EthCC 2: Francois Branciard

okay so hi everyone so I am Francois Vasya i working at i exac and today i'm really pleased to present to you what we experiment with a substrate today I will present two things first I will present what we are doing with bridges we have seen in the previous presentation that bridges are part of the ecosystem coming and then I will present our first proof of concept with substrate so the title of this presentation is how to build a domain-specific token-based chain and I will not read the abstract because it will spoil all my presentation okay so the plan for this presentation I will briefly introduce you exactly ecosystem and I will put it into web three perspective and then I will focus on showing what we are delivering today we plan to release for next May what we call consortium bridge and I will explain what is about and then I will present our substrate talk we have done for a few few weeks now so briefly I will introduce the ecosystem exactly custom for those who don't do didn't know we use blockchain for two things the first thing thing is to to create a marketplace so the the blockchain is used to seal deals between actors into the blockchain so when we sing active there is root three main actors the first one is a resource providers so that means that service can monetize their resources so this is the first actor the the other actors our application providers and also that I said providers so we we see that all these are talk and since orders between them to to sell or buy their resources and then this these deals are sealed into the blockchain and we can see that for the resource providers we have Walker pulls and walkers so the network topology is like you have minor pools and minors and we we have entity that yes and Walker pool with this topology can have quality of service tour provide different kind of services of our resources so this is the first part of the blog this is the first part the marketplace and then the blockchain is also used when the deal is done it's trigger execution option exit option compute and so we have what we call the Paco it's it's basically a batch introduction that proves that the execution it's it's correct and so you have a trust of the result thanks to this protocol so it's an all over view of the ecosystem and to sum up with option compute you can extend u-dub scape as capacity and with on the top of that we add in an economy game to preserve the trust of the result and we use the blockchain to to create this kind of marketplace for data set application and self I do so in a way three perspective we are some second layer protocol focusing on Ava competition and so this is an extract from the wave draw with three stack so we deploy our smart contract for marketplace and prove contribution on my net but we also plan to to allow every allow others to deploy us my contract on their own chain and this is what we for the next release in May we plan to to propose for consumption to deploy this kind of smoke on track on the unchain and we deliver Internet the entire solution with existing technology that's mean that we plan to to have bridges between the main net and PHN to preach assets and allow people to to build their consumption and and I will explain this in in the next slides so this is for the delivery part fund or next release and in parallel we have research and we experiment with substrate and the polka dot ecosystem and the idea is instead of using bridge under a proof of authority we we can delegate this relying messaging between chain to polka dot and polka dot will will have this own incentive to do this relay so it's part of our idea but that and we also need what has the internet in the in the previous presentation all this kind of governance models and also other thing like that so our ultimate goal is to to have a not autonomous chain within an incentive and also all the government's parts that is a enroll Schaffner research now so this is a two aspect I will now would present so the first thing is consumption bridge we start from the imperative region then we use PA network talk token bridge so they introduced the possibility to to bridge ERC 20 to 20 I don't know if you are familiar with breech baby I can explain in one once one second so Bridget you have two contracts you have a contract on the main the main net and a contract on the on upo a chain and you can allow people to transfer asset in this contract and some software watching the transaction on the main contract and create the asset on the other side of the bridge and the other side is it's your proof of Authority chain so there is no supply creation you keep the the supplies of you as a set but you you can breach you new value on the other on the other chain and so you can think like deposits and then you can also withdraw and coming back to the minute so it's the basic overview of what a bridge can do but we needed also for consumption to to do not support everybody coming in the chain beacon because as you cannot block somebody to transfer token we introduced a feature proposal to have a white list of people that can cross the bridge and so we currently contribute ins adding this new feature to to the pure network bridge so I will illustrate what kind of use case it can do so imagine with with a consortium you you have application verticals you have a company that can be the data provider they have data that can be trained to that can be used to train models and you have another company that have application and the third company can have for instance I'll tell resources of Japan GPO and for instance when proof of walk will be story maybe there is alder GPU that can be used for all those other tack tasks so it's company can represent an authority and they can join together to automate their business so this is the first case they can automate the their business and its providers will deploy the data set application or worker pool it's like there are reference they deploy there are reference for their resources and they gain the shared audit ability of the usage between them and that means that at the end of the month they can say they can say they can see what usage they have to pay each other so this is the first level building this POA without bridge and we will see that we can add other levels of features so when you entered you you cannot reduce a bridge under the the the same authority and this can automate to containment between them so it's it's time and one Authority can deposit and withdraw and then all the audit ability and contraction are in the PIO in the home chain and the feature that is they can add the dynamic prices for their resources so we this is the second level we can add come feature and then they also can average on customers that means that if the formula is like this if this user our customer trust this Authority and like the efficiency of the services provided by those company and if this Authority whitelist the user the the user is able to to to use this asset in into the home chain of the P way so this is the kind of use case we we see useful form for consortium and so what's happen if you are not whitelist did in to cross the bridge so there is a like a sonic loop that as we cannot block transfer FC Twente transfer in fact the bridge opted to to reform the the user so the asset is given back to the user and today in Bridgeview you can add also fizzes visa smart contract so that means that yeah only if this will will be removed from the from the transfer so it was for the breach part of what we are currently working on and now I will present you what we experiment with substrate so the idea is we start with the substrate template node modules it's it's like a minimal block chain with some basic modules and we add we try to add custom modules on it and we focusing on the the prove