Blockchain 3.0 — Innovations in The Distributed Ledger Technology Part 3
In part 3 of our series, we will explore Hashgraph as the newest type of DLT. We will also draw some conclusions regarding the current standing of distributed ledger technology and address both the concerns and opportunities related to the new technologies.
Hashgraph — The Newest DLT
Besides the Tangle, the term “Hashgraph” is also making quite a stir on the market. This newly developed technology also fits into the category of distributed ledger technologies (DLT). The idea for Hashgraph was developed by Leemon Baird in 2016 and was originally intended for the private corporate sector. The intellectual property in Hashgraph is owned by Swirlds, a company founded by Baird. Swirlds distributes a software development program that enables anyone to test and experiment with the “Hashgraph Consensus Library.” With CULedger, a consortium of 6,000 cooperative banks in North America, Hashgraph has already found a potential customer who uses their private Hashgraph software and has even found it to be more reliable than other alternatives such as Hyperledger.
Due to this success in the corporate sector, Swirlds has now launched the “Hedera Hashgraph Platform” with aims to drive forward Swirlds’ patented Hashgraph technology for the development of a public Hashgraph network.
While the source code of the Hedera Hashgraph is publicly accessible and anyone can join the Hedera Hashgraph ecosystem as a network node, the project will still have a centralized governance model similar to that of Visa. This means that there will be 39 organizations that will constitute a kind of leadership council. The exact terms are currently being finalized and the 39 members will be announced. Due to this centralized structure with a management body, splitting the source code to make an alternative project using a hard-fork will not be possible.
As with the Tangle, the Hashgraph concept is no longer based on blocks that are chronologically put together to form a chain. Instead, we have so-called events, which are hashed to each other — hence the name “Hashgraph.” The following information is contained in these “events” — a timestamp, two separate parent hashes, and one or more transactions.
While in a blockchain the winning node can add the new block with transactions to the existing chain, in the Hashgraph all nodes within the whole network inform each other about the current status and “exchange” their information with each other. Similar to a Tangle, a connection diagram of “events” or transactions is made, and transactions are put together in a chronological time sequence. The transaction history enables a consensus on the progression of individual transactions.
With the Hashgraph concept, the key information within the network is likewise transferred via the so-called Gossip protocol, a communication protocol. To distribute information within a network, the Gossip protocol is considered the fastest and most efficient method of communicating between different computers. Each computer passes the received information forward, to a randomly selected machine. This leads to an exponential distribution of information throughout the network.
However, the mere propagation of information within the network is not enough to achieve a consensus on the shared information. For this purpose, each network participant needs to know the precise transaction history and the exact sequence of transactions, which is guaranteed by the timestamps. So, every machine within the network shares all its knowledge about which network accounts addressed to what, to whom and when. This circumstance allows so-called “virtual voting” because all nodes in the network have a copy of the transaction history and information about who received the information at which point in time, every participant can determine how others will behave. Hence, each node knows the decision of the other, without an active decision, i.e., a “vote”, having been made. Based on this “voting without voting” there is a consensus among the network participants.
As described at the beginning of this blog series, innovative approaches such as the Tangle or Hashgraph are seen as the next generation of DLT technology. In the case of a DAG, there is no global network state, since a DAG (Tangle and Hashgraph) has no blocks and is based to some extent on the principle of regional consensus. This indicates that network participants no longer store all transactions, but only “local” data and rely on “other regions” to do the same carefully. This could be a weaker defense against double spending than the blockchain.
There are also fears that the Tangle and Hashgraph could assume a considerable data size due to their scalability and that this will lead to centralization among those network nodes which keep the network running. IOTA and Hedera Hashgraph seem to have a solution for this problem: they announced to shorten the Tangle or Hashgraph regularly. Although that solves a problem, there are some things we need to see first. The projects must first achieve the promised scaling. Although both the Tangle and Hashgraph technologies appear promising, they have yet to provide the practical proof for what the claim.