Lightning Network AMA: Three Giants Dispel Doubts and Mistrust
Lightning Network, a “Layer 2” payment protocol that operates on top of a blockchain-based cryptocurrency (like Bitcoin). It enables fast transactions among participating nodes and has been touted as a solution to the Bitcoin scalability problem.
Although Lightning Network has been running for several years, it has not been fully put into use as some still question its complexity. The help of nodes is required as coins loss occurs. However, the increase in the amount of capital in specific nodes may lead to centralization. In addition, there is an upper limit on the channel of lightning network as the amount of bitcoin stored in the wallet is the maximum amount of capital in the channel when two users establish it.
On March 14, 8btc, the earliest bitcoin social media in China, invited CSO of Blockstream Samson Mow, CEO of ACINQ Pierre-Marie, together with Head of Infrastructure of Lightning Labs Alex Bosworth to hold an online AMA (Ask Me Anything) with the Chinese crypto community answered questions about Lightning Network (LN) in ChainNode’s AMA. Here are some of the excerpts from the AMA.
Q1. What are the major flaws of LN? Can they be solved?
Pierre-Marie：I don’t know about flaws but there are certainly a few challenges ahead. The major challenge right now to me is reliability. It doesn’t sound like a very exciting problem to solve, but what we are building is a payment network and it has to be reliable, all the time. We need to track down all the nasty corner cases and improve the UX for end users. Other usual suspects are liquidity and routing calculations, I think the former isn’t really a flaw but a resource limitation that will be solved with fees, and there are promising work done for the latter, e.g. with trampoline routing.
Alex Bosworth: In Lightning, network nodes must commit sufficient capital in the correct directions. If they do not, payments will fail. This can only be solved by building a market to incentivize capital deployment and building tools to make it easier to deploy capital correctly. Some of these solutions can be added to the protocol and others may be products and services, like the Lightning Loop project that I work on at Lightning Labs that offers a non-custodial method to adjust the liquidity on your channels to react to market conditions.
For Lightning to be needed as a scaling solution and drive development on it, there must be sufficient demand to move money around to justify the additional complexity. Currently the Blockchain’s native methods of moving coins works very well and other scaling improvements like SegWit have taken a lot of the pressure off. By timing my on-chain transactions and using SegWit and the Lightning Network, I rarely pay more than $0.01 for an on-chain transaction.
The Lightning protocol was not designed in a way for it to be set in stone and it does require additional planned changes to address various deficiencies. Problems need to be studied and solutions tried, so the biggest issues may yet be unknown.
Q2. The growth of the number of bitcoins locked in LN seems like in stagnation for a long time. What is the reason behind that?
Pierre-Marie : That’s because that number only includes the funds locked in public channels. More and more, we see people using private (not-announced) channels, especially for mobile non-custodial wallets, which do not appear in explorers. I believe that going forward this trend is going to amplify and the growth of the private/unannounced network will largely outpace the growth of the public/announced network.
Alex Bosworth: There are multiple reasons for the capital allocation in Lightning not growing as quickly as before. The main reason is that in the past there was one person who added so much capital by themselves that they overshadowed everyone else, so even though others were adding capital the amount that they added was hidden behind one person.
Another reason growth has been slow is that there are protocol limits that prevent people from adding capital, as a safety precaution. As the software matures, these safety limits can be relaxed, but so far LND has not relaxed any of the safety limits.
Capital growth is not a goal of the system. In fact we want to have as little capital in the network as possible, because more capital means more risk. A lot of technical improvements have been added to allow for sending volume but with less capital allocated.
The Lightning Loop project that I work on at Lightning Labs is actually designed to help remove capital from the Lightning network. Merchants who allow capital to stack up in their channels want to get that capital out of their higher-risk Lightning nodes and into their lower-risk cold wallets, and Lightning Loop is a service that facilitates that transfer.