of contribution part when we we mobilize tasks and contribution for option compute and it is economic game incentive as part of our platforms our objectives was to familiarise ourselves with wisdom and the next technology to come and also see the difference between working with smart contract and runtimes and also all the governance model we need today for something like if you have a domain domain chain the blog production policy of great ability and all these kind of governance that need to be solved so this is another graphic from the stacking you will see that the PUC only covered the contribution part of the older stack so I will briefly introduce the contribution we will schema to understand later what the PUC is about so if you have enough chain tasks the the workers have to do 2 to 2 calls the first call is what we call contributes that mean when you contribute you have to stake and for the contribution you have to only two parameters the first parameter is your result boat that means that at some point many workers will come and vote and at the end there will be a consensus of the vote of the result so this is the first part the vote on the contribution but this what is some way item so this is what the second parameter is contributions still and this contribution SIL will be used later to to further revealed files so every workers contribute and at some one there is a concern and that means that old Walker I'm invited to reveal and at this time they only provide one parameters what we call contribution unsealed and when they reveal they get reward of / / according to the valid rom contribution to achieve this option task so the next is it's the same thing but just just look at the colors you see that when you vote there is two levels and when so the contribution is is SIL and when you reveal you basically reveal the one part of the previous level and in fact the runtime or the smackin and the runtime of the smart contract check that results what was valid and also the contribution sill was valid and then so after that they can trigger payment to reward actors like Walker's nope so it was just an introduction of our economic game on for the option compute and so we try to model eyes this this economic game and and code into substrate today we we have developed this into smart contract and we wanted to see you what the difference between smart contract and writing this kind of code in a substrate runtimes so substrates come with some basic modules so you have what basic module of blockchain like balances consensus systems and of course you can customize it but also you can add new modules for your own logic and Here I am seduced a new module exact module into the runtime so you can define your mod models some models not here I only describe the task and contribution so simple model is Asian and in the contribution you see the two parameters I have talked about the vote and and this result SIL when you have a runtime you also define a storage basically is a mapping of all the tasks and all contribution into the runtime storage and like in smoke on track you have function you can call here you define you your functions that can be called bye-bye wallets and yeah there is a function to create task definition and then to contribute on those tasks and to reveal on those those tasks there is a subscribe interface that allows you to quickly test your new runtime so you you run you blockchain and then you connect your interface to this runtime and you see that in the console you can access the runtime then your module and then all storage of a function of this module so here I create a interface for threating tasks and then in service to to contribute and reveal so the contribution you need the task ID to on which you want to contribute and the two parameters we have seen the vote and the silh of your contribution when you contribute you take in contribute and when you reveal you will reveal your one parameters and you get reward if you are part of the consensus so we create icons scheduler is like the manager of worker pool and we also create a work identity worker one work of two so the in reality the replication is a more complex than only one number the this is a there is repetition and a score but yeah I simplified the replication so I when you create tasks you you ask to the number of replication you want to achieve other consensus and then we can see that the task is created and no contribution has been received yet so there is no task consensus after that workers contribute on those this task they have the they have the same vote so that we will reach the consensus of the option task and the silh part the last parameter is different for each workers so the consensus it reached the task consensus is revealed and that means that all workers can have incentive to rebuild to get reward and the workers really reveal and so at the end as he is his part of the consensus it gets rewarded so basically it's what we experiment and it was quite easy to to implement new logic and new with a substrate so I think it's really great tools to to quickly bootstrap some some some chain logic and I will I will finish on speculative ID with what we can do for further so I see that in fact you can real the specific logic and pollution and at some point you you will have also difference performance needed between use case or security our governance will be different so it can be easily customized for each use case and but at some point you want to enter a pair ability between between you or your your domains and it's where I think polka dot is a key part that allows this cross send message passing and you can imagine that if you have a domain specific chain like exact that the ocean compute with this optimized chain for doing that like we have seen with tasks application on game incentive for option tasks and on the up other path there is another domain like donation we can imagine that giveth elise o have all different security models and different logic to do forest and the raise money they have milestone and this is a different topic but I don't at some point if or if we if we can bridge that them together we can create this what we call it we can call this you this donate computation offering such mean that when you you will do a donation into Givet Oh Alice dimension you want to be sure that your donation used to compute this kind of of application only and you can track that you don't donation is well well used so to do that we can do bridge between one solutions to doing bridge between the two Parchin or to chain but if you have an external actor like the government and you have to have tasked the addiction because you have done this donation maybe the government do not trust this power chain may be the because of the the security is not as I of oto-chan and if you do not trust both power chain you will do not trust the bridge between them and maybe it will be okay if you have a proof in this polka dot really chain that is a third party and you can imagine that you you will prove into this relation some non-reusable proof like that you have done the donation and then the government will be okay to to give you tax deduction so it was a speculative ID but but to introduce what what interpretability can can any power so I will finish with this slide so we we see in the substrate roadmap that option compute is the topic or arriving so we really happy to to see this option topic coming and we see that we we are for the economic game we are working on a proof of contribution and for organized group of exactly executed task we have this worker pool of network topology so I start with software meta exactly I would finish with executive threat and we really happy to have this common vision and synergy bit on this topic so yes thank you for your attention you have any question