Post-Roadmap Comments and Some Reddit Questions

[Music] all right good evening everybody this is Charles Hoskinson broadcasting live from warm sunny Colorado it's currently one o'clock in the morning here but unfortunately this was the only time I really had to make a video it's been a long busy day and the last two weeks have been pretty long and busy but that's no excuse not to have a nice conversation about the recently released Road map and there's lots of questions floating about comments and other such things and it's morning and afternoon in Asia so I figured why not forego a little bit more sleep and let you guys know a few things okay so we spent about a month and a half in deep and detailed discussions about what would be the best scaffolding and skeleton for the roadmap and we came to the decision to release what we just recently released and as you can see it's broken down into different sections the firing section the Shelley section the bacio section the Gogan section the basho section and the Voltaire section and each of them covers a different scrolling of the project briefly the purpose of Byron was basically to release something to build a community around it to learn how to be on exchanges to learn how to write API is that the cryptocurrency space would like to use to work with third-party integrators to build all the development life cycles and also threaten or space that we would then be able to at some point come back to and say okay well what didn't we do so optimally and for the most part Byron as success in that respect an overwhelming success when you compare it to a normal cryptocurrency I didn't live up to my expectations in terms of how quickly we were able to get software out or how quickly we were able to move through the Byron era but of all the e resisting to be the slowest because that's the one where the least and that's the one where everybody's doing something for the first time so we exited the buyer an era and we're now on the Shelley era and things are moving pretty quickly now the self node test not is being prepped and some people have started already independently vest vetting it and playing around with it because we did a release on the github repo just a few days ago and now we're prepping some documentation on the website and on a lot of other things to get people along in addition to that about 150 people have filled out the form that we sent out and give us an ample amount of information about what they intend on doing their technical background their geographic location and also their relationship to the ADA ecosystem as well as the broader cryptocurrency ecosystem the state pool task force channel has well more than 1,500 people in it and our hope is that that will continue to grow so the opening up of the Shelly world to the general public is as well underway and we're very happy about that moving beyond that hackathons that are underway and we're just about to hold one in Israel I'll be traveling to Israel the next few weeks and we're going to be holding another hackathon in Argentina we're planning probably towards the end of July and then we'll be holding a fairly large hackathon hopefully up in Wyoming in September every time we do this this exposes both Plutus and marlow to the general public we've already gotten an enormous amount of feedback a lot of advice about the types of things we need to do to make what we've written a bit more developer-friendly so Gobin progress is looking quite great and we're going from an RD product management cycle to now more of a actual consumer product an actual product useful for developers so that's nice to see so anyway the roadmap that we released it's it's written in a way that allows us to add a lot more detail or links to a lot more detail in a very systematic manner so over time all of those little parts you see about features and narratives and other things will eventually link to videos a link to blog posts it'll link to specialized content so our content teams our comms teams our marketing teams can continuously go back to that same very simple interface and just keep adding and adding and adding in addition to that it meant very well to continuous updates so I'm going on the card no effect with David tomorrow but the recording probably won't be uploaded general public until Monday or Tuesday but on that show we're going to cover some of the plans that we we've come up with about how we're going to be a bit more proactive and giving weekly and monthly updates to the general community and also doing more interactive updates so for example I do a MAS but it's very important that people involved product management also do a mas and hopefully developers start participating in a MAS as well so as a consequence I'd like for David on a monthly basis hopefully monthly maybe every six weeks to do an AMA with you guys and give you some ample timing on the schedule so you can ask questions in advance I tend to do surprising amazed because they're a bit more fun for me anyway so the roadmap will become kind of an archive of these events and it's just a good place to keep going to time it again and I'd highly encourage people to keep going there and we'll be adding more content on a pretty regular basis we put a lot of golden related information and that will be filled out in the next four to six weeks and then we're gonna move on to Basho and Voltaire I can continue to fill out information there so I'm pretty happy with the redesign I think it's a really good big first step in the right direction and it's a much more mature step and a realistic step given the nature of the project see when we first came in and we looked at this we tried to look at it from a very traditional project management viewpoint so we had Gantt charts and all these percentages and all these things that normally do if you're writing a well-defined well scoped well understood piece of software but that's not really where we