EOS is still under development, and the official version is launched, which may be different from what we know so far. However, I have led the team to successfully develop the Ethereum project. I am also very clear about the basic theory of hyperledger fabric, and I am a blockchain enthusiast. Below I'll do some comparisons of what I've learned.
Comparison and discussion of EOS, ETH and Hyperledger fabric.
Not long after Bitcoin was released, many visionaries realized that the technology behind Bitcoin had more potential than just being the cornerstone of electronic money. In fact, within a few years of Bitcoin's development, many decentralized applications have been created, all of which are based on the blockchain technology of the public ledger, which is also the technology adopted by Bitcoin. These decentralized applications include: PKI encrypted communication, decentralized exchanges (Bitshares), non-repudiation trust transactions, cloud computing (Golem), and social media (Steem/Steemit). In this new blockchain economic system One of the challenges facing investors and developers is that they need to build a new blockchain application from scratch, which is difficult and complex. And the hardest part of this — with traditional proof-of-work (POW) and proof-of-stake (POS) consensus mechanisms — is that the security of the network and its applications rely on a lot of hashing power and the widespread availability of network tokens. distribution. For small business owners or startups, this creates a relatively high bar. For a small start-up company, it is impossible to independently establish a widely distributed and powerful computer network to ensure the security of its applications, but can only access their public network, which is very congested.
Although Hyperledger fabric does not adopt the consensus of POW and POS, but adopts BTFS, kafka and Solo consensus mechanism, Fabric is a blockchain alliance chain technology based on real-name transactions. He provided us with a super ledger, providing rich experience in real-time exchange, electronic points, product source, and other scenarios, and he really improved the informatization construction within the enterprise to the informatization construction between enterprises. For the first time, discrete silos of information are brought together. Solve practical problems for alliance enterprises. Based on this, many major institutions around the world have become his members to jointly promote the development of fabric.
There is only one symbol in the Bitcoin world, and he cannot describe other meanings. This is also why blockchain technology has not been favored by the industry, especially the financial community. In order to solve these problems, smart contract platforms came into being, and the most successful one at present is Ethereum. Think of Ethereum as a decentralized platform for developing and running decentralized applications (DAPPs) on which users can run their DAPPs (decentralized applications) with confidence. At present, the market value of Ethereum is close to more than 100 billion US dollars, which proves the value of the smart contract platform, but the emergence of smart contracts has made the industry truly see its value, especially the financial world. However, smart contracts are composed of code, and there are bound to be loopholes, which lead to the incomplete function of Ethereum.
Recently, Dan Larimer (inventor of Bitshares, Graphene and Steem/Steemit) and the EOS.io team announced development plans for EOS, a blockchain operating system that provides application developers with databases, account permissions, Process scheduling, authentication, and Internet application communication functions. Therefore, EOS provides the necessary tools for developers to focus on the business logic of their applications without worrying about issues like encryption implementation or communicating with decentralized computers. Moreover, EOS will use parallel technology, which makes it possible for the scalability of the blockchain to reach millions of communications per second.
Below, I will analyze the technical capabilities and limitations of EOS and Ethereum, as well as the differences in their design philosophies.
Chapter 1: What is a Smart Contract
For those new to cryptocurrencies and blockchain technology, the most important thing is to understand exactly what a blockchain is. A blockchain is essentially a decentralized system whose core is a public ledger. The ledger records the current state of the system (such as how many cryptocurrencies are in each account). In addition to the public ledger, blockchain technology includes a consensus mechanism that describes how the decentralized computer (the network of computers running the blockchain) updates the current state of the public ledger.
Interestingly, in 1994, a cryptographer named Nick Szabo discovered that a decentralized ledger system could be used to execute smart contracts (also called self-executing contracts). Mr. Szabo coined the term "smart contract", He hopes to incorporate the practice of contract law into running electronic business agreements with strangers on the Internet. Smart contracts enable a transparent way to transfer and exchange funds or property without going through an intermediary. Like traditional contracts, smart contracts also define all obligations and potential penalties in this agreement, but smart contract platforms can automatically enforce all of these obligations and penalties. These smart contract platforms allow these decentralized applications to run on its network. Ethereum is currently the largest and most successful smart contract platform, but the new platform EOS will seek to solve several problems facing Ethereum.
Chapter 2: Design Philosophy
The most critical difference between the EOS and Ethereum networks is the design philosophy behind them. It can be said that Ethereum is application agnostic, which means that it is designed to be a neutral platform for all potential applications. Ethereum's design rationale document on github says this: Ethereum "has no features" and refuses to include even common high-level use cases in the protocol. This makes the application less bloated, but it still requires different applications to reuse code. But if the platform itself provides those commonly used functions, it will improve the efficiency of application developers.
On the contrary, EOS realizes that different applications require some of the same functions, and EOS provides these functions, such as the implementation of encryption, the application of blockchain communication tools. Based on this philosophy, EOS will include the following features: role-based licensing capabilities, a suite of web tools for developing interfaces, self-describing interfaces, self-describing databases, and a declarative licensing scheme.
As far as I understand, these functions provided by EOS will greatly simplify the generation and management of user accounts, as well as security issues such as account recovery.