Finally, capital allocation in Lightning must grow organically through a market process. Direct investment in the network without studying where capital is needed will result in the total capital going up, but this capital will not be useful to the network or the investor. Capital invested must be put where it is needed, and the main way to figure out where capital is needed is to try a limited amount of capital first and then increase it where it is being used and decrease it where it is not being used.
Q3: Some people think that the lethal problem of LN is that it’s difficult to use. We need to lock the coins and most of the channels are one-way. Channels between exchanges and payment gateway are also one-way in general. The funds will flow from exchanges to the payment gateway, that means we’ll run out of BTC in channels in a short period of time, and every liquidation will cause a channel reopening and closing, this will increase the transaction fees, thereby hindering its adoption by merchants. Do you think that makes sense?
Pierre-Marie: UX is certainly a major challenge, with our newest wallet Phoenix we have tried to address it the best we could, and come up with an UX that is the same as a regular, pre-Lightning bitcoin wallet. It does include features like on-the-fly channels which solve the inbound liquidity issue.
The fact that the network flow may be unbalanced is not necessarily a bad thing because it can lower the liquidity requirements. E.g. payment gateway will mostly *receive* funds, which save them from having large liquidity requirements beforehand. Users may receive their salary, but then they mostly *spend*, which means that they don’t need to worry too much about inbound liquidity. The real pressure regarding liquidity are routing nodes, but that’s precisely their job to manage it and that’s what they should be competing on.
Alex Bosworth: Lightning was not designed to be a user-friendly application, it was designed as a privacy-preserving, fast and scalable protocol to transfer money. It would be great if there were easy ways to use it but that is a separate problem from the protocol itself which is designed more generally just to move money. The Lightning protocol has been designed for people who need payments to be cheap or private or fast or all of the above.
Traffic on Lightning is definitely highly one-way and that reflects a reality of a small network where the receiver may not necessarily have a Lightning accepting destination to send their funds after receiving them. As the network grows the receivers will have more reasons to send the funds they received back out into the network. Also the aggregation of payments even in a single direction can still be an enormous savings. If you are buying articles on yalls.org and using a maximum protocol limit channel, even with traffic entirely in one direction you can perform over 100,000 transactions before you have to go back to the Blockchain with another channel open transaction.
Q4: When do you think we can see LN’s mass adoption? What stops it from being widely used?
Pierre-Marie: I would say the real question is when we can see Bitcoin’s mass adoption. I see LN as just a tool for that. Your guess is as good as mine.
Alex Bosworth: I don’t know when we would see LN mass adoption, that would be a great outcome of the project but it depends on market adoption which is unpredictable.
There are many obstacles but the main driver of adoption would be market demand. If there were to be profit in using Lightning or developing Lightning, then more people would use it or work on it. Figuring out those business models and efficiencies is the job of the market which requires investment and then trial and error.
Q5: Can LN really solve bitcoin’s congestion problem?
Pierre-Marie: It improves it a lot, and not only the overall throughput but also the speed of payments (“instant confirmations”). Will that be enough? Time will tell, but if we max out LN it’s a good problem to have.
Alex Bosworth: No, Lightning is not an equal replacement to the Blockchain and it cannot replace the Blockchain. If you view the full utilization of the chain space as a problem then Lightning cannot solve that problem because it doesn’t replace all of the functions of the Blockchain and so it cannot replace all of the traffic of the Blockchain.
Instead Lightning can act as an alternative to the Blockchain for the types of transfers where its differences are acceptable to the people transacting. So if you want to transfer a small amount and the Blockchain is too expensive, you can transfer the same amount of Bitcoin on Lightning and pay a lower fee.
Q6: Is LN decentralized? If most nodes sending transactions need to go through some super hubs, will these hubs be censored by governments?
Pierre-Marie: There is a fundamental difference between centralization in LN and centralization in e.g. mining. On LN, you can always bypass a node you don’t like, which isn’t true on the base layer. Economic incentives are such that there will be larger nodes and smaller nodes. What we need to make sure is that you can still easily use LN even if you are a small operator, which makes it difficult to globally censor.
