1. PoW+PoS

secondary title

A person named QuantumMechanic first proposed Proof of Stake in the Bitcointalk forum on July 11, 2011. At that time, there was no abbreviation of PoS, and it was proposed for Proof of Work. The article predictably proposed many solutions that are still in use today, such as replacing the weight of computing power by currency holdings, entrusting rights and interests, reward distribution, etc., and some words such as Delegate, Vote, Voting Power, etc., in 2018 today We can still see it.

At that time, people were very excited about the proposal of PoS, and this excitement has continued to this day. Many later projects have begun to study the blockchain of the PoS algorithm, and even many existing projects have begun to consider transforming PoS. Fang is the most typical example. It took about a year of development from proposal to implementation. On August 19, 2012, the first blockchain project based on PoS consensus (Peercoin) was born, and the Token code was PPC. The emergence of Peercoin truly represents the milestone of PoS from paper to engineering realization.

Hybrid PoW+PoS consensus is a transitional stage from PoW to PoS. During the period from 2009 to 2014, Bitcoin’s PoW still had high authority. Most people’s impression of PoW was that there was no major fatal problem except waste of resources and a little lower efficiency, and many Second-generation PoW based on Bitcoin, like Litecoin, is highly regarded simply because of the reduced block time. It can be said that the design of Peercoin still stands on the shoulders of Bitcoin PoW, and has inherited many good strategies, such as incentives, fork protection, consensus implementation, distribution strategies and so on. However, due to the fact that the practical process of PoS consensus is too short, PoS is mostly an exploration component in the realization of Peercoin, so this mixed consensus is actually a transitional method.

secondary title

2. Pure PoS

PoS is constantly developing. Following the birth of Dotcoin, the pure PoS consensus has become the focus of research. About a year after the release of Dotcoin, on September 28, 2013, a person with an account name of BCNext announced on Bitcointalk the start of the crowdfunding of the Nxt project (it was called Fundraise at that time, and it was not called ICO).

Nxt is a 100% PoS project, no PoW consensus exists at the same time. The highlight of Nxt is that it does not simply copy the Bitcoin code like other altcoins (altcoins), and then change the hash mechanism, shorten the time between blocks, or change the difficulty of starting, but is based on its own newly written algorithm. blockchain. Emphasize 100% PoS mining, no pre-mining. The equity (Stake) required in PoS is also completely determined by donations, and the distribution of PoW is directly distributed by donations.

The crowdfunding lasted for 21 days, 73 users participated, donated a total of 21 BTC, and shared 1 billion Nxt together. On November 24, 2013, Nxt released the main network, indicating the first 100% PoS-based consensus creation World blocks are born on Nxt.

The development of Nxt is not smooth, the forum’s history of Nxt stays at May 12, 2014. The decline of Nxt is due to too centralized development, code quality and speed have been unable to keep up, only 3 developers can not meet the expectations of many investors, although the development sequence of the main network function has been voted by the community, but Nxt is still slow Slowly, it has become criticized by many investors. After the market value reached a peak of 80 million US dollars on June 5, 2014, it left a lot of feathers. But no matter how you say it, compared with the 21 BTCs that were crowdfunded at the beginning, it is like a sky and an earth. Compared with the projects in the same period, it can be said that it is really old to welcome the new year. Looking at the dotcoin and blackcoin of the same period, they are not so good.

Blackcoin first announced its plan on February 24, 2014, claiming that it is based on pure PoS, but it is actually a mixed consensus of PoW+PoS. Then why the mixed consensus is mentioned here in the pure PoS, because the PoW running on the black currency only lasts to a height of 10,000 blocks. According to one block per minute, the mining time is about a week (6.9 days), and then the entire block The verification will no longer accept PoW mining, but instead form PoS mining. From this perspective, coupled with the chronological order, I attribute black coins to the pure PoS camp.

