Jointly produced by Tongzhengtong Research Institute × FENBUSHI DIGITAL
Text: Song Shuangjie, CFA; Sun Hanru
Special Advisor: Bo Shen; Rin; JX
guide
Text: Song Shuangjie, CFA; Sun Hanru
Summary
Special Advisor: Bo Shen; Rin; JX
guide
Since the release of the first white paper in 2017, EOS has received widespread attention. However, there have been doubts about its consensus mechanism and governance methods in the industry, and there are also opinions that EOS can catch up with ETH and become the next-generation public chain. This special report will discuss the consensus mechanism and governance model of the blockchain, and analyze the three levels of blockchain consensus with EOS as an example.
Vitalik Buterin, the founder of ETH, once talked about three levels of "decentralization" in an article: the architectural layer, the political layer, and the logical layer. Drawing on Vitalik's ideas, we can divide the "consensus" of the blockchain into three similar levels: the architectural level, the political level, and the ecological level. The architectural layer refers to: the blockchain relies on cryptography and consensus algorithms to enable each node forming the network to reach a consensus on the historical state data of the blockchain; the political layer refers to: the actual controller of the blockchain, including the operation of the project by the community , governance, development route, etc., and the development team reached a consensus on the realization of the architecture layer technology; the ecological layer refers to: the consensus formed by all participants in the blockchain ecology and economic system on the value of the blockchain. We can evaluate a consensus mechanism from three perspectives: fault tolerance, anti-attack, and anti-collusion.
Table of contents
This article starts from the three levels of consensus and the EOS election bribery event, briefly reviews the EOS architecture layer consensus - the DPoS-BFT consensus mechanism, and analyzes the economic model of EOS, including its main functions, distribution methods, supply and demand relations, Cluster and divide EOS ecological participants, analyze the demand of participants with different identities for EOS certificates, and speculate on the root cause of EOS election bribery and the defects of its governance model.
Finally, this article compares PoW, PoS and DPoS consensus mechanisms and public chain governance models, analyzes the root causes of EOS governance problems, and looks forward to the future development direction of public chain governance.
Risk Warning: Centralized Risks of Public Chain Governance
Table of contents
1 Three levels of consensus
1.1 Consensus on "architecture layer"
1.2 Consensus at the "political level"
1.3 "Ecological Layer" Consensus
2.1 Brief Analysis of EOS Architecture Layer Consensus
text
2.2 Brief Analysis of EOS Economic Model
2.3 Governance model and consensus of mainstream public chains
3 Where is the root of EOS governance problems?
text
In the series of technical reports, we introduced PoW (Proof of Work) consensus, PoS (Proof of Stake) consensus, and indirect consensus formation through elections——Delegated (delegated voting) consensus mechanism. As a public chain using the improved DPoS-BFT consensus mechanism, EOS has received widespread attention since the release of the first white paper in 2017. However, there have been doubts about its consensus mechanism and governance methods in the industry. At the same time, there are also opinions that EOS can catch up with ETH and become the next-generation public chain.
The PoW consensus allows nodes to freely join or withdraw from the blockchain network, and nodes compete for bookkeeping rights by consuming electric energy for a large number of hash value calculations. In the PoS consensus, the basis for nodes to compete for bookkeeping rights is changed from computing power to holding rights, which reduces energy consumption. The DPoS-BFT consensus used by EOS only allows a small number of nodes to participate in the consensus process. These nodes are voted by ordinary nodes holding rights and interests, and adopt a system of dynamic update and accounting in turn. The degree of centralization is higher than that of PoW and PoS, and the network efficiency is also greatly improved accordingly.
In this special report, we will discuss the consensus mechanism and governance model of the blockchain, and take EOS as an example to analyze the delegated voting consensus.
1 Three levels of consensus
Vitalik Buterin, the founder of ETH, once talked about three levels of "decentralization" in an article: the architectural layer, the political layer, and the logical layer. The architectural layer refers to: how many nodes with full functions a system is composed of, and how many nodes can be tolerated to fail at the same time during the operation of the system without affecting the normal functions of the system; the political layer refers to how many independent individuals or organizations The node has the actual control right; the logical layer refers to: if the internal structure of the system is not considered, but from its performance and overall function, it is more like a complete single device or a cluster of scattered nodes.
