Let's Talk Basho

hi everybody this is Charles hoskinson broadcasting live from warm sunny Colorado always warm Always Sunny sometimes Colorado today is October 3rd 2022. It's a Spooky month the month of Halloween the month of Canadian Thanksgiving for all you connects up north and a lot of cool things so there's cnft con I will be there and then there's also a big pays event rare Bloom that's coming on and a lot of other cool stuff so it's going to be a busy month but I did want to make a video slightly technical but not too bad about Basho because a lot of people ask about it and these are the last two eras of cardano so wow as always the nice screen let's get right to it so there are two eras left for cardano there's Voltaire and Basho okay and they they're being done concurrently so like all things we try to make things run as quickly as possible and the Basho team is very different than the Voltaire team so the Voltaire team is all about governance and there's an enormous amount of progress that's under the hood and this is things the members-based organization a constitution [Music] there's lots of sips that are currently being drafted and soon to be announced and we'll have a lot to say in November about all of this it'll drip its way out so not too long and that's going to be at the cardano summit let's start a little bit about Basho so basio has a lot of stuff in it so there was diffusion pipelining was an example and this is a spectrum so we launched version one but there's all kinds of optimizations that can be done there and this came with vossel we've already seen a huge impact there are improvements to foreign plutus so version two is already out and next year there'll probably be some sort of version three and we've already seen a huge impact some people say a 10x reduction in script size and all kinds of things so lots of cool stuff there these are known things and we've already seen some of these things come out in vossel so we're actually already starting the Basho era we're in it right now but then there's some things that are on the horizon so there's mithril for example in mithril is rapidly approaching version one whether that happens this year or not remains to be seen about the development velocity but spos are already testing it we already have a pretty nice open source project materializing pretty good documentation and there's a lot of open questions of well what does mithril do for scalability now if you're on cryptocurrency Reddit they say absolutely nothing and it achieves nothing and it's just all made up marketing speak and that's because they're idiots so we won't pay much attention to them the reality is that mithril is one of the biggest pillars of Basho so what mithril does in a nutshell is that you have your layer one so this is cardano and really what it's doing is it's creating this kind of this overlay and layer 1.x ish and that overlay basically gives you the ability to construct proof certificates and that's a super useful capability like for example when we think about sips for on-chain governance just as an example now the naive way would be that everybody votes okay well that's fine but what happens when you have three million votes occurring for each thing wouldn't it be really cool instead to construct all of that off chain and a network and then just aggregate it together as a threshold signature and have a single transaction occur on chain and that has the same impact and the same trust assumption as if you've aggregated all three million of those well that's effectively what mithril does in the special case of mithril right now is for like clients and for fast bootstrap okay so basically bootstrapping a full node quickly and then giving like clients full node Security even though they are a light client well this is a framework and it's a networking framework and a proof construction framework and a commit framework to get it onto the cardano main chain and that framework is basically going to allow many to one or some part of scalability is saying well hey we want to do lots of stuff in just one thing if you kind of think about what's happening with version two of plutus it's doing the same thing with reference inputs and reference skips scripts and inline datums and these types of things it allows you to basically have reusability it allows you to compress a lot of things that are happening so you're going from many to one so instead of doing something in 500 transactions you do it one transaction and it's just constructed a little differently the nice part about mithril is it's moving very quickly and the nice part about q1 and Q4 to q1 this time Horizon is that we get this fast bootstrap and like client application but we intend on using mithril in in just a moment we'll talk about it input endorsers we intend on using mithril for voting and we intend on using mithril for other applications that come up and that many to one concept okay then we have Hydra and Hydra is also fast evolving and version one is one another one of those Q4 q1 situations and these are public road maps you can actually go to GitHub and look at the repos forum and you can actually see the road maps real time and they're on different release Cycles I think it's three weeks here for the release Cycles it's a very agile project two weeks here for mithril nice proof of Concepts as they go through and basically Hydra a lot of people think that Hydra is kind of like this because there's conversations about it that Hydra is like this overlay Network that's going to sit on top that the spos run and they run all these heads and the heads communicate with each other and somehow it's some sort of accelerator Network you could certainly imagine a world where that's the case and for things like microtransactions