This is epicenter episode 340 with guest Daniel Wang. Hi, welcome to epicenter. My name is Sebastian cuchillo today, our guest is Daniel Wang. He is the founder and CEO of loop ring. Loop ring is a decentralized trading platform, which makes use of ZK Roll-Ups to achieve impressive transaction throughput on chain.
So if you want to learn more about ZK Roll-Ups, you should definitely go back and Listen to her episodes where we've talked about these Technologies. So there was episode 310 with elements of saint of Stark where or our recent interview with Jin Ling Wang and Carl flourished at sex 03:36 where we talked about optimistic Roll-Ups, a few things to know about Loop ring. It is an order book based exchange which sets it apart from protocols like Eunice Whoppers of gnosis protocol.
However, the order book is centralized it's non-custodial though which means that users can claim ownership of their tokens at any point. Like I said, it leverages ZK Roll-Ups to achieve scaling on ethereum and loop bring claims that they can achieve over 2,000 transactions per second on
chain. And they have a strong focus on the Chinese market, which is where the team is based sunny and Federica that this interview and in the days after the recorded this in early, May a security vulnerability was identified by the star core team and report it to the loop bring team, the post detailing that bug are linked in the show notes, the vulnerability, In the wallet front end. And as far as we know the smart contract code was not at fault.
So, in short the wallet, generates a key pair to conduct snark related operations, and these keys were generated in 32 bits space, which is quite insecure because one could compute all of the possible account keys in the system. Daniel mentions in the interview that they audit every new release. But I wonder how this sort of thing could possibly be missed. Perhaps the front end wallet
wasn't part of that audit. In any case, Case, once again, I think this shows the fragility of defy infrastructure. I don't know how we get around this, but every so often a vulnerability is discovered or a defy protocol gets attacked and if trillions of dollars of financial value, our 21 Day, live on defy, the ecosystem needs to be much, much better at security. Here's what you'll learn in this interview, Daniels background and how he became involved.
Grip toe loop Rings, 2017 token sale, and why they needed to Current funds to investors the concept of ring matching which existed in the early Loop ring protocol, loopings, use of ZK Roll-Ups to achieve scalability, current throughput, and transaction costs Loop ring compared to buy Nance, and other
centralized exchanges. Why they chose to build on a centralized order book model, the difference between optimistic roll up and ZK Roll-Ups, the loop ring token lrc and how it's used in the protocol. And Daniel's views on the future of the decks ecosystem, just a
few housekeeping items. I am participating in to conferences, in June, from June 1st, to June 3rd, May net is happening in Maine. That is a conference organized by Masari. And on June 2nd, I am really excited to moderate a fireside chat with Joe Lubin of consensus. And Danny Ryan of the theorem Foundation. As this comes out, you can still get your early bird tickets. For the next two days, they're only 50 bucks. Yeah, thanks really worth it.
It's going to be a great conference head over to main net dot events for all the details. And then a few days later, I am going to be moderating, a panel at the web 34, mm, which is part of COG X. That's from June 8th, to June 10th, and we're still putting the panel together, but it will be on the topic of Freedom, versus Civic duties. So that should be a lot of fun. And once the panel is officially announced and on the website, I'll let everybody know.
So you can register for that conference at wed 34mm dot-org. Last week, I made a plea to all of you to give us more iTunes reviews. Well, you have come through and answered my plea. We receive a handful of new reviews, which is really great, really excited to see them coming in. I have a little slack notification. Get new reviews and every time that thing Rings. Unlike and you review and I go read it. So thank you to those who left
this review. If you haven't left us an apple podcast review, please do so please. Take two minutes. Going to your Apple podcast app on your Mac or on your phone and just leave us a review. Tell us what you think of the podcast, how long you been listening? What you've learned your favorite guests, it can sometimes be difficult to think about what you want to write in a review. You can write But all these things.
So yeah, I would really appreciate it if more of you would leave us reviews and you can do so really easily by going to epicenter, got rocks /, apple, and that takes you right to the page. And as a gift, as a small token of my appreciation for leaving us an iTunes review, I'll give you a discount code for a free key. Key hardware wallet, just send me an e-mail. Sebastian at epicenter dot TV. Tell me like, hey I left you review and I will send you a discount code for 100 percent
off. On your keep, your Hardware wallet. So, for the last couple weeks I've been telling you about least Authority and they had a webinar in late April to educate people about security Audits and what that could mean for your project, well that webinar is now available on their YouTube channel and on the website if you'd like to take a look at it and it's been really great to work with them because there's such a mission-driven team and they're working on privacy-preserving Technologies.
And their latest project is one that I find really interesting. It's called Z Caps or zero knowledge access passes. Now I'm not going to go into detail about how that works under the hood because I probably wouldn't be able to do a very good job but I'll tell you what they do. So your knowledge access passes allow developers to create Services where the payment data is disconnected from the customer data or the service data.
So let's say you're running a password manager application and your customer data is encrypted on your servers by Zeke apps you could allow for customers to use that service effectively take credit card payment for that service and you issue a token. And this is a zero-knowledge token that they can then use to access the service but you don't know which token is used with which account so effectively you disconnect the service data from the customer data. So it opens up.
Lots of possibilities for companies to provide even more privacy to their customers and perhaps, Even censorship resistant because there's no way for you. The service provider to connect the payment data and the customer. Dave, these are the kinds of privacy-preserving technologies that least Authority or working on. And if you want to know more about C caps, go to least Authority.com / C caps, that's Z Ka. PS. And with that, here's our interview with Daniel Wang.
So we're on today with Daniel Wang, who is the CEO and founder of loop ring. So, Dana was great to have you on. Could you tell us a little bit about your background and how you got involved in this space? So I have a technical background, I graduated from University of Science and Technology of China. And then I got my master from Arizona State University after that, I had been working for Google for Or Chinese e-commerce
company called JD.com. So basically just working on the backend systems back in 2013, I purchased my first fake coin after reading their wet papers, I got this news from the CCT way, you know, the Chinese Central television pretty big TV in China, they covered Bitcoin a little bit so I got a really interested in that Imani concept. So I read a the white paper X3 times that night. And then the next morning, I purchase a coin, I still have
those Bitcoin in my wallet. Then I talked to a VC say, you know, the Bitcoin is really fascinating. I want to create a centralized exchange at that time. It was not called centralized exchange, it's just crypto exchange and then I raised the about two million u.s. dollars and I launched the small exchange where people can treat Bitcoin some other altcoins like moss. Stir coin repo as well, but the platform really didn't go. Well, we have a really terrible period of time where Bitcoin the
prices collapsed. So I closed the company. I joined another high-tech company called joann.com which is an insurance company. So internally, we use blockchain technology to do some internal cool stuff, but I don't really enjoy time, you know, working with corporations, so in 2017, I think maybe it's time for me to, you know, start using etherium or smart contract to build something cool.
Something really to trading, by the way, I'm one of the early investors in etherium, I still see how those easier as well. So basically I my career Stars crypto career starts from buying Bitcoins and then invest in Easter and build centralized exchange and then the desert decentralized exchange. So that's Where, you know, I started the crypto life what made you transition from the concept of a centralized exchange to a decentralized exchange.
People are talking about security but at that time really is the pinpoint is to maintain a centralized exchange is the continuous worrying about your capability of. Keep the users asset secure because you don't want to wake up in the morning and then suddenly realize out asset are gone. And then you have You know, take hold accountable for everything happened. We consider from our own former operational perspective.
We don't want to have that pressure to work against hikers because, you know, you have to invest a lot of money into the infrastructure to make sure the security of your exchange is good enough. Security is from my perspective is always tight with cost. I think it's still true for many centralized exchanges and then Then we also need to consider from the users perspective. Most crypto exchanges, especially those, you know, back in 2014, 13.
They're not regulated at all. So if as a user, your asset got stolen from a centralized exchange, there's nothing you can do. You can just complain with the operators of the exchange and no more, the eyes are gone. And, you know, that's the final ending of your story. So, I think maybe we can use Smart contract. Do something that hiker cannot You know, steal from exchanges.
So, you know, back in 2016. I think early 2017, we had this idea of looping for make sure, you know, some work can be still done of chain, but the settlement part especially, you know, the fund movement between different users can be done unchain. So, I think our project is a little bit later than 0 x. And then at that time, people are talking about this concept of peace and Relax exchange. So I think in general security is something we really want to
achieve to make sure. You know people. They use blockchain a Smart contract Technologies to treat some I said that are backed by those Technologies as well. I always say this is it Raonic to say, you believe in blockchain technology about you have to treat functionalized exchanges which the traditional technology that just means the blockchain. Technology is not that ready yet, right? So, Need to figure out a way to push this technology further to make sure it's usable.
First for our crypto holders ourselves, not to just talk to some industry to say, hey I want to revolutionize your industry by this cool technology. Right. We have to solve our own problems first. So that's why I focused on each change instead of other applications. We are deep dive into this in a bit. Let's just get the rest of Your
back story on tape. So basically, you, you actually did a token sale in 2017. Just before China actually had this blanket ban on icos and you were actually forced to refund most of your Ico proceeds. Correct. So tell us about that time. That must have been crazy. Well, it's a story really, I will recall many times during my life. So we did the do I co back in 2017, I think in June, or July? I'm looking at only had like three or four full-time personal we raised 120,000 ether.
We didn't really expect that many. We set a hard cap. So after that nobody can get more money there. And we reached that hard cap. I think we are also one of the first projects who performed the Ico using smart contract. So we didn't really have any platform to use. We're at a smart contract. We have a one-page website and people I can just check the the tokens they received and how many tokens are still there for sale. I think that experience is
pretty cool. It makes me feel, you know, in the future maybe fundraising can be as simple as what we have done. But later on, you know, people have gone to another approach using some big platforms like Finance to raise money, I think it makes more sense because there are more users, but I think with our Pros, you know it's just very cool one week after the Close of the Ico. We got the notification from Regulators.
Say you have to refund and I was told my name is on the list, like a two-page list of people who have conducted Ico. And if I didn't do the refund, then I will be in serious trouble. I cannot leave the country. I cannot do blah, blah. You know, a lot of things. I was really terrified to be honest. I actually individual in China and if you get that kind of like a pressure from The government, you will have the pressure, it's not like in other countries, you can hire a lawyer.
There's no way you can fight back you how to be responsible for that? So as I okay, I don't want to be put into jail. I don't want my staff to be put into jail. So, let's do the refund, right. So, we how to write another smart contract. We how to tell our media friends to say, hey, we want to refine just how as many people as possible. So, we collect the token and then we At least The Ether back to many users I think is 80% or something like that or 70%.
I cannot recall exactly. The users have to trigger this. So did they actually have to send back the lrc tokens to you and they were then refunded? Their ether, the most cases. Yes, but there are some cases where it's not an individual's is some small companies who helped us to do the marketing and we have to fight with fact them to get the token back. They would don't want to give the token bag. They just want to hold the token as and in the meanwhile, they
want to get the money back. So there are some Backpack, you know, some fighting's between those Partners but most people, most individuals will just give back the token and then get the easier back. We also rewarded some individuals who helped us to coordinate the process so we didn't ask everything back but in general having a large portion of our tokens are returned. And as a result we returned about 70 to 80% of the Deezer.
I think most people who didn't really Lee written easier as you individuals are from countries outside of China. So they are not Chinese. So they not scared by the media. Right now, I think it's a lock for us. Otherwise, this blueprint project will certainly go down at that time. We just finalized one idea in terms of operation and marketing, we don't want to spend those as much money as other project team will do a
marketing. So most money are spent in D. I think that's why you will hear some people say. Hey why do print doesn't really do any marketing or stuff like that? And people say, hey why don't you? It's better those people in China will say, why don't you pump your price? And then do something right? We have been holding this talking for so long and you every time you have you just talked about, you have a new version. You have a new feature. We really don't care.