Vitalik expounded his views on the blockchain: architecture, decentralization of the political layer, and centralization of the logical layer. The reason why the blockchain is logically centralized is because each blockchain network has its own consensus mechanism to maintain the same state (book), and the function of the blockchain network composed of nodes is more like a overall.
Drawing on Vitalik's ideas, we can divide the "consensus" of the blockchain into three similar levels: the architectural level, the political level, and the ecological level. The architectural layer refers to: the blockchain relies on cryptography and consensus algorithms to enable each node forming the network to agree on the historical state data recorded by the blockchain system; the political layer: the actual controller of the blockchain, including the community’s control over the project. Operations, governance, development routes, etc. have reached agreement, and the development team has reached an agreement on the realization of the architecture layer technology; ecological layer: all participants in the blockchain ecology and economic system have reached a consensus on the value of the blockchain. Consensus at the architectural level is the foundation of the blockchain network. Consensus at the political level is the guarantee for the realization and iteration of consensus at the architectural level, while consensus at the ecological level depends on the architecture and consensus layer.
1.1 Consensus on "architecture layer"
Simply put, the consensus at the architectural level is the consensus formed between the nodes that make up the blockchain network at the physical level and guaranteed by algorithms. Vitalik believes that the advantages of decentralization include fault tolerance, anti-attack, and anti-collusion. Similarly, the "architecture layer" consensus can also be evaluated from these three perspectives.
"Fault tolerance" refers to the proportion that this consensus algorithm can tolerate the existence of Byzantine nodes in theory, and it varies with different consensus mechanisms. For example, the fault tolerance of PoW is 1/2, and the fault tolerance of pBFT is 1/3. "Anti-attack" refers to the mechanism by which the consensus mechanism resists various common attacks such as double payment and sybil attack. "Anti-collusion" is to prevent nodes from adopting strategies to cooperate with each other to seek improper benefits through the incentive and punishment measures in the consensus algorithm.
In the "Blockchain Technology Introduction" series of topics, the characteristics of the architecture layer of mainstream consensus mechanisms such as PoW and PoS were briefly described. But even if a blockchain system achieves fault tolerance, anti-attack, and anti-collusion in the consensus of the architecture layer, its final performance at the ecological layer, that is, the stability and security that users ultimately feel may not be satisfactory. This involves the political consensus of the blockchain system.
1.2 Consensus at the "political level"
Regarding fault tolerance, there is a common failure mode in the engineering field called "Common mode failure". Assuming that in a system composed of multiple sub-parts, the probability of each sub-part being wrong is p, and they are independent of each other. If N sub-parts in the system fail at the same time, the theoretical failure rate of the system is p^n. But in practice, it is often observed that the failure rate of similar systems is much higher than the theoretical value, even close to p. This is due to the fact that multiple possible errors are not statistically independent from each other, which means that the causes of these failures may be related. For example, these sub-components may have the same design flaw, and when a failure of one sub-component is observed, it is likely that other sub-components will also have the same failure.
Although the blockchain system can guarantee a certain fault tolerance at the architectural layer, it is difficult to resist common mode faults. Assuming that the development team of the blockchain has left a bug in the released client and has not found it, and most nodes have installed this version of the mining program, when the bug is triggered, the proportion of faulty nodes can easily exceed the fault tolerance limit .
Consensus at the political level is also an important part of consensus, and it cannot be measured by algorithms or codes. If the development community has differences on the development route and concept of the blockchain, it is likely to cause a split in the consensus of the architecture layer. For example, because some BTC miners and community members did not agree with the SegWit2x expansion plan of the Core team, BTC was forked on August 1, 2017 at a blockchain height of 478559, using 8M large blocks instead of SegWit to expand the original chain, creating a split The first of its kind to fork BTC.
Whether the consensus at the political level is sufficiently resistant to attack and collusion is also worth exploring. There are some common ways to launch attacks on blockchain networks, such as double-spending transactions, Sybil attacks, 51% attacks, and so on. Some of them can reduce the possibility of similar attacks by optimizing the consensus mechanism of the architecture layer, while others are limited by the consensus mechanism of the blockchain, and these possibilities are further amplified at the political level. For example, the centralization of the political layer (centralization of mines, centralization of computing power, monopoly of mining equipment manufacturers, etc.) will make it easier to implement 51% attacks; provide convenience.
