 So, guys, thanks for joining. We're very happy to have Jordan Mack, Senior Software Engineer at the Nervous Foundation, and Cheng Wang, founder and CTO of Alephium. I added that he's a Scala Master, but a lot of other things on top. And Hou Chao is joining us. He's the core dev at Alephium, and he will answer some of the questions. We'll have a wide-ranging exchange of proof of work, tech subjects like sharding, UTXO, cross-chain interoperability. But we'll also play higher with a discussion about decentralization, what is the genes and how to get there, and we'll also let the magic of the conversation just take over sometimes. So, I want to first thank Jordan and Michelle for a flawless and pleasant organization on Nervous side and Marina on our side, which has done that perfectly. I hope it has been such a cool experience for you to Jordan, and I suggest that... Yeah, yeah, it's been good, yeah. I suggest that here we go directly. Like, we are here with two different audiences from two different blockchains, so I think the first thing is to really introduce yourselves. So, Jordan, let's start with you, Jordan. It's a pleasure to have you with us today. Your Twitter bio says you're a Senior Software Engineer at Nervous and that you love Rust. It also says you're an OG in crypto since 2011 and that you have many interests in life, from crypto, coding, tech, health, and meme. Can you tell us a bit more about yourself? Since when are you with Nervous and what do you do there, exactly? Yeah, sure, sure. So, a real quick background on myself. I've been writing software ever since I was a teenager. I spent my early career working in various startup companies, primarily web 2.0 companies and mobile application companies. Then in 2011, I joined the Bitcoin community after hearing about it and began contributing to that. But back then, it was much smaller space, much different than it is now. We didn't really know if this crazy blockchain idea would ever actually catch on anywhere and there was no money in the industry. So, I didn't immediately drop everything and just shift all my efforts directly over there. I was starting a business and I continued on with my own business. Then a few years later, Ethereum came out and I realized this was really the start of something bigger. So, it took a couple of years. But by 2016, I'd begun to completely shift all my efforts towards the blockchain industry. Then at Nervous Foundation, I joined them in 2020 as a senior software engineer and I'm currently working in the developer relations department. My goal is to facilitate developer growth on the platform which that translates to a lot of different things. Some days I'm writing code, other days I'm writing standards and documentation or assisting other development teams with building on the platforms and some other days I'm not writing code at all. I'm writing articles, making videos or speaking like I am here today. Thank you, that's quite an introduction. So, you really have lived a few different lives already in crypto and that's really cool. Can I ask you, Cheng, to present yourself too? The Aleph emergence knows you pretty well now but the Nervous Committee doesn't. So, for those who don't know, can you share a little bit more about yourself? What's your itinerary life in crypto? And what's your day job at Alephium? Yeah, hi everyone, I'm Cheng. I'm the code of Alephium. My past two blockchains started in 2013 when I started my PhD in EPFL to research on consensus algorithm. And back then I think consensus algorithm was not really a thing and I did not even heard about Bitcoin back then and I was doing pure theoretical research. And in 2015, I proposed the first million times asynchronous understanding consensus algorithm which was a bigger theoretical results. And at the same year, I started to hear more about Bitcoin and Ethereum and especially Vitalik, they found my paper online and then they can contact me and invited me to talk about my algorithm in the Ethereum Foundation. And that's how I get kind of into the crypto space. And then I started to follow the development of blockchain. I created it's Alephium in 2018 trying to build a scalable and secure blockchain for the apps with a new algorithm called BlockFlow. And since then I pretty much focused on it for a time. In besides crypto, in general, I'm very tech new guy. I do coding and a mass space time as well. So pretty much a boring person. I would strongly disagree with this, Cheng. But I should also have asked the question a bit differently what's your day and night job? Because that's actually how much we see you working. Like I wanted to ask Gong Chao also to present himself briefly, but I think he has a little bit of a trouble to actually request speakership. So we will come back to him a little bit later. If that's fine with you guys. So that's cool. Thanks for introducing yourself. Now I want to know a little bit more about the blockchains you're working for. Like Jordan, can you introduce us about Nervos? It's been around for a while now. But can you tell us a little bit like what it is, what it comes from and from when? Yeah, so you would think after me being with the Nervos Foundation for three years, this would be an easy question to answer. But for me it still actually isn't. Every time I try to give the elevator pitch it turns into a 15 minute discussion. So I'm going to do my best to keep it brief. So Nervos is a platform that's born out of a necessity for new solutions for many of the problems that exist on smart contract blockchains today. It places an emphasis on long-term sustainable decentralization and security. Whereas a lot of the platforms today, they've achieved smart contracts and decentralization. Yes, hello? Can you hear me? Hello, can you hear me? Yes, I can hear you. Did you hear any of the beginning of that? That's funny. I've been talking for the last minute or two here and nothing's been coming through. Okay, so let me start all over again. So Nervos is a platform that's built out of necessity for new solutions to many of the problems that exist on smart contract blockchains today. It places an emphasis on long-term sustainable decentralization and security. And it uses a multi-layer design to attain the scalability portion of that. Flexibility is another cornerstone of the platform of its design, and that's which it allows smart contracts to execute on a risk-5 based virtual machine, which is actually emulating risk-5 real-world hardware, which means that the smart contracts that are executing are basically executing on bare metal at the lowest possible level, which gives the maximum amount of flexibility possible to the developer. And the reason that it's doing that is this enables it to adopt new features in the future, in the future, faster, because there are actually less limitations to the platform. So I believe that the development of the platform was started in 2018, and the Mainnet was launched in 2019, and it's approaching the first halving. It halves every four years, just like Bitcoin does, and that's gonna be in late 2023 this year. Oh, Rath, it seems we cannot hear you anymore. Can you guys hear me? Yes, here we go. I'm really sorry, I have a connection problem. So I don't know where you were. Did you finish telling the story of Nervos? Yes, I did. Hopefully everybody was able to hear that. We're no stranger. We've used Twitter spaces for a couple of years, and this is unfortunately the regular scenario where half the people are not functioning properly. So anyway, we'll just have to do the best to get to the end. Thank you. It's one of our firsts. We've had a few already, but yeah, it's a bit of an experiment. I'm really sorry for this. I hope it's gonna work well now. So now that you have introduced Nervos, and I've read quite a bit about it, so I think now, Ken, could you introduce Alephium, please? Yes, so Alephium studied as a project to try to solve the scaling problem of blockchain. As you know, that's the Eastern Foundation, they have been working on the scaling problem since 2015, and they initially came up with a very complicated proposal and very ambitious goals. And