ever went we were building crypto currencies based on research and all these protocols have never seen the light of day so we really needed a more agile approach and it took quite a bit of time for the organization to kind of move in that direction but one of the problems was that two of the development arms one being the formal methods group the other being a research done immense so well - an agile process it said a little bit more water falling so it kind of took us some time to figure out well how do we keep our principles but move was a degree of agility and the old road map reflected the old way of thinking the new road map kind of reflects the new way of thinking and what's really nice is we're starting to get down to what why Cardinal was talking about back in 2017 where we'd like to have these small competing agile teams of three to seven people doing a very focused thing in that team by aggregate will kind of build up to emergently a whole ecosystem so the biggest epitome of this is is the haskell wallet back-end we build that we did we started a complete redesign of that team last year in November December timeframe and now literally every single week we get a wonderful release of something and they've just been moving in blinding speeds now that they've been moving it's blending speeds with great processes and also great testing in fact Cardno is adopting beck 32 as a as an address structure for Shelly and while we were integrating the Haskell back 32 library we found not only bugs but we also found some design defects in the specification itself for beck 32 and this is a consequence of our property based testing that we were doing so the fact that we're able to follow an extreme programming model in that team it's modular software so it can be plugged into both the rust and the haskell client or even other software stacks potentially and week by week we're able to move very quickly but yeah we're still using formal methods is a actually a really remarkable achievement and I'm super proud of that and not to be outdone the other teams are starting to now converge to that we're starting to get a lot more demos and it's it's nice to see this so the road map is kind of a reflection of this new process this new organizational way of doing things and we're quite a bit more agile what it means is you're going to see less that these big dumps where massive things get released and then it takes a long time for people to parse it and you're gonna see a lot more iterative stuff so every day every week every month things are just gonna start organic coming out now there happened some questions about prioritization zorda prioritizations and also kind of the architectural direction of Cardno so one of the things that's come up is a discussion of well what are we going to do with KVM yella and the Kay framework in general and so this was one of those high-risk high-return pieces of research we had several high-risk high-return pieces of research in the Cardinal agenda so one was or Boris so the whole concept of can you do proof of stake that was very high risk it wasn't completely decided in the ecosystem whether it was even theoretically possible and if you asked the proof-of-work people on the Bitcoin side they actually say even today is still not so back in 2015 this was a big open question and there had been some projects that had done things but no one hadn't really done think things was rigor or to a massive scale now obviously certain communities will disagree but the fact that none of the top 15 top 20 crypto currencies in 2015 were reliably running with a proof of state system outside of a few at Barratt examples that would come in and come out really tilt you the state of things so we decided to invest an enormous amount of money time and effort into trying to do that high-risk high-return research and that part of the project was overwhelmingly successful not only has or Boris borne fruit we now are ahead of everyone in terms of a stable beautiful design for proof of stake and we're in the fine-tuning and fine details phase where as a lot of people are still kind of exploring their design space and trying to figure out optimal ways of doing things so mission accomplished I'd say with that research thread and there's a degree of inevitability and determinism that there where the things we want to do will get done so great there the second piece of high-risk high-return research was this concept of sidechains and interoperability through structures like NEPA piles and proof of stake side chains and this was another one of those high-risk high-return ideas because you could end up spending years just thinking about various ways this and it does take a lot of deep thought and it's it's a very broad moving target because basically what you wish to accomplish with this is this concept of you've sent me a transaction from either a local or foreign chain so either the cryptocurrency itself or from another cryptocurrency I don't have the blockchain I just have some small amount of information and yet somehow I'd still like to have the same security as a person who possesses the whole blockchain of that asset and I can verify that the asset has not been double spent and that the asset exists now if you can solve this problem you get awesome like clients you really help the scalability side of things and furthermore you lot of systems to talk and work with each other in a much more natural way so this has been one of the major pillars of the project and something we've explored a lot and it's an again high-risk high-return research and thanks to the anisa says engrosses hard work and aguilas hard work and a lot of our partners who we've done research with and have deep deep and detailed discussions with we also feel that that direction has a degree of inevitability to it just the proof of stake research has a degree of inevitability so I'm quite happy with that the third high-risk high-return