The maximum total amount of PoW mining is 10,000,000, and after entering PoS, the annual inflation is 1% (like Nxt, there is no maximum limit). Pure PoS has a shorter block interval and a more reasonable coinage method, which has attracted many early adopters to mine, and even many specialized mining pools have begun to collect coins on the forum as Validators. At that time, PoS, a new type of consensus, did attract the attention of many people. It did not require a computer or a GPU, and you could get rewards as long as you held coins.

But the good times of HeiCoin did not last long. On April 14, 2014, in less than 2 months, the market value was nearly 20 million US dollars. After that, it fell all the way. After several twists and turns, it still failed to escape The bottom of the doom. And because of the decline in the market value, the 1% additional issuance reward seems to be a drop in the bucket for PoS Staking people.

The first fair-launch pure Proof of Stake coin. No premine. No ICO. Community focused.

Back again, better than ever, stronger than before.

Although, in 2016, a new group of developers restarted the development of HeiCoin, and spent a lot of money to invite the regular army, corporate, and comprehensive publicity, but HeiCoin and its PoS did not return to the vision of the mainstream world. come. I still remember the classic slogan of Heibi, which exists like a textbook: the first one based on pure PoS coins, fair start, no pre-mining, no ICO, community-centric, returning, better than ever, more than before powerful.

On July 18, 2014, a coin similar to the black coin consensus, called Shadowcoin (Shadowcoin), is an application coin based on chat anonymity, and also based on the PoW+PoS hybrid consensus. PoW only digs 31,000 blocks, and then enters PoS, but unlike Blackcoin, Shadowcoin sees a lack of incentives, and it increases the annual inflation rate, bringing Blackcoin's 1% per year to 2% per year. But after entering 2015, the value of the application currency is far from being seen, and with the arrival of the bear market, the shadow currency will soon disappear from people's vision.

I summed up the entire history of POS and found that many PoS-based projects basically appeared in 2012~2014. The exploration paid the bill, but generally speaking, during that period of time, everyone looked good on paper. From 2015 to 2016, the currency circle entered a big bear market, and many projects and developers entered the hibernation stage, and few projects appeared to call for donations, IPO or ICO, and PoS research also disappeared in the market a period of time.

Excluding the reasons for the large market cycle, we attribute the reasons for the failure of the PoS mechanism currency during that period. Bitcoin, Nxt, Blackcoin, and Shadowcoin are all early exploration projects of PoS, which have not been fully considered technically (in the ninth In the chapter we will introduce the technical defects of PoS), in addition, the incentives for Staking in PoS are too small to support the cost of validators, and the application of the decentralized operating mode on PoS is not as perfect as that on PoW .

The annual increase ratio of 1%~2% cannot allow miners to maintain, especially in the bear market stage. This start-up PoS additional issuance model is still fresh enough for many people. From the point of view of the currency price that may explode initially, the initial benefits of additional issuance are only a very important incentive, but when investors face that the currency price cannot bring After the expected benefits come, low incentives become a tasteless one. Miners can't see the benefits, so they can't participate in network verification, so the network will either be blocked or die slowly, and the potential attack cost will be lower and lower.

So far, this is the second stage of the development of PoS.

secondary title

3. Delegated PoS (DPoS)

Delegated PoS, also known as Delegate Proof of Stake, DPoS, its prototype was born on December 8, 2013. Daniel Larimer first talked about using voting to select block producers in bitsharetalk, instead of random numbers that may appear in PoS. The problem of manipulation is described like this:

I recently came up with another system that could work, though I am not fully sold on it.

What if the share holders elected people who would sign blocks?   You vote for the signers with CDD. When they sign they spend the CDD they accumulated from people voting for them.  No one would be allowed to sign more than 1 in 100 blocks.  

This plan has many potential problems, but I thought I would add it to the idea pool.

I recently had an idea that worked, but I'm not entirely sure if it will work yet. Why not allow token holders to elect validators for signed blocks? Holders use CDD coin-day-destroyed to vote for the signer, and the coin age of the voter will be consumed when the signer signs. No one will be allowed to sign more than 1 block in 100 blocks. This idea may have potential problems, but I will add it to my library of ideas.

