In my last article “Crypto: Tulip Mania 2.0 or more?” I was looking at the current state of Crypto. In fact I was specifially highlighting the trade-offs between decentralization and
I was concluding that the future of Crypto is dependent on if or how we solve these issues. About 2/3 of the article I was writing about the Speed / Scale problem of today’s blockchains as I see this as the most pressing issue. I promised to write another article about alternative approaches to distributed ledgers and consensus. One of these alternatives is called Directed Acyclic Graph (DAG) and is being implemented by a couple of startups. Another one is called Hashgraph and is a patented solution developed by Swirlds.
In the following I want to elaborate in more detail how they are different from a blockchain and how they could potentially solve the throughput problem with substantially higher transactions/second.
The concept of blockchain as a standalone technology started gaining popularity in 2015. Prior to that, it was just known as a data structure underlying Bitcoin technology.
In Satoshi Nakamoto’s white paper, the two words “block” and “chain” appeared together. It was only called “a chain of blocks.”
Bitcoin’s rise into popularity resulted in it being categorized as Blockchain 1.0. With Ethereum making waves as a decentralized platform for applications that run exactly as programmed, more and more people began to categorize Ethereum as Blockchain 2.0. In the meantime we have witnessed the limitation of Ethereum’s blockchain with the hype around Cryptokitties. Now the market is battling to see who will be named Blockchain 3.0.
Some startups even claim to be building Blockchain 4.0 or 5.0, this is ridiculous marketing of course.
One should be aware of the fact that most people use the term “Blockchain” interchangeable with “Distributed Ledger Technology” (DLT) which is not really correct as a DLT does not need to be a chain of blocks.
Blockchain is just one category of DLT.
However the term Blockchain has gained such a high popularity that we are talking about the Blockchain 3.0 although it mostly likely won’t be a Blockchain.
DAG is a directed graph data structure that uses a topological ordering. The sequence can only go from earlier to later. DAG is often applied to problems related to data processing, scheduling, finding the best route in navigation, and data compression.
The Bitcoin Blockchain has always been inefficient due to the proof-of-work (POW) system. Blocks can’t be created simultaneously. The linked storage structure allows for only one chain on the whole network. All the transactions occurring around the same time are kept in the same block. Miners then compete for the block validation. One single block is created about every 10 minutes.
The first community to come up with the idea of changing the chain-like storage structure into a DAG of blocks was NXT. If the time of mining remains unchanged, the storage could be extended by X times with X blocks on the network at the same time.
The blockchain combination with DAG still comes from the idea of chains, so called side-chains. Different types of transactions are running on different chains simultaneously. DAG of blocks still relies on the concept of blocks.
IoT Chain (ITC), IOTA, Nano (former Raiblocks)and Byteball are the blockless projects currently shining in the market. With Bitcoin or Ethereum, the block creation speed is a bottleneck. Bitcoin generates a new block every 10 minutes. Ethereum is better, but it takes around 15–20 seconds for block validation.
But why do we even need a block? On the bitcoin network, many transactions are mined into blocks and the transaction sequence is maintained by the prehashes between blocks. What if you combine blocks and transactions together? Make every transaction directly involved in maintaining the sequences. After the transaction is placed, you can skip the process of mining. This makes it blockless and more efficient.
DAGs have no mining, no blocks, no transaction fees. The security and consensus of the network is not divided among miners, validators, and users. Users of the network validate a number of old transactions (via proof of work) in order to be able to conduct one of their own. No one receives a reward and no one has to pay transaction fees. A miner-centralization like in Bitcoins or in Ethereums network is, therefore, not possible.
Its field of application is set in the IoT, as the technology for data integrity and industrial appliances. Furthermore pay on demand, micro-payments, and machine to machine communication like sensor technology, smart cities, adaptive systems etc.
As a settlement layer, it aims for interoperability between many existing systems. Common blockchains cannot be adjusted to a no-fee settlement layer because the fees are integrated by design: as a monetary incentive for miners and thus: as protection for the network.
Due to the engineering of the DAG and how it is implemented, its scalability is near infinite.
In contrast to blockchain-based networks, which slow as more users are added, the DAG actually becomes faster with a growing user base. This is because, an increasing number of users results in faster validation times. The lower hashing power required to transact over the DAG network is also an added incentive for new users to use the DAG.
Hashgraph’s consensus algorithm provides a new platform for distributed consensus. Some of the attributes commonly used to refer or describe Blockchain are distributed, transparent, consensus-based, transactional and flexible. Hashgraph bears all these features. However, it is a data structure and consensus algorithm that is much faster, fairer, and more secure than a blockchain.
It uses two special techniques to achieve fast, fair and secure consensus:
Gossip about Gossip basically means attaching a small additional amount of information to this Gossip, which are two hashes containing the last two people talked to. Using this information, a Hashgraph can be built and regularly updated when more information is gossiped, on each node.
Once the Hashgraph is ready, it is easy to know what a node would vote, since we are aware of information that each node has and when they knew it. This data can thus be used as an input to the voting algorithm and to find which transactions have reached consensus quickly. The process described is called Virtual Voting.
Unlike other DLT’s, Hashgraph is proven to be fully asynchronous Byzantine Fault Tolerant (aBFT). This means it makes no assumptions about how fast messages are passed over the internet. This capability makes it resilient against DDoS attacks, botnets, and firewalls. Bitcoin for instance is not BFT, not even under bad assumptions. In Bitcoin, there is never a moment in time where you know that you have consensus.
If you would like to learn more about Hashgraph then I would suggest to watch the below video where Dr. Leemon Baird explains it in simple English and with lots of visuals.
As time goes on, we do see some problems with the blockchain. Though it still offers decentralization and immutability, a huge problem that has been coming up is the issue of transaction times due to the nature of how the blockchain works. A new contender, with new technology, has entered the scene known as DAG and is offering the solution to where blockchain went wrong.
The technology used by a DAG and a Hashgraph is fundamentally the same. There are nuances of course, but that is like describing the difference between Mac OS X Sierra vs. Mac OS X High Sierra.
The deciding factor maybe commercial. Hashgraph has proprietary, patented technology. This is the opposite of the open source approach most of the DAG solutions follow. The proprietary, patented technology approach is perfectly suited to enterprise permissioned networks. That is a big market and Hashgraph is really an alternative to solutions such as Ripple or R3 or Hyperledger. One can imagine a big inter-enterprise network such as SWIFT looking at Hashgraph DAG as an alternative to enterprise Blockchain solutions.
What really differentiates a private permissioned system from a public permissionless is the incentives built into the network.
A private permissioned system does not need an external incentive system. The node is owned by a corporation that has a business model.
If the DAG is picked up for broader use, it really does show some advantages over the blockchain at nearly every level. As problems keep popping up within the blockchain system, users may attempt to migrate to a newer more technologically advanced system such as the DAG. Will this be the way of the future for DLT? Who knows, however, it is probably unwise to bet against either at the moment. What we do know though is, that a DAG (permissionless) has not been tested in a real world evironment at scale. The old man, the Bitcoin Blockchain, is doing it’s job since 2009 with the highest reliability.