Louis Bellet & Maksym Naichuk I Yellow Workshop I ETHGlobal Prague 2025

[Music] Come everyone. Um, thanks for thank you for joining the uh, hackathon. I see that I have my taipe tie shirt. It was not even on purpose. This is where we have started uh, this project.

So uh let's get right into it. Uh so on the agenda uh small introduction to yellow and then we will uh speak about clear what is it our motivation how does it work um from the user perspective the user journey and uh then we will go over some demo apps that we have built uh these are open source already published I believe and uh we will be sharing the links and you can start hacking right away from this uh call. Um and uh you would be eligible to the grant program if you want if you are going into the Ethereum Prague hackathon. Uh just bear in mind that you will have to start Yes. when the hackathon start.

Exactly. Um though you can do two apps you know you can do some practice. Uh so introduction to yellow network. Um why why are we building clear? Um it's because we are a clearing network leveraging state channel for high performance web free application and in our case our high performance application is a global network for high frequency trading.

Uh so we needed this infrastructure layer in order to um let's say move to production our um our high performance application that we call nodax. Stay tuned. Uh this release is due to next month. Uh so might be interesting for you hackers as well. So what what is this application is decentralizing trading through B2B or you can think of backend to backend server.

Uh so here you should think of data center high speeded uh TCP IP localiz localized latency and the vision is not to decentralize like unis swap but rather have businesses being interconnected. The core protocol is a smart contract called clear sync. Uh you will see a lot of clear prefix in our ecosystem. Um and so clear sync is really the core smart contract and um it's really built on top of state channel uh framework nitrolyte. We are the initiator of the ERC 7824 and um we are looking for contributors and uh Prague is going to be an event for us for the initial release of the SDK that goes with the ERC.

Set myself a timer. So what is clear? It's a chain abstraction layer and also a network. So the protocol express um a new logical layer that abstract yourself from using directly blockchains. So you have like a new sets of entities and you do not need to speak you know G API or transactions uh smart contracts etc to to leverage this network.

uh so it's fully it's yet another way of abstraction and I would say it's one step further than uh account abstraction because uh account abstraction is really attached to EVM ecosystem and does not necessarily exist elsewhere and our uh chain abstraction is really crosschain so uh it could be possible to have it on top of Bitcoin of course Solana and the for the ecosystem uh by let's say migrating and deploying nitrolyte primitive chain layer and then clearet will unify this into a unified accounting layer. So this is not a blockchain once again I do not believe we can fix problem created by blockchain with more blockchains. This is a mesh network built on top of all blockchain. So I really the mindset here is we couldn't need another blockchain. there is plenty out there and uh let's build on top of it so we can have a new layer between our application and and blockchain which should uh let's say facilitate the [Music] um having my application live across multiple chain without having to worry about bridging or redeploying applications or sending transactions.

sponsoring transactions etc. And you have the added benefit of having the super power of state channels. Um that means able to drastically uh lower the transaction cost because it operates offchain most of the time. So for example your game don't need to pay a transaction at every action and basically you can offchain uh important action that you want to secure uh without having a a blockchain consensus under the hood. So to summarize it, it enables you a web to style application development uh client server restful architecture the things that we are very productive in development um and also using legacy application.

So a web to games could be leveraging blockchain now without too much difficulty and without entire rewrite. Uh so this should be specifically interesting for us later to engage with audience that does not specifically develop web 3 application. Our motivation in creating clear net. So for the story is that state channel has been quite difficult over time. Um they have started I would say in 2014 uh even earlier uh lightning network is one of them.

I would say it's a success story but there have been a lot of research groups on top of let's say Etherum dominantly which have made state channel framework and design but they never really went in production with their project. Um state channel.org or was really the leading um and being very close to production grade. I believe today we are the the next project going to production with state channels. Uh this is this has been a four years journey for us and this is why we have created a layer.

uh this is to gain productivity and reliably reliability in our development and team. So we wanted to uh create those intermediate layer to make our software stack more reliable and so we can scale it update it independently without having to rewrite always our NODAX application our high frequency trading applications. You can imagine that as well by creating those layers um uh we are letting layers evolve independently. So for example on the chain layer we can add more chains and you wouldn't need to rewrite or change anything on your application layer. So they can it's a bit like the internet and let's say Wi-Fi.

You can upgrade your uh ISP. You can go to fiber but you don't need to change your phone or whatever. So this is the same idea here. Um I think everything is said. So, Nitrolite was born in the previous hackathon if Taipei and there is a YouTube video that we can share in the chat room here uh if you want to catch up and see uh the introduction of nitrolyte.

So, that's really the core component for managing state channel. And now we go into the how does it work. So as I said before this is a multi-layered stack. The channel layer is onchain. Uh this is a bit different point of view uh than traditional state channel if some are aware.