I was following it for quite a long time and was not really happy about the approach. Because it's just too complicated. And I personally really liked the simple and elegant approach by Bitcoin. So I started to research into a kind of a possible shorting solution that is more like Bitcoin design philosophy. That's how I came up with a new algorithm called BlockFlow. It's based on the proof of work and the UTXO basically scaling Bitcoin, but don't really satisfy the decentralization, satisfy the security, the good properties of Bitcoin. So that's how the project started. I started working on it in four times since 2018. And then the project kind of started to evolve from the very starting points. Initially, we mostly focused on doing scaling. And then as in 2019, DeFi started to grow and then we see the huge potential of DeFi. So I also started to research into that direction and turns out that the UTXO model staff is a great fit for building secure applications on top of blockchain. So in 2019 and 2020, the main focus was on the retro machine and the app features for RFM blockchain. So right now I think we pretty much have done all of the core design and all of the important technical designs. So in general, I feel like RFM's main motivation and the main goal is to build a scalable and secure platform for this application. And this year we're going to prove that we have achieved it. Thanks, Cenk. So we've read with great interest the latest Dev Update published by Cenk on the Nervoste Talk forums about metabolism and this crazy analogy that I think was very interesting and also very well illustrated. Can you tell us a bit more about your team and the ecosystem job and just to give you, give us an idea of how much people are working and how is where you are now? Yeah, yeah, I can give a real quick background on that. So Nervoste has five founders, all of which are blockchain veterans in the industry. The chief architect is Jansi, who's a former Ethereum core developer. And while working at Ethereum, Jans realized that there were a number of problems that he didn't feel were fully solvable without radically breaking design changes. So he actually ended up creating a brand new blockchain which ended up being Nervoste. So Nervoste is designed to be both a technology and government, both decentralized in technology and government. So right now there are actually five or six separate companies that are dedicated exclusively to building out Nervoste's technology and many more teams building on top of the platform, primarily through an accelerator called Build Club. And let's see on the network side, it actually has two layers right now. The CKB chain is layer one and Godwoken is the EVM compatible layer two. We also have multiple bridges to Ethereum and many other chains. Godwoken has a brand itself as a game plus blockchain which is actively focusing on the gaming market and CKB, the layer one is rolling out a bunch of new features and infrastructure this year along with new tooling, documentation and a whole bunch of new design solutions which we hope it's gonna make it a lot easier for developers in the ecosystem. The last product I should probably mention is Axon which is a brand new side chain framework that's coming out. It's more for like application specific blockchains kind of like Cosmo if you're familiar with that. It was recently, very recently deemed ready for production and it's being used by a couple of teams that are building on it right now. So we should see the first products this year. Well, that's a lot. Thanks a lot, Jordan. And like since now I think it's working like can I invite Hong Chao to give us an update on where we are at Ethereum and like tell us more about the LEMO upgrade. Hey, everyone. Can you guys hear? Yes, thank you. Perfect. Oh, great. So I will briefly introduce myself a bit. My name is Hong Chao. I'm a core dev at LFVM. I'm generally fascinated about the asymmetric technologies meaning that the technology that can sort of give some kind of a leverage or age to the weaker players or the stronger players. And that is one of the reasons why I was interested in blockchain in the first place. And I want to see like I will kind of power we can potentially give to the ordinary people. When it comes to LEMO upgrade, it is the first network upgrade in our history like in LFVM history. We've been working on it for quite a while now and includes many important updates to the functionality, security and performance of LFVM. It's kind of difficult to come up with a complete list of all the things that we have changed. But I can give you a few examples of what LEMO upgrade includes. So we have introduced a new set of VM instructions and building functions to make smart contract developments more efficient. And that includes, for example, the dynamic array indexing, subcontract systems, which is like a more secure sort of a map data structure. We also introduced the building functions to facilitate, for example, debugging and logging and so on. We also have introduced like new features, both at the language and the VM level to make smart contract development more secure. And that includes, for example, a new assets commission system where you can basically dictates what kind of assets and how much of the asset can be used before you call a contract, before you invoke a function called a smart contract at the code level. There's also a system to check external costs. So like by default, you need to be explicit about who can actually call a public method of a smart contract that can mutate the state of the contract. We do that because that's traditionally a source of many security issues. We have also improved the node APIs and SDK to make the development of the smart contracts a lot easier. And that includes like a coding, unit tests, integration tests and deployment and so on. You can think about like travel and hard heads and the development experience is pretty similar to that. So with the LEM upgrade, it enables us to build the bridge and some of the prototypes for our depths such as like NFT and a DEX. So these projects were not possible before the LEM and upgrades or at least significantly less efficient and more error prone. So yeah, I'll stop here. Thank you Hong-Chao. So now that we've made the introductions, let's get into the debate itself like and discuss a little bit some interesting topics we have in common. So for example, let's start with proof of work. Both chains are proof of work which is an unusual choice in today's, let's say, layer one's landscape. So it's not really cool or popular to be a proof of work partisan these days. So there must be strong reasons for anyone to build or contribute to build a proof of our chain. Can both of you, Jordan and Cheng, can you explain why you chose proof of work when conceiving your respective blockchains? Yeah, well, I agree with that sentiment completely that these days, proof of work is not the, it's not seen as the green option. So we get a lot of flack. And if I'm being completely honest, there was a point in time where I was concerned about it as well. But I no longer believe that that's actually the case. And that's because when you look at the reality of what's going on and especially looking into the future of where the technology is headed and what the innovations are coming, ultimately the proof of work by utilizing energy in the most efficient way possible is the cheapest. And since it is a competitive platform it naturally gravitates towards green energy of all kinds. So I no longer see it as a problem. And the reasons why we would select proof of work for a chain is some of the most common reasons which are, it's absolutely the most well understood. It's a very simple to understand type of system where you can really understand every detail about it whereas some of the other things like proof of stake are actually very, very complicated once you actually start looking at the implementation. And it's also very much battle tested and we've seen, it's the only method that actually has a long history behind it since Bitcoin started with it in 2009. And our observations are that it is