We just care about a price. So that's the fact in China, most people are speculating, Thursday really don't believe in anything. So making a profit is the most important thing. This is a pressure for us during the last two years. I get that sense too. That like you know there's sort of very few technically focused legitimate process coming out of China. I think you guys are definitely one of them. Let's jump in on till the some of the details of the product.
So first question would be, why did you guys decide to build sort of a decks protocol? So why not instead of like it's going out and building a decks? Why take sort of the more 0x like approach and build a platform for dex's The first two versions of routing protocol is really similar to 0ax. The one huge difference is that we have this concept of ring matching where 0 x only has like two other matching, right?
So this ring matching is a distinct feature that we offered but besides that there's no difference between those two projects. Could you describe a little bit about what ring matching is? In a traditional trading pair, you have two tokens a swapping with each other. Right? With this ring matching in our implementation, we can have up to 16 different tokens of 16, different accounts they can, you know, swap tokens in a ring,
right? So people can all the relayer can look into different markets and to achieve these arbitrators in one Atomic operation. So, Let's say Alice and Bob and Charlie these three person can be put into one ring so that they can do swap in as am operations. There's you don't have to do like two traits or three trees. So like Alice wants to buy Bitcoin. Yeah. Bob wants to buy ether and Charlie wants to buy Litecoin and know. Maybe they're not selling what so you can plan and make a
triangle trade. Exactly. That will be effective. It has to Arbitrage righteous very much like arbitrators the Outside of that is the matching is really complicated. Is it would take more time to find the best deal. When the number of participants increase in the ring, is it like some sort of np-complete problem? I'm not good at that theory. So I'm not sure about that, but the time should be exponential. It's an np-hard problem. All right, thank you for that.
I didn't know that. So later, I we limited The number of participants into 8, is that of 16 didn't really support that ring matching in production because a backhand or the relay, er, Team really. Has a different time to find a good real go for to support that. That feature really didn't justify grouping as a independent separate protocol, right? So we should do something different to make sure lubricant can still survive. So that's the Chinese.
We had like two years ago or one year and a Right now, we have the third version of the routing protocol and the ring matching is totally gone from this version, because it's not something we want to achieve, but we keep the project name there. In principle, The Ring matching is a very powerful concept and that it kind of pulls your liquidity across different token path. So basically especially in a world where there isn't one
dominant trading partner. So for instance in a world where you have many different stable coins. For instance, this ring matching is super powerful so do you intend on picking up on this eventually again? Or do you think it's just given up on it? I think we are going to give up on that and I do agree for a staple clients. This ring matching really will bring a lot of value especially when the trading volume is going to be huge. So that application I think is
still very usable. The reason we want to give up the ring matching is because in terms of zero, knowledge proof, there's really no way to implement the ring matching because we need a lot more constraints than Those can be supported by etherium. So technically we cannot support ring matching more than maybe, three or four participants in the ring. So we gave up that because we ran out, I just focusing on your knowledge proof based solution. Now, to using smart contract, a lot on chin.
What was the problem that ring matching was trying to solve? It seems to me that most exchanges even decentralized exchanges. Usually take an approach where they use like Straight A so for even you to shop. For example, if you require all markets to use ether as a base trench, a somewhat solve this fragmented, liquidity problem. You know, it has this higher overhead of people may have to do two trades instead of one. But why did you decide that?
You know, ring matching was something that needed to be done? And why would you think it was sort of superior over just the base trading pair of model? Back in 2017, we think the ring matching may help user to find
the best deal. If you don't support ring matching is not the user who will benefit from best price is the Arbiter. The people who do the arbitrators who will get the best there, get the differences between the best price and the price that the normal user can get from The Exchange. So if a centralized exchange can support ring matching, then the user will certainly benefit from that. Reality most centralized exchange don't do that because it's going to make the system
more complicated. It can output match engine as per trading pair, right? At they have to consider all the Liquidators across treating pairs into one, huge, liquidity pool, and then make sure they can find the best ring among up to hundreds of treating pairs. That's a problem. The is not going to bring more benefit to. To The Exchange cash flow that the investment into that is more
than they can learn. But with smart contract wasting a maybe in the future Autobot, trading is not going to be dominant because I really don't like the Wall Street being dominated by treating all girls. I think it should just be users which we had not to high-speed trading. I think smart contract because the trading fee is you cannot just ignore the street and V so probably all go Trading.
Not going to be popular. So real user trading, they can use this ring matching, maybe they can benefit from that. And X is obviously also don't lend themselves to high frequency trading, right? Because basically the minimum time you have as the block time and basically everything that happens in a block happens defacto instantaneously exactly. Yes. So, speaking of Wall Street, obviously, like you mentioned crypto exchanges, don't really
make use of like, ring trading. But, you know, if like, is this ring, trading sort of Paradigm something that's used, maybe in, traditional Financial exchanges, like, NASDAQ or things like that. Is it used anywhere else that I don't know? I think the overall concept is to find the best price, right? To make sure the price you find the password. Salt in traditional markets pages to arbitrators. They use more than multiple trees to make sure past price is fine right with ring matching.
Basically we just want to make sure, you know, in one autonomy Corporation, they can achieve this. But the heart that the challenging part is that if you talk to a professional Market maker to say, hey, we have this ring matching cool stuff. They need to learn about that. I think those people are not Really good at accepting new ideas, they just want to make sure they're treating.
All those can be migrated to you know, crypto and then make a profit when I can imagine is. It's probably very useful in sort of OTC desks, rather than order book system. Like, if you're running OTC that which is sort of, you know, lower volume, they might make sense that ring training might help there. Yes, I think in some treating all goes, they have this ring matching, I'll go there, but they don't call them ring matching. They just try to find across two markets the opportunity to
Arbitrage, right? And they break those opportunities into multiple orders to make sure they have the profit. So if you bring those all goes into a ring matching relayer and make sure in the original part book, you can find a ring then it's a ring. Matching. Right? So I think the overall concept is really just find Opportunities to find the best price for the user event. The best opportunity to make money out of those spreads.
