PoA 2.0-SURFACE algorithm is both performance and commercial landing options

A purely utopian blockchain has no future.

Or to be more precise, in addition to bitcoin, the latecomers who want to occupy the business highland of blockchain must consider the actual soil of application landing, which has the inevitable trend of regulation, the tide of enterprise-class services, and the inevitable demand to lower the threshold of users.

When blockchain develops to today, both practitioners and ordinary investors should be soberly aware that blockchain can no longer be treated in the light of technology alone. This is a complex new reform involving multiple political and economic interests.

Unlike some projects that are obsessed with discussing technical perfection, smartlink is directly focused on the opportunity to actually land at the beginning of the project, trying to combine the cold blockchain technology with a realistic landing scenario. BlockBeats learned that wirechain had recently proposed a PoA 2.0-surface&cake for real-world scenarios; SURFACE is an acronym for Secure, use-ase adaptive, and Relatively fork-free Approach of Chain Extension, which stands for security, ease of Use, and high extensibility to enhance chain-only security and stability, and to meet the growing needs of business scenarios.

To this end, BlockBeats was fortunate enough to invite Dr. Zhou ziheng, the chief scientist of weijie, and Dr. Ren jiji, the senior researcher of blockchain technology.

Dr. Zhou ziheng received CS bachelor's degree from Shanghai jiao tong university in 2002 and CS doctor's degree from the university of Southampton in 2007. After graduation, he worked as a postdoctoral research fellow and senior research fellow at the university of Kent in the UK and the university of oulu in Finland, engaged in the research of computer vision and machine learning. In July 2017, I joined weichang as a partner and served as the chief scientist, responsible for leading the research and development of the underlying technology of the block chain and the cooperative research with external research institutions.

He received his bachelor's degree in communication engineering from beihang university and his master's and doctor's degrees in information theory from delft university of technology in the Netherlands. After graduation, he joined the block chain laboratory of delft university of technology as a postdoctoral researcher in 2016, and his main research interests include block chain consensus algorithm, capacity expansion and application. Dr. Ren has published a number of blockchain papers at international academic conferences, and wrote an in-depth technical column titled "the history of blockchain" on zhihu by ID: maxdeath, and opened the first blockchain consensus algorithm course in China. In 2019, I joined weilian ecology as a senior researcher of block chain.

BlockBeats would like to hear how this special team conceived the blockchain consensus algorithm and how they see the logic between blockchain technology, regulation and landing. The following is the transcript of the interview.

BlockBeats: as we know, the most common consensus algorithms in blockchain projects are PoW and PoS, and there are also the classification of the consensus and Paxos algorithms. Could you please explain what is the PoA 2.0-surface algorithm in the popular words?

Only link: the prototype of the SURFACE algorithm is PoA, which stands for Proof of Authority, and does not always refer to the same technology in different situations. Generally, people are familiar with PoW and PoS, which take work and stake as a method of proving consensus, while PoA does not take work and stake as judgment criteria for the access mechanism of nodes, but chooses some qualified nodes with a certain reputation and conforming to the only link ecological standard in reality.

Second, PoA 2.0-surface is not judged according to calculation force or equity when it comes to the block. Instead, all selected nodes take turns to block, following the longest chain rule. In fact, the implementation of each authoritative consensus algorithm is slightly different, with its own methods. In a broad sense, it is to change the consensus access mechanism to achieve as safe and efficient as possible through KYC authentication of nodes.

In simple terms, PoA 2.0-surface is to make the access mechanism more efficient and strict, and replace the form of work and stake with authority.

BlockBeats: some argue that the lower cost, faster PoA consensus algorithm didn't seem to get the market's approval at first. In one case, the ethereum-based POA project did a lot of code work, but still didn't have a product that would make the market hot. What do you think of this point of view? Do traditional PoA consensus algorithms have some fundamental shortcomings?

Chainonly: as mentioned just now, when we talk about PoW or PoS, we are not talking about the specific algorithm behind it. There are also differences in the implementation of PoA. Based on the success of ethereum's PoA project, I think the topic is not only technical, but also commercial and marketing considerations.

Specifically speaking, PoA 2.0-surface has a strict general license pledge system, KYC certification, and committee review to determine a series of node access standards. A pure PoW or PoS measure is also a measure of trustworthiness, the willingness to maintain a mine, and we think it's a better way to do a full assessment with authority.