pillar was this whole concept of can we build an idea of a programming model for crypto currencies is something that makes a lot of sense not just for smart contracts but also for how smart contracts would interact with the broader world you see it's not just about can I write some solidity code and get it to work on aetherium that is the heart of AD app it's also the fact that that application is going to be talking to things that live outside of a blockchain there is something that lives on your computer on a trusted server on a Federation of servers or perhaps a confederation of blockchains all talking to each other so there's this concept of on chain and off chain code that you somehow have to reconcile and you have to figure out how do I build applications where I have determinism and predictability where I need it and I also actually wrote is not significant it's it's fairly small and that's a huge problem and when you don't get it right you have things the dow hack and when you don't get it right you have things the parity hack furthermore it also yields issues of performance and scalability in these types of things so ok how are you actually going to build a system worried not only can you sort all these things out and you can get symmetrical correctness but also the system is reasonably performant for all users not just a particular group or for a particular window of time but when you go from a thousand to ten thousand to a hundred thousand to a million and so forth that things are things are actually moving in the right direction everybody is basically getting reasonable performance now we implicitly know this for example we have Netflix and we have Google and Amazon and Facebook and these giant services that despite the fact that they have millions to billions of users and despite the fact that these users are really pushing these systems to their limits they're being very intensive they're uploading videos they're streaming content they're doing tens of thousands of searches all the time yet somehow you get a fairly good experience so the key is can we take those ideas and those lessons and put them into a model that somehow is still decentralized and preserves your trust and your privacy and these types of things so is it over is a huge challenge so the bedrock of all of this though is your computational model and in your programming languages that you write your smart contracts in so we took a very pragmatic approach you can either be on the functional side of the world or you can be on the imperative side of the world ok so you can look at languages like Java and C++ and JavaScript or Python or you look at languages like Lisp and Haskell and oh camel and closure and so forth ok so what we said is look we have enough money let's explore both of these paths and let's try to innovate within those spaces with an eye on first formal correctness and second having a nice path to the broad scale interoperability okay because it and then add a developers are gonna want to write language and the languages that they're comfortable and familiar with within their domain so we'd like them to be able to reuse tooling we'd like them to be able to reuse the things they care about so one team led by Phil Wadler and later minimal chunk of already alongside Phil Wadler and many others explored the functional side of the world and they're among the most qualified people in the world to explore that because they created functional programming languages like Haskell and they've been in this space for 130 years and collectively the team has more than a hundred years of experience so that team went off and said let's use ideas like template Haskell let's implement dsls to do the on chain thing so in case of marlo it's a turing incomplete DSL for modeling financial contracts and in the case of Plutus it's a turing complete fully powerful programming language but a lot lighter than Haskell and then let's allow your off chain code 2 to be a full older programming language with lots of history and legacy and good tools around it and that the idea is that this two layer model is something that runs on chain something runs off chain the odd chain being a DSL the off chain being a general-purpose programming language would be a nice template that we could then play around with within that domain and we get a lot of power we'd be able to write formal specifications for contracts designed by contract would be somewhat easy to approach and we'd be able to use property based testing and model checking and all these other tools that you really need to use or should be using when you're writing high shares code and smart contracts are probably the definition of code that ought to be high assurance if you're talking about code that you're gonna run a lot in a very hostile setting that's concise with a well-defined business logic behind it that's code you should verify so that side of the project was overwhelmingly successful it took about two years for us to really get to a point where we knew the exact threat and path of how we wanted to go through it but now we see consistent progress week by week month by month and not only are we innovating there we're actually now improving the state of affairs in the functional programming world not just for crypto currencies but in general for example I which K is actively contributing developers to the haskell - web assembly effort so you can take Haskell code and pluto's code and you can compile that to run in web assembly we're also actively contributing to the GHC Jas program which means you can take Haskell code and convert it to JavaScript so we've that these are vital efforts so that you can deploy your off chain code and node and the browser and other places and that infrastructure we're building can work in web applications and browser-based applications without having to be translated or reported it just works through a compilation chain you can experience