Arthur Breitman - Tezos X: Opening a New Frontier for All Developers - TOKEN2049 Singapore 2024

today I'm going to be talking about tesos x uh I'll get to the next slid of what it is but uh what it does is that it opens New Frontier for all developers um so what is tesos X so tesos X is a proposal and that's important to understand um you know tesos does the centralized governance we actually have decentralized governance is going to um dozens of upgrades now or sorry over a dozen grade and all of them have been voted by um the stakeholders of the chain you know it doesn't evolve by hard Fork it doesn't evolve by a centralized team you know dictating what should happen um it's all decentralized and so tesos act is a proposal for um tesos so what is it um well first it's modul litic I'm not a fan of the neologism but it says it um it's B modular in approach so it's a modular design but you get all the benefits of having a monolithic uh approach all the composability so you keep the you know you keep the performance of being modular uh but you keep the compossibility of being monolithic you get the best of B world you keep the interoperability and what do you get in the end well the goal is to get something that's almost like a cloud-like backand where any application you you know you could build on a cloud you could just go and get your um cloud account and and run your software um the chain if it's just so performant and so cheap you can just treat it as a cloud and that's that's the goal that's where we want to go with um TX so um how does this work um what are the different layers how is it mod montic so you start with a platform and the platform on the L1 that's your base that's where you have your governance for deciding what's going to happen for deciding how the chain should evolve that's super important I don't think you can have a decentralized chain today without onchain governance you know otherwise what you have is a uh is a you know like is a software company miserating as like a uh a ledger and it's not a good foundation you have your consensus super important at the end of the day why do we have blockchains you know we have blockchains because we want to have a record a ledger and you need to have consensus to record transactions um sticking so sticking is what allows the consensus the settlement um and you need to have fast finality if you want to have really good settlement guarantees uh and potentially bridging also with other ecosystems because you know there are now hundreds of blockchains out there um and bridging has become quite important on top of that you build a data avability layer so data avability layer of course there's been a lot of hype around them with many people different like building different type of data avability layers you've seen them with celissa you've seen them with like the blobs in ethereum and you know the problems of dank charting though we have a data avability layer on tesos you can try it today um it's currently being uh it's currently optional for the for the validators of of tesos to run it but it's becoming a full part of the protocol very soon and it's based on you know the full sharded data sampling uh approach you can build on top of that the low latency sequencers I've given a lot of talks on what sequencers can do uh the long and short of it is if you look today at sequencers I think the T has been bastardized because a lot of people use them as a crutch for centralization you know you'll they'll run a rollup they'll say like oh come in you know go on Shane go on Shan but you're not going on Shane at all you're going on some corporate companies private database and their sequencer does everything you know it controls the security it controls your asset it can steal your asset it can do everything so those are not a sequencer we're talking about we're talking about very very narrow sequencers that do one thing and one thing only which is decide on the order of the transaction so that you can get confirmations very very fast so that in less than a second you know that your transaction is going to be sequenced is going to be integrated in the chain and they have no other powers than that um they cannot steal money they cannot control the security of the chain so um be wored there's a lot of L2 Solutions which are just private databases masting and l2s that's not what we're talking about here then you have a smart rollup engine so we started um with a Mumbai upgrade in tesos last year uh with wasm so anything that compil to wasm and many languages compiled to wasm you know assembly script rust anything that compar to wasm today you can run as a smart rollup on tesos but we're adding more so we're adding uh risk five support which I think is an even broader compilation Target than uh wasm there was a while back if you follow the trends and Vibes a little bit where the crypto industry was all about wasm now it's like shifting more towards um risk five and you know we we're following this it's not just because you know it's not just because of hype there are a lot of benefits to this architecture uh fraud proofs obviously you know we wouldn't run any of this without fra proof I think it's a complete travesty when people are doing l2s and they don't have fraud proofs uh and also zero knowledge proofs love zero knowledge um you know I wrote a blog post in 2017 I encourage you to check it out it like has held really really well on uh how you could build thek rollups essentially to scale uh uh and to uh and to provide a lot of stut I think it holds to this day but the cost of the cost of DK is still very very high I think it's Fallen to the point where if all you're doing is just like regular transactions you're going to be fine uh even like some smart contracts but if you want to build a true cloudline backend where you can just run any application where you can compile it and you don't have to like gas optimize everything then we're not quite there yet but I think it'll be a drop in replacement if and when we get there run times that's you know what do you build on top of that so you know melson is the original runtime on you know for for tesos it's a VM it's an awesome VM if you want to do formal verification of your contract it's super important but what happened well the first thing is like people really don't like doing formal verification it's very very time consuming and second a lot of applications have been quite centralized people have been happy having centralized applications so if your application is centralized and you have keys where you can upgrade your contract and if formal verification is still very timec consuming um then maybe you're not going to bother but I think that's changing one is because AI is making formal verification much more automatable now um the long and short of it is that um fortification is tedious but not intellectually very complex um it's just that you really need to like slog through a lot of of effort to do it that can be automated it is being automated and the second thing is like I think over time we're going to see things Chang out um system which are centralizer more brittle the keys get stolen um you know they get U they get fined um they have to comply whereas system which are actually decentralized are more useful so I think we'll see some of that but for now um the evm is also a very very popular uh uh system and we've built on top of tesos eink which is a layer two solution it's 100% um ethereum compatible you can use it today it's not you know properly launch launched but it's on Main net um it works so try it out um and it's fully evm compatible lots of demand for evm of course in our industry but you can go much much much beyond that and I'll talk about that um including for example JavaScript as a runtime what do you build on top of that languages so JavaScript it's a language in itself so it can be a VM and a language but languages that we've known in love in the ecosystem like ligo smart pie and also more popular languages out there um like solidity a little bit a little bit more popular and massively more popular languages like python Java those have orders of magnitude more Developers than any of our you