still the most decentralized for numerous reasons. And when you look at what's going on with proof of stake today, there were predictions about its centralization and especially with Ethereum where unfortunately seeing those would be true. And I'm not necessarily bashing proof of stake here. There are definitely certain merits to it. But when you're trying to create an ecosystem that is truly decentralized, especially on layer one, proof of stake doesn't really seem like that's necessarily the best option for that. And so Nervos, we have a lot of different options, a lot of different layers. And we continue to use proof of stake on the higher layers, but for layer one we really saw there was no alternative here. We needed the absolute best and at this point in time in history proof of work is still the best. It's interesting that you're calling Michael Saylor when he said there's no second best. But I think we, I agree with the sentiment and I would like to know what Cheng, do you have anything to add on the choice of proof of work versus proof of stake? Yeah, that's what I said. I want to add a bit more technical perspective. So first, proof of work is really a very simple consensus algorithm. As John said, it's very easy to read about, it has a long history and also it's very easy to implement and easy to validate. On that hand, proof of stake, it requires more CPU and a network to check the messages from the different validators. And also you need to have some own chain resources to manage the validators. And that's cause is not trivial if you can see running for node in a very large scale decentralized system. A second, because RFU is a shadowy blockchain. So to compare to the approach proposed by Eastman, they have a beacon chain and then they have to manage all of the validators in the beacon chain. And also they need to shuffle all of the validators to randomly distribute the validators to different charts. And that protocol is very, very complicated. But in RFU, we have block flow algorithm and this does not need a beacon chain. All of the block chains, they can basically operate in a stateless mode. You don't need to know the states of the miners. That's the greater advantage of proof of work over proof of staking. So pretty much we have no choice, we have to use proof of work. Yeah, and last, I want to also echo Jordan's point that proof of stake is still very new and there are quite some open problems that we still need to wait a bit more time to see how it's evolved, especially the network effects of staking, which is very, which I think is kind of a big trouble nowadays. Even though like the protocol give people incentives to run for now, people don't run, people just really want to delegate the stake to some providers and this is not good for decentralization. Yeah, that's all. Thank you, Cheng. Like I want to ask you both briefly, if you can, to detail a little bit, like how your respective proof of work algorithm works. So in the case of Nervos, it's called NCMax, if I'm not wrong, I must say that your docs are really amazing. Like it makes it really easy to dive in the project. And then I would ask you, Cheng, to tell us a bit more about the proof of less work, which is a different way of achieving the same thing, I think, but in a different way. So Jordan, can you tell us a bit more about NCMax? And I think I'm going to merge like also, can you in the same answer tell us like what kind of TPS and what kind of block time does it allow NCMax? Yeah, sure. Thank you. So NCMax, the NC part of that is actually stands for Nakamoto consensus. So it's actually similar to Bitcoin's Nakamoto consensus. It's based off of that, but it's been modified in certain ways to improve certain problems that Nakamoto consensus has. Like the first one is that it actually uses a two step process of proposing blocks and then, or proposing transactions and then confirming transactions, which is used to improve the block propagation time, which helps prevent selfish mining attacks. And then the second thing that's much different is it uses a dynamic block time, which automatically adjusts to network conditions to give the shortest possible block time without actually compromising security at all. So the next part of that is like, what is the TPS? Well, because of the dynamic block time, it can actually change up and down, but it generally runs at under 10 seconds per block, which gets us approximately 200 TPS, which is about 10 times of what Ethereum is today. And all of this is possible while running on a Raspberry Pi device, just like Bitcoin does. And that's just L1. That doesn't include the other layers. If you include other layers, of course, it's much higher. It's a similar approach in some ways to multi-layer and sharding actually work similarly in that approach, and that's how you actually increase the amount of TPS. But strictly speaking on just the layer one, it is approximately 200 TPS. And the block size is, I believe it's 597,000 bytes, which is, it's a number that was chosen, which it's the estimated amount to pack in 1,000 basic transactions per block. So anyway, it's one of those every 10 seconds, and it ends up equating to what I had said before, approximately 200 TPS. What about you, Cheng, can you tell us a bit about proof of less work, the TPS, and the block propagation and block time, et cetera? Yeah, so we use proof of less work to mitigate the energy consumption of the classical proof of work. So one of the observations is like, if it's considered to be secure to consume like 0.1% of electricity, globally electricity to get a good security, does it still make sense to spend 0.3% of electricity for Bitcoin? So I personally think that 0.1% sounds pretty much secure. The extra energy can be avoided if we can design the protocol differently. So that's what we do with proof of less work. We shift part of the energy consumption from the physical world to the network, to the internal network by burning coins basically. So it means miners, they need to both consume electricity and also consume, and also burn coins to my new blocks. In this way, the costs still kind of the same. So the security is the same, but the total energy consumption is reduced a lot. So that's the proof of less work parts. And then regarding TPS, because we are shorty blockchain, it really the final TPS really depends on the number of shots in the network. Right now with 16, with 16 shots and with very conservative parameters, we can easily get to four to 500 transaction per second. And if in the future there's a huge demand of transaction throughputs, we can increase the number of shots. And the block time, we run out the setup is 64 seconds. So we try to find the balance between very long block time and the very short block time because if the block time is very short, it means the blockchain is going to have a high often rates. For example, with 15 seconds block time, the anchor rate is more than, I think it's more than 10% so we try to, because we want the protocol to be very lightweight and efficient, so we choose a bit higher block time than 15 seconds, so we end up with 64 seconds. But of course longer block time has some issues with the user experience, but I personally don't think that's a problem because ultimately for proof of work blockchain, the security is about the confirmation time is not about the block time. So it really depends on how many blocks in total, how much it works you in total, the transaction has accumulated since it's sub-imitted to the blockchain. So I think Wall is an explorer can really mitigate this user experience too. So that's how we end up with these 64 seconds. Interesting, so it's two approaches to the same problem and we're in the same kind of order of magnitude, but it's very interesting to see different ways of solving the same problem. I think this is very healthy for the space and we learn from everything. Now I want to ask you one more question about the proof of work element of the block chains is that neither Alephia nor Nervos are ASIC resistance. Both have let's say an uncommon or newer hashing algo. So can you explain first why you decided both