Linked List

Data structures are everywhere. Every developer should know how they work under the hood, starting from the most common ones. You should pick the data structure that is the most suitable for your…

独家优惠奖金 100% 高达 1 BTC + 180 免费旋转

Introduction to Nervos Network

Nervos Network is a layered network for crypto-economy. Nervos CKB (Common Knowledge Base) is the Layer 1 blockchain that is optimized for Layer 2 solutions. There is no consensus faster than ‘no consensus’, Layer 1 is hard to scale for a reason. We need to accept the fact that the underlying blockchain is slow and expensive, and this is what actually makes it secure.

Given this constraint, we need to make the best use of the trust provided by the blockchain at Layer 1. This is exactly what Nervos CKB does. The CKB design lays out three fundamental requirements for Layer 1:

1. Instead of a store of value, CKB needs to be a store of assets;

2. CKB should be capable of being a Crypto Court when necessary;

3. CKB should be Layer 2 friendly.

To meet these requirements, the Nervos CKB has employed a unique design in terms of protocol and engineering:

1. A RISC-V based simple VM, which is capable of crypto-based arbitration and crypto-primitives friendly;

3. An architecture that is a generalization of Bitcoin

Here we’ve compiled a summary of questions and discussions from this panel:

Eric: Yes. Layer 2 provides features such as privacy-preserving and batch transactions.

Patrick: Solutions like Sharding are basically trying to increase the throughput of the blockchain, while Layer 2 is actually for reducing the load on the network. There are basically no transaction fees and you can get instant finality on Layer 2.

Lisa: In the case of state channels, when considering the watch tower problem, in certain cases the necessity of Internet connectivity can be eliminated as well (two parties can connect to each other directly), which could enable another category of applications.

Liam: Time-to-Finality on a Layer 1 system is far worse than it is on a Layer 2 system.

Mo: The final target of scalability is to make value transfer as fluid as data transfer on the Internet. This is simply not possible with only a Layer 1 blockchain for various reasons, for example, even with Sharding, there’s still a small portion of transactions that can not be executed parallel.

Tom: On a blockchain, you will get distributed global consensus. At Layer 2, you are trading this for local consensus, which is always going to be quicker and less demanding.

Patrick: In a case study (implementing a Battleship), I found that if an application is not reasonable to build on Layer 1, then it’s not really reasonable to build on Layer 2 as well. For example in the case of Lightning Network, we don’t need to care so much about safety. We can always get our money back upon dispute. However, in a state channel, we need to consider liveness. So if we get a better Layer 1, we will also need to have a better Layer 2 for it.

Mo: As Layer 2 researcher, I have a long wishlist for Layer 1. The first thing is handling multi-signature more efficiently. Another wish is for the ability to directly inject state into a blockchain. (Watch video for a detailed explanation of these two points) A key thing I always consider is ‘Should this part of work be done on Layer 1 or Layer 2?’

Liam: I think the problem is about whether the cost of converting off-chain state on to the blockchain is really significant.

Patrick: In a workshop in Berlin, we discussed the problem of ‘Should we minimize on-chain state or off-chain state?’ In the battleship case, we found that the state we send to the blockchain should be minimized so that it can actually fit into a block.

Mo: On the other hand, I think it would be very beneficial if we can directly inject a multi-sig signed state into a blockchain that can handle it perfectly.

Patrick: In the dispute process of state channels, there is a congestion problem that can arise in the dispute window. The blockchain could be congested and the fees could go up making it expensive to send things to the blockchain. I think it would be great if there was an op-code that can help determine how many blocks are reasonable for settling a dispute by looking at previous blocks.

Lisa: When we relay transactions, some of our money is locked for these transactions for quite a while. I would hope for this to be fixed, but I guess this is actually the cost of trust.

Eric: I have a question regarding Virtual Channels: could the channels reform themselves to make the whole network more efficient?

Lisa: The idea of a Virtual Channel involves renting a path for sometime which you can use freely during this period. It makes more sense for high frequency and small payments for two parties who do not want to use intermediates. However for a single transaction or transactions that go over a lot of hops, it might be better to open a new channel to cut the path shorter and make it more efficient.

Tom: Considering the incentives in the economic model: if the liquidity reaches a lower level, then the cost of transactions would go up, which in turn would reduce the number of transactions. The system should be able to balance itself.

Mo: We have a Liquidity Backing Auction project for this problem. It allows liquidity providers to lend liquidity to peers. Of course it has a trade-off of trust to some extent, you need to trust the liquidity provider in this case. This solution is for lowering the barrier to becoming an operator or a node in the state channel network.

Mo: The protocol is constantly updating its peers on “pressures”, which are propagated among the network, instead of based on a routing table. This kind of hop-by-hop routing is relatively efficient. There is indeed a communication overhead, because theoretically for every payment you deliver you have to update the pressure information to all your peers. So this kind of implementation is an engineering challenge and we are handling it right now.