that would make a lot of sense but the current version of Hydra that's under construction is more like basically a Dap embedded software so the way that we design plutus is that you have an unchain component and that's your Pluto script and then you have this idea of off-chain infrastructure that coordinates with the DAP to run things and this is very common in the ethereum model as well and they have all these things to combine these two and then off-chain infrastructure sometimes is centralized sometimes it's not centralized it depends on the specific Gap the idea of Hydra is that it would be part of that off-chain kit so it would be embedded adapts that require a lot of scale and basically it would do something similar to what's occurring with mithril where you go many 2 1 or sum so you can imagine a lot of use cases again like voting like dexes things for payments etc etc where you want to batch a lot of stuff together and you use Pluto scripts on chains sparingly so you only have a few and then you do a lot of stuff off chain that is kind of what the version one of Hydra is looking at having this acceleration Network that runs off chain for super high throughput for lots of transactions it seems it would be a super useful super nice thing to have similar to a lightning Network or something like that and we are exploring it there's something called the Hydra tail protocol but the reality is that if you want to achieve in next year and the year after high throughput you're really thinking about the whole banana the whole architecture so you're thinking about the design of plutus itself and how that off-chain and launching communication Works you're thinking a lot about the extended utxo model and how to most efficiently use it but then you're also thinking about having specific off-chain infrastructure that can work on chain that has an on-chain representation in this case a Hydra head and by embedding that inside the DAP then basically you can take lots of stuff and you can make it one thing okay so what does that mean it means your cost goes down your operating cost your up cost and that means your speed effectively goes up because these can be run with high efficiency infrastructure they can do tons of things lots of TPS and you also Confederate them as well so this is kind of the design goal of version one and there's already several dap developers in the cardano ecosystem working directly with the hydro team and actually some of them have started making open source commitments to it obsidian is one and so that's really exciting and they've made some commits to Hydra so Hydra is becoming rapidly a full open source program and what's cool about it is that everybody has an incentive to add to it because as they embed it in the guts of their application then basically the future versions will be done by not just the cardano core team so the foundation and I O and others but actually by the community as well and the goal is to reduce your operational cost and improve your transaction throughput and speed mithril is also closely related to Hydra in that components of mithril could connect to and work with Hydra which is why they have shared teams so as mithril advances Hydra advances these two things are actually going to be massive for improving scalability now the biggest single contributor to our ability to have throughput will be input endorsers I don't think input endorsers are necessary in 2023 because if you just look at improvements to diffusion pipelining and what we're doing with plutus as well as mithril and hydro the reality is we'll have more throughput than we'll have use because the effective TPS rate is is quite high already again cryptocurrency Reddit they just have an incredibly naive stupid way of doing things and and so they're they say oh you're nine TPS but the real impact is in the dozens or hundreds of TPS when you actually look at the blocks and how people are using the scripts and how things can be chained together so input endorsers what we're doing right now is we're writing the sips for them so we had a humongous design process it started in January of this year and all the way to last month in September in California we had a huge workshop with all of the researchers and in June we had a big thing when Edinburgh with all of the researchers and basically this allowed us to get the design figured out so now Ben Beckman the head of architecture at i o Ben what he's working on is actually translating this end result into sips and there's a lot of simulation that's going on so input endorsers are not a small thing they're actually a complete reimagining of the blockchain structure okay so what happens is that you go from this stand process of a chain of blocks and there's only one canonical view to basically doing a lot of these things asynchronously and in parallel and they just kind of get batched together and then asynchronously and in parallel lots of stuff happens and then they get batched together so this concept of input blocks and key blocks as discussed in our parallel chains paper now this batching you think of mithril certificates where you talk about batching transactions events together you also could reuse this technology in the extended utxo model to patch blocks together so it gives you a way of figuring out conflict-free history and potentially what this means is that you have asynchronous parallel processing of an enormous amount of events so the throughput gains assuming new data structures and Assuming New Designs are gargantuan for this because you basically can be real time continuously running computations right now in the current design only 0.25 of your block time budget is for computation so running execution of scripts so that's incredibly limiting 0.25 not 25 but 0.

