Harmony (ONE) is an open infrastructure platform for scaling consensus and is the latest company to carry out their IEO on the Binance Launchpad system. Today, they opened for public trading on the main Binance exchange platform.
As IEOs are very similar to ICOs in several regards, it is important for potential investors to do their own research and understand projects before investing in them.
The Harmony Platform
Harmony is a public blockchain network designed to scale as the infrastructure for ‘the decentralized economies of the future.’ Focusing on high-throughput and low-latency, Harmony deploys several innovative, and experimental, technologies when applied to a public blockchain.
Harmony is part of an emerging class of next-generation blockchain networks focusing on scalability, so the best method to evaluate the network separate from the suite of new projects on the horizon is where it applies unique technology. Overall, we can break down Harmony’s primary technical components into the following:
- PoS Consensus Using BLS Signature Scheme
- Adaptive State Sharding
- Networking and Distribution
PoS Consensus and BLS Signature Algorithm
Similar to most of the platforms striving to improve on Bitcoin’s on-chain scalability, Harmony is a proof-of-stake (PoS) consensus cryptocurrency network that relies on a linearly scalable BFT algorithm. Harmony calls their consensus ‘Fast Byzantine Fault Tolerant’ (FBFT) as opposed to the conventional ‘Practical Byzantine Fault Tolerance’ algorithm.
While PoS consensus mechanisms are widely known and understood at this point, Harmony throws a unique twist into their mechanism. Drawing inspiration from ByzCoin’s BFT algorithm that uses Schnorr multi-signatures, Harmony uses BLS signatures, another multi-signature protocol for constant-sized signature aggregation.
Using BLS actually makes Harmony’s FBFT faster than a traditional BFT algorithm, which they quote as being 100 times faster.
Adaptive State Sharding
Akin to a similar on-chain scaling platform, Zilliqa, Harmony uses sharding to help the network scale at the protocol layer. In particular, Harmony shards network and transaction validation — like Zilliqa — but goes a step further into uncharted territory with state sharding.
State sharding is considered a highly sophisticated field that is still experimental when applied to a distributed network like a public blockchain. Zilliqa has opted not to deploy state sharding at this point due to the immaturity of state sharding and its unproven track record on a public blockchain network. To understand its relevance to Harmony, the team uses Zilliqa’s approach as a pillar to reference the difference between transaction, network, and state sharding.
Essentially, network and transaction sharding enable the on-chain processing of transactions to be performed in subsets of nodes in the networks, called shards, which are then consolidated into the network’s broader PoS consensus. Processing occurs in parallel, enabling the network’s TPS to surge far beyond the limited nature of proof-of-work networks. However, in such a system without state sharding, all the nodes in the network hold the blockchain’s entire state to ensure secure and authentic validation of transactions — as with Zilliqa.
Harmony actually splits the network’s state into subsets of shards, where groups of shards only hold a portion of the state and not the entire state.
Harmony achieves this using a ‘Beacon Chain,’ which is similar to what is employed in DFINITY, and uses a verifiable random function (VRF) to serve as the source of randomness in the PoS validation mechanism.
One of the key obstacles in sharded blockchains is the issues of reconciling cross-shard communication. Essentially, this encompasses the concept of how shards can send messages between each other without sacrificing the security or integrity of the validation/state of the network. Harmony uses a shard-driven model, which means that each node broadcasts messages to the network independently, and Harmony uses the Kademlia routing protocol to make communication between shards more straightforward.
It is also important to note that Harmony uses an account-based transaction model like Ethereum, rather than the UTXO design employed by Bitcoin. Each shard chain has its own account state, which makes the overall design less complex than if a UTXO system was put in place.
Networking and Distribution
Cross-shard communication is complicated and largely viewed as one of the major barriers to a securely sharded network. Issues such as the ‘train-and-hotel’ problem are well-established and reveal the underlying problem of atomic database transactions.
Kademlia is the routing protocol that Harmony uses for cross-shard communication. Kademlia is a distributed hash table invented in 2002 and was designed to improve the exchange of information via an overlay network for node lookups. The consideration for using Kademlia in Harmony is based on the fact that Kademlia does not require as much overhead as standard gossip protocols, and messages only travel and explicit distance before reaching the destination node — making the communication less cumbersome.
Leaner communication can lead to faster consensus and reduced networking load for nodes in the network — supplementing the decentralization that Harmony strives for.
Part of the core concepts Harmony is going for is reconciling the ostensible mutual exclusivity of scaling and decentralization. One often comes at the expense of the other. Harmony uses a suite of innovative technologies to try and achieve both, but time is the ultimate arbiter of success in the technological space.
A Crowded Field of Scalability
Following the recent launch of Cosmos and the looming inauguration of Cardano, the competition for on-chain scalable networks is ramping up. Add in Ethereum’s ongoing transition to Serenity, and it seems PoS, sharding, and new routing protocols are the popular trend among development teams for on-chain scaling.
However, concerns over the long-term security and decentralization of PoS persist, and it is unclear how these networks will overcome some of the established predicaments regarding smart contracts (i.e., the Oracle problem) and blockchains. Quite simply, scalability is not the limiting characteristic that is precluding the mainstream from using blockchains. Rather, it is that centralized systems operate at scale better, have better UX/UI, and are more familiar to the mainstream.
Until public smart contracts platforms can furnish a value proposition that is luring enough to encourage a widespread transition to decentralized networks instead of centralized alternatives, their user numbers will continue to be lackluster.
Conclusion
Harmony’s team comes from a background in major tech companies, including Google, Apple, Microsoft, and Amazon. While impressive, that is not a guarantee of sustainable success in the rapidly evolving crypto sphere. They have developed an intriguing confluence of technical advances, and the launch of the IEO on Binance is sure to attract attention.
As an investor, it is important to realize the differences between upcoming projects, and especially their core value proposition. Many projects in 2017 promised ‘scaling to billions’ of users for the ‘future economy,’ but the reality has tarried far behind that.
Considering the meteoric rise and fall of ICOs and altcoins, it is prudent to take the idealistic promises of upcoming blockchain networks and cryptocurrencies with a grain of salt.
