Jan 29, 2020
If 2019 Was the Waterloo of Blockchain, Then How Can Public Chain Rise Again In 2020?
QuarkChain CMO’s Sharing on New Year’s Eve
I am Anthurine Xiang, the CMO of QuarkChain, also one of the co-founders of QuarkChain. I graduated from Shanghai Jiaotong University and Johns Hopkins University. Prior to founding QuarkChain, I spent two years working in finance on Wall Street and 6 years in the tech industry in Silicon Valley. Thus, you can call me a veteran of the Internet. In the past, I mainly worked in the area of data science and product development. Two years ago, together with Dr. Qi Zhou and Mr. Ting Du, we founded QuarkChain. It is hard to believe that two years have already passed and our project has matured.
Some of you may have little information about QuarkChain. Therefore, please allow me to give a brief overview here. QuarkChain is a blockchain infrastructure design that is based on the idea of sharding, which bears high flexibility and scalability. Sharding may be a familiar term to some, mainly because of Ethereum 2.0 and also of the 7 or 8 projects that sprang up after us. When we first started QuarkChain at the beginning of 2018, many people have questioned whether sharding was a sounding technique for blockchain. Currently, we are the first public chain to successfully implement state sharding. Also in terms of throughput, we are the first project which focuses on heterogeneous sharding.
Heterogenous sharding means that for each shard, its consensus mechanism, ledger, transaction model, and token economy may not be the same. Users can configure each shard based on their needs while all the shards can still interact with each other efficiently. In this way, such a design can adapt to the changes in the industry rapidly and satisfy the diversifying needs in the industry as well.
About QuarkChain
QuarkChain is the first public chain to implement state sharding. Thanks to its great design of sharding function, QuarkChain has realized a high TPS. At the same time, it makes customization available for four factors: consensus, token economy, virtual machines, and ledger model. QuarkChain has become a flexible, scalable, and user-friendly blockchain with the highly secure, decentralized, and highly efficient underlying architecture.
Today, I would like share my thoughts about public chains with you all. In truth, public chain was a hot topic back in 2017 and it has been gradually cooling down since then. Till now, some may think 2019 was the year of Waterloo for public chains. Is it really like that? Based on some observations of the industry, let me share a few viewpoints.
Before 2018, or more like the second half of 2017, it was the hottest time for public chain projects. Basically, if you have a whitepaper, you can raise capital easily since many had high expectations and imagination for public chain; the investors and the public thought it was omnipotent and could be used for everything. Our project also got started around the beginning of 2018 and we launched our mainnet in 2019, together with many other projects. By 2019, many supporters may think they have read the preface of the story of the public chains, a long and magnificent one. However, after the initial excitement, they realized that for many projects in 2019, after the launch of mainnet, the story became blank pages with no continuation. Such an impression left many in question.
Let’s first take a look at the projects who have launched the mainnet. Their market performance was not comparable to the glory of the previous year. Like I said, the hype of public chains is cooling down while IEO, mode, and DiFi stepped up to become the key words in 2019. When I told someone I worked on a public chain, it would seem as if he or she may look down on me, which contrasted the dreamy, curious face I would see back in 2018.
Currently, the market performance is quiet and unimpressive, since no applications with massive adoption was born amidst high expectations from the public. If one concludes that this is the end of blockchain based on these two facts, then I would think this view may lack consideration and accuracy. I think in 2019, the direction of public chains has changed from what was in 2017 or 2018. Its development is still under way and is undergoing a transformation. I would like to analyze its behavior and share with you its future directions.
Let’s establish a few baseline understandings of blockchain. Currently, all the structure of blockchain infrastructure includes all the tokens and public chains we heard of, such as Bitcoin, and Ethereum, anonymous tokens, ZCash, Grin, EOS, and all public chains. Blockchain infrastructure consists of four important components, the first of which is called consensus mechanism. We often talk about Proof of Work (PoW), or Proof of Stake (PoS), or Delegated Proof-of-Stake Consensus (DPoS). Like what others shared previously, Conflux and Algorand were working on their own consensus mechanisms. In fact, a lot of public chain projects are innovating on consensus mechanism and market that as a selling point.
