, new financial primitives, and new interoperable blockchains.
What will it allow to do?
- Launch customized blockchains, private & public : With Avalanche, one can deploy blockchains that fit any application needs. Build with their own virtual machine and dictate exactly how the blockchain should operate.
- Create and trade digital assets : Avalanche enables the launching of smart assets or digital representations of real-world-assets that obey special parameters and trading restrictions. A recent partnership presents a Decentralized exchange for NFTs!
- Build scalable smart contracts and dapps : By leveraging the Avalanche protocol, developers can launch smart contracts that confirm in a second and scale to thousands or millions of validators.
The platform has not yet been launched, the mainnet is TBA before end of august. Although the team needs to deliver and prove the system works in a real world environment, the testnet and underlying tech show good promises. For now.
To get a full picture of the team and context of the avalanche platform development :