PoA is a development trend. We see that some projects on the market, such as Polkadot, Algorand and EOS, have adopted some relatively centralized methods. That is to say, from the perspective of reality, although PoW and PoS have achieved the self-consistency of logic in academic theory, they will still rely on individuals or institutions in reality in their actual operation, even if they are evaluated according to calculation power or amount of money held. However, we can also see the existence of PoW mining pool and Staking mining pool (entrusting) in the end.

We believe that the market recognition of a project cannot be determined by a simple consensus algorithm, and the ease with which a business is launched depends on the building of modules on the block chain, the promotion of the market, and whether it can bring value to the enterprise itself. This is a multi-dimensional and multi-level problem. In terms of chain-only, our block chain design can help enterprises to land quickly and provide them with a complete set of solutions.

The design philosophy of PoA 2.0-SURFACE is worth pondering. Since PoW is bound to move toward force centralization, the end of large mine clusters, and even gradually evolve the upstream of the industrial chain into the business logic of traditional manufacturing, it is only a matter of time before it becomes centralization. Similarly, PoS seems to solve the problem of PoW centralization, but in fact, large nodes still rely on offline reputation and resource relations to maintain each other. It is also a matter of time for this business logic similar to the traditional financial industry to evolve into a centralized system.

& have spent

So, since pure node decentralization can't be done, can we think about it in a different way? In other words, the contribution and stability of block chain nodes can be guaranteed to the greatest extent without sacrificing basic decentralization, which is just the case with chain.

& have spent

BlockBeats: the VRF function selects the committee, thus reducing the node's abuse of accounting rights to interfere with the system. Can you talk about the specific thinking of this technology design?

& have spent

Chainonly: the application of VRF functions in blockchain projects is not uncommon. Similar designs exist for projects such as Algorand and Dfinity, all of which are used to select committees randomly.

In fact, an ideal blockchain is one in which everyone takes turns to block, but malicious nodes can cause bifurcation, which wastes bandwidth and reduces system performance. Now, one way to solve the bifurcation is to use VRF to select a large enough committee, let the committee conduct BFT, and then the consensus reached is the final consensus, such as Algorand adopted this idea. However, such a design achieves theoretical perfection, but does not consider the practical application and the demand for system performance.

The only way to think about it is, we use the committee mechanism to make bifurcation as unlikely as possible in practice, but at the same time we don't pursue the theory that it's impossible to bifurcate. So the size of the committee is set according to the actual testing in the network. This scheme, which lies between the traditional mechanism without committee and the large committee mechanism, on the one hand reduces the probability of bifurcation, and on the other hand introduces as little delay and redundancy in committee communication as possible. In general, this gives us the flexibility to optimize the output and latency for normal use by adjusting the size of the committee, depending on network conditions.

VRF functions have been one of the technology hotspots in recent years, and we often see the committee's choice to use such technologies in the white papers of top projects. The chain-only design still embodies the principle of serving the actual business, not pursuing the absolute perfection of the academic, but preferring the use effect of the actual scene.

BlockBeats: block finalization is the guarantee of security on the chain. What are the considerations of PoA 2.0-surface?

Chainonly: in terms of finality, there is a difference between the Byzantine fault-tolerant consensus and the traditional satoshi nakamoto consensus. The satoshi nakamoto consensus does not have finality, only that if there are enough block confirmations behind it, then mathematically speaking, there is a high probability that no longer chain will replace the existing chain. This design was criticized by many people at first, but it has been accepted gradually.

According to wang, the two understandings are not contradictory and coexist in PoA 2.0-surface.

Finalization problems are actually internal and external. Internally, the emergence of finalization can reduce storage costs, eliminate the need for some master nodes to store all blockchain data, and make it easier to verify transactions and cross chains. Externally, finalization makes it easier to connect the blockchain to the real world. For example, in a real-world application scenario such as legal deposit, it is not enough to satisfy only a high probability; the system must be able to provide the finality of the block.

Many current projects, such as Polkadot, introduce finalization considerations. The only solution is to refer to the HotstuffBFT consensus mechanism to achieve non-tamperable asynchronous conditions.

Comments on BlockBeats: it is impossible to use blockchain technology in reality and break away from finalization requirements. Especially in the financial and legal fields, the irreversibility of probability cannot meet the requirements of high-precision scenarios in specific industries. Only absolute security can adapt to these scenarios.