This is the rudiment of Delegate PoS. This system has been successfully applied to Bitshares. The slogan of the BitShares community for DPoS is: DPoS not only solves the problem of wasting resources in PoW, but also solves the problem of Bitshares. And the PoS problem of Future Coin. At the time, DPoS was indeed a solution. Although the development of BitShares was not smooth, and Dan was eventually kicked out of the BitShares community, it is undeniable that the delegated stake DPoS consensus will lay the foundation for the further development of PoS in the future.

BitShares' DPoS is built on the Graphene platform. This set of graphene technology was later developed into a dedicated bottom layer for DPoS. Blockchain projects based on Graphene can directly use the DPoS consensus mechanism, eliminating the need for repeated development. Distressed, Steemit, EOS, Lisk, etc. are all DPoS public chains built on the graphene platform. Until now in 2018, we can still see that in Bitshares DEX, many gateway assets are based on graphene-based DPoS, and the number of projects based on PoS consensus has reached an order of magnitude.

Indeed, DPoS has many advantages. DPoS solves the problem of untimely use of rights and interests. It is a basic principle for professionals to do professional things. It can be understood as a representative system in national governance, or the system of people's congress representatives. Residents in towns and villages elect township heads through elections , and then the mayor, county head, governor, etc. The biggest advantage of this system is that it solves the problem of inefficiency caused by too many validators. Of course, this system also has obvious shortcomings, which is the criticism of centralization.

The number of initial node verifiers of BitShares is about 100, and then gradually reduced to more than 30 (because it is not profitable), and the number of node verifiers of EOS is fixed at 21 people. In the blockchain world where the concept of decentralization is highly respected, projects that tend to be centralized are easily criticized, and many developers will not agree with this practice of entrusting rights and interests. Another project that originated in 2014, Tezos, better combines decentralization and delegated rights.

Tezos’ PoS, the development team prefers to call it LPoS (Liquid Proof of Stake), but it is not an exaggeration to call it DPoS, because the delegated stake is an extremely important part of Tezos’ PoS. In order to ensure system security, PoS hopes to increase the Staking amount of system tokens, so that the cost for attackers to launch a 51% attack will become very high.

Tezos solves the problems of insufficient initial incentives and low willingness of users to staking, and rewards staking users with an annual increase of 5.5%. At the same time, users can entrust their rights to validators (called bakers in Tezos). Let validators help delegators exercise their rights. The verifier can charge a little handling fee from the user as a service fee, so that the verifier node can also get the reward for accepting the entrustment in addition to the reward of its own staking. This kind of incentive is 1%~2% of the future currency and black currency. , much higher.

In addition, the delegation of Tezos is different from that of EOS in many aspects. In principle, each delegator can choose any candidate to vote, but the voting incentives of Tezos and EOS are different, resulting in different results. On Tezos, the delegator can select any validator, vote and obtain benefits, while the validators selected by EOS delegates only have the result of being elected or not, and there is no incentive.

Tezos launched its main network on June 30, 2018. Currently, there are more than 400 main network nodes and more than 100 validators. EOS launched the main network on June 15, 2018, with 21 main nodes and 100 candidate nodes.

4. BFT+PoS

secondary title

The combination of Byzantine Fault Tolerance (BFT for short) and PoS was first proposed in 2014 by the Tendermint team, Jae Kwon. The predecessor of Tendermint is based on the proven BFT algorithm developed at the Massachusetts Institute of Technology in 1988, and Tendermint has been optimized and improved. However, since Tendermint came out, it has not been operated into the project. Jae Kwon later created the Cosmos project, which can be said to be the first project based on Tendermin.