to conceive of an ASIC resistant blockchain? And then maybe if we have time we can talk about the hashing algorithm. Yeah, so we definitely did not want to make something that was ASIC resistant, even though as somebody I've done a lot of mining in the past and I've had mining rates on both GPUs and ASICs and I can say that from a hobbyist standpoint I'm actually a fan of GPU mining, I like it better but there's no real good way to prevent the eventuality that ASICs come out and generally replace GPUs other than constantly rotating your algorithm. So the important thing is to look at like with ASICs is it actually bad? And from our standpoint we actually think it might be a good thing because when you get specialized hardware built onto your ecosystem it's really a show of support in certain ways because people are investing in hardware that can only be used on a single platform. So that kind of goes hand in hand with using an uncommon or unique hashing algorithm. When you pair those two together it's companies investing in your ecosystem to produce this hardware for it and then it's your miners investing also in your chain as well. And so they're much less likely to attack it because if they did so they would render their hardware obsolete and it would take quite a bit of hardware to attack it so it really doesn't make sense. And from that standpoint we see it as a safer option to allow ASICs to be on the platform and also to use a unique hashing algorithm. Cenk, do you have anything to add on the ASIC resistance? I agree totally with Jordan and also ASIC friendly has been proven by Bitcoin that it really works well in practice. So nothing more dried. Okay, so let's move on to UTXO actually. Both projects also have that in common to have like a variation on the UTXO flavor of accounting model. This is also not so common those days as the account model seems to kind of have won the market of the accounting model exactly a little bit the same way than proof of stake seems to have taken proof of work. So can you explain each of you respectively the cell model for nervous and the state for UTXO for Alephium? So Jordan. You're right, so this could be a pretty deep topic but I'm gonna try to keep it really, really simple. In short, when you look at certain things about the account model and certain things about UTXO they're very different in their structure but ultimately what you're trying to achieve is usually very similar types of smart contract functionality. So the biggest difference between the account model and the UTXO is the way that I've typically explained it to developers who are trying to learn it for the first time is the account model is closer to a single core processor and the UTXO is actually similar in some ways to a multi-core processor. And if you look at what's going on in hardware today we all started in single core but now even your cell phone is multi-core and every desktop processor is multi-core. It's necessary for scaling and we see that same type of a thing going on with UTXOs whereas the account model it is very difficult to scale that because the state is one single element and it runs basically in a single thread which means it's only single core and while that's for the developer experience that's actually preferred, it's a much easier to work with. You end up with problems of scalability whereas UTXOs being akin to something like a multi-core processor allows things to happen in parallel. So you can be executing multiple smart contracts at the exact same time and this is something that really scales horizontally and we see it as one of the main reasons that we wanted UTXOs is it really does scale a lot, lot better. And then there's some less, some minor, there's some other things that are a little less apparent which is also UTXOs because they're individual. They actually fit better for state rent a lot better than the account model whereas we know that Ethereum's had a lot of proposals out there for various methods of state rent and none of them are implemented and we don't actually know if they ever will implement something because it's very difficult on the account model because everything is owned by a smart contract and how do you divvy up the smart contract? But UTXOs, every individual state unit, little piece of data can be owned by an actual user and because of that subtle difference there it makes certain implementations of state rent to be much, much easier. So even though they're less intuitive to the developers just in general, UTXOs seem like they are a much better option for the future when you're looking forward to blockchain and trying to scale globally to encompass all users on the globe. Well, that's a really great analogy. I love it a lot. Like the single core, multi core is a really interesting analogy. Hong Chao actually wanted to talk to us a little bit but explaining state for UTXO and how and why we did these choices. Yeah, I agree with many of the things that Jordan said. We picked UTXO because UTXO is a proven technology from Bitcoin and it's very good abstraction to manage assets and it's immutable data structure and for that it's actually a little bit easier for us to do our starting algorithm. As you know, it's pretty difficult due to have a shorting algorithm for a cut model because of the immutable state. And it turns out that UTXO is a pretty crucial data structure for us to implement the block flow algorithm in other things. At high level UTXO, it has inputs and outputs. So like transaction spend inputs and spend outputs and creates outputs. So visually it's established like a graph for all the transactions and in the traditional blockchain, unshotted blockchains all those transactions are structured into one chain which was a lot of unnecessary dependencies because the block after this current block, those dependencies in these two blocks do not necessarily have dependencies to each other but structurally in this chain it has dependencies to each other. So these dependencies might not be necessary and the block flow algorithm at a high level basically find a way to structure these graphs into multiple chains and it manages all these dependencies in a better way. That's how we implemented our 16 charts as we have right now in the production. So this is number one. UTXO is crucial for us to implement shorting and number two is that we feel that UTXO is pretty good. It's a very good data structure to abstraction to manage assets. So we actually managed the token in the UTXO compared to like the Ethereum managed tokens in a contract. So token in other things is the first class season and we feel that we have a better people as a security for our tokens. And also UTXO has better... It has an invariance when it comes to input samples so it's easier for us to verify in the transaction. And because of the input and output model some types of MEV is a bit harder for in our LFU. Some of the MEV transactions like arbitrage and so on it requires like multiple transactions on Ethereum. So it's a little bit difficult for the detector for the searcher to basically execute this successfully of this MEV. So, yeah. But the problem with UTXO is that it's a little bit difficult for developers to program since it's not very intuitive. And the COMP model is actually a lot more developer-friendly. And UTXO model also has a problem of concurrent execution constantly access the same like UTXO at the same time. The COMP model doesn't have that problem. So for us to... We actually combine the UTXO model and the COMP model so that we can take advantage of both. And so for the... So in the combined step or UTXO model we use the UTXO models managed... Still we use the UTXO models managed to access like the native ELF token, native ELF and other tokens. So the token transfer would have a much higher scalability but we use the COMP models managed smart contracts which has a global state. And those smart contracts has partial scalability but we can have a lot higher expressiveness than the UTXO model. So we have our own VM and our own language which we don't feel that we sacrificed our expressiveness when it comes to programming the smart