BlockBeats: in the PoA consensus, the bookkeeping right has nothing to do with the amount of assets held. What do you think about the motivation of the node to hold the original tokens? Or besides relying on the soft conditions such as reputation, could you introduce to us the economic hard and fast anti-evil measures and emergency mechanism of chain-only PoA 2.0-surface?

Chainonly: we do not think that PoA and PoW/PoS are fundamentally different. In fact, all block chain consensus algorithms can achieve system design security as long as they meet their security assumptions (for example, the number of malicious nodes does not exceed 1/3 of the total number of points, or the computing power/currency holding capacity of malicious nodes does not exceed 1/2). The same goes for PoA 2.0-surface. If there are too many malicious nodes, no algorithm can guarantee security. So in the end, it's the economics of the nodes themselves that keep the system safe.

For chained nodes, the losses from attacking the system are in four ways. Firstly, the tokens pledged by the node will depreciate and suffer serious economic losses. Second, the node's reputation across the whole chainless network and ecosystem, as well as its physical business, is impaired. Third, because the node will sign some agreements at the beginning of the audit, the node may face legal liabilities; Fourth, the mining income of this node on the only link also decreases correspondingly. In fact, for the existing consensus algorithms (such as POW and POS), once the attack occurs, its nodes will only suffer the above 2 or 3 losses at most, so we believe that the chainless nodes will be more motivated to maintain the security of the system.

For a block chain system, how to deal with emergencies is not considered at the level of consensus, but at the level of governance. There is only one set of contingency mechanisms in place, as detailed in our governance model design white paper. In short, only the chain block chain has its governance of the central body, that is, its strategic decision-making committee. In the event of an emergency, the committee is able to act quickly to minimize losses. Bitcoin and ethereum lack that possibility. At the same time, a chain-only governance mechanism guarantees that the commission will propose temporary emergency measures, whose finality will ultimately be determined by a full online vote. So in the long run, it's still a process of governing by vote. In this way, there is both efficiency and democracy.

Generally, the block chain system is guaranteed by the consensus algorithm, and the emergency mechanism is not necessary. However, the only link for commercial scenarios has been designed accordingly. At the same time, we also believe that no software is 100% safe, and any manipulation can lead to emergencies.

BlockBeats comments: putting humanity into the institutional cage is one of our powerful means to solve problems in the traditional world, while in the blockchain world, the use of economic interests of nodes, reputation on or off the chain, and business binding can fundamentally reduce the probability of black swan attacks.

& have spent

BlockBeats: in ethereum's POA project, the verification node must be a us citizen. Is there any geographical requirement for the selection of verification node?

Chain-only: chain-only PoA 2.0-surface has no regional requirements for node selection, mainly depending on the contribution of the node to the whole ecology.

BlockBeats: some people think that PoA is too centralized, violates the spirit of blockchain and does not conform to the technology concept of decentralization; Others argue that PoA's efficiency, low cost and security are consistent with regulatory accountability, that so-called decentralized algorithms are still just words, and that PoA is more efficient and land-based. What do you think of this problem? How does PoA 2.0-SURFACE optimize or solve these problems?

Chainonly: in fact, when we discuss this issue, we need to consider the scope of decentralization of each blockchain design. The range of designs varies from the large chain to the small chain. Decentralization is not a word, but it should be realized within a defined scope. Blockchain landing is not just a technical issue, but connected to the real business world. We see the popularity of the PoA algorithm because it is easier for companies to recognize, and cost, efficiency, and security can meet their current needs.

We believe that the block chain based on PoA algorithm, including the block chain based on PoA 2.0-surface, is not a more centralized block chain but more decentralized. There are more chain-only existing outblock nodes than most items, and there are very high requirements for the selected nodes. On the contrary, although PoW and PoS can satisfy the logic consistency of decentralization in theory, in practice, we find that whether it is mining pool or staking, people usually judge their credibility by the consensus outside the chain, such as reputation.

Comment on BlockBeats: many PoS chains that advertise themselves as decentralized actually do not have many active block nodes, and the huge amount of mortgage of some large nodes has already gone against the idealized decentralized characteristics. The only technical path chosen by the chain is PoA 2.0-surface, which has its own advantages.

& have spent

BlockBeats: as security and finalization improve dramatically, the next most important issue is how to get blockchain off the ground. How do you see the relationship between PoA 2.0-SURFACE technology innovation and regulation?

Chainonly: for supervision, the first thing to be considered is the supervision of block chain application, because many problems are not only the problem of chain, but also the supervision of the whole application layer, while the bottom block chain only accounts for a part of the application, and the interaction with users is the focus of supervision.

