EOSC Cyborg: The Public Chain Needs Computing with High Performance and Low Latency Based on DPoS Consensus| Chain Interview
The EOS public chain utilizes DPoS consensus mechanism, which supports computing with high performance and low latency. While DPoS mechanism itself is not decentralized enough. As the first forked chain of EOSC mainnet, EOSC has optimized the election mechanism of EOSIO and launched the EOSC mainnet at the genesis block height of one. Through persistently upgrading EOSC mainnet that makes EOSC Network evolve toward the decentralized high-performance smart contract platform.
At this edition of Chain Interview, Chain Teahouse invited Cyborg, the founder of EOSC. Cyborg will tell the stories happened during the process of EOSC development and operation, and interpret the existence of the future computing public chain with high performance and low latency.
Technology R&D Team Supported by Internal Funds
Keep Optimizing DPoS Consensus
Founded in early 2018, EOSC team currently has about 20 members, and is based in Hangzhou. At first, the team felt that the industry competition lied on the protocol and the bottom public chain, and the entrepreneurial direction was based on the market, so there is no need to create a new consensus and ecology. At that time, EOSC chose the most advanced ecology EOS and is committed to the optimizations based on DPoS consensus.
EOSC team has technical talents from famous Internet companies such as Huawei, Tencent, Baidu, etc. Cyborg also worked on Internet related work before. Currently, about 8 people in the team are responsible for technology development, mainly the upgrade and iteration of EOSC. The maintenance of EOSC mainnet is in the charge of the community. The technical team is mainly responsible for the following three directions of work: first, the improvements based on the community’s opinions; second, study the latest technologies, such as cross-chain, privacy computing, oracles and virtual machines; third, it will help peers with some blockchain development.
At present, the EOSC team does not raise funds. It is entirely supported by its own internal funds. The revenue of the public chain business in the business model cannot cover the cost. Therefore, providing some other projects with some development help is also one of the sources of income.
According to Cyborg, DPoS consensus is reasonable in its existence and development. Because the computing with high performance and low latency will be a must in the future blockchain industry, which can only be reached by public chains with DPoS consensus. In EOS, 6 blocks can be produced in every second, this concept is very advanced. Besides, the implementation of multi-threading technology EOS is exploring will bring a qualitative improvement for the high-performance computing in the blockchain industry.
EOSC has encountered two difficulties on the its entrepreneurial path. The first is that EOSC was not recognized in the beginning. EOS is the beginning of the 3.0 era. People don’t understand that there are several blockchains in one ecology, so they naturally reject chains with fork tendency and think that fork is unjust. In 2017, BTC and ETH’s forked token ran away and left a lot of mess to the community, leading to a bad impression of the fork behavior. The EOSC team realized that in the blockchain world, actions speak louder than words. EOSC insisted developing for a year and a half, while ETH and BTC’s fork development did not take so long. Users in the EOSC community can feel that the team is hard-working by the upgrades and iterations of EOSC.
Second is the shortage of team funds. EOSC cannot spend money on the ecology and marketing like other financing teams. In Cyborg’s view, the POW public chain can be financed, because holding POW tokens can only receive dividends, but DPOS sells both dividends and voting rights, and this network may not able to achieve the community governance. Profitable people took away dividends and voting rights, but they are not able to do these things. For example, EOS investors can’t participate in governance and voting, so EOSC will not raise funds for DPOS tokens, and maybe will raise funds for the company’s purposes in the future. EOSC token is not going to raise funds. The dividend is only the profit of the legal currency. These are two different things.
EOSC Node Cost is 1% of EOS
TPS is Adjusted According to Market Demand
EOSC enables blocks to be produced every 0.5 seconds with 6 consecutive blocks, which is different from EOSIO’S model of producing blocks every 0.5 second with 6 consecutive blocks produced by 1 node. Even though consecutive block-producing can reduce the waiting time of unpacked trades, it may affect the chain stability because of the unsatisfactory network environment, resulting in a large number of microforks.
In addition, Cyborg also believes that TPS is not a very rigid measure of the performance of the public chain. TPS in EOS currently reaches more than 5,000 and can support 432 million transactions per day. However, the daily order of Meituan APP is only of million level, and Taobao is only of ten million level, EOS’s TPS completely overflows the normal trading needs of users. The reason why EOSC reduces the speed of the block producing is that the market demand is not so big, and the node does not need to provide such a high configuration. EOS has a large number of microforks, and the synchronization cost is very high. When the market demand is high, all nodes just need to provide configuration to transform. Compared with the synchronization of 0.5 seconds and 3 seconds, the network cost is definitely low, so the EOSC node cost is one percent of EOS. If the network is very blocked, the nodes just need to improve the configuration, which means that the market demand for this chain has increased.
Use Cross-Chain Technology to Make Up the Disadvantage that DPoS is Not Decentralized Enough
Cyborg said that working on cross-chain is because the EOSC team feels that DPoS is not decentralized enough, but high performance requires DPoS, so it is impossible to give it up completely. How to realize the decentralization of the public chain that is not so decentralized, what needs to be done is to use another POS chain as the sub-chain of EOSC. At present, the team is working hard to develop in the test network.
One important thing that both EOSC and EOS lack is Account Balance Status Verification, which is not on the mainnet. In EOS, there are two ways to alter the balance without the private key, the node private sign and the change of super privilege. Although the core users of the EOS ecosystem don’t care about this, everyone has different opinions on how to define the blockchain. The Bitcoin community believes that the user’s currency cannot be altered, EOS Ecology feels that if the private key of ETH and BTC is lost the asset will be 0. EOS sets node private sign and super privilege to better protect user asset security with a community consensus. But in extreme cases, if the node is controlled, there is a risk of changing the balance.