Anti-collusion is reflected in how to prevent nodes, and more importantly, how to prevent the actual controllers behind the nodes from cheating, inducing voting, etc. Manipulation of election results, and how to avoid collusion with each other to obtain illegitimate benefits while ensuring that accounting nodes maintain a good cooperative relationship.
1.3 "Ecological Layer" Consensus
The "ecological layer" consensus is the consensus formed by all participants in the blockchain ecology and economic system on the value of the blockchain. It covers various aspects such as blockchain network stability, security, and network efficiency, and is ultimately reflected in the recognition of the value symbol in the system—the value of the token. A merchant is willing to accept BTC payment because he expects that the BTC he holds will be accepted by others with a certain value in the future. This is because BTC has a long-term verified and reliable architectural consensus, a stable developer team, and an active community. People believe that BTC can function as a "means of circulation" to a certain extent.
But even for blockchain systems with similar architectural layer consensus, the performance of their "ecological layer" may be quite different. For example, after the original ETH chain experienced the hacking incident of The DAO, there was a disagreement within the community on whether the transaction that the hacker initiated the attack should be rolled back, and the fork became the ETH chain that rolled back the hacker transaction and the ETC chain that was not rolled back. Since the ETH core development team has a high reputation in the community, most miners choose to mine on the ETH chain. The structure of the two at that time was exactly the same. It was the political and ecological consensus that enabled ETH to overwhelm ETC and develop to the present level: whether it is price, transaction volume on the chain, or number of DApps, ETH is significantly ahead of ETC. At the beginning of January 2019, due to the continuous decline of computing power and the continuous downturn in rental prices, ETC was also attacked by 51% of hackers. Such attacks will inevitably reduce users' trust in the ETC network, and even form a positive feedback mechanism, causing users Loss, the network computing power and the transaction volume on the chain are further reduced, weakening the system's ability to resist computing power attacks.
2EOS Governance Puzzle
In October 2018, a screenshot circulating on the Internet that implied that Huobi Exchange had bribery and other improper behavior in the election of EOS super nodes caused an uproar in the blockchain industry. The data shows that there may be mutual solicitation between Huobi and more than a dozen candidate nodes. This incident has also aroused people's criticism of the EOS governance model.
The "governance" of the blockchain world is actually the maintenance of the "ecological layer" consensus. To understand the governance model of EOS, we need to understand its consensus algorithm and economic model.
2.1 Brief Analysis of EOS Architecture Layer Consensus
EOS adopts the DPoS-BFT consensus algorithm. For the analysis of this kind of consensus, please refer to the topic of "Blockchain Technology Introduction Volume 3".
PoW and most PoS consensuses are open, and nodes can join or withdraw freely. However, DPoS is restricted by its own algorithm, and the number of nodes that can directly participate in the consensus process must be limited and fixed. In EOS, nodes that directly participate in the consensus process are called super nodes (BP, Block Producer for short). 21 BPs are elected by EOS token holders, dynamically updated at regular intervals, and produce blocks in turn in a randomly generated order.
In the PoW consensus, the interval between each block is relatively long, which is usually adjusted by an algorithm and kept within a relatively fixed interval, generally ranging from tens of seconds to several minutes. The block interval of BTC is 10 minutes. As the scale of the PoW network expands, the delay between nodes also increases accordingly. If the block time is too short, it will cause the nodes to fail to synchronize the latest block in time, resulting in a network fork. Because PoS does not need to ensure the security of the network through the computing power competition of nodes, the block time can be further shortened.
The block time in the DPoS consensus is shorter. The production of each new block of EOS takes 0.5 seconds, and a transaction can be confirmed in only 1 second. The DPoS-BFT consensus has the advantages of low latency and high performance, and can guarantee large-scale transaction throughput.
PoW and chain-based PoS consensus are not final, the network may fork, and the latest one or two blocks may still be rolled back.
DPoS, and "Byzantine-like" PoS consensus is deterministic. After it is the turn of the node that produces the block to broadcast the packaged block, if more than 2/3 of the nodes sign the block, the block becomes the final consensus. During the block production process, if a node cheats and produces a block on a forked chain, it will be voted out and the node's power will be deprived. Some PoS consensus will also impose economic penalties on cheating nodes.
EOS and some PoS consensus also use transaction proof, namely TaPoS (Transaction as Proof of Stake). TaPoS requires that each transaction must include the hash of the previous block. Such a consensus mechanism greatly limits attacks such as maliciously forged transactions and replay attacks.
This article will compare the governance models of DPoS with PoW and PoS in the third part.
2.2 Brief Analysis of EOS Economic Model
The EOS system is built on the basis of "Contracts". Many underlying protocols of the system, including the realization of basic functions such as EOS tokens, are implemented through contracts. The concept of contract in EOS is more basic and close to the bottom layer than the concept of "Smart Contract" in Ethereum. It is a program that can be executed by BP, registered in the network and stored in the EOS database, and the execution of these The program needs to consume certain computing resources. The interaction between the contract and the EOS account is called "Actions", and the execution process of one or more actions is called Transactions.
The resources in the EOS system are divided into three types: bandwidth (NET), CPU, and memory (RAM). Bandwidth and CPU are divided into short-term and long-term usage. EOS stores the behavior generated by the contract in the form of a log and distributes it to all nodes through the network. It can be used to restore the state of the contract or application, which consumes network bandwidth resources; the process of executing the contract or restoring the state of the application It will consume CPU computing time; the data that the contract or application needs to access will be loaded into the memory, consuming RAM resources. The three types of resources are provided by BP, and its available resources are disclosed at any time.
The initial issuance of EOS tokens is 1 billion, and there is an annual inflation rate cap of 5%. 1% of the tokens issued each year are used to reward bookkeeping nodes to replace the fees paid by users when they initiate transactions. Among them, 0.25% is allocated to BPs, and 0.75% is allocated to candidate nodes with more than 100 EOS votes. An additional 4% will be deposited into the EOS Worker Proposal Fund to support project development and community development.
2.2.1 Functions of EOS Token
The functions of EOS token include:
Election of BPs. Each EOS account can vote for up to 30 candidate nodes, and no matter how many candidate nodes an account votes for, the number of votes obtained by all nodes that get votes is the entire effective balance of the account. The effective balance is the EOS token used for the election mortgaged by the account, and it needs to be mortgaged for at least three days.
Get network resources. EOS accounts can obtain NET and CPU resources by mortgaging a certain amount of EOS tokens. The proportion of resources that can be obtained is determined by the ratio of the number of EOS tokens mortgaged by the user to the total amount of EOS mortgaged in the entire network to obtain this resource. NET and CPU resources can be leased and transferred to other accounts, but RAM can only be obtained through EOS token exchange. EOS has a market dedicated to trading RAM resources, and its handling fee also needs to be paid through EOS.
2.2.2 The supply and demand relationship of EOS token
The initial issuance of EOS tokens is 1 billion, and there are 5% additional issuances every year, of which 4% are deposited into the foundation. In addition, some tokens are mortgaged to obtain network resources, and these tokens will not enter the market circulation. RAM market transactions will also charge 1% EOS as a service fee, and this part of EOS will be destroyed to resist speculation and offset some inflation. Purchasing an EOS domain name will also permanently destroy some tokens.
Therefore, the supply of EOS is relatively stable. When the transaction volume in the network increases, the demand for NET, CPU and other resources increases, the number of pledged tokens will increase, and the circulation will decrease; if EOS passes a resolution to upgrade BP hardware, the supply of resources will increase, and the required amount of resources to obtain the same amount will increase. The number of EOS drops, thereby increasing the number of tokens in circulation.
The current application scenarios that consume EOS system resources mainly include: creating accounts, executing transactions, deploying smart contracts, etc. Therefore, the growth of the total number of accounts on the EOS chain and the activity of DApp will increase the demand for network resources.
The EOSRAM market is a relatively special market. On the one hand, almost all network operations require memory (RAM), and memory resources can only be exchanged for EOS tokens in the RAM market; There is room for speculation.
The Bancor protocol is a set of tokens that are fully or partially prepared to be issued with a certain anchor token. Its price is linked to the anchor token, and is determined by an algorithm based on the supply of the token and the reserve of the anchor token. RAM is issued under the Bancor protocol anchored to EOS. RAM transactions do not require a counterparty, which is equivalent to directly trading with the EOS system. When the total supply of memory remains unchanged, the increase in demand will drive up the price of RAM. Based on this feature of the Bancor protocol, many speculators hoard RAM maliciously, causing the price of RAM to rise sharply in a short period of time, increasing the memory cost required by users and developers for normal needs, and EOS officials have to expand memory capacity.
2.2.3 Participants in the EOS economic system
In the EOS economic system, accounts holding EOS tokens are called "Stakeholders". They are divided into several different roles: ordinary users, DApp developers, BPs, and candidate BPs.
One of the concepts of EOS is to save the handling fees that ordinary users need to pay for using DApps. The complicated calculation process of handling fees is an important factor that affects the experience of using DApps for ETH users. EOS rewards bookkeeping nodes by setting a fixed inflation rate, which is equivalent to collecting a fixed percentage of "seigniorage" from all EOS share holders.
In the credit currency era, the cost of the central bank issuing currency is very low. The Organization for International Settlements defines seigniorage as the profit a central bank receives from monopolizing the issuance of money. In the EOS ecosystem, the pre-set algorithm rewards BPs by issuing new tokens in exchange for hardware resources and accounting services provided by BPs, and passes them on to all EOS token holders through inflation. If the demand for tokens in the EOS economic system does not change, the value of the unit quantity of EOS tokens will decrease. It is equivalent to the certificate holders bearing the system resource overhead according to the proportion of the certificate held, and rewarding BP and candidate nodes. In the case of no transaction fee, the only factor that motivates ordinary users to hold EOS tokens is the right to vote for BPs.
For ordinary users, the amount of EOS they hold is not enough to affect the election results, and at the current stage, the election of EOS super nodes is not equivalent to elections in real life. Super nodes are basically useless for ordinary users. Differently, if there are candidates willing to pay for their votes, the rational choice for the average user would be to accept it. However, EOS has no reasonable mechanism to prevent canvassing, bribery, collusion and mutual voting.
For DApp developers, because contracts and DApps deployed in the network require a certain amount of network resources to operate normally, in order to ensure user experience, a certain amount of EOS tokens need to be mortgaged. But developers may adopt another method, that is, users can mortgage a certain amount of EOS to use the application, or charge users a certain fee in the application. This model may be used by some gambling DApps.
BPs and candidate BPs will be the group with the greatest demand for EOS. The main economic incentive that 21 BPs can obtain is an annual total of 0.25% of additional tokens. In order to maintain their identities, they need to hold a large number of tokens to vote for themselves. Candidate BPs will have a stronger willingness to hold EOS, especially nodes with a similar number of votes to the last BP. The existence of candidate BPs is also a due diligence supervision of BPs. They can obtain 0.75% of the total annual issuance of additional tokens.
BP's powers include: freezing accounts by voting, changing contract codes that are considered harmful, and voting to make changes or upgrades to the protocol. EOS BPs have great power, so unfair competition is inevitable in the process of running for BPs. Since the distribution of EOS tokens is relatively concentrated, and a single account can vote for multiple candidate nodes, the mechanism design allows BPs to collude with each other and canvass each other to ensure their BP identity. And under the mechanism that allows "multiple votes for one vote", the probability of getting enough votes for nodes that do not collude with each other will be significantly lower than those nodes that vote for each other. Under the existing mechanism, if a proposal is approved by 15 out of 21 BPs, then the proposal is considered passed, and there is reason to believe that the BPs elected in this way have the ability to influence the voting results.
Although it is written in the EOS white paper that any BP who neglects his duty or exhibits malicious behavior will be voted out, and an alternate BP will replace the expelled BP position. But any consensus algorithm solves the problem of "finality", that is, the state of the system that reaches a consensus, but cannot solve whether this "state" is reasonable or justified. If the definition of "malicious behavior" itself requires the "voting" of share holders, then how to detect these BPs who may be elected through bribery? EOS did not give a satisfactory answer.