First of all, we believe in regulatory preparation at the application system level. Second, the underlying regulation is not entirely a technical issue, but should be classified under the category of block chain governance. Thirdly, block chain governance is a problem that needs us to study seriously and continuously. It is relatively easy to make relevant innovations based on PoA 2.0-surface, with better security and stability guarantees.

Comment by BlockBeats: the healthy development of blockchain cannot be achieved without embracing regulation, and the choice between centralized efficiency and decentralized trust is the king. Blockchain is not an end in itself, but a means to a real problem.

Block rhythm BlockBeats: in fact, the ⾥ baba design regulation in the 2018 league chain nodes, right to the government's regulatory nodes ⼒ ⼤ from other nodes. If possible, will there be some reference or modification in the future design of wirechain in this respect?

Chainonly: the regulatory node is derived from the government's requirements for regulation, so according to our understanding, alibaba's alliance chain design will definitely meet such requirements. For only chain, it is still ready to increase supervision at the application level, while supervision at the public chain level will follow the governance model detailed in the white paper. In the future, only on the chain of any governance changes, will be approved by the entire network vote. At the same time, in the future, chainonly may deploy its own alliance chain system if the chain cannot meet specific requirements or application scenarios.

As national policies gradually tilt towards blockchain technology, especially in the alliance chain sector, we expect to see a more prosperous competitive landscape in the future.

& have spent

BlockBeats: a sound suggests chainless PoA algorithms & NBSP; Some, like Libra, also have union nodes. What do you think of this?

Chain-only: the earliest chain-only PoA algorithm was proposed prior to Libra, and has since been extended to today's PoA 2.0-surface, and even to the extent that Libra may have used our 101 consensus nodes and other project design ideas. At the same time, it is possible that the design of Libra's governance architecture has borrowed from only chaining, or at least coincided with it. Of course, in PoA 2.0-surface, we refer to the idea of HotStuff BFT algorithm, and Libra also adopts HotStuff BFT to make a consensus, but only because HotStuff BFT is recognized as a very good algorithm and suitable for blockchain.

From another perspective, this also shows that the PoA algorithm is a trend. Full disclosure and non-licensing does not necessarily lead to perfect decentralization, nor does it necessarily have the security of today's licenses. Libra's business scenario has a similar focus to that of chained, which may be why we've all chosen such a technical path.

The Hedera Hashgraph project, which bought headlines in western newspapers accusing Libra of copying its design, is a blockchain project that predates the Hedera Hashgraph project. In fact, rather than copying, blockchain projects that focus on commercial landing scenarios have taken a similar technical path.

BlockBeats: at the beginning of the project, BlockBeats attaches great importance to enterprise-level services, and this PoA algorithm upgrade can still reflect this theme. Others, of course, see the biggest opportunity in blockchain for consumers. How do you see the future blockchain as toB or toC?

Chain-only: chain-only will stick to the current development policy in the future. On the basis of developing enterprise-level blockchain services, we actually have a lot of cooperation or products for c-end users. For example, vichain.com and pricewaterhousecoopers have helped wal-mart China implement its product traceability strategy. By scanning the qr code on the commodity, customers can learn the detailed information of commodity supply source, geographical location and logistics process received by walmart China, which is the block chain application for the end consumer. In the future, we believe that block chain will also have more and more applications for c-terminal.

We cannot just say or assert that the block chain must be oriented towards B terminal or C terminal. No matter what the answer is, the security and efficiency at the bottom of the block chain are the foundation of all applications. Therefore, weijie has invested a lot of human and financial resources in the research and development of PoA2.0 algorithm.

Soon, the interview was over, and BlockBeats was still thinking about the topic.

Back to basics, what exactly is the tipping point blockchain needs? Landing, landing, landing.

So how to combine the commercial scene landing? In BlockBeats' view, at least, killer apps will only emerge if regulation allows them to, with the help of bricks-and-mortar companies, target areas with real demand changes. Under such thinking logic, the only link that has been completed and is waiting for the wind to rise and the Surface consensus algorithm that it has been developing with great efforts deserve long-term attention and exploration.

Reading:

New white paper of wionly:http://www.vechain.org/qfy-content/uploads/2020/01/VeChainWhitepaper_2.0_zh.pdf

SURFACE: & have spentA Practical Blockchain Consensus Algorithm for Real - World Networkshttps://arxiv.org/abs/2002.07565

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia