On Wednesday night, Mr. Jan, chief architect of nervos, shared with you eight payment channel solutions in nervos live studio
- Nakamoto「high frequency trades」
- Spilman Channel
- Duplex Channel
- Timelock Channel
- Duplex Micropayment Channel
- Poon-Dryja Channel
- Eltoo Channel
- Generalized Bitcoin-compatible Channel
The whole live broadcast lasted nearly two hours and was full of dry goods. Click on the video to view it!
Jan and you talk about various payment channels (Part 1)
Eight payment channels
Although on the surface, the state channel is a generalized version of the payment channel, but in fact, their differences in some core areas lead to their great differences in usability and construction. Today, we see that most of the more interesting payment channels are built on bitcoin, so Jan’s sharing will be mainly based on the payment channels on bitcoin.
First of all, Jan introduced to youNakamoto「high frequency trades」。 Nakamoto’s previous communication with the community has actually mentioned the idea of payment channel, but the scheme is relatively simple and has many defects.
Later,Spilman ChennelImproved Nakamoto’s proposal and made a channel with low cost and no challenge period. But the problem is that it is still one-way payment, there is no way to achieve two-way transmission.
So there was a follow-upDuplex channelIt changes a one-way channel into a two-way channel by directly generating two channels. However, such a scheme will also face some problems: the life cycle of the channel is limited, and the use of funds is very inefficient. If every time there is no money in the channel, another channel must be opened.
So, there it isTimelock ChaneelIt realizes two-way channel through time lock. The advantage of this is that he has no challenge period. But the problem is that the number of times it can update is limited, and the time granularity of payment will be relatively large.
After that, it appearedDecker-Wattenhofer Duplex Micropayment Channel (DMC)A hybrid structure. DMC is characterized by the addition of an invalidation tree. It is not only a two-way channel, but also has no challenge period and many restrictions like the previous channels. However, it still has three problems: limited life cycle, limited reset times, and high cost of unilateral channel closure.
Lightning networkPoon-Dryja ChannelThis paper proposes a new framework to solve the problems faced before, and punishes the party who sends out the old transaction through the punishment mechanism. This is an innovation that has not been achieved in the bitcoin channel in the past. At the same time, the problems considered by the Poon dryja channel are very considerate and solve many problems encountered by predecessors. Therefore, it is the current solution of bitcoin lightning network. It can be updated unlimited, and the life cycle of the channel is infinite. The penalty mechanism also prevents both sides of the channel from sending old transactions. However, the Poon dryja channel has also encountered new problems: first, it has a high complexity and needs to construct a lot of transactions; secondly, the punishment mechanism will take up a large space (because the previous trading state must be preserved); and there is no portfolio (because the transaction is asymmetric). But Jan thinks that the lightning network is really amazing because it solves a lot of problems, and even considers the design difficulty and large span of privacy.
Jan and you talk about various payment channels (Part 2)
Then, Jan and you introduce a new construction schemeEltoo。 Eltoo was put forward after the PON dryja channel, which is characterized by the mechanism of replace by version, that is, each updated transaction is brought with a version number to replace the punishment mechanism, and the latest version number is used to achieve the final consistency of the state. This design will be relatively simple, low complexity, and good combination, but at present he needs a sight_ The structure of noinput is not available in bitcoin at present. It needs to be supported after bifurcation. However, Jan thinks that this is also a very interesting design, and the design of achieving state consistency is very similar to the idea of bitcoin reaching a final consensus.
Finally, Jan introducedGenerelized Bitcoin-compatible ChannelIt is also a two-way channel, and has an infinite life cycle, which can be updated infinitely. The unilateral closing of the channel is also a constant level, which eliminates the asymmetry in the lightning network construction, so it has better composability. But the problem is that because it is based on LN punishment, there will still be o (n) space cost and there will be challenges.
Join the channel webinar together
Why is nervos studying payment channels? Everyone who is familiar with nervos knows that nervos is a layered architecture. At present, layer 1 has been launched, butThis is only the first step in the long march。
Layer 2 is a very open question. How to combine them, how to develop layer 2, how to achieve the best layer 2, and what the future situation of layer 2 are all worthy of further discussion.
There are already a lot of teams in this industry who are doing their own exploration, and so are we. Besides, we are still doing itChannel network Webinar, which has been held for two times. We will discuss virtual channel, routing and other solutions together. We hope that through this form, more people can participate in the research of channel network.
At the end of the course, Jan also left three questions for everyone,The first three students who answered correctly will be invited to attend the nervos channel network webinar that is being held every Thursday：
1. Which channels do not need a challenge period?
2. What are the disadvantages of LN?
3. How many transactions does it take to close the eltoo channel?
Chinese buddies are welcome to leave your answers in the background of the official account of the Nervos Chinese community, and explore more channels and network solutions.
- A general payment channel structure and its composability
- Fork it ා 6: talking about the technological evolution of lightning network with Jan
- Fork it ා 7: in depth analysis of lightning network technical details with Jan