The second component is called transaction mode. Actually, transaction mode includes virtual machines. As we know though, Bitcoin does not have virtual machines and its own transaction mode is called Bitcoin transaction model. For those programmable public chains, its transaction mode refers to virtual machines. For example, the transaction mode of Ethereum is EVM and Libra to Move, which is a virtual machine that Facebook has invented.
The third component is the ledger model. Currently there are two popular models. The first type is UTXO and the second type is account base.
The last component is the token economics. Since anyone can join a public chain freely unlike under the scenario of alliance chain or private chain. Thus, token economics plays a critical role in balancing the dynamic of users. For example, the number of Bitcoins released is halved from cycle to cycle while that of Ethereum is expanding. Each project would have its own logic designed for token economics.
The essence of blockchain technology comes from the arrangement and combination of the following four components：
- Consensus (PoW, PoS, DPoS, PBFT),
- Transaction model (BTC transaction model, different virtual machine, privacy transaction model),
- Ledger model (UTXO, Account model) and
- Token economics.
After this introduction, I will use these four terms again and again in the following part of the discussion and you will be very familiar with it.
The first problem of public chains is speculation. Some public chains did little for the industry and achieved minimal progress. Back in 2017 and 2018, as I mentioned earlier, while public chain was “hot” and it was easy to raise funds, some people just created new chains by creating a hodgepodge of different components of existing chains. Let’s give an example here: I can take the A component of Bitcoin, the B component of Ethereum, and the C component from elsewhere to combine into a new “chain”. After copying and pasting these components together, then one could launch a coin and call that a new project. Such mishmash of components was simply a rehash of old ideas with little innovative ideas. Instead of energizing the industry with new directions, such behavior brought a lot of noise to the industry.
Here comes the second problem, which is the obsession chasing after TPS. QuarkChain started out also to solve the TPS bottleneck. A lot of us shared similar thoughts because we thought Ethereum was too slow to be commercialized. If we can solve the TPS bottleneck and let blockchain run faster, then the industry would transform into something different. Currently, the community have proposed a plethora of ideas for TPS enhancement. For instance, vertical expansion can speed up a single chain by adopting a new consensus mechanism like DPoS. At present, some new consensus mechanisms are being developed with the aim of enhancing TPS and speeding up a single chain. Our approach is different: we expand horizontally by adding multiple chains, which is a viable solution as well.
Even though there exists substantial room for improvement in TPS, the solutions proposed by different projects would serve to overcome the TPS bottleneck eventually. I am not saying that TPS is not important; rather, improving TPS only would not singlehandedly solve the problem of the entire industry. Some public chains market themselves as having a high TPS, which is great in and of itself but then it does little to invigorating the industry with new ideas, building the ecosystem of the industry, or building landing applications.
The third problem harks back to the four components that I just introduced before. Some projects were just shuffling different components and did some mix-and-matches with little innovations. However, many public chains realized innovations on one particular component, particularly regarding consensus mechanism. I hope that more resources will be dedicated to innovate on all four components. Be it virtual machine or consensus mechanism, or token economy, any innovations are great developments to push the industry forward.
The real question at hand is that, if one achieves some breakthroughs in just one of the four components, is it advisable to develop a completely new chain by tweaking that one single component using the innovation? After much deep thought, I find such proposal unnecessary. From the innovator’s perspective, he or she may think that what he or she has built is the most cutting-edge technology in the entire industry; however, as new ideas continue to appear, the “not-so-new” chain would not be able to incorporate these ideas. In essence, this kind of framework has little practicality.
With only a single point of innovation, there is no need to create a new chain. I will substantiate further in the later part of my speech. In my opinion, a real public chain should be a framework, which is flexible to incorporate new features as more technology breakthroughs come along. If I have a new consensus mechanism and then I create a new public chain. Soon after, as newer consensus mechanisms come along, then my public chain will become obsolete. It is great that many of us are interested in enhancing different components, but we should distinguish these single-point innovations from distinctive, new “framework”.