know blockchain uh uh programming languages and on top of that what do you build uh great applications in the defy space gaming space in art so many possibilities um so what are the key deliveries of is tesos Xing I went a little bit through the stack uh inter interoperability is a huge factor and so I've mentioned you know support for solidity which we introduce now on tesos VI eink but solidity is small you're talking about thousands of develop uh python Java JavaScript c those have millions tens of millions of developers they have libraries they have these deep tool sh these deep ecosystems and despite you know the billion that have been spent uh trying to make etherum work and trying to make solidity work like the this ecosystem of python JavaScript has had way more way more development uh at it way more support um integration so interoperability means interoperability in terms of code you want to be able to reuse those libraries you want to be able to reuse this human capital that exists out there but it also can mean integration with other chains but also non-blockchain system now what powers this interoperability is performance I would say a lot of early languages a lot of early VM like melson are very very concerned with performance because blockchains were extremely uh throughput constraint you couldn't do all that much and so your language had to be like really tailor fit to doing smart contracts and and and doing these operations but if you can com you know if you want to be able to just compile some code and run it and not have to worry too much about performance then you need to have a lot of performance in your backend so the target is a million um transaction per second CP uh throughput We've demonstrated that already so last year um at test Dev the T developer conference we demonstrated 1 million transaction per second on chain but that was in parallel across a thousand rollups so that's a modular approach the modelistic approach is we want 1 million TPS in a single rollup with subsequent latency and less than 10-second finality so that's um the key delivery B for tesos SE finally composability no fragmentation you know no one wants to be jumping between 10 different rollups uh moving their assets around their Bridges um no one wants to uh have to decide where is they going to be their contract because if I build my contract here it's not going to be able to access this other ecosystem and so you want the flexibility of just being able to compose with any other contract that's being deployed uh and enforce Atomic operations across those contracts so we could call it tsos Broadband moment you know it's a difference between the dialup where you can go to a few websites and you can get some information you can read a few things online to broadband where the web stops being just an information content delivery and become an application content delivery where essentially the browser has replaced the operating system as a way to deliver applications today you know very rarely now unless for like some very very specific like um resource heavy applications really do you imply install applications on on a computer anymore it's it's all on the internet so the way that the internet shifted from just content to Applications I you can see a radical shift um in blockchains in the same way in terms of the application that is supports um if you give them this support if you give them the ability to use all these languages if you give them all this throughput so what does this mean in practice for um for users and developers so again going in the old Paradigm of blockchain I would say censorship resistance is the bread and butter why is that well that's why they were invented for you know the whole point of the invention of Bitcoin is you want to have a decentralized cryptocurrency and it needs to be senstive resilient and it needs to be very very resistant because people are going to try to corrupt it and that comes at the time with those systems with a huge cost right there's a huge cost in running those systems they're slow they have high latency and whenever people have tried to build you know applications I would say a very very good rule of thumb to know if a good applic an application was good for a blockchain was like does this actually need censorship resistance are there nation states colluding around the world to stop it and if not why on Earth are you paying the cost of being on a blockchain it's it's it's a huge cost it doesn't have that much advantages compared to a centralized database so why would you why would you pay that cost but if you reduce the cost if the cost becomes minimal if basically you are within maybe one order of magnitude you know a factor of like three let's say or four from cloud cost then it becomes very different then that question goes away then it's like well you know maybe I don't strictly care about you know resistance but I care about convenience I care about the fact that payments are natively supported I care about the fact that people are already un boarded um I care about the fact that they have their identity I care about the fact that I don't necessarily have to pay the bills uh for a server right um it can be paid by customers directly so there's a lot of like small but benefits that completely add up and if you can get the cost low enough it TR transform the type of applications that can be built and that can be deployed on the systems and so mention programming languages is I think is super important piece of that um they're used by tens of millions developers not tens of thousands they have better documentation better tooling way more libraries way more additional material and also and that's I think it's you know there's a dawn of absolutely meaningless Dil and you know like Dil hype around uh Ai and blockchain that doesn't make sense but I will say this AI systems are very very good you know they're getting very very good at coding and they're much better at coding when they have large corpuses so they are better programmers of python and JavaScript than they are of uh smaller languages and so the pressure you know everything I said about the benefits of using JavaScript and python is going to keep increasing and increasing as much more and more code is being built by AI so that I think is a one of the most important reason for moving to those languages is uh the the tooling is getting so much better and so much more important um so because supports matter we're starting with JavaScript uh to open tesos to tens of millions of Builders and when I say um JavaScript so right now the project to do it is Justice and you mention you see I mentioned a few projects etherlink Justice right now we have a few initiatives and the goal is for those to coales into this big project that is tesos X so jsz pronounced Justice is an experimental rollup that you can try with a JavaScript execution environment um it has a web like interfaces between contracts so talking to another contract feels like making a call to we web server um except you know with atomicity and without the latency and all of that you can try it now there's a link and there a QR code if you're a developer please um try gstz and let us know what you think um one thing uh real JavaScript hasn't been tried so you've probably heard of like oh yeah but you know didn't try JavaScript in that project and that project and that project so yeah I wonder because I thought like look JavaScript would be huge we need to do it but you know didn't they do it in uh didn't they do it in like way back in what was it called like links or something like that no no they only supported like a few contracts and you had to do governance toy it didn't they try it in near no your contract had to carry the entire interpreter with you so like every single blockchain that has done like JS there's always like a little fine print um it's not full support we're going for full support of JavaScript uh without the caveat without the fine print so JavaScript python melson all this VM all in one big rollup cheap fast always on with easy integration to existing code a lot more on logic allowing millions of users to have real-time experiences thank you