contracts. Whereas if you look at the Bitcoin, the Bitcoin script is actually much more limited. Yeah, that's... Thank you. Thank you. It's really... I agree. We could make a whole session only on UTXOs and different design choices. It's really interesting. And that leads us directly to sharding actually because both chains have that as a setting argument that they can scale. And whether it is at the layer one level or is it in general, this is a strong selling proposition in this market that you need to be secure but you also need to be scalable. Both chains accomplish that very differently. And I wanted to ask first, Cheng, can you explain a little bit the approach to sharding? We're all like, let's try to stick to short answers now so we can get a bit fast to the end. But can you explain the Alephium's approach to sharding, please, Cheng? Yes. So RFI Sharding is based on new algorithm called Brokthrow. And Brokthrow is based on proof of work and UTXO model as we have discussed earlier. So basically sharding is you have multiple blockchains that can process blocks and transactions in parallel so you get the highest throughput. But here, all of the chains, they are similar to each other. So users, they don't have the learning curve to adapt to a new blockchain. And also all of the transactions between those blockchains are atomic. So compared to other approaches, like you have one blockchain connected to another completely different blockchain, the bridge experience is also much better. Users don't even need to feel bothered. Yeah, so that's basically what we're doing right now. Yeah. And Jordan, can you explain like the multiple ways that the service is doing sharding too, please? Yeah, well, we focus primarily on using multi-layer, which is L2s, which is a different approach to actually doing the same thing. When you talk about what are you doing for scalability? It's ultimately breaking up the data set. Oh, no. Oh, here, we're doing this again. Are you able to hear me now? Hello, hello? Yes, I hear you very well. Okay, yeah, I was doing that thing. I unmuted, started speaking and nobody could hear me. Yes, again. Okay, so Nervos uses L2s to achieve scalability instead of sharding and this is similar. I mean, when you look at sharding and you look at L2s, they're actually trying to do the same thing. They're taking the data set and the execution and they're breaking it up between multiple computers. So it's really kind of in certain ways it's a philosophical difference. Like I can say, one of the advantages of sharding is that when you create multiple shards, they're generally pretty much identical in their environment, which is great for developers who want to move from shard to shard. On the flip side, L2s have advantages and every single L2 does not necessarily have to be the same. They can have different attributes. And so when a developer needs a different type of environment for a project, even if their existing project starts in one state and grows and changes over time, they can actually move between different environments and have different attributes to them. So and L2s in general is kind of a little bit more of a track record in history of being understood and working very well. So when you look at the actual experience for the user, it's typically very similar type of thing. You do use some type of a bridge interface for both. And eventually we hope in the future that these interfaces are actually simplified, but yeah, Nervos went on the L2 route. I think it achieves pretty much a very similar thing. We can add many, many L2s on top of our L1, which in certain ways inherit the security of L1, which is the reason why it was so important for it to be really decentralized and secure. And then if the L2s at some point in the future are so popular that they fill up, we can just simply add L3s or even L4s to continue that scalability indefinitely. So that's the approach that Nervos has taken. Thanks Jordan. So I'm gonna speed read a little bit because I want to talk about governance and decentralization. So Nervos had his first hard fork last year and it was really interesting to read in the metabolism updates that you, that Jensi said the most difficult challenge was explaining what would happen after the hard fork to ecosystem players. Since Alephium is gonna have his first network upgrade in the next few weeks, which advice could you give us? Yeah, I could give you a couple pieces of advice on that. Well, first, hard fork is a scary term in our industry. So you need to reassure your community on what the real situation is and what the actual benefits are. I mean, we say hard fork, but in more realistic terms, we're just saying it's a network upgrade. So just make sure that that message is very clear to your users. Also, make sure you get the word out to everybody who's going to need to upgrade and get them the updated tooling as soon as possible because we've learned the hard way that you cannot get your teams out there from development teams to exchanges. You cannot get them to upgrade immediately. It sometimes takes time and oftentimes they're going to know well in advance but they're still going to miss that deadline. So you need to have contingency plans basically built in for that. And then the third thing is you really need to leverage your... Wow, that was strange. I just got a really loud tone in my ear. Yes. Leverage your community the best you can to amplify your message to development teams and exchanges because sometimes they just don't update and going out there and telling them, hey, you got to update or this is going to happen isn't enough. They actually need to hear it from their customers. Their customers have to go out there and complain and say, hey, this isn't working because you didn't update properly. So that's the way that we've actually found is best to get these things like exchanges to update when there are breaking changes because they simply just don't listen to us. We don't know why this is but they will listen to their customers. So make sure that that message is getting out there. Thanks, that's actually a really interesting tidbit and I'm sure we keep that in mind like that. To be honest, we're also talking about the network upgrade because we feel it reflects a little bit more what we're doing. And so, but this points out in general to a more, let's say more general question, which is governance. Based on your experience in crypto, both of you, this is a question for both of you and also your experience in your particular blockchain. What is your opinion on the ideal governance and decision-making system? Because at the end of the day, blockchain's are huge governance experiments. Cheng, maybe can you start with this question? Yeah, this is a very hard topic. I think it's kind of research in progress. A lot of different projects are still trying and experimenting. So personally, I don't think there's a perfect solution and we have seen many issues with token-based voting, which is kind of very popular nowadays. And token-based voting can be manipulated by whales and it can even be exploited by hikers. So in general, I don't see a good solution right now, but I think for some very limited scenarios or specific use cases, governance can still work. And in the case of lay one blockchain, which is very complicated and at the same time, very critical system, I personally believe that the Bitcoin approach is the best, basically no governance is the best governance. And in the case that if your blockchain has a big community and the community have two different group people, they don't agree with each other and they want the protocol to go in different directions, I think governance does not really solve the problem because they have different choices. By making a decision with the governance, basically separates the minority rights. So, yeah, I think governance does not really help in this case, but forking could really help. If they are sufficiently big enough people, a big enough group of people, they really want the protocol to go in some new direction by introducing some new innovations, they can basically fork the protocol. Yeah, this