with rust and other languages like that so that work is coming along quite well and we're validating that that model works through hackathons so we're taking what we have and we're giving it to developers and getting feedback from developers about where we go next and we're seeing great acceleration there okay second the other side the imperative side it's not too hard to innovate there because it's a very broad space and there's a lot to do but we didn't want to do something just like for example come up with a new virtual machine and say okay look we have a new virtual machine or take a subset of an existing virtual machine it's a Java Virtual Machine as a Yan has done or look at web assembly and just somehow translate it with a gas model and say okay well now we have something like what theorems trying to do with E was M okay so in that particular case we wanted to explore something a bit more general and broader and this is why we were really in love with the K framework and ideas that Gregorio shoes team had come up with it runtime verification so again being high-risk high-return research we took a step back and we said let's go roll the dice and see how far along we can take this research so we put a few million dollars of capital down to enhancing K so rewriting K in Haskell exploring techniques and concepts like semantics based compilation as well as trying to improve the performance of the machine generated code and there was some great success there but it wasn't far enough along for us as a project to really feel comfortable saying within the 2020 timeframe that this collection of technology would be viable for consumerization and would definitely deliver to market a great experience for developers so we had to take a step back and say that the strategy has to be changed a bit so the concepts were work rate so this concept of the thing any programming language you can write the case semantics of that programming language and then suddenly through the magic of semantic space compilation all programs written into that language could just somehow be compiled to run on the yellow virtual machine that would just be magical because it means to get support for your programming language all I'd have to do is just simply write the semantics of that language once and store them let's say on the blockchain and then suddenly all those programs can work in my system so it's a game where over time I can support hundreds and eventually thousands of programming languages furthermore I never have to rewrite my compilers when I upgrade the core virtual machine I just have to upgrade the case semantics for that virtual machine and somehow everything just takes care of itself so this is still ongoing an active research in academia at run time verification and there are other parties who have witnessed the power of K quad stamp for example is using K defined formally formally specified things the ERC 20 specification is 721 specification for oddity and furthermore elrond which is a project I believe based in romania has recently written a go k backend so that they can explore the cave virtual machine for EVM and or yella and so forth so we're on good terms with these projects and we talked to them and while we deprioritize that stream we're going to be following that along but to be clear interoperability with the KVM interoperability with yella interoperability with solidity wasum this is not a hard thing to do and we've explored it with the card on our CL test net that we ran last year we've written mantis and we know how to build an assyrian virtual machine from scratch we've done it twice it's it would not be difficult for us to pull that into the stack for the sake of interoperability and if this is a desire that the community has it's not going to really take too much effort money or time so we decided to focus on where we're getting the biggest bang for our buck in terms of innovation so currently on the functional side and bring something new to the space and bring a whole new group of developers into the space so we can get great experiences and great applications out of that and then the other side the KVL and yellow side not going away they just basically have been deprioritized and we're going to examine them at a later date and the whole space is moving rather rapidly on the imperative side there's a lot of options we have and frankly we don't actually have to write new code we can simply take some best practice off the market and bring it in for the sake of interoperability if this is a high priority so that's the nature of these projects is that sometimes when you're doing high-risk high-return research some of the research works really well for you for example the case of side chains and proof of stake other research it takes two years plus to get to a point where you have a clear thread of where to go which is what happened with Plutus then it works really well with you and other research is just so powerful and broad that some cases you don't quite get where you want to go and you realize that you're probably not going to hit your trajectory so that's where we're at there now in terms of Rena that's another example of high-risk high-return research that we've been doing and what we did is we down scope Trina to something that we felt we can implement within the 2020 timeframe and actually we've warned a lot of grapefruit the network team at i/o hk's worked for over a year now implementing this mini protocol concept and at some point we'll release a nice piece of documentation explaining how this design works but what we've done is isolated a lot of the things that we felt were real big innovations in the arena protocol and try to apply them to the needs of the cryptocurrency space we also very carefully examine the needs of Ora Boris and the particulars that you're going to have to do there to make that performant and work well and we were able to extract those and put those into