The cross-chain of EOSC can solve this problem. It can verify the balance regularly, and then put the data into the block. Unless the transaction is rolled back, otherwise the asset can be truly protected. When developing cross-chain technology, EOSC found that when the market is talking about the importance of heterogeneous cross-chain, but everyone is doing isomorphic cross-chain. All chains in the isomorphic chain are developed based on the same protocol and can naturally be cross-chained.
Cyborg believes that cross-chain changes the competition between public chains from the account book capacity to the accounting method. Excellent accounting methods can be used by everyone. There is a network effect, which is different from a single book. Before EOSC develops cross-chain, it investigated who can form a market effect in the market. At present, the team feels that COSMOS can. COSMOS has produced a good accounting method, ecologically chooses the GO language, and many traditional Internet developers familiar with the language. In addition, COSMOS sdk is very simple, developers can start in ten minutes. At present, there are hundreds of teams developing based on COSMOS. These hundreds of projects can be linked in the future, which is an ecology.
1-Token-1-Vote Guarantees the Fairness
BP Reward & Penalty Mechanism and Heartbeat Mechanism Guarantee the Block-Producing Stability
In EOS, only elected nodes have profits, and the voting users have no income, so there is bribery. EOSC adopted “1 token 1 vote” that as long as you vote, you will gain dividends. But if the node does not produce blocks, the voting users cannot gain dividends. So the voting users will not vote randomly. They will choose the nodes that make contribution to the community and produce blocks stably.
The main optimization direction of the EOSC’s election mechanism is to separate the voting rights and the dividend rights. This requires the user to vote for the node and ensure that the voting rate network is safe. As long as there are dividends, there will be bribes. This will be adjusted by the adjustment of voting rights. The longer you deposit, the more you will vote. The fairest thing about POW is computing power. EOSC believes that the fairest thing in POS consensus is time.
But there are problems in practice. For example, a big holder directly locks his tokens for two years to gain more discourse power. In fact, big holders are also part of the community, and they are loyalty computing power. However, the gap between the rich and the poor exists in both POW and DPOS. In POW, computing power means the discourse power. In POS, the more tokens, the discourse power. The gap between the rich and the poor is inevitable. which is created by private ownership. And it is unrealistic to eliminate private ownership. The team is currently concerned about the unequal discourse power caused by the gap between the rich and the poor. Only by solving this problem will POS really break out. Cyborg believes that the blockchain is a new thing, we need to solve the problem first, and then improve it. The impact of the gap between the rich and the poor on the discourse power is a long-term issue.
The nodes in EOSC can be divided into BPs and BP candidates. BPs just need to canvass and deposit a small amount of tokens. The work of BPs is to maintain the stability of block-producing. According to changes in market demand, the high market demand requires higher node configuration. The functions of mainnet can be updated and planned with more than 2/3 of BPs voting through. Users are voting in EOSC all the time. The eliminated BPs will be reduced to BP candidates, and the candidates need to prove that they have real ability to produce blocks. In the future, EOSC will also introduce light nodes that cross-chain needs, and they will verify whether the entire network is normally producing blocks or not
EOSC has established two mechanisms to ensure the stability of the mainnet. First, BP Reward & Penalty Mechanism. The stable block-producing is the first goal of network security. For BPs which don’t produce blocks, EOSC will impose very severe penalties. The stable block-producing will gain more profits. The first task of the node is placed on network security. The BP has to pay the deposit. If the BP doesn’t work, it will not gain profits, and its deposit will be deducted. Anyone can initiate a proposal to punish the BP that doesn’t produce blocks. If the proposal is approved by the multi-sig, this part of the penalty will be distributed as reward to the proposer and the nodes that agree to the multi-sig.
Second, the Node Heartbeat Mechanism for BP candidates. The high probability of BP candidate is that there is no outbound task. If it does not have a real server, but connects to someone else’s server, then you can get the benefit without paying the cost. EOSC requires that the BP candidates needs to have a real block-producing ability. When it is the turn for a BP candidate, it needs to be able to produce blocks. The Heartbeat Mechanism is to check whether the candidate has the ability to produce blocks online. If the candidate node is checked out not online, the heartbeat mechanism can make the BP candidate gain no profit.
Single Public Chain Carrying Too Many DApps Does Not Conform to Development Trend
EOSC is Committed to Provide More Public Chains with High-Performance Computing
EOSC is very compatible with EOSIO developers, and DApp can be migrated directly. Cyborg believes that DApp developers are currently facing three problems. First, from the physical bottleneck, a public chain can’t carry too many DApps. If a chain runs too many DApps, it will be blocked. In the future, the from will be multiple DApps on one public chain which cannot carry too much. It is not in line with the future trend to develop too many DApps in a single public chain.
Second, it’s about the problem of development tools and development resources. Traditional game developers can invoke a different library to form a game, but the development tools for the public chain are few. This requires the community to do tolls and tutorials together.
Third, the use of tokens. In the market, no more than 10 public chains have established the economy model. Few teams that are able to design an economy model. At present, no DApp can make continuous profit without the economy model, and developers don’t know how to get started. For the above points, EOSC will work on tutorials, application governance and economic models, and study without prejudice to the development progress of the mainnet.
On EOSC mainnet, at present, there are nearly 600,000 registered addresses, more than 12,000 accounts participating in the vote, 23 BPs, and more than 30 active nodes including BP candidates. EOSC defines itself as a computing chain with high-performance and low-latency. Whether it is available to EOS or COSMOS ecosystem, EOSC will work with other public chains to provide high-performance computing protocols for other public chains. Moreover, the code will be open source in the future. The team plans to complete cross-chain R&D in half a year, and the high-performance computing provided in the second phase will be faster.
Copied from Chain Teahouse