has happened in the past. That's basically my point. What's your opinion on this Jordan? I share a lot of the sentiments that he just expressed there. Governance is a really, really a huge topic. And when I say governance, I'm speaking both not just of technology, but also of people because ultimately that's all that blockchains are trying to do, they're trying to find a ways that we can use technology to get everybody to agree with each other in certain ways. Or at least if they disagree, it's organizing it so that they can peacefully disagree in respective chains. Now, forking is a very important thing, hard forking in particular, because in my opinion, this is absolutely necessary for any type of a platform to truly be decentralized. And it's a natural thing to try different, to experiment and try different directions. It's a natural for people to disagree, different teams to disagree with each other. And so if there's any project out there that forbids the ability to actually fork, that is a centralized product. And in my opinion, that is not really at all something that should even be considered for any project that is serious about decentralization. But if you wanna say in just a very broad and general sense, what does the future of governance look like? I think that DAOs in particular do have a role in that future. Definitely that's one of the best solutions we have today. I don't think that it's by any means the end solution, but it's a part of it. That combined with certain things may be a proof of authority, meaning essentially that somebody's your reputation as a person, a non-anonymous, not an anonymous person is actually out there and that you're using that and during your voting is probably another component because ultimately there are advantages to remaining anonymous in voting and it's clear advantages to remaining well known into the community in your voting. And so these different aspects actually working together I think can really help shape what we'll see in the future of governance. Thank you, Jordan. I just wanna say to you guys, I see that you're requesting to ask some questions. If you don't mind, I have a few more and then I'll open the floor to everyone if that's okay with you. Jordan, you opened the door to my next question, which is a tricky one because it's not something that I've seen that has been, let's say solved so far. You're talking about decentralization. So my question is like, how do you measure decentralization? What are the interesting metrics that you think help us define what decentralization is? Is the Nakamoto coefficient or do you have other measures or do you have, how do you define decentralized enough system? Yeah, that's a very, very complicated question. And it's one that we've had internally at Nervos too about how do we actually measure this? And it's, we've looked at multiple metrics out there and there's some interesting different ways of doing it. But ultimately when you really get to the nitty gritty it's such a nuanced topic that any single metric doesn't really reflect the entirety of each individual project. So something like the Nakamoto coefficient, like I believe that this is a good starting step, right? This is something to look at but I have absolutely have problems with that number because it was basically written by a Bitcoin maximalist to specifically show one side that they wanted it to show. They guaranteed their selves on top. Whereas like I said, it's a complicated subject with a lot of nuance in it. So I don't think that that's necessarily a single good number to go with but the general direction that they were going for, definitely there's some merit to that. But the general question that you asked like what is a decentralized enough system? That is also a multifaceted answer to that. Depends on what your project is. For example, there are certain projects like for example, a game or something doesn't necessarily need the same level of decentralization as a currency, a global currency does. It's perfectly fine for it to exist on just a handful of proof of authority computers. In most cases that completely fine. But something like a currency which is directly tied to the well-being of individuals in the world. That is something that is really fundamental to the way that humans operate these days. And because of that importance, it needs something much stronger. So my own personal definition of what does that mean? Decentralized enough system. When you're talking about something like a currency or a financial application and maybe daps in the future, things that are really truly critical, decentralized enough means that it's able to withstand a prolonged attack from any central large party or even a government. Any single or large entity. And I know that that's a very vague answer. But again, this is such a complicated topic that it's really difficult to quantify these into actual numbers. It's more just scenarios that you can describe that say this is what would be decentralized enough. Thank you. I like this answer that raises really the good questions that are the answers to the question. Cheng, do you want to add something to this answer or should we move to the next question? I pretty much agree that it's not easy to quantify. But I'd like to share my personal perspective when I want to evaluate whether a project is decentralized enough. I only look at two aspects. The first aspect is the cost to decentralize the network. So here is mostly about the cost to run for notes and also the cost to run the core infrastructures. Bitcoin is the base in this part because the funnel is really the most lightweight one and the infrastructure side, they are basically like no infrastructure is needed to run it. And even on the other hand, it's not doing great because most of the DeFi stuff is relying on MetaMask relying on Inferior and the Isis again. The second way I usually use to value a protocol is the cost to decentralize the ecosystem. So this part means is it very easy to onboard the new devs? Is it easy to build the new applications on top of it? Because ultimately the protocol is very bounded to its ecosystem. If the ecosystem is kind of controlled by a small group of entities, then it's not really decentralized. So here, they've experienced tools, infrastructure, all of these, all of these metrics and it's very important for the future of the blockchain. Yeah, that's basically how I value the decentralization of a protocol in a more kind of a doable way. But it's a very technical perspective. Thank you, Cheng. I think it's always interesting to have like the clear criteria that someone uses to evaluate something. So I really like it. And the next question for me comes inspired by Jameson Lopp, who, as you know, is a passionate Bitcoiner. He has written recently a very compelling article on the death of decentralized email. In it, he really shows how convenience and centralization have triumphed of the technical yet decentralized beginnings of email. So I want to make an analogy here and ask you if you think that blockchains, such as Nervos and Elefium, can result, can triumph and resist these really centripetal forces of convenience that it just works under regulations versus the ideals of decentralization and self-submitting. Okay, well, first I'm gonna have to break those two apart. Regulation and convenience are two different topics. And of course, just like all of your other questions, these are extremely deep topics on how we're actually gonna accomplish this. But ultimately, to answer your question directly, are we going to triumph over convenience? I think the answer is no. We cannot triumph over convenience. We cannot win by having an experience that is unfavorable to users. So we have to make our systems just as convenient. And in the future, things aren't getting better. We are watching these things slowly but surely evolve into ways that they will be easier to use. Some of this will be completely self-custody. Some of this in the future will be custodial-based services. That's kind of the way it's going to be. But this is a workable problem. And so the convenience of blockchains