Patrick: In payment networks like Lightning Network we have a privacy-preserving feature that only relevant people have the information about a transaction. In the case of Celer’s routing algorithm, could it be compromising this privacy feature? As the Celer network lets every node know about the “pressure” information, could people potentially figure out information about transactions based on that?

Mo: That’s definitely an interesting question, but I think that will probably not be the case. Because this kind of hop-by-hop routing is actually state-link based, it won’t propagate the actual topology of the entire network. From the perspective of a single node, the node will only know the pressure of delivering a transaction to another specific node, but not necessarily know the path of each of the hops. This is just my intuition for now, but we will conduct more scientific research on this in the future.

Mo: Celer Network has a State Guardian Network, which is a distributed state guardian system that anyone can participate in. It can better solve the problem of the availability of state data. The State Guardian Network uses cryptoeconomics to maximize liquidity. Celer solves liquidity problems by introducing two mechanisms, Proof of Liquidity Commitment (PoLC) and Liquidity Backing Auction (LiBA). In the PoLC mechanism, those who have a large number of native tokens can obtain CELR tokens as incentives by locking their currency in the system. In the LiBA mechanism, CELR tokens can be borrowed from holders with a large number of tokens in the form of an auction, and the lender can obtain interest. CELR tokens are a priority document when borrowing, and are also used in the State Guardian Network to maintain system security and reliability.

(Jeff showed up on the side of the panel and decided to create a bit of drama lol)

Jeff: I think with sidechains, some additional “trust assumptions” are introduced to the network. In the case of Celer’s State Guardian Network, I think there is an assumption that there are people in the market who actually value this token so that they would help ensure the network succeeds. I think it is ok if we can use this in an optimistic way, but we should not simply rely on it, because you are then relying on a trust assumption based on something that does not exist yet.

In this Panel, we have 7 researchers and CEOs as well as Mo Dong as Moderator.

On this question, the panelists generally agree that Layer 2 can indeed provide user experience benefits. These benefits include: instant transaction finality, privacy preservation and lower transaction fees. James pointed out that even though Layer 2 tends to require more complicated implementations, this complexity is hidden from the user.

Nathan: I think Plasma still has a lot of research to be done. Each design has its own pain points that need to be solved.

Georgios: Original Plasma doesn’t work. Plasma MVP doesn’t work. Plasma Cash is what I’m working on and I think it works. Plasma Cash allows you to have a safe construction, however it has the pain point that you cannot split a coin. Also you have a growing linear history, which I need to provide to my counterpart when I do a transaction. This could stack up to a gigabyte of data. We have implemented Plasma Cash with Loom software suite and have provided solutions for these two problems.

Nathan: I think if your path doesn’t work, you fail, game over. However, I think it’s also very hard to find something that actually works at this moment, especially when you can combine some of the things that don’t work and make one thing that works. So at the moment, we should try different new things.

Jacob: I think we need to find a way to collaborate to some extent. It’s not something we are doing at the moment, but I think we should consider it in the future, when we are close to getting things to really work.

James: Do whatever Spankchain does for state channels and do whatever Loom does for Plasma.

Arjun: If you look into these Layer 2 technologies you will find they have many similarities in their core design, so I think in the future these technologies could be combined together to make something more efficient. The only thing we need to compete on right now is mindshare. It makes more sense to compete on that instead of working on a standard. Having distinct tokens also makes it really hard to collaborate, because you have the incentive to make your token win in the end. Although we have these challenges, I believe we can work it out.

Georgios: It’s so experimental to have a standard on interoperability. Things change every week. My suggestion is to get one thing done and then people will build on it.

Nathan: If we can find things that are good we can make it standard but it would be hard to standardize general things.

Arjun: What I think will change is we will have more things in research. Plasma is very new and I think it’s getting close. I hope people can form groups to work on specific use cases.

Johann: I hope we don’t only have just payments but also have smart contracts on Layer 2.

Jeff: I think there’s a misunderstanding that research is only for researchers. I hope everyone can get involved in this public open space. I believe anyone can make a significant research contribution on these topics.

Mo: One thing I think you can get involved in is to combine different Layer 2 technologies. For example, combining state channels and Plasma, or combining the dispute process with the concept of interactive computing. This is also something we are trying to achieve in the next year or two.

It was fun to host this technical discussion with our Layer 2 friends in Prague. We at Nervos are also looking forward to seeing this community thrive. We are constantly listening to the advice and feedback from Layer 2 folks and we believe eventually we will together build the future of the fundamental crypto economic protocols.

We are committed to Layer 2 and will be on this ride all the way to widely used, successful implementations. So if you have any ideas or just anything you want to let us know, feel free to find us at:

Linked List

Introduction to Nervos Network

Add a comment

Related posts:

The Rise of Drones for Commercial Use

Making Share My Trip Feature much more robust.

Kittysires