I think that's the general idea. If you look at the arbitrageurs who actually spot these rings and execute them, they net the difference, right? Anyway they actually transfer liquidity from other book to order book but they are paid for this yearly by the user. By actually joining the split liquidity across different order books in sort of this ring trading world. You actually make the arbitrageurs the market makers Superfluous to a certain extent.
No yeah, I think I I would agree to what you said and so that's the eighth. Some centralized exchange or decentralized ones. They offer this ring matching all go in there. Match engine then the users will
benefit right? Not the Arbitrage because the there's no Arbitrage opportunity at all because all those opportunities detected and used internally by The Exchange benefiting the users now to external Market Matrix or arbitrageurs To me this is super interesting because as the listeners may know, is I'm the CEO at knossos and we built this protocol named the gnosis protocol and the nose is particle when the exact other way.
So basically, we also looked at scaling and bring Fades and we also decided that we could only Implement one of these at this time and we implemented the ring trades and instead of the scaling but you actually went the other way and you implement the scaling. So what you're actually doing is your life version of Luke, bring three actually. Uses, ZK Roll-Ups tell us about that.
It's probably because I used to have this small centralized exchange and I always have this goal to create a direct that can have the potential to compete with like Finance. It's my belief that it is much easier to convince a user to switch to a new project with a similar user experience. Then to tell them at very different business model to say, hey this Is where you can do the trees. The user experience is different but the goal is the same Ray, is
it more? Like you talk to a normal user to use? You Nisswa units. What works is great. But for normal users, they just, hey, this is different. This is not exchange, right? So they have to take time to learn the new stuff with most people, especially those in China don't want to do, they just want to invite us make money and leave market, right? So my interest like a very A of looping is to create exchange to compete with like Finance or okay, Yaks, I cannot say, I
don't like them. They certainly play a very important role in the ecosystem but I think going forward if those exchanges are still there, those sad stories will happen again and that means we haven't done a good job. So my goal is really clear just to computer with them. So throughput and reducing the cost because my first priority by creating something, New and interesting is not so from our experience at least I think right now we are on the right track. I'm not saying focusing on the
ring match. It's not a good idea, but they certainly will be challenging even for you to make normal users to use. Can you put numbers on this? So what's your current throughput and what's the cost per order on the current version of lubricant. So the current version is reporting. One is throughput is about 2020 plus trees per second on the Syrian midnight and the segment upper the cost per settlement is about 0.0002 US dollars. So three zeros after the point,
I think it's okay. Is our new versions 3.5 which is just code complete today because is going to be lowered a little bit but not too far. The reason is that we have been optimizing the protocol so hard that there's little room to for improvement but I don't think is a problem. The cost and the performance or the soup would are no longer issues for us in near term. As you can see in some statistic our Production throughput is 0.128 per second.
There's a long way to go before. We can reach our throughput, upper bound, the cause is. Okay, our cost for zero knowledge, proof generation for the instrument case is only about 15 to 20 percent of our payment to Amazon. So we pay for five times more to Amazon than to like etherium gas. So I think those are now no longer problem, the biggest problem. I like user experience and user onboarding, those are more challenging right now. So, just for a reference point.
So you mentioned that your protocol concurrent to get about 2,000 traitor second. How much do does an exchange like Finance sort of what's their average throughput? When I owned my own exchange, the throughput is about 100 to 300 trees per second for small centralized exchange. Because in reality, you don't want really put too much because that means more cost you Cement in the infrastructure, right?
So for finance I think it will be a couple of sovereigns normally and then maybe maybe tens of thousands at the peak I think by nice is okay because they make money. They can you know, scale pretty easily with more investment but for looping I think the upper bound is the upper bound right now. There's no way to go even further than 2,000 per trees per second.
What I'm saying is that A loop ring is as of now can compete with small or even middle-sized centralized exchanges but to compute this by now as you need maybe 10x or 100x throughput, and you may want to reduce the cost per trade even further. So right now is really not ready to compete it out, replaced by net. And there's also other challenges such as cross chain, right? We are only supporting tokens years are Tokens or and either not across chin assets. So, that's the challenging part.
So, the number of orders that you currently process on, looping is orders of magnitude below what you could in principle, do what in your mind is the principal challenge to adoption for you guys because it's not cost, right? It's not caused. First is user experience. So we right now only support matter mask, we support a wallet connect. That's the only to approach to a lock wallets and especially in China that people don't really use. Chrome browser.
They don't have access to Chrome Store which means they don't install metal mask at all. So the use onboarding part is really challenging. People have to get a VPN to access other the internet so we access the inter China, the Chinese Only Internet. So that part is really challenging for us. Even for people who have access to Chrome and The Meta mask. I think they still have to learn something about looping about Why should I create a treating password? What if I lost the training
paths, were things like that? Some some Corner cases scenario. They have to learn how to handle that and then the most challenging part A liquidity. So right now liquid, you has two sides, the word on one side. We don't support Bitcoin. We don't support those top 20, you know, other blockchain projects coins, but luckily we are going to support TPT see very soon but the other side is that. Initial liquidity is not good enough.
We have to work with market makers, to make sure they are willing to put their assets into a looping and to the market making, we have prepared at this liquidity or market-making competition to attract Professionals for calm, but it will take some time and they will have to tell us whether our profit sharing program works or not. For those people, I think making money is important because they are putting their asset even in decentralized exchange.
It also means risk, right? Nothing is safer than putting your sat to our own wallet, so we have to justify the risk and profit for them. Liquidity is very much a chicken and egg problem. So basically if you look at the Loop bring architecture. So basically you we talked about this in the very beginning, so you are non-custodial exchange, but I don't need to tell you guys this, but X is not a deck.
So basically, there's many ways of actually architecting a decks and you are not cursed, hold you, but you are a centralized order Book Exchange. Why did you opt for this centralized order book? And for Do you think about fully decentralized exchanges? Well, I do lie, I prefer to use fully decentralized exchanges because that's what a tax is supposed to be right. In reality is hard because of the performance of the underlying tag blockchain
technology. So hopefully when the theorem to comes out we can have a facts which is fully on chain and can also provide You know, a better user experience, the all the book management, you know, others are created a canceled, you know, these operations, if put on chain will cause a lot of gas than purely the settlement part because the sediment is all the trees have been done and just move the money around.
So I think even in the future even if serum has been scaled like a thousand times more, I think moving the all the book management, order matching Orange, Ian is still quite challenging so going forward and maybe some like am an exchange like a uni swap it's going to be more feasible in terms of implementation and operational so I don't see in my lifetime. The other Book Exchange were our be moved fully unction.
I think that's just not going to happen but I do if they engage in a lot of new Trading. Models like you Nisswa like your auction-based pass processing. Ring matching stuff is going to become more popular and get some user from the centralized few changes. I'm not really optimistic about the later one scaling, because I would say, even if the blockchain skills a lot more, there will be a lot more applications deployed on top of the layer one system.
So that the gas price will be, you know, adjusted according to you know, the consumer and and you know the minor. So the guys price will not be 1,000 times lower than it is now after a serum to. So I think scalability is not a thing. You can rely on layer 1. So, why do you think that most of the on train exchanges or decentralized exchanges in the broad sense? Like the order books? Haven't sort of gotten a lot of
popularity yet. So if you look at things like zero or X is probably the most popular one. It's volume is like, much lower. I don't know. Actually, I don't even know if it's lower die, maybe it's higher, I don't know. But at the very least like, you know, the Mind share is much lower than things like Yoona swab or Khyber or, you know, things that are taking non order
broke approaches. And so why is that The thumb side of older book based exchange with my knees other books often is that there's no easy way to achieve this composability with other function. If I product Khyber, and Yuna swap, they are really friendly. So it's easy for people to integrate with Khyber, with Yuna swap, but it's not easy at all to integrate with Xerox.
Exile, grouping, because when you submit order, you don't know what's going to happen after the submission of the other, right? You don't know whether the trade is happening or not. And when, but with units will have, you are certain that and there's a high latency as well. Yes, yes. So basically the composability is one issue for any Layer. Two Solutions, I think 0x is not exception.
The other reason that the racks is not are going Haven't been popular, is that for normal users, in that the user experience of using Xerox is far from being ideal, right? So most people accurate don't care about security, most people, they care about user experience. The security part is, you know, if you can offer the same user experience and and you are much more secure, then of course you are a very good. Good solution.
If you sacrifice user experience, no matter how secure Your solution is, if people don't get it, people don't want to pay for security at. They want to pay for user experience. They want to pay for being able to make a profit. So I think is 0 ax and new parents early versions, We just cannot meet users expectation at all. So to be fair to say that order book exchanges are sort of have to be a product in and of themselves, while maybe more, you know, swap or Bachelor
auctions. And these things are lesser products in and of themselves are sort of more tools like as part of a larger defy toolkit and it's how they've differentiated themselves is the composability on chain while order books changes can't Use that as a differentiator. So they need to find different way to differentiate themselves. It is true for now but I think it doesn't has to be true in the long run. So right now with Loop running, there's no way you can compose with layer 1.
If I products, but with our 3.5 version, we opened the door. So that people would deposit their tokens or Easter into our exchange contract. Those ICS can actually Work with other defy products. So we allow people or the operator of the exchange to make use of those tokens with the users permission and on the other side, you know, on layer 1, layer 2 side, maybe going forward, there are more layer to composability so that we can have more layer to D5 products.
They interact with each other so we have interaction or composability on layer 2 as well. And without Lear, they are one composability. So maybe that's another thing, maybe will happen soon, but not soon enough because their knowledge proof stuff is still, is not very developer-friendly right now, but going forward, maybe with even, even better technology or implementation of zero, knowledge proof, this can happen. So yeah, let's talk about that. So, how did you guys end up sort
of using Z Cairo? Roll ups and you know you I guess you guys were using them before they were even called GK Roll-Ups. So for the point that you decided, okay, you know what, let's sort of pivot to focus on scaling these exchanges, what are some of the Avenues you explored? And how do you end up on the
current solution? The reason we started investigating in your knowledge proof is that you know, one day like three years ago, I talked to my team to the hey, are you in optimistic about what we are doing? Are you excited? And they most people saying, no, I don't think this is going to work so. So here's some people say I pretend to be, you know, excited because I don't want to make you sad.
So I said, okay, thank you for telling me the truth that we need to. To figure out a way to make some change. Otherwise you know we are all going down, right? So I did some researchers, I talked to my Chief Architect to say, Hey, you know, you know what, spare yourself from the daily life, the daily work, focus on zero, knowledge proof to make sure, you know, we understand the technology and we can, we can try to use the technology to make sure we don't use the blockchain itself for
competition at all. Just about this. So racked, our Chief Architect, spend like three months for three months and then come up with a design. So I told the team that we are going to switch Direction, not focusing on Xerox, like exchange at all. We want to make sure skating is our priority, not some, some other features. It's a candle is a path. We didn't really know it. It's going to work out.
As a matter of fact, we didn't really have this Tiki roll up idea, you know, in our design at all. Because maybe the seek a rollup term has been brought up, but will it really didn't focus on that. We we just want to make sure, you know we have our own solution without be without relying on any other you know ideas it turns out our solution is TK Rolla.