Um and usually there is like virtual state channel on top and uh we do not call them virtual state channel anymore. We call them ledger. Um so on top essentially a channel is an escro smart contract. Um so a channel is less elaborate than a smart contract. We have created and decided to name it the ledger layer.

Uh and you can think of it as like accounting lay ledger. Uh so really uh a chart of account ids and entries into each account uh and some of these account are individual funding. So Alice can have her own account on the ledger layer. Uh we call it the unified account. um that would be for example USDC all merged together in one balance but from multiple ecosystem.

So this could be USDC from Tron network. I don't even know if USDC is on Tron actually um USDT more popular. Uh so you Alice could have $1,000 but that's the combination of free channel on three different blockchain that would be her unified let's say current account and then she can do suballocation sub account that are shared among multiple participant or technically shared doesn't mean a participant can just withdraw the money. It means that they're going to engage themsel into a new business logic that could be a poker game for example. So when you play a poker game, it's not because Ali sit on the table with funds that another player can withdraw it.

There is rules. Um and that is basically your application layer this idea defining those rules. Uh so we call them V app a bit by opposition to distributed application. So virtual apps uh they really sit on the virtual uh ledger layer and those apps really don't need to understand blockchain just the simple accounting layer that we offer and what we call the corum so the signature threshold for unlocking funds. Um so this is like a sort of um port and adapter design uh between you know we could say blockchains are plugged into this clearet layer and on the other side apps can just connect on the ledger layer this abstraction layer.

Uh so technically we have a go implementation of the clear node. There is much to do on this side. Currently they are not peer-to-peer connected. They are independent. Uh and there is like a expectation of trust in the node you connect to.

Uh a bit like you should trust your GE node when you're using it to be honest. Um doesn't mean always deploy your own. Uh so this is available on GitHub and uh feel free to browse the source code and this is essentially using websocket RPC protocol for consensus but uh at this stage of the project we do not need very advanced consensus for our demo applications. So the user flow, the abstraction is really uh the Las Vegas type of casino. You just come in, you go to the cure, you tokenize your your fiat, right?

So you just uh deposit cash or credit card and you're going to get chips, poker chips that are only valid within the casino. So here this is only your ledger is credited on the clear net network and um and then you can start using it. So you can just sit down at any table that will be an app and you can just play uh unlimited numbers of times without you know manipulating your credit card or your cash anymore and you just have this virtual balance of poker of casino chips. Uh, and then you can withdraw on any um on any form. So you could say to the cashier, I want cash, I want euro, I want credit card deposit, I don't know what's available in Vegas to be honest.

Maybe a QR code or something else. So here, yes, Alis could withdraw on a target chain and uh um removing the need of bridging. So we will go into uh some business cases. I think uh maybe Maxim or Andrew can take over and present those. I don't know if I can give control on the slide from here.

Uh but I don't know I mean we can I can just tell you when we yeah it's only free slides. So yeah, so we prepared uh by the protocols three business cases. I would say the from simplest one go to hardest one. They are leveraging different types of the implementation of the app. So the first one is the no back end application named by the yellow pay.

Uh in easy in easy words is the lightweight payment system. So you as usual duplicity and as Louis said before you locking the funds and then you can instantly uh send those funds to the other other users be based on this one you can do for example some merchants uh merchants apps etc stores uh and the cool things of it that it's like really I mean you don't need to wait for the for execution of production or something it's fully offchain uh we will have also the example of it And uh yeah and it's like yeah you you will be able already tried it soon and uh and yeah I guess that's all for this one. Uh second one is the classic game tic tac. Uh we have few points. So yeah it's classic turnbased game I guess you all know.

Uh this is uh this app is showcasing how to build some competitive apps uh with the minimal back end effort. So here you can leverage our RPC for the for the terms and the final results. Uh Louis I guess yes we can showcase here. Uh just click the the and uh also we need a more better quality. Yes I guess you can stop and Yep.

Can you uh make it from start I will explain the flow. So from the start uh if user one is the connecting to metam mask u locking the I guess default funds we didn't yeah locking some funds uh making the few signatures then it's automatically created a room for you other users that already existing channel uh join the game uh they play it one of them will win and it will it's automatically sending the result to our clear node and making the entry in the ledger table that user A is the I mean win some funds and user B will need to give those funds and uh yeah all of those APIs accessible so you can build I mean it's and it's really flexible from the apps point of view so you can build a really cool apps and uh and UIs and I guess yeah I mean the let's go with the short case it's the hardest one I mean most complex so it's the snake battle uh this app is showcasing that you can have the back end server uh and you can for example build all your business logic uh all your business logics on your sorry we have that uh you can build all your business logic on your servers and then just leverage in the I would say state channel management and channels management with us uh this game is requireing the real real time and good servers and so yeah you can So with the with the demo it's the same flow. User one is open the channel. Uh user two is already have the channel. Uh few signatures uh depositing.