According to the mainnet launch time, the application of BFT+PoS or BFT+PoS-like projects first started in 2018. First, EOS brought BFT+DPoS online, and then Tezos’ BFT+PoS-like algorithm was launched one month after EOS is also online. Cosmos, the first project based on Tendermint, will launch the main network in Q1 of 2019. In addition, what I know now is that Ethereum and Algorand will adopt the BFT+PoS consensus algorithm, and the specific launch time is unknown. Although Tendermint came out first, Cosmos took the lead, but it has to be said that the latecomers basically refer to the design of Tendermint, so Tendermint still plays a very good role in the application of BFT in the project. From 2018 to 2019, there will be a consensus that BFT will launch a large number of projects, and a hundred flowers will bloom.

BFT fault tolerance can bring considerable anti-risk capabilities to distributed networks, and can also bring faster block confirmation speeds, solve the problem of blockchain forks, and bring a certain degree of final certainty in transactions, indirectly improved performance. If pure PoS only solves energy consumption compared with pure PoW, then compared with pure PoS, BFT+PoS not only solves energy consumption, but also solves the problems of PoW transaction confirmation, transaction fork, and lower performance. , the leap brought by it cannot be ignored, especially in terms of security, let’s look at Tendermint:

A very important part of Tendermint is its consensus engine part (Tendermint Core). This consensus engine combines BFT and PoS consensus. Token holders get the opportunity to verify the block through stake. After the signature is completed, the verification of the new block passes 3 Round submission, 2 rounds of 2/3 voting process to reach a final consensus. Since the voting starts after the block is broadcast, when it completes the verification, the new block obtains instant finality (Instant Finality), which is different from Bitcoin transactions that require 6 block confirmations, Tendermint greatly shortens The verification time is shortened, so that the transaction can take effect in time, and the problems such as transaction rollback and double spending are avoided in advance.

In addition to the transaction confirmation, Byzantine Fault Tolerance can tolerate less than 1/3 of the verification weight downtime in the system at the most, which can ensure the operation of the system. It can be simply understood that as long as most people in the system are honest, regardless of other verifications No matter what people do, the system can still run, and there will be no stop or downtime.

The first project based on the Tendermint Core engine is called Cosmos. Cosmos is a public chain project aimed at providing cross-chain functions. The Cosmos hub is the most important part of this project, and it is also the general coordinator of the cross-chain. If the cross-chain ecology is prosperous and multiple links are connected to the Cosmos center, then the security of this center can be seen how important it is. When it comes to Cosmos, we have to mention another cross-chain project, Polkadot. Polkadot is also a BFT-based PoS project. It is led by Gavin Wood, the author of the Ethereum Yellow Paper, and has gathered many technical experts. The team lineup can be described as strong . However, when the author wrote the series of articles, Polkadot's GRANDPA consensus was still being improved, and the specific analysis needs to wait until Polkadot is about to go online.

In 2014, BFT began to be introduced into the blockchain distributed network, and then PoS became popular. The mixed consensus of BFT+PoS has been valued by many projects, not only because the security of BFT algorithm has been studied for nearly 30 years Time, and because this algorithm improves the anti-risk capability and performance of the system, I predict that this consensus will be popular for a long time in the future.

secondary title

5. A New Era of PoS

PoS research is still advancing, and security, performance, and decentralization have always been the pursuit of PoS. Some new projects launched after 2018, I see that they have more or less the shadow of the predecessors, based on this, either optimize the performance, or optimize the security, or the ultimate optimization is not possible iron triangle. Algorand and Thunderalla are new projects that have entered my field of vision since 2018. I think these two projects can represent the new PoS era to a certain extent. If the development of PoS is divided into a 4-year cycle, it happens that these two projects will start to take shape around 2022.

Both Algorand and Thunderalla stand on the shoulders of giants. Algorand, led by Turing Award winner and MIT professor Sivio Micali, focuses on the new PoS+BFT consensus algorithm (referred to as BA, Byzantine Agreement), which aims to make the election of validators more random and unpredictable, so as to better guarantee The security of the system is the research direction; Thunderella, initiated by two Cornell University computer professors Rafael Pass and Elaine Shi, aims to solve the performance expansion problem in PoS. At present, the theoretical performance can reach up to 1000~2000Tps.