But you may not know that our solution also Has another option to tint the roll-up part of so that we can only have the TK part on chain and use a like ipfs or some other decentralized storage system to store the data. So if that option is is on the throughput is going to be much much higher. So it's a gamble. It's a desperate need for make a change. Otherwise the project is going to go down, so I don't want to to fail. So so luckily we found the right
direction. We didn't make mistakes in terms of implementation. So I think I would say we are lucky and just both to tools and New Direction without consensus from our token colors. They don't really know what's going on. Yeah, so we've been throwing around the term quite a bit. You gave her all up. Let's Deep dive into. What. Actually is it? How does it work? What Is the architecture. So, basically a previously, we at the other book is maintained of change of the ethereum
blockchain. And then, if we find two others can match with each other, we bring those two others unchain. So dungeon part is smart. Contract will verify our the trade, the others can treat with each other. We also verify the balances of those two accounts house, house of fate and proteins, and then they do the actual Pope in Swati. So a lot of computations on chain, the talking swapping is only the only a small portion of
the computation. But even with our optimization is still takes like 200 to 300,000 gas per settlement. So the throughput is really not good enough. So we figure maybe we don't want to use the serum for the competition. We move the competition often. So the idea is to create c'etait often or layer to a concert system, structured in a binary Merkle tree.
So the leaf of the Merkle tree will be the violence of the trading account for each user and whenever we do the trees we just modify the often Merkle tree and make sure the modification is propagated all the way up to the down to the root of the tree. And then to make sure the theorem Has the information about the often computation of the account modification. We have to put the root of the, you know, Merkle tree angry, theorem with a proof to make
sure the way we updated. The trees are exactly falling, our protocol, right? And the protocol is accurate, the zero knowledge proof because the circuit those are actual protocol, which makes Jurors in others, assigned by the user. And the accounts are the two trading accounts in a country, has the balance. And after the settlement, you know, those accounts balance are still have their tokens, not somebody else who get their book in Balance.
So all the rules are actually still there, but not on chain on layer. One is accurate often on layer 2 to make sure people are really understanding. People are really sure about the rules. I installed a followed we have to do something called a trust set up. So we open source the zero knowledge proof, related circus and ask people to make it to run the setup to generate a random number into the system and make
sure it's linked with a theorem. So that all the often Parts open-sourced audited verified by Third parties. And then we have after a set of settlement, we called a batch or a block, a new Merkle, tree root is generated new corresponding proof is generated. We then put those two pieces of data on chain and onion, theorem. What we do is to just to verify the zero, knowledge proof. There is a pre-compiled verification is tracked,
instruction that we can use. It costs very little about the 22k gas and we can settle thousands of trees within that little transaction. So in general is a layer to a consists of multiplication instead of layer 1, right? Okay, there's a lot to unpack now, so maybe let me, let me recap this to make sure that I've understood it correctly. So basically, you collect the data current you on chain, but it could also be on ipfs or somewhere else.
And then, basically, the computation is done off chain, but the proof is then put on chains, so that everyone can make sure that you guys are not cheating. And the setup is done in a trust way. So as to make sure that the proof Just also above board. So a couple of weeks ago we had on the optimism team.
So Colin Ning Nang. And they talked about optimistic roll up which is the version of roll up that I'd X is going to use a centralized version of. Could you explain some of the differences between optimistic roll up and the ZK, wrote up that you guys use? I think the TK roll-up. Have you described as proactive? Approve based and the optimistic roll-up is more passive proof based or fraud proof based.
So, in order for the optimistic roll up to work, you have to figure out a way to incentivize serve party to make sure they can continuously monitor the functioning transaction to detect any. Potential fraud, proofs, right Auto detect the inconsistency of data and then generate the fraud proof to make sure, you know, there is a watchdog who are now to really member of your alliance who can, you know, just fight against you. Right.
The challenging part is that you have to design the way to incentivize them because if the, if the Education has been performing really, honestly, then there's no way, there's no fraud at all, right. So, the other party will just feel. Hey, you know, I have been wasting my time here because I didn't really detect anything wrong so and I get paid. If I'm not get paid, I will
leave, right? So if nobody is actually watching the application then suddenly you know, the operator will have a chance to create a fraud to behave, you know, this artist. And then because there's no active watch was dark there then later you know after the period after a certain time then everything is finalized. You know what stock even they can find some inconsistency that there's nothing they can do because, you know, there's also
this always a time period. You know, the fraud proof can be submitted after that. There's nothing you can do. So, I like the idea of optimistic roll up, but if there's Is no way to create such a token incentivize mechanism. Then I will be not waiting to put a lot of money into such application, but with zero knowledge proof, the cause is bigger, but every submission of data and the proof makes sure everything works before that timestamp.
So I'm 100% sure that all the data is is cracked and I can trust. So that's Just the my maybe I'm biased because I chose to take a roll up. I hope there is a way people can create that to mechanism to incentivize people who always watching of this Lola application. But I doubt So what are some of the trade-offs here? So it sounds like, you know, ZK Roll-Ups are giving us the scalability and security and all this amazing stuff. What are we sacrificing by using
them? First of all, the till now is proof is not really optimized but this is a technology of ten years old but really not how been used in application in real world fall for some time. So we are bound by the performance of zero knowledge proof. So what is the current zero knowledge proof libraries we have to pay for extra cost. In production for generating the proof right, optimistic. Roll-up is more like a economic
design. It doesn't have really any understand the underlying, you know, mathematical modeling or some other Technologies to rely on to depends on. So, if you you are good at modeling, the the risk and you know, make sure people can use your data to create a fraud, you can do that without Hang on other stuff.
So so for us, we how to make sure, you know, going forward if people are working on zero knowledge proof and you know, bring as a biter Library, we can just raise to the other library, but we are not scientists. We are not a, you know, good at math. So we cannot create that underlying library, right? So that's our our limitation here so we are basic this use tichenor, right? We are not capable of Improving that they could start a lot. Well optimistic roll up. You don't have to rely on
anything, right? So it's a modeling, it's a mentality that you allow people to challenge you and that that's it. That makes sense. So let's switch gears a little bit. We talked about this earlier, at least, we alluded to it. So you guys did a token sale where you actually had to refund a lot of The Ether that you raised, but the fact remains that you guys have a token. So what does your token do?