Since the demand for EOS brought about by campaigning for BP and RAM transactions will cause fluctuations in the price of EOS, ordinary users will also have speculative demand. Except for obtaining corresponding network resources, since the cost of operating supernodes is anchored in legal currency, BP has also become the most motivated group to speculate on the price of EOS tokens.
In the BTC network, the incentive of block rewards exists to encourage computing power to compete for bookkeeping rights, thereby maintaining the security of the network and the reliability of the consensus. However, the inflation strategy of EOS did not significantly improve the reliability of the network, but caused a series of negative effects such as speculation and unfair elections
2.3 Governance model and consensus of mainstream public chains
The consensus of the blockchain architecture layer, political layer, and ecological layer is closely related, and the community autonomy model of the public chain should also be selected according to the consensus mechanism of the architecture layer. The blockchain does not necessarily represent decentralization, but a way to achieve decentralized autonomy.
Public chains using PoW consensus represented by BTC and ETH usually adopt community governance. Improvement proposals (referred to as BIP, EIP) are initiated by the community, and the core development team meeting decides which proposals will be implemented on the main chain. Although computing power does not directly have voting rights, the computing power support of large mining pools can also affect the decision of the development team. Even when there is a disagreement among the development members, some development members can fork the original chain with the support of computing power. Stand on your own.
DPoS, represented by EOS, can adopt a combination of on-chain governance and community governance due to the existence of an on-chain voting mechanism. EOS has an independent arbitration committee (ECAF) and an EOS constitutional system. According to the constitution, six arbitrators make rulings on various matters arising from the operation of the blockchain, and the ruling results are generally implemented by 21 BPs, so the actual execution power is still in the hands of BPs.
Early PoS blockchains generally still adopt community governance models, such as Peercoin. Blockchains that adopt the new PoS consensus, such as Cosmos, also adopt a combination of on-chain governance and community governance. The on-chain governance model of Cosmos is roughly the same as that of EOS. The difference is that ordinary nodes mortgage tokens at verification nodes in the form of entrustment, and verification nodes need to distribute the block rewards obtained from bookkeeping to ordinary nodes that mortgage tokens in proportion. Ordinary nodes will also partly bear the punishment for failure of verification nodes to fulfill their due obligations. Therefore, Cosmos ordinary nodes also need to have a certain awareness of participating in community governance, shoulder certain responsibilities, and share benefits, instead of only voting rights like in the DPoS consensus.
The main contradiction of the EOS governance model is that the consensus at the architectural level is essentially centralized, while the consensus at the political level hopes to be reached in a decentralized manner. The following are the main areas where the EOS governance model can be improved.
There are problems with the electoral package. One vote can be cast multiple times, leaving room for some nodes to vote or bribe each other.
BP has too much power, and the arbitration organization is centralized. In the eyes of some blockchain enthusiasts, "code is law", but the EOS Arbitration Committee can issue rulings such as modifying account private keys and rolling back transactions for BP to execute, which is against the blockchain's "immutable". sex".
Incentive programs are inadequate. At present, the inflation rate of EOS is relatively high, and the incentives for ordinary users to hold tokens are insufficient. The token rewards given to BPs are not to improve the security and efficiency of the network, but instead induce node bribery.
The mainnet was launched on January 17, 2019, and BOS, which was officially activated on January 18, brought us a new perspective on these issues. BOS is a side chain based on EOS. BOS officials believe that EOS has excessive governance. BOS hopes to balance the free market mechanism of BTC and the current centralized status of EOS by being guided by actual needs, and at the same time give full play to the characteristics of efficiency and decentralization .
BOS has made improvements to the above aspects of EOS.
Improvements to the "architecture layer" consensus. The consensus mechanism of BOS is based on the PBFT theory, combined with the improvement of the EOS consensus, and under the premise of ensuring the realization of Byzantine fault tolerance, the time required for a transaction to become "irreversible" is shortened from minutes to seconds.