Alex Bosworth: Lightning is a protocol for interoperability, so it can be either centralized or decentralized, depending on how people use the protocol. You can use the protocol to setup a centralized “intranet” Lightning to exchange funds between known entities, and these entities can work with governments. You can use the protocol to interact with unknown people and have these people run their nodes in privacy-preserving ways and avoid working with governments. Because the protocol is only describing the way that peers interact with each other, it doesn’t dictate what the usage will look like, that is up to people who run the software.
Q7: The very first LN based DEX Sparkswap was shut down earlier this month. The founder Trey Griffith cited the reason as not having enough users. Have you ever faced the same challenge?
Pierre-Marie: We (ACINQ) do not communicate on numbers, but we have seen tremendous growth of LN usage in the past months.
Alex Bosworth: I know Trey Griffith and talked with him about his projects before he started his company, I think he did a great job exploring the space but not every exploration will find results even if the explorer is doing a perfect job.
With Lightning Labs and Lightning Loop we have not hit mass adoption yet but in order for us to grow to mass adoption we need to start at the beginning and keep growing “一口吃不成胖子”. The network and usage of our software does continue to grow. We are prepared for this growth and development to take time, and the market definitely is still small.
Q8: The very first LN based DEX Sparkswap was shut down earlier this month. The founder Trey Griffith said what they did was way ahead of current time. Do you think we are too early to develop LApps?
Samson: For Sparkswap, I don’t think it was that they were too early but their model was flawed. The LN is best suited for small transactions, so I don’t think that they would be able to generate significant revenue from their exchange. Also, Lightning wallets are online so there is hot wallet risk, which also precludes any users trading with a large amount of funds.
The main issue with Lightning is the UX right now. However, a lot of the UX relies on some specific new tech for Lightning to be developed and adopted, such as splicing and dual-funded channels. Once that happens the UX will be greatly improved.
Q9: Some people believe that LN is DeFi in bitcoin. What do you think of this opinion? Do you think bitcoin will have DeFi?
Samson: The term “DeFi” is somewhat vague and undefined, so it’s difficult to determine what is or is not considered DeFi. Lightning is trustless and decentralized, and does open the door to much more complex solutions to problems that could not be solved using the Bitcoin main chain. The Liquid Network will also be implementing the Simplicity smart contract language soon, which means even more innovative projects can be built that open the door to the same functionality as what we generally consider to be “DeFi”.
Q10: What do you think of the recent plunge of bitcoin?
Samson: The price drop was caused by a combination of panic and the markets and over-leveraged traders. We saw it stop after BitMEX stopped trading. It took some time for the price to recover, but there was no reason for the Bitcoin price to be that low, so it was inevitable. It always takes time for money to come in and bring us back to normal. I think we should be back at 10k by the time of the halving this May.
Q11: How did you guys join in bitcoin? Why choose LN?
Samson: Blockstream was founded by early Bitcoin developers and Dr. Adam Back who is well known for inventing Hashcash, which is a technology used in Bitcoin’s proof-of-work algorithm. So it’s natural for us as a company to be focused on Bitcoin.
As for why LN, it is the only way to get Bitcoin to be usable for retail operations at a planetary scale. We started investing in LN R&D in 2015 so we saw its potential from very early on. In fact, one of our developers, Dr. Christian Decker, came up with the idea of Lightning in a paper called “A Fast and Scalable Payment Network with Bitcoin Duplex Micropayment Channels.” It was under peer review when the more well known LN whitepaper was published, so most don’t know about his contribution.
Q12: How are your countries affected by the COVID2019? What are you doing now?
Samson: I’m in Canada now and it’s not being handled well at all. It doesn’t seem like the Canadian health authorities are very competent and the government is ill prepared. We don’t have enough testing and it’s very difficult to get testing. Masks are hard to buy and the government isn’t providing them either. There are no procedures for isolating cases or even suspected cases, only suggestions to stay at home. Also, we closed borders far too late as we were following the lead of the US. However, we’re lucky because we don’t have very many dense population centers so our R0 is probably lower than that of other countries.