Well, that's a good question. To be honest, we have been always struggling with the token economy. We certainly don't want to turn our token into a security token, right? And we don't know how to design a better token economy. We have been talking about this even like this week to our own Engineers that will say, Hey,
you know, this is not good. Great, the current staking model so right as Of. Now if you have ever see token, you can stick ever see to earn the protocol feeds protocol fee is the portion of the treating fee is accurate. The six basis points of the trading volume every trees on
every looping based exchange. You know, six, visit points of the treating fees is going to be taken into a protocol fee full, and if your steak for more than Two days, you will claim your portion based on, you know, how long and how many tokens you have staked their so. So you stake, you get the protocol feet, that's it. So it's a passive staking meaning that you don't have to do anything else besides steak
which is not good. Because passel means, you know, it's more like a security, right? You like stuff, you did, you get the learning? So going forward, we may want to incorporate with This doll to make sure people have to stake, and they have to vote or something like that. To make sure active users really have more will benefit more from the protocol fee e-stop is that of just need passive staking but I'm not 100% sure.
I'm a big fan of bow going forward, I think we will also create a looping down to make sure a lot of parameters governed by looping token holders not ourselves. But for the token, I even think maybe a security token. Malta is the best. If you if you really believe something you should be just passive why should a project team require you to do some to certain thing to make a profit?
Right? You have already invested in the project with ho tokens, it should just do nothing but it's going to work against the security law. So I'm still debating internally with our team but going forward I think we will try different things to make sure that the adverse is taking is going to be improved. Yeah, so that's the current status of the looping staking but as of now, I think this staking rewarded like 2%. I knew I knew Ali right now but going forward.
I don't know how mass is going to be 2 percent is good enough. I appreciate your candidness. I think this is commendable. So we're talk about this in a little bit. So I know you guys also took part in the D X Tau and a fairly large weight. But so basically, if you are talking about the lrc token, kind of becoming governance, token of some sort, what kind of parameters? Do you see being set by the lrc token holders? Is it things like which tokens to list of what Friday trading
fees to take or something? Entirety. So there are two sides of parameters. One side is the protocol wise of parameters such as the protocol feed percent is from right now is 6. This is points, maybe it will change to a different value. The other parameters will be like, how many overseas should a sex operator hold before they can open a looping tax, right? And there are also a lot of products parameters like commas. Treating fees, they will charge at that party is totally
controlled by the tax. Operator is not controlled by us or by any other parties. We need to give sex operators or owners the flexibility to control those parameters. Otherwise they will not choose blueprint, they will choose a looping clone who allows them to, you know, control those parameters. So we figured you know what are they can do with looping, they should what are they? Do with like Finance like sexist, they can do with blueprint so that part we are not going to control at all.
So the protocol wires parameters will and should be governed by the, the towel. The one thing I really don't like doll or have concern on doll, is that most people are not really interested in participating in voting or they'll a communist, right? You know, people just most Just late included, be if I don't vote, if it's not voting is not bring me a benefit, I would tend to, you know, just watch some
TV, right? So it turns out most down users they don't contribute we as a as a bad example, we are a like one of the top 10 token holder in the ax down, by the way, it did in really being, we haven't been really active. We didn't really participate in A lot of voting. So we feel sorry about that. So as a result, we propose to gave our down token to add her token holders. So say, hey, we are not really active.
We feel sorry about that. There were too busy with our own project, we are waiting to give this happens to anyone who wants, you know, be active, right? So not being active is a big downside of Dow. Maybe there are ways to to make changes but I I don't know, the term my English is not native, there's a term that, you know, if some some Elise they govern, small economy, entity and there are like a normal users, everybody to Calvin, The Entity
or organization, right? I tend to believe that Elise they are better at governing, is that of, you know, normal users, like one vote per user, right? If you look into any, a lot of democracy country These people have to elect some, some people to represent them to make better judgment call right, not themselves, right? I totally appreciate what you're
saying. So basically what you're saying, it's perfectly legitimate to have an interest in being a stakeholder without wanting to visit a participating in the governance and just saying, I want to invest in you as a project but I don't want to say or do I don't need to exercise my say and basically having that be legalized under security if that didn't trigger all kinds of security laws, you would be in favor of that for do bring as well.
Yes, that's true. I have one last question on the protocol before we move on to like the wider ecosystem. So obviously the smart contract is Central to your protocol and tell us about the security level that you have on that. So basically has it been audited but your process in that area. I guess so for every protocol version we put into production. We have a surprise party who have done the Oddity for us for the second major version would even have a third party to run
this formal verification. So I don't know whether that works because that really didn't find any back at all for us. The manual inspection of the coda is actually helping a lot because I can the users and their experience in a bring a lot of lessons. They learned experience that they, they have to us through to improve the code base, but formal verification, we tried once it didn't really help us. Maybe it's because we had the mining inspector in first and
then the anyway. So for for this 3.5 version, this new version, we are not going to have a independent auditor because we don't want to put the new version into production. We Aim to deliver looping 4.0 directly maybe in a few months and put those into diversity into production, but he's, but in general, I think security is up most. We cannot lose security. We can sacrifice anything else but not security.
If we have one big issue in security, if peoples token or Easter or last, it will take us. Maybe we just we will go down and maybe we will take Gus like minds work, you know, to recover where we are, so I don't want to sacrifice security at all. So there's this is the bottom Okay.
So I mean and as you said, that's what if your eyes, the higher Mark of a decentralized exchange is namely that people self custody there's is rather than giving the custody to a third party who may then lose it or no be hacked or so on. So basically yeah that basically sacrificing on security obviously would not be the right move. So, with our newer Pros, we allow the operator of the exchange to use people's fun but only when they also open source,
they're smart contract. They also, you know, audit their smart contract. So the simply the most simple way to utilize peoples fund is to, you know, put most funds people deposit into a routine based exchange into. Like let's say you nicoise And then you know they can't they can make some money there, right? And then they can't claim back their tokens proportionally back
from the deposit contract. But again, I agree with you, you know, user how to choose, you know, whether they want to stay conservative by not allowing their token to be moved around, or they want to be more like, profit-driven to make sure they're falcon. Traded ought to be traded in looking basic, change will be accurately, earning some interest for them. So they have the flexibility, they have to make the call. We just make sure it's possible
there. And if, you know, we adapt a new version into our own decentralized exchange, we may stay very conservative not to do a lot of deeper integration. Because defy, right now, has a take risk of being too complicated. They are stacking on each other and you know, You don't know where the weakest link is right. And the system is only as secure as the weakest link. So I would strongly, you know, suggest against doing a lot of complicated if I integration. So what do you see?