The arbitration organization is decentralized. BOS issues an additional 0.2% every year to BOS governance organizations or volunteers. Anyone in BOS can issue arbitrations. The more governance organizations seconding an arbitration, the higher its reliability. If arbitration takes effect, governance organizations or volunteers can receive governance rewards. There are two methods of BOS arbitration award: 1. No less than 15 BPs agree, 2. Community referendum. There will not be a single "ECAF" agency in BOS governance, and it will be replaced by multiple neutral governance organizations or volunteers.
Improve economic models. The initial circulation of BOS tokens is the same as that of EOS, which is 1 billion pieces. Among them, 100 million is used for airdrops, of which 50 million are directly allocated to EOS accounts in proportion; 100 million ecological funds; 400 million ecological incentives are used to subsidize the payment and BOS transaction business generated on the BOS chain; 200 million are allocated by the founding team Locked; 200 million private equity quota. BOS issues 2% more each year, 1% is used for node rewards, 0.8% is used for community developer rewards, and 0.2% is used to reward community autonomous organizations.
Although BOS has made many improvements in governance programs and economic models, its influence is still limited. At present, the support rate of the proposal on "abolishing the core arbitration committee" on EOS has reached 99%, and ECAF is likely to disappear.
3 Where is the root of the EOS governance problem?
In addition to the criticism of the EOS governance method, the EOS infrastructure has also been questioned by researchers. A paper researching EOS infrastructure pointed out that EOS lacks cryptographic security. The thesis of the paper is that EOS does not have a process for valid cryptographic verification of transactions.
The basis of blockchain technology is a distributed database. The difference between blockchain and traditional distributed databases is that blockchain uses verifiable and secure cryptographic algorithms, which can maintain the "shared" system under the premise that Byzantine nodes may exist. State" consistency, including maintaining the legitimacy of transactions, system consistency, etc., and preserving cryptographic evidence.
BTC will calculate the Hash of each transaction, and the Merkle tree ensures that the transactions are traceable and verified by the nodes of the entire network. The PoW and the longest chain mechanism ensure that the block history cannot be easily tampered with, and the certainty of the consensus .
The article points out that non-current production block BP does not have the ability to verify the transactions in the block, so there is no cryptographic evidence to verify the validity of the block.
Regardless of the scientific nature of the paper's views, the root of the problem with the EOS governance model is not the so-called "lack of cryptographic verification".
The infrastructure layer consensus of EOS is not the main cause of governance problems. We contrast the DPoS consensus of EOS with the PoW consensus.
The PoW consensus has maintained the safe and stable operation of BTC for nearly ten years, and its stability is guaranteed by sufficient computing power. If malicious nodes want to attack or tamper with the history of the blockchain, they must make the length of the attack chain exceed The main chain needs to have more than 51% of the computing power of the entire network, and such high-cost malicious attacks often outweigh the gains, thus ensuring that its blockchain history is difficult to tamper with. But this will inevitably lead to a waste of resources.
The DPoS consensus mechanism adopted by EOS does not require nodes to compete for computing power, so there is no waste of mining machines and energy, but DPoS also has certain hidden dangers in network security. The number of super nodes is small, and they are vulnerable to distributed denial of service attacks (DDOS), which may seriously affect the stability of the network system. In addition, there may also be loopholes in the DApp smart contract, which threatens the security of users' funds.
In the ETH DAO incident in June 2016, 60 million dollars of ETH was stolen because PoW did not have a complete repair mechanism for vulnerabilities. In order to roll back the hacking attack, a hard fork has to be carried out. At the same time, a large number of competing chains may appear, weakening the ecological consensus of the original chain.
The root of many disadvantages of the EOS governance model is also a common problem of the Delegated consensus mechanism. Some blockchain projects use BFT algorithms in order to pursue transaction efficiency and consensus finality, which limits the number of nodes that directly participate in the consensus process. Consensus nodes are selected by the method.
Note:
Due to some reasons, some nouns in this article are not very accurate, mainly such as: general certificate, digital certificate, digital currency, currency, token, crowdsale, etc. If readers have any questions, they can call or write to discuss together.
This article was originally created by TokenRoll Research Institute (ID: TokenRoll). Unauthorized reprinting is prohibited. For reprint, please reply to keywords in the background【Reprint】