Q13: Is LN really worth a try? Since bitcoin’s blocksize problem is still here, can we do things like Craig Wright – to build a new, widely recognized bitcoin and kill the old one – instead of building LN?
Samson: LN is usable today for many use cases. However, we need to remember that Lightning is still in beta with ongoing development efforts from many different companies along with Blockstream. Most, if not all, of the current issues with LN have solutions in the pipeline too. Lightning is definitely a rapidly evolving part of the Bitcoin ecosystem, and new tech developments don’t have to wait for consensus unlike on-chain improvements like SegWit or Taproot..
Q14: How will Schnorr signature and Taproot/Graftroot affect LN?
Alex Bosworth: I don’t think Schnorr/Taproot will have a huge impact on Lightning, but it will allow things to be somewhat more efficient and more private. Because signature aggregation is easier to accomplish with Schnorr, channels can require less on-chain space to close, they will be more difficult to identify as channels.
LN can also use Schnorr signatures more generally in the internal protocol itself to get some efficiencies for its peer to peer Lightning protocol, and that does not require a soft-fork but these improvements have yet to be developed.
Q15: LN is a sustainable solution for blockchain scalability, but also raised questions like: will it be a new source of centralization?
Samson：Centralization in the LN is unlikely because anyone can start and run their own LN node. You can look at a graph of the entire LN right now and see no signs of centralization. There are some larger nodes with many channels, but the vast majority of nodes are interconnected in a decentralized manner.
Q16: Most exchanges and wallets in China don’t integrate with LN. Do you have any plans to do that? What is the development environment like in LN? How can we build applications based on LN? What will LN’s plan be in the next few years?
Alex Bosworth: I have had discussions with Chinese exchanges and wallets about LN. I know Cobo wallet does support LN. I think it’s more up to users to demand access to LN and then the developers and exchanges will support it. We recently had Bitfinex exchange integrate Lightning using LND. They are one of the largest exchanges in the world and they are headquartered in Hong Kong.
Development in Lightning is much easier than developing on Bitcoin or other chains, because you get instant feedback with minimal waiting for a block. There is also no need to worry about multiple confirmations.
I have published an open source MIT licensed node.js library for building applications on LN and built some of my own applications as well, I would say it is a great environment for building applications and in the next few years LND will open up a lot more options for developers in the next year or two.
Q17: What if the price didn’t go up after halving? Miners will collect less fees, will that threaten bitcoin’s security? What can LN do? Like compensate miners with extra fees?
Pierre-Marie: That problem is completely independent of Lightning in my opinion. If miners are not profitable, then they’ll stop mining and difficulty goes down, benefitting remaining miners.
Q18：Do you support the Taproot/Schnorr upgrade proposed by Pieter Wuille? Will it benefit LN?
Alex Bosworth: I generally support the proposal, I think there is always room for improvement and it will be fixing some mistakes that were made in SegWit which goes to show that nothing is perfect.
I answered above that it will benefit LN but only in a limited way, there is no critical feature of LN or a necessary future feature that depends on Schnorr/Taproot.
Q19: LN develops slowly compared with competition from new solutions like sharding. What can LN do to stand out?
Alex Bosworth: I don’t believe that sharding is moving quickly since I’ve heard about it for a very long time but I haven’t seen any major results, but I’m not an expert in that.
The biggest thing that Lightning can do to stand out over time is to help more businesses and individuals realize solutions to their problems with Lightning. Our focus is on helping more merchants and exchanges move more money more privately, cheaply, and quickly and we are steadily making progress.
Q20: Is there any demand for bitcoin payment solutions? What impact will it have when exchanges implement LN?
Samson：There are already many payment processors on the market and they all seem to be doing quite well. Most of them that I know already support LN too – for example, Globee and OpenNode.
Pierre-Marie：Bitcoin isn’t currently widely used for payments, we are making the bet that it will be in the future. As for exchanges, good news is we will soon find out, given that Bitfinex, Bitstamp and BitMEx (possibly more?) have all started integrating it.