will, in my opinion, rival what is out there today in the web 2.0 world and then in the traditional world in many respects. And then we're also competing on costs there. And that's where I see that the innovations that are coming out today are the very beginning, the first step on what's gonna be a long climb. But eventually, we're going to surpass the costs of the traditional system. And we're going to have lower costs involved. So it's really going to be a change of the tide at some point in which everybody wants to go to the decentralized systems because they're just as convenient and they're actually less expensive in the long run. Now, when you talk about regulation, that is another really big topic. And it's one that it actually scares me sometimes because of the way that I see certain entities like the SEC trying to reach out well beyond their, what should be their scope of interest. But ultimately, blockchains are something that cannot really truly be stopped. Regulation would not stop anything, but it could really slow it down by decades. So this is not something that I want to challenge. It's something that I hope and I push for actively that regulators will actually see that this is technology is something that we need to embrace because this is something that is truly, it's something that has never existed in the history of mankind before. A platform that eliminates the reliance on each other because we've seen throughout the history of the world that we can't really trust each other. So finally, we have a system that we can work on even ground. This is something truly remarkable and really it's something that we need just as a people to continue existing with each other. So anyway, it's kind of a philosophical thing, of course, but ultimately regulators, there are good guys out there. These guys don't get the spotlight. When you see on the news, of course, they're focusing on outrage to get people pissed off. But what you gotta realize is there are people out there that we should be supporting that have our interests in mind. So that's kind of my opinion on it is that we need to just make sure that we are diligent and supporting those who do actually support decentralization efforts because this is something that if we go to war on this, there is a chance that we would lose. And while it wouldn't necessarily shut down projects permanently, it would really set it back. So let's just continue to make our efforts the best that we can. Amazing answer. I would almost want to ask you if there's a precedent of decentralized services or tools and that eventually ended up in the decentralized version kind of winning even in the long run. But maybe Cheng wants to add something to that answer. Not too much, but I first think the centralized email kind of failed as well because right now about half of the emails are actually spamming emails, right? So it's not doing a good job. I think the initial decentralized design failed because it likes a good incentive design for the system. And if we do it properly with blockchain, probably not layer one because layer one probably is not good enough for emails, but with layer two and with some services on top of layer one could do that well. And I think we could completely solve the spamming issues with proper incentive design. And regarding convenience, I think right now crypto in general is not really that user friendly because we are still early and the building really takes a lot of time. Yeah, I build, I code every day and I use a lot of stuff. So I can definitely say that technical and the infrastructure side we are still very, very early compared to the web two stuff. It took many years for the web two email system and other services to be so smooth, so user friendly. And I personally feel like we can do the same thing with for crypto for blockchain, especially if you imagine a multi-chain future with different chains with different foxes and different design trade-offs. So yeah, convenience, like user experience, I don't see it as a big problem. It's really a matter of time and also matter of adoption. And regarding regulation, this is not really something I'm good at because I pretty much focus on building. I personally feel like it's gonna be evolve similar like internet at the very beginning of internet regulations do not know how to properly live with internet, but I think right now still it's kind of a problem, but I think it's much better than like 10 or 20 years ago. So I think it's going to be the same for blockchain and for crypto. Thanks, Cheng. So I mean, I really love these answers. I have two more questions, but then I'm gonna open the questions to the audience. So you can start requesting if you wanna ask a question to Jordan or Cheng. I wanted to ask you like, okay, let's give a short answer to if you're dead and you wanna start experimenting the experience of Alephia Monervos, where do you go? Short question, short answer would be perfect. Yeah, so for us, it's probably docs.nervos.org or our second site is actually startwithnervos.com. It depends on what you're actually wanna build on. If you wanna build on our layer one, then that's docs.nervos.org or startwithnervos.com if you're interested more and just EVM development on our chain. So if you're looking to get started, like this is gonna be absolutely a, we've actually written a recent article by Travis, who's my coworker who put together what layer do you actually wanna build on? Which is a very heavy topic, but it depends on a lot of the different things because we've said like with UTXOs are you definitely more difficult, but they're possibly the future. So anyway, I would recommend that article. That's Travis, he's nervous and Ninja is actually in the chat room right now. And if you're on EVM for sure, it's Godwoken. That's a really easy solution for you. That's the Ethereum compatible layer. What about you, Chang, Hongchou? If you wanna build today, where do you go? Chang? Ah, okay, I thought Hongchou is for this. Yeah, so I think it's similar to Jordan's answer. Right now a good starting point is the websites so the doc website is very easy to find and the doc is docs.aliflion.org. And I think if you are a developer, the best way is to be ready to build something and try to, for example, build a simple D app, try to play with the SDK, try to run a phone notes and try to run also the whole stack to see if it's kind of decentralized enough. Yeah, so basically that's, and we are still very early and there are a lot of things to build. Almost all of the D5 stuff we're gonna to build with our virtual machine and our language. It's a new model with very unique security features. It's going to be very fun. So yeah, if you are devs, I would recommend just to try it and to enjoy it. Thank you, Chang. This leads me to my last question that I actually didn't prepare, but like now I want to ask you. What, so you've both been in crypto for a long time. Now you both have like quite an interesting track record. What gets you excited now in the morning when you wake up and you wanna start building? What is it that keeps you on fire about the space, Jordan? Well, it varies from day to day, but ultimately I get excited about just watching so many things be built. It reminds me of looking back to the beginning of the internet where back then I remember when I first got on the internet like the most amazing site was I think Coca-Cola's website, which it was silly. It was this bear characters, animated bear like sliding down a mountain. And there was, but that was the best that was available. Like that was, there was very little out there. And then over the years, we saw so much be built on the internet. We watched it change so much and actually improve the lives of almost everybody on the planet. I see that same type of potential in blockchain. It might not be as a parent. It might be something that happens a little bit more behind the scenes over time. But this is something that eventually everybody on the planet might have a chance to be actually using this technology. And when they go into the store, they're not gonna go into the store and say, hey, can I pay with