Chapter 3: Consensus Mechanism and Governance
Another notable difference between EOS and Ethereum is the blockchain consensus mechanism, and the blockchain’s governance approach. Ethereum uses a proof-of-work mechanism (which will soon be converted to a hybrid POW and POS mechanism), while EOS will use graphene technology, which uses a DPOS-authorized proof-of-stake mechanism. Both options are important for business scalability, an issue that will be discussed in the next chapter.
One of the problems with the current POW mechanism is how to fix a buggy and broken application. For example, recently the DAO had a serious bug or was hacked. Those with a "code is law" ideal see the DAO hack as a "feature" rather than a failure. In any case, the failure of the DAO caused two problems: a broken application running on Ethereum led to heavy losses for investors, and a panicked hard fork. Based on the current POW consensus mechanism of Ethereum, a large number of competing chains may appear every time a hard fork, just like Ethereum and Ethereum Classic after the DAO event. What’s more, a panicked hard fork to fix a broken application would disrupt the entire Ethereum network.
Instead, EOS includes a mechanism to freeze and repair broken applications. For example, if this DAO event happened on EOS, it would be frozen, repaired, and then updated without interfering with other EOS applications. Moreover, based on the DPOS consensus mechanism of EOS, even a hard fork will not generate a large number of competing chains. This has been proven by the Steem network, which has undergone 18 hard forks, which is also based on graphene technology. Also, EOS will include a legally bound constitution for dispute resolution, as well as a community voted based on stake weights.
Chapter 4: Extensibility
For a platform to be commercially viable, scalability is extremely important. In this regard, EOS is very different from Ethereum. The current Ethereum network is limited by the single-threaded performance of the CPU. The early testnet reached 25 transactions per second (still optimized), and with optimizations it can be increased to 50 or 100 transactions per second. However, in practice, transactions on the Ethereum network are around 10 transactions per second, or even less. Previously, the network was even flooded so that only those transactions with the highest transaction fees could be executed, and the others were rejected. This is especially evident in recent ICOs, such as Status’s ICO, in which the ether network was completely blocked and eth crashed on a large scale. Vitalik Buterin has come up with a roadmap to achieve "infinite scaling", however, the technology of database sharding will undoubtedly increase the complexity of the network, and at the same time may reduce the security of the network. I don't think the technology of database sharding can expand the Ethereum network, but it will increase the process of database sharing and reintegration.
In terms of scalability, EOS has two significant advantages over Ethereum. Once implemented, EOS will be the only platform that can handle truly commercial-grade decentralized applications. First, EOS will rely on graphene technology, which has been able to reach 10,000-100,000 transactions per second in stress tests. Second, EOS will use parallel technology to expand the network, and the processing capacity will reach millions of transactions per second. If these can be achieved, EOS will be able to support thousands of commercial-grade DAPPs. EOS will use anonymous communication technology, while separating verification and execution to increase speed, and since it has no transaction fees, EOS will not require counting operations.
Chapter 5: DOS Attacks
As with scalability, the issue of attack is equally important. In this chapter, I will briefly discuss potential DoS type attacks. DoS attacks refer to malicious attackers blocking the network and paralyzing the network so that legitimate and regular requests cannot be processed. EOS is not vulnerable to DOS attacks. In the Ethereum network, miners tend to prioritize transactions with high fees. Since bandwidth and computing power are limited, it is foreseeable that if the network is flooded with high-fee transactions, this will kick out legitimate low-fee transactions. You might think it would be expensive to carry out such an attack on the network, but the rewards are even greater. For example, the recent Status ICO, which is really a competition, if you can send funds to the ICO smart contract address, then you can get ICO tokens at a very low price. This induces rich players to clog the network with high-fee transactions to keep their transactions executed. This became a serious flaw in the Ethereum network, where a single application or smart contract could freeze the entire network. Conversely, EOS is not vulnerable to DOS attacks. Holders of EOS tokens will give users a certain percentage of equity in network bandwidth, storage and computing power. Therefore, malicious attackers will only consume the portion of the network that their EOS tokens give them. An application may be subject to DOS attacks, but these attacks do not disrupt the entire network. Startups with only a small stake can have reliable bandwidth and computing power, even as other malicious attackers try to attack those large network applications.
Chapter 6: Network Economics: Gas Burning Fees vs Equity Ownership
Finally, I would like to briefly discuss the economic models of EOS and Ethereum. In fact, the difference between them is the difference between the ownership model and the lease model. In Ethereum, every computation of transactions, storage, operations, bandwidth usage costs gas. Also, gas fees are volatile and can be set to very high values, as miners tend to choose transactions with high fees to process. This is evident in the recent Status ico, where the gas fee of $100 is still low even for transactions of little value. In the previous chapters, there was a situation in this economic model: wealthy players could send a large number of high-fee transactions into the network that would bring the network down. Moreover, this model requires developers and startups to continuously burn gas during the development and deployment of their applications.
Instead, EOS will use an ownership model where holding EOS tokens will be given a share in network bandwidth, storage, processing power. This means that if someone owns 1% of EOS tokens, he will always only get 1% of the network bandwidth, regardless of the load on the rest of the network. In this way, small startups and developers can buy a relatively small share of the network, so they can get a stable and reliable predictable network bandwidth and computing power, and if they need to expand their application, they only need to buy more of EOS tokens. Also, since the EOS network has no transaction fees, there are no network development costs, other than the cost of purchasing EOS tokens to begin with. Of course, you can also sell the tokens if you want to get your investment back.
Summary
I am more optimistic about EOS technology. Although the current market value of Ethereum is about the same