a bespoke Network stack for Cardinal so so if you're interested in that we have a github repository where that code is being written I believe it's some orb or network or card out on network I can't recall the exact repo but if you look through our repositories it should be fairly straightforward to find it and we've seen a lot of great progress there and a lot of great ideas there so well we didn't get the whole thing in Rena we didn't get to rewrite the whole internet we did get a lot of value out of that exploration and it's nice to be able to pull that into an actual production system and see that in real life because actually no one did that arena has been around since 2008 and no one really implemented Reena into a production system so this is really the first time where anything even it or inspired by ideas from it is being pulled into something that's actually fairly large such insignificant now there are other areas that we have been exploring innovation at for example we've had to think a lot about incentives we've had to think a lot about state pools we've had to think a lot about rational actors in a system Fran this is where we have things the CIP process we need to roll out things like voting things the Treasury system we've written several papers on the Treasury system we think a lot about voting and there are specialized teams within ioj Kay that are doing exactly what we were doing with Plutus where we have a product manager on the R&D side basically getting it to a point where we have a clear execution threat and Ben Chang is that equivalent to Manuel chunk of RT on the on the voting side and on the Treasury side and great progress has been made there as with all things all of these threats regardless of we're talking about a done stuff like proof of stake to wide blue open greenfield Rd it's a long-term effort and you release something but there's always more to do there's always the next step for example the Samsung s10 is a phenomenal product it's a probably the best phone at the time it was released certainly that was powerful 12 gigabytes of RAM a terabyte of storage I see I think six core processor very powerful phone but yet there's going to be an S 11 and s 12 and protocols are the same way so whatever version we create there's always going to be something more to do in something more to push out so then that's really the point of the roadmap is it's kind of a living document where it collects the facts of the past and puts them there so new entrants in the project can see how far we've gone and how much progress has been made and it gives you a increasingly higher resolution over time view point of where the project is going and the types of things that you worry about our less broad and far more specific over time okay so so that's kind of how to use the roadmap and just keep coming back to it and we'll keep pushing up updates pretty happy overall with the quality of the research and and also I just love the fact that code is really starting to cascade and be written in a much faster rate and I really do honestly believe Shelley is coming out this year and I really honestly do believe staking is coming out this year and I really do honestly believe that Gogan is also going to be hit sometime this year there's the velocity we have and the developers we have the understanding we have a these particular problems is quite good and there's not a lot of technological barriers I'd say to prevent this from pushing in that direction unforeseen things can happen budget our developers could be on a plane in the plane and could crash I could die you never know a life is life and events can occur but barring disaster or barring unforeseen events the velocity does suggest that we will be able to deliver a fairly significant portion of Shelly and Gogan this year but even Shelly is never quite done because decentralisation is not an achievable target it's an aspiration no matter how decentralized you are by definition you can always be more decentralized no matter how well you are writing smart contracts or how high quality or languages that are you're always going to have the next version of the language you're always going to have the next library you're always going to have the next tool because problems change scope changes you learn things over time innovations occur competition occurs so the goals of gouken the goals of Shelly will continue on well past 2019 and 2020 and well into the future for either us or future generations of developers to aspire to achieve and that's exactly the point that said what really gets me excited is now that we've rewritten code now that we're starting to see features come in that we really need to have what's taking we are now able to start talking about completely new features for example contingent settlement or embracing the dead standard and bringing identity into the protocol or looking at various services that or Boris can start providing to smart contract developers for example concepts like horse chronos which will provide time stamping services or other things like or porthos which can actually allow a card out and become a cryptographic beacon and so forth so these are the things that really excite me because they're new or it's a new way of doing them and it adds a lot of value to everybody in the entire ecosystem and also it really excites me that we have a lot of high quality ways now for people to build great applications so and that's everything about the roadmap now there have been a few things said about well what is the difference between cardano's CL and Cardinal SL now you may have noticed that we were the first project to really start talking about this multi-layered setup where some things handle just settlement and that are slow and methodical and other things are more ephemeral and they handle concepts like data and computation and so forth so so when you look at those types of things for example the etherium 2.