Bitcoin? They're probably just gonna say, hey, can I pay with my cell phone? It's gonna be built in just like Apple Pay or something else. And it will be one of those, one of those, the main currencies out there be a Bitcoin or Ethereum or Nervos or Aletheum. It's like, one of these things will be out there. And maybe it'll be all of them that are out there. And maybe you actually walk into a store and you don't say, can I pay with Bitcoin? You actually say, can I just pay with crypto? And you'll pay with whatever you want in the recipient. The shop owner will receive whatever they want. You know, just be instantly exchanged. I mean, there's so much potential here on what happens here. There's so much potential for collaboration with decentralized systems, which are truly the only way that many corporations would ever see a way to collaborate on a platform. Something like, for example, video games is a good one because there's multiple different platforms out there. And all these big companies, these big publishers are in competition with each other. There's no way that they're ever gonna collaborate on the opponent's platform. But if there's a single platform that is decentralized, all of a sudden that type of collaboration becomes possible again, because they can actually move to something that is guaranteed to be neutral. So I mean, this is what gets me up in the morning, right? This is what I see out there. I just see so much potential. I see the innovation that continues to happen and just surprises me. Most recently it was NFTs with artists getting out there, thousands and thousands of artists and being able to finally monetize their work in new ways. Just, it's like, I never saw that company coming. It is something that I actively, at the very beginning, I didn't think it was gonna catch on. And then all of a sudden, this gigantic boom, all of these artists out here and so many of them are absolutely fantastic. It's just, it excites me because I don't know what's next. You can only see so far and the next big innovation is still to come. Well, what a passionate answer. I love it. And I really concur in most of the sentiment here. Cheng and Hongxiao, let's start with Cheng. Like what gets you up in the morning? Yeah, I pretty much, I would be Jordan. Gribito and blockchain decentralized technology has a huge potential. One of the most exciting part for me, I think it improves the freedom of the world. In general, I treat freedom as options. When people have more options, usually they are more freedom. So, blockchain and the crypto gives people more options when it comes to finance. Right now with the traditional finances, I think like, we are not sure what's going to happen with people's assets. Like if fate announced something new in the next week, probably the stock price is going to change a lot just because of the announcement. I personally think that kind of system is a bit weird for me. And with blockchain and with crypto, we're going to have different options. That's the first part. The second part is, because Gribito and the blockchain is still very, very early and there are a lot of things to build. I basically like, every day I have a lot of things and a lot of new ideas in mind too. I want to work on them. I only wish that I could work in parallel. I have several brands to program me in parallel. Yeah, so that's also, that's one of the big reasons. Like we can build a lot of things in such a young and potential industry. Thank you, Cheng. So, same question for Hong Cha. What drives you up when you wake up in the morning? And just so the others can, if you have questions, you can start requesting by clicking on yourself. So, Hong Cha, what gets you up in the morning? I very much echo Cheng's point about options. It's very important to have alternatives. And I think crypto and blockchain provides alternatives to the current system. That's one of the most important reasons why, yeah, I'm interested in this space. And also it's extremely, this space is very multidisciplinary and it serves almost like a driving force where you learn so many different kinds of things. And that I find very fascinating as well. So, yeah, that's the major reasons. Cool, thank you. So now I'm gonna open the floor. So if anyone here has a question, now is the time. Otherwise, I'm gonna free our speakers because we've already passed the time. So, Seto has a question. Let me make him speaker. There you go, Set. You have to unmute yourself though. Hi, I'm Seto and I'm a bit curious about this selfish mining you mentioned. Could you explain a little bit what it is and how do you address the risk? Okay, well, I can answer that on a very high level which is so selfish mining is a process where a miner doesn't actually share the block that they've or they manipulate the transmission of the winning block so that they have an advantage over others. There's actually a couple of different ways that it's done but it basically refers to cheating in the system. Now, I don't actually see if this is something that should be like, for example, if these were not decentralized system, you would say it should this be illegal. I actually don't see it that way. I see it as an inherent flaw which is in some of the early implementations of proof of work which is why nervous change there's around specifically to make this to de-incentivize this process because you're supposed to ultimately with mining it's to decentralize the pools. You don't actually know who's going to win but if somebody uses specific techniques with selfish mining, they can actually gain an advantage where they gain additional rewards that they were not supposed to get. So, yeah, I mean, that's the reason why NERVO's created NCMATS and specifically included a two step feature in there because it alleviates that by aiding the block propagation by separating transmission from actually confirmation. And there's a paper out there on our website, nervous.org, if you really wanted to get into the nitty gritty of that but yeah, that's the basics of selfish mining is it's kind of like cheating within mining. Thank you very much. Thanks, is there any other question for our host today? Otherwise, I'm gonna close this because we've extended the time already quite some. So I'll start the conclusion. Thanks a lot, Jordan. It's been a really awesome experience to have you on. Personally, I've really enjoyed the clarity of your explanations and the strength of your first principle talk and also the technicality of it is quite a very interesting mix that we love at Alephium and also that I love personally in general. Thanks, Cheng and Hongchao for participating for Alephium. We love to have this discussion with blockchains with which we share things. We think this dialogue is crucial and interesting and leads to better outcomes for everyone. I hope you enjoyed your time here, Jordan and that we can do this again someday. So thank you. Yeah, yeah, absolutely. Thanks for having me. This is great. And yeah, I'm always happy to collaborate again in the future and for anybody who's curious like how did this actually start? This is all part of the UTXO allowance is how we got introduced. And so anybody who's from the Nervos ecosystem, they know I speak very highly of that. That's a collaboration between all the UTXO chains and I look forward to more of these types of things in the future. Yeah, this is great. We've had already like a few discussions with Ergo, with them and we're gonna probably talk with some others. So like the UTXO Alliance is a perfect association in which we are very involved and we think it has a great future actually. Lots of work though. So thanks everyone. Thanks, Cheng, do you want to say something else? Or I can conclude the session for today. All right, that's all. Thank you for the great host as always. Thank you. Thanks Hongxia also for taking the time and to everyone for listening. We'll do more of this because obviously it's a very nice exercise and we're lucky to have such amazing hosts. So thank you a guest and that's what I meant. Thank you and I'm gonna close this now. Have a good day or evening depending where you are.