Is the future of the decks ecosystem going forward. And what do you think? What do you make of some of the more recent changes that happened ecosystem? I think one of the most interesting one of the ones especially is like, how a lot of centralized exchanges are building their own dex's and in quotes and like the fact that Finance has this Finance chain or like and will be is creating their home. He changed off which are like you know, pseudo decentralized things.
But do you think this might hurt the adoption story for you? No more truer more pure decentralized exchanges. Yes or no. And how do you see this going forward? Well, in terms of taxes on their one, I would love to see a lot of different trading, very different trading models, who become, you know, products, especially amm on layer 2. I really just see right now. TK Roll Up is probably the best solution to scale attacks on earlier to as to binaries chain bags or okay.
Yeah, like school be, you know, if If you really understand what security means, you know, they are not secure, right? They can argue, you know, they are also decentralized. They also Block Chain but not all blockchains us secure, right? But if you look at the code base, yeah, it's decentralized. You know, it's very similar to the theorem whatever at who are behind the consensus. Right. What matters to security is the consensus has to be erased among different parties more parties.
They know each other. They cannot easily affect each other's decision making with a lot of blockchains. You know? They're, they're minors. Their consensus is really just centralized, right? They have six more scenes, maybe even if one data center, right. All controlled, by the mob, controlled by people who they can influence or control. So if the underlying consensus is not be centralized enough in everything on top of them. On top of it is not a secret at all.
So I'm not worried about taxes build on top of. Okay. Yeah, question, financed in whatever chin, it doesn't matter. How do you educate users to be able to do this sort of differentiation? That's a good question because most people just don't understand that, right? So, is it takes a lot of education's to make sure normal people really understand security? I think what you are doing, I think is amazing photo.
We don't have time to fall for that and we don't have the influence for that, but I think over a gradual, a people will realize people will follow people. They they believe in like right now, botanic it really Pro Roll-Ups and there are Of opinion leader. They talk about sticky, roll up if you talk about security, so I
think that's that's good. What then to you is the last thing that we need to overcome to help dex's reach the popularity's that Sexes currently do have because if you look at the, if you look at the global trade volume is like on the order of 1% or something that happens on dex's and all the rest is on centralized exchanges, right? First of all, you have to, you know, the mate those trading volumes aren't centralized exchanges. A lot of you cannot say they are real, right?
Maybe a person with about real, some exchanges are more real than others but yes I agree that, you know, Sexes are still dominating. We help you thinking about this and our answer to the current challenge is a smart Wallet app so that people don't have to learn about their private keys. There are phrases. They don't they just need a secure network to make sure they can help recover, their private
keys. So a smart contract, that speaks the same language that normal users, speak with a good user experience with sufficient defy integration, but not too complicated evaluation so that they can hold the wallet and
then make money, right? So that will In my opinion, be a killer app for massive users and actually, that's what we are working on. So, to offer, especially for the Chinese market, a mobile app with a small smart wallet will go app which looping built in so they don't have to use meta mask, use import their wallets, right? So the user experience is going to be like, very smooth, but we this is just my My answer, but not necessarily to be in a
successful in the future. We need to give it a try. I think, you know, a lot of different projects that are just trying, right? Our approach is different, not, not in terms of, we are providing a smart contract but we are really targeting the people, the user base. Who, what not probes, right? They are just normal user. They don't want to learn a lot. They don't listen to what.
Guys I'll free, right? They don't listen to my podcast of my life streaming, they just want to download something to make money, that's it. You know, they are not Pros. So once this smart wallet coming, we are testing internal version right now. So hopefully later this year, in property, in July, is going to come out. But mainly for the Chinese market I think regulation is Thing we really want to focus in the future so we don't think we have the body. The resource is to get regulated
in other countries. I think defy blockchain crypto eventually it will be part of Finance so it's going to be highly regulated locally. We want to get locally regulated in in China but the technology is going to open source to other two everywhere. Well ain't this the smart wallet that you launch later this year, is it going to have? I mean, you said you want to what you wanted to be secured by the network. So does it have like social recovery features or if I lose my mobile phone?
How do you make sure that I don't also lose my assets? Yes, that's the way we got going to provide the recovery. So it's based on social network relationship right now in China, they use WeChat. So we are going to integrate which we chat. So that if you lost something, you can use, we try to reach to your friends, your secure network to make sure they know you are you, and then they can
help you to recover. So, away also going to support, Hardware's at as garden and we also going to offer a centralized Garden service to make sure you know we are going to act as the first you know responder to recovery. A lot of feature will be added but I think we are going to make sure this is a still a decent decentralized wallet than custodial wallet which can integrate with looping Exchange natively. To make sure people when they get on board, exchange is going to be much easier.
They don't have to learn a lot of terms. Daniel, thank you. This was super fascinating and I think the Chinese user poor obviously is enormous so we wish you the best of luck. And I hope this becomes a success because this sounds super cool. Thank you for being on. Thank you, hopefully going forward. We can also integrate the ring style, unchain, you know, trading experience into our wallet, so hopefully, we can work there. Thank you for joining us on this week's episode.
We release new episodes every week. You can find And subscribe to the show on iTunes Spotify, YouTube SoundCloud or wherever you listen to podcast. And if you have a Google home or Alexa device, you can tell it to listen to the latest episode of the epicenter podcast, go to
epicenter, .t V /, subscribe for a full list of places where you can watch and listen, while you're there, be sure to sign up for the newsletter so you get new episodes in your inbox as they're released if you want to interact with us guests or other podcast listeners, you can Also on Twitter and please leave us a review on iTunes helps people find the show and we're always happy to read them but thanks so much and we look forward to being back next week.