Last problem that I would mention is that the rhetoric that many projects have been using is failing. The rhetoric goes like this: enhance TPS, develop community, launch mainet, and push out dApps. It does not matter whether those dApps have a large user base or not since the dApps are quite homogeneous with few users. The public rejects this rhetoric because it see few benefits in supporting Dapps with no users. On top of that, the public is also disappointed by the public chains who failed to develop the killer dApps promised.
The answer to these two questions are here. Firstly, we have to acknowledge that there are few application scenarios for dApp besides DiFi or gambling. Secondly, developers who are capable of developing dApps do not enjoy making dApp. I learnt of this by talking to the developer community. They have this view since they have to work with the constraints imposed by smart contract coins that the dApp generates. For example, the coin created in dApp for Ethereum is called ERC20token, TRC20 for TRON. These coins are problematic because it can only serve as a proof of asset with no application values and cannot deploy smart contracts. For gas fees, one would still need to rely on the native tokens so these smart contract tokens bear no practical values besides a proof of wealth.
Because of that, those dApp developers have to start building a new chain from scratch, such as CryptoKitties. CrytoKitties develop its own chain instead of purely using ETH because it thinks that developing its own chain gives itself more say in direction and its own tokens bear actual usage besides as a proof of value. This train of thought yields more public chains inundating the market, in addition to those who are reshuffling the four components and those who achieved single-point innovations.
Returning to the topic of killer dApp, I think that in the future, killer dApp will be born out of existing Apps where blockchain will enable new functionalities. For users, they do not need to know or understand the backend logic. Therefore, dApp is more company-facing. For a company, it would prefer having its own chain because it wants to dictate its own ecosystem. ERC20, or smart contract tokens similar to that, would not satisfy demands like that, which circles back to the reason why projects prefer building their own chains. Currently, public chains only allows developers to build simple dApp but not to construct another chain on top of its infrastructure. The reasoning behind such decision is because constructing another chain will translate into starting another new public chain project. Also, with that permission rule, such dApps can only use tokens that serve as a proof of assets.
All in all, there are four main reasons why the current sentiment of public chain seems a bit down and out. As I said before, public chain have been selling the idea of high TPS, launch of mainnet, and launch of dApps. Such idea is no longer working and we are expecting public chains to head to a new direction. Instead of launching a powerful dApp or innovating on one of the four components, the public is expecting something more.
In this regard, what are some of the characteristics that a good public chain should have? Firstly, it should have high throughput. Earlier I said that high TPS is not the only factor in determining whether it is a good public chain project yet it is still a critical factor in and of itself. Secondly, the chain should be flexible and can continuously absorb new technologies for upgrades. Thirdly, it should serve as an open platform for the community to launch a variety of projects, such as dApp or building chain.
Before, when none of us knew how the cloud works, companies would build all the infrastructure themselves. Nowadays companies simply host the infrastructure layer on the cloud and build applications on top of it. Similarly, one can apply the same concept for the blockchain industry: since not all companies have the ability to develop a blockchain infrastructure from scratch and solve the problem of scalability, storage, and usability. No one should reinvent the wheel all over again. Rather, one should focus on tweaking a few aspects on top of an existing infrastructure and focus on its competitive advantages in any of the four components. But the question at hand is that current public chains cannot satisfy this usage and that it allows only dApp to be built on top of its infrastructure.
Despite all those problems, I see some solutions out there that can address the problems aforementioned. I hope that these solutions would flourish in 2020.
The first solution is sharding and the second solution is cross-chain. My idea of sharding may be a bit different from what you may have. So is my idea of cross-chain. I will delineate further here for you.
What I am referring to as sharding is heterogenous sharding. Heterogenous sharding treats each shard as one chain and each chain can configure the four components we aforementioned based on its needs. Such design allows new technologies to be incorporated into a chain and such chain can be embedded into the overall system design easily. So different chains can host different consensus mechanisms, token economics, and ledger models.