25 percent of your block time is for script execution so it's a very scarce resource here it's asynchronous so it's continuous execution foreign so we expect to see not just a 10x but a very very very very very significant increase in speed and a reduction in cost that's what we're looking for now what's really cool about input endorsers as well is how do you do this where the network doesn't stall Solana Style okay and then the other thing is do you preserve or reduce decentralization so one of the key metrics that we have with writing these sips is a model where your set of spos increases meaning you actually have a larger k despite the fact that you move to continuous execution of scripts much higher throughput of transactions and a completely different structure now mithril is essential for this because as you get here and you talk about that kind of throughput your blockchain is going to grow the amount of that's going on is going to grow considerably so if you want to do a bootstrap you need to have an accelerated bootstrap mechanism so fast bootstrap and better support for like clients so the mithril program will continue all throughout 2023 a ton of things that are going to happen the Hydra program will continue all throughout 2023 tons of things need to happen we need to keep pushing forward on that because these are components if you think about how they get structured together into how one would approach building out input endorsers now input endorsers aren't the only part of the story their design we consider to be so the sips we consider to be the the final milestone of Basho because if you think about all the stuff up here is effectively enough for cardano to be competitive against all the top 15 cryptocurrencies including Solana that are in circulation when we look at what's happening with input endorsers this Evolution combined with the other open source programs basically is something that allow cardano to scale over the next five to ten years of use and it's closely related also to other innovations that we have like Chronos for ntp replacement things like Fast finality so The Ledger Redux program and then ultimately what we're planning with or Boris Omega this this entire setup so basically what we're going to do in November is announce the sips and have a big Community conversation about their implementation because they touch everything and it goes beyond the scope of what IO can offer because it has to be a community discussion meaning that Voltaire is required in order for input endorsers final design to be ratified because it's going to change the logic for spos we would argue for the better it's also potentially going to introduce new reward mechanisms so the reward functions need to be adjusted as well it's going to change how staking Works a little bit these types of things and there needs to be some competition in the ideas like for example DC spark has expressed some interest in examining the sips for it so there needs to be a long Community review because it's such a huge component and the question is how do you avoid Network stalls how do you ensure further decentralization of the system how does this system be more fair to spos and do good things for them and then also how do we reuse and leverage existing Technologies like mithril for example and Hydra for example so this is all a big big part of the Basho agenda the last thing is side chains and the idea here is that you use oroborus to bootstrap a side chain foreign so that side chain will take from oroborus some set K of spos and that becomes the quorum with a bft protocol of the side chain and there's also a bridge to move value back and forth oh by the way this also is talking about Babel fees and tiered pricing as well so there's a huge reset of the pricing and fee model as well with input endorsers so anyway so there's a bridge component and then there is a bft component and so you have your main chain and then you'll have this ecosystem of side chains so why well because side chains are going to have their own business logic their own application domain and that's a lot of TX volume if that TX volume is being held by a side chain then all of these things are being done in parallel meaning that the processing of those side chains doesn't reconcile on the main chain so it doesn't consume the resources of the main chain so it's basically like super Hydra hydra's dap specific side chains or ecosystem specific and ideally the side Chains would be doing things like different computational models like evm or perhaps privacy maybe supply chain it could be loyalty points or nfts and you could build these things bespoke specifically for that logic but users and value and other things through a canonical bridge will be able to move back and forth and represent it as a cardano native asset on the main chain and then take advantage of Ledger support treasure support exchange listing dexes all of these types of things it uses the main chain for security liquidity and other such things and infrastructure like wallets these types of things now what's really cool about side chains and cardano is if you hold Ada okay right now you get staking rewards with Ada okay well these spos that are selected to be part of a subset well they're doing work and you're partnered when you're staking with people when you delegate when you set up a stake pool you compensate it if you do work so if they're maintaining the state of these systems they actually pay a reward for that work that's being done well that reward would then go to the SPO via their fees but also the people who have delegated so one side chains are implemented in the future you're staking Rewards will go from just Ada to all the coins that those spos have contributed to and made blocks in so you'll actually get a flow of these so there does need to be a little bit more work on how to set that up and that's contingent on how the bft protocol is set up and also there's a consent component because so the staking certificate has to basically be updated a little bit so some sips are going to need to be written here and we'll do that through the members based organization process when it's available but side chains are definitely a great component of Basho because they're a special case of the Hydra where there's so much stuff going on and it's an ecosystem play that you want people to build on top of so a lot of TX volumes here so you can see how complicated Basho is it's very interconnected you you have some basic stuff like diffusion pipelining and there's always optimizations there there's also utxo HD that's a component there's obviously all kinds of things that can be done to plutus to make scripts more efficient we've already seen that with V2 a 10x in one year is pretty good there's mithril and Hydra which are interconnected and that they go from many to one and then you have input endorsers which is kind of the big Daddy and the input endorser's side is basically saying hey we're going to go to this parallel architecture that we've been thinking about for a long time the parallel chains paper and other papers that we've written and we have these beautiful ideas and we finally figured out how to string them all together and it's great because it's going to make it more decentralized and it's going to prevent Network stalls we don't have them but when you go to these systems you can create them and ultimately you get your speed without compromising decentralization and you get real-time continuous computation so you go from 0.