Babbitt column 丨Chen Zhiwei: Libra technical papers make people feel full of sincerity

We were the first to send Libra technical papers to the Gemlink blockchain group. The next morning, we saw someone translating the paper. After a cursory reading, there are only two words "sweat". The mistakes are all out, actually translating the transaction into things, obviously machine translation. Of course, it also reflects the lack of professionalism, so it is not competent.

The academic circle and the business community are a biological chain. I will talk about it when I have time. In the face of a new technology, the academic world has great advantages, especially when this technology has high thresholds, such as cryptography, it is difficult to shake without special learning and cultivation.

Having said that, in the past two days, we have published an interpretation of Libra (Interpretation 1, Interpretation 2), which has aroused widespread consensus and discussion. Next, we want to interpret the technology so that everyone can have a clear understanding of Libra. This thing is estimated to be the best we can do (refer to our Bitcoin white paper analysis), because we are in the academic circle, but also in the industry, Gemlink's consistent thinking is to do the development of cryptography on the blockchain, It is also the first company in China to focus on the development of cryptography on the blockchain.

For the general public, the most concerned is that Libra is a blockchain, can mine, how about the incentive mechanism. Friends at a higher level are concerned about what the consensus algorithm uses, how it performs, what programming language it uses, and so on. Explain one by one below.

• Resource aggregation in the PoS era, cloud service provider contract node work
• The first 5G+ blockchain traceability platform landed, and a century-old multinational company Cargill took the lead in testing water?
• The king of money returns to $10,000, what makes it stand still?

After reading the entire technical paper, I realized the sincerity of Facebook. Because the whole paper refers to Libra as a database, a certifiable database based on cryptography. Switched to other white papers, I am afraid that others will say that they are not blockchains, and dare to use the database three words.

However, Facebook is very honest. Because they want to put the project to the ground, please see the first sentence of the second paragraph of the last part of the paper:

We are working toward an initial launch of the system, and to keep it within a manageable scope, we plan to make several simplifcations in the frst version.

What this sentence means is that in order to land Libra, some things are simplified in the first edition.

So this is why I call this technical document a thesis. The whole article is written according to the style of the paper, and it is well-founded and completely different from the white paper you have seen.

Since Libra is a decentralized database, what type of blockchain is it?

It is a chain of licenses, and the popular one is the chain of alliances. Because it is to be used, only the alliance chain is technically the most feasible. The paper also pointed out that in the future, it will gradually develop towards the public chain. No one knows what month. In short, at present, if you want to really apply, only the alliance chain is the most appropriate.

Then many people may ask: Is there an incentive mechanism? Can you mine?

Since there is no incentive mechanism for the alliance chain, it is impossible to mine. Who is going to book?

Libra gives a detailed explanation in this regard, for the performance of the system, the accounting members are from the Foundation members. In the future, Libra will extend the billing rights to members with Libra currency.

So is there an incentive mechanism? Tell you exactly: no. It’s all about self-initiation. There are no outsiders in the Libra system, only their own members.

Someone has booked it, so how to reach a consensus?

The system uses a consensus algorithm called LibraBFT, which is an improved version of the HotStuff consensus protocol. It has Byzantine fault tolerance and can reach a consensus between 3f + 1 nodes, as long as f nodes are honest. This is a technical need and prevention. In fact, you think that the nodes are shareholders and foundation members. It is impossible to make a lot of mischief.

Libra is not building a consensus algorithm from scratch, which is quite different from some white papers (using other people's algorithms, but it is a revolutionary creation). In choosing a consensus algorithm, Libra chose to support at least 100 nodes and expand to 500-1000 nodes in the future.

In view of the above criteria, Libra chose the HotStuff consensus protocol. The protocol's security proves to be simple, easy to implement, and faster. The Libra system can reach a consensus through 3 rounds of network communication.

All of this again shows that Libra is not playing virtual, everything is for landing applications. Think about EOS, you can see the difference when you compare.

What is the performance of Libra?

The paper clearly states that it supports 1000 transactions per second, and each transaction takes 10 seconds from production to final confirmation. Students who are familiar with Bitcoin must think of Bitcoin to make a block every 10 minutes, and confirm that there are still 6 blocks. Libra assumes that transactions need to be done under the chain and can also be done through the payment channel.

So how does Libra come out?

Regret to tell you that Libra does not have the concept of block. Because it is a database from start to finish, only the concept of a transaction list.

What development language is Libra built in?

Use RUST. It is also wise to choose this language. In addition to security, the language has a big feature that is highly compatible with payment transactions. The choice of language shows that Facebook is indeed paying for the transaction scene.

In addition to the above features, the paper repeatedly said that it is only a system prototype, and some aspects have been simplified. However, there are several main components, and the rest of the work can be developed on this basis. In the early stages of Libra, the use of externally recognized Foundation members as bookkeepers simplifies incentives and increases system responsiveness. In order to apply the application, the user cannot define the rules by himself. After the MOVE language parameter specification, the Libra system will be gradually opened in the future.

Article source: WeChat public number btc201800

We will continue to update Blocking; if you have any questions or suggestions, please contact us!