Let me use a diagram to explain the four components that we just talked about. For each public chain, I have to define one token economics, one consensus mechanism, one transaction mode, and so on. For those chains with a single point of breakthrough, it modifies only one of the components of existing blockchain. With sharding, one can add more shards to the root chain and configure each shard differently based on the needs using the latest technology. Shards can interact among themselves and transfer data across the chains.
Some asked me this question before: would there appear a new public chain that rules the space? I replied by asking the person how new the new chain would be? I asked that because I think there were no ending points in technological advancement. There will always be new technology surfacing during upgrades and new releases. For example, from Ethereum 1.0 to 2.0, the community discovered that it was difficult to upgrade from 1.0 to 2.0, and thus decided to rewrite it from scratch. However, if you have a flexible infrastructure that allows you to just add new shards to implement the upgrades, then the infrastructure would stay popular instead of being irrelevant during the currents of competitions.
In terms of cross-chain, I am not talking about cross-chain between different public chains such as between EOS and Ethereum whereby the two have completely different infrastructure. It would be a daunting task to realize. What I am referring to as cross-chain is a form of API that allows different infrastructures to communicate across each other using the same data structure. The exemplary projects in this space would be Polkadot and Cosmos.
In fact, the idea of polkadot and cosmos are similar to that of sharding. Polkadot provides a hub that allows users to release a new chain easily. Among the released chains, one can cross the chains for communications. At this stage, each chain can pick its consensus mechanism out of the three options that Polkadot provides. For protection, Polkadot’s hub will provide hashing power to protect the chains launched under the hub.
Cosmos has a similar strategy: users can leverage its API to launch a new chain easily but one would need to take its own risk in protecting its from attacks. Users are welcome to customize the configuration of each of the four components.
Structurally, heterogeneous sharding bears resemblance to Cosmos and Polkadot. There will be a root chain on which shard chains are added. For each shard chain, one can configure each of the four components flexibly. Each shard chain can be launched as a new chain and one may also launch a group of them all at the same time, which is the key differentiator here. Among shards, one can carry out cross-shard operations freely, which is also called cross-chain in some scenarios.
To summarize, I want to highlight that the public’s understanding of public chain and its expectation has changed significantly since 2017. In the conventional sense, public chains may only have innovations in one single component or just build a dApp and so on. The new concept of public chain is different: it is an infrastructure framework on which one can build public chain and dApp. Such framework can also adapt to new technologies sailing through the current of time. Its components are not fixed with one consensus mechanism, one token economics, one ledger model, or one transaction mode. Its components continue to evolve and adapt in order to survive the competitive and ever-changing environment.
As we said before that after a public chain project has been open-sourced, another project can fork and create a new branch. If there are new projects appearing, does that mean we would need to fork every single one of them? This kind of ecosystem is incomplete and discontinuous. Like the development of cloud computing, if one can do well in building the back-end infrastructure, then one can safely pass the torch to those who are good at building the front-end, be it business development or in other ventures. In the same way, not every one excels at building the infrastructure. It is more logical to entrust such construction to the back-end experts who are capable of designing a flexible and interoperable structure to satisfy the demands of the future and adapt to new technologies.
Lastly, I would wish that in the future, public chains can play a role like what Amazon does with cloud computing. From what we know in history, in the empire, long divided, must unite; long united, must divide. This is an inevitable phenomenon. With all the splintering public chain projects in space, we foresee some form of consolidation will happen. While it is gratifying to see all the passion poured into innovating some building blocks of the blockchain, it is questionable whether each little improvement requires one to launch a completely new chain.
I hope that in 2020 our heterogeneous sharding and cross-chain will be further improved. Soon we can witness how such an infrastructure framework allows interoperability and flexibility and fit the needs for all to build dApps or launch a new chain. Thank you and I look forward to exchanging ideas with you all after the talk.
Website
Telegram
https://twitter.com/Quark_Chain
Medium
https://medium.com/quarkchain-official
https://www.reddit.com/r/quarkchainio/
https://www.facebook.com/quarkchainofficial/
Discord