Yes. Waiting for that. And then yeah we just sped up a bit the game itself. But yeah, the cool thing in this protocol and in our idea that you can play in a row. For example, you can open one state channel, lock funds and do uh continuously.

You pay for one guy, then you play another game, then you buy some in a merchant and then you play third game. And uh you don't need to do any blockchain interactions with it. you have already those funds and we are leveraging all of this and leveraging games results or again merchants uh in our system. So literally you could play all night without a single blockchain operation and maybe on the morning cash out your earnings. Of course we not focused on um games necessarily.

Our core business is trading. Uh but game offers a very good uh easy to learn, easy to understand business rule that everyone knows. That's why we chose tic tac toy and snake as as everyone usually knows the rule of the game. Um, and snake is is chosen because it's high frequency, because it's uh it requires speed and usually the more um the game is going, the faster it goes for the players and the faster you move your snake, the better you will play. So, this is really a typical case where blockchain finality will be a huge mess.

And here there could be a very um uh yeah, that's a good question. Let's think about your question in a bit. Um uh yeah, I forgot my thought. Um the game could be uh uh high frequency uh with uh with no latency while you still leverage blockchain and rule enforcement. So the the game could be high stake could think of the Olympics of snake battle where really there is an incentive for cheating but we can have referee and third party that will control that nobody has been cheating.

Uh so that's possible with the protocol. Uh so let's go on the question. Uh Arando I think uh check the chat. Uh so I think my answer would be um I don't think those types of game would require you to get a gambling license because they are really decentralized. They are user to user.

Uh so let me take the US uh for example. In the US it is legal to play with money on a poker game with your friends in any state. I mean I in most state have been let's say maybe there is a state where it's illegal. So you can go to a bar pay um your deposit for the game and do a tournament in a bar with new people that you met. Looks like gambling.

Feels like casino. uh but the bar is only earning from drinks. So this is very common for the Americans maybe we have some in the audience who go to the local bar for a Tuesday poker tuna tournament. uh this is entirely legal and I would say since this is your private keys and there is no uh middleman or if you're not taking fees on the game for example and you your model will be maybe advertising uh you could very well operate uh a legal uh poker application that might even work in the US well let's see how how they u agree with that uh so that's all I can say uh but in many case like we're thinking probably sport betting also may not require a license because this is between friends um the protocol is really not a centralized authority and you can really implement it in a decentralized way like money games with friends. So for those going to uh if Prague um we have done two prizes for the best application and two prizes for GitHub contributions.

Uh so they could be profiles that rather hack and uh learn and help us improving this protocol. We have to-do list feature to add and uh improvements or bug to fix. I mean to fix it's you to find it actually but I'm sure you can find the bug and fix it. Um so we will uh list GitHub issues and uh we'll be very happy to be uh in the in the Prague event and be brainstorming how we could do protocol improvement uh iterate on this ERC um uh and maybe improve the design especially I think we will be focused on ability to challenge feature uh challenge change is uh the disaster fund recovery, the escape hatch or state channel. So this is when everything goes wrong, the poker game is dead and Alice want to take back her chips from the table at any time.

This is called challenge and you can do it on chain when all backend services are down. Um so this should be an interesting topic to have during the hackathon otherwise building application. So if for us if you have an existing application you want to migrate it it is fine just I'm not sure it works for the if global rules uh just don't tell anyone um but we have designed it for you to be able to use existing code and um and port it to this new framework and it should be easier than integrating a new blockchain I would say like moving an app from uh let's say EVM to Solana is quite tedious. Uh here our API is simpler and the primitive are simpler. You just need to deal with this cash accounting.

Finally, a word about the grant program. So the Yellow Foundation is offering up to $50,000 in funding. Uh we providing mentorship accelerator incubation service. We have a big incubator in Chungme Thailand. Um feel free to visit anytime and with funding startup as well.

We offering seed funding since many years with yellow capital. Um we would like to focus this program on trading businesses but also now with this clear net release we open it to more use case. Uh so it could be payment industry, it could be anything that requires escrow, could be betting, could be gaming, you name it. Uh that should be fine. And uh it could be real world asset, it could be gamified tokens, it it could even be one day NFCs, NFTs.

Um and uh this is all eligible by our program. So we're looking for uh app developers, businesses uh and of course hackers to help us uh maybe with security evaluations and other improvements. Uh so that's it for the presentation I believe and [Music] um uh yes I think the audience can can ask a few questions now. Uh but we will make the video short so maybe we can close the recording. [Music]