찰스 호스킨슨 - Cardano Update 영어 스크립트 (25 May 2018)

금일 트위터를 통해 찰스 호스킨슨이 카르다노 업데이트에 대한 내용을 알려주었습니다.
유튜브의 영어 자막을 보면서 영어 스크립트 내용을 아래와 같이 정리하였으니 참고해 보시기 바랍니다.

유튜브보면서 영어로만 정리하는 것만도 몇 시간 걸려서 한국어로 번역할 엄두가 나지 않네요 ^^;
영어 스크립트 내용 중 단어 및 문법에 오류가 있을 수 있는 점은 감안해 주시기 바랍니다

[Charles Hoskinson on Twitter - Cardano Update 25 May 2018]

Hi this is Charles Hoskinson, chief executive officer of input output, I wanted to talk about few things today in particular I wanted to talk about our release process and the upcoming Cardano 1.2
Release and then a little bit information about stake pool and delegation as well things like the Cardano CL tesnet and so forth.
Broadcasing live from Colorado. I just came back from my vacation from undisclosed location. Yes I went fishing and I certainly did catch a few things which was quit fun and I finally had time for a shave and haircut.

Anyway next week we hope to get out the cardano 1.2 to release. There’s been a bit of confusion about dates and times and how these things work and so I wanted to elaborate a little bit more about how our release process works.

So in April when we finished writing all of the code for Cardano 1.2, this includes paper wallets and activating the version one APIs alongside a litany a little corrective things here and there from performance improvements to security enhancements to network improvements and so forth. Once that was done we turned it over to a part of the company called release manager yeah led by girl named Tatiana and the release manager works with our QA director Anatoly as well two external firms, one is called Allied Testing and the other one called Cubic and these firms basically run that release that we’re planning through large collection of tests. these tests can substitute things like smoke tests, feature tests as well as regression testing. Now basically ideas is that it’s like a funnel and as you go through it, you get more and more assurance that the release is going to work for as many people as possible and not introduce new bugs or new problems because we change quite a bit of code, so we went through seven release candidates with the 1.2 release and we’re rapidly converging to the final form and our hope is sometime next week to be able to release the 1.2 version of Cardano.

Well it’s the update to Daedalus. This includes paper wallets, the V1 APIs, and litany of other enhancement so release notes are coming and release in imminent.

So thank you for your patience and thank you for waiting. This is quite long release pipeline. A lot of things still need to be automated and we need to make some improvements in that pipeline. So our hope is every time we do it, the release pipeline from when we stopped active development to when we actually release the update to end-users will shrink overtime and will get better and better and every time we do a release, we do a post mortem on it and try to find ways to improve and automate that process.

Okay so that’s coming very soon. Thank you for your patience. You’re gonna like paper wallet a lot. I do and I think it’s a great feature and I think there are a lot of other great things underneath the hood. It probably won’t fix everything for every single user. In fact we’ve made a lot of improvements to our QA pipeline and our testing and bug reporting pipeline and we’ve received I think almost 8,000 support tickets since January and we’ve been working real hard to filter them, categorize them and be able to get a better sense of where people are having trouble with the product and we’re getting a greater mastery of that and over time I think it’ll lead to considerable usability improvement. Unfortunately software is software so this things do take time know that we’re working real hard and definitely have made huge amount of progress relative to where we were just a few months ago and we’re getting to a point where we can actually better measure things and better estimate things and we know exactly where need to add more resources.

Okay so the other thing is the release of the Cardano CL Test net. This is the very first test net we’re going to release. It’s the K-EVM test net, so this is the test net that will ensure backwards compatibility with Ethereum. So basically what happened was that last year we worked with a firm called Runtime Verification and RV wrote the first formal semantics of the Ethereum virtual machine, so they took the yellow paper and they took the reference C++ client and methodically they were able to engineer a collection of formal semantics and if you’re interested in reading the semantics and learning more about K, you can google runtime verification github and you can actually take a look at the K-EVM semantics on their github repository as well as readme file there. The advantage of expressing formal semantics using the K framework is that you can then actually generate the entire virtual machine from the semantics itself. This is the first time this has ever been done in the cryptocurrency space and actually it’s seldom done in the in the regular development world for a variety of reasons usually stemming from the difficulty of the task. So we decided given that we feel a smart contracts deserve to run in a high assurance setting that it would be important to purse this avenue. So this is the first time it’s ever been done in the cryptocurrency space and come next week I believe March 28th we’re going to release the very first test net so this will be basically running backward compatible with Ethereum. So it’ll be command line client, command-line client will be able to deploy smart contracts to this test net in tandem with the release of the test net, will be a bunch of materials for people and documentation for people and feedback loops, so that you can go ahead and write your solidity code, deploy it and take and look and see how that performs and what the networks like. It’s not really a good test for performance nor is good test for a full featured test net although it’s pretty powerful, it’s based on our Mantis technology which is quite mature , it’s more of a test to see how the K framework works and give us a nice foundation so that we can iteratively improve K. we’re working on a with our RV a lot of improvements like K2 LLVM back end things like semantic space compilation and so forth as well as just general improvements that the K framework itself which overtime should dramatically improve performance and usability for people who wishing to express their own languages and do cool and interesting things.

The more interesting test net which is coming on in the coming months hopefully July but may slip a little is the IELE (Yella) test net which is our in-house virtual machine that we had developed specifically for Cardano based on LOVM with a lot of really special things that have been baked in specifically for smart contracts given the year of knowledge that RV has learned from the smart contract space. Like the K-EVM, IELE will also be generated from the forma semantics which is another first in class and our hope is that people can play with the IELE language and we’ll have a solidity to IELE compiler so you can write a solidity program and either deploy to K-EVM or you can deploy it to IELE. So this is the first generation of our test nets and over time we’ll have more and cool and sexy things to talk about very specialized things specifically for Cardano, so look out for that but these things are coming. so we should have an update for 1.2 and we should have that test net coming within the next week and that’s gonna be a lot of fun.

Okay so last thing I’ll mention is the incentives and delegation. So this has been an incredibly difficult research dream and it’s one that we took very seriously so we put our best scientists on it from Elias Koutsoupias to large branches to Aggelos and others and the end result of this has been production of two papers, one paper specifically for incentive scheme for Cardano and one paper specifically for how delegation works with Cardano including all the business requirement like cold staking and dealing with what happens when exchanges get your stake so creating segregated addresses and so forth. So we plan on submitting the incentive paper to a conference in July and we’ll probably it on e-print around that period of time and there’s going to be a collection of three blog posts along with some videos released explaining what is the monetary policy of Cardano meaning what should be the amount of ADA generated every year up to the cap of 45 billion as well as dynamics how to decentralize the network on A B and things like the amount of stake pools and incentives running stake pools and so forth. It’s a very complicated topic so we’re gonna approach it very systematically and release a collection of blog posts, three of them each one addressing something in kind so things like how do we ensure civil resistance, what is the optimal level of decentralization for the system and finally what will the final monetary policy before the launch. So we’re rapidly converging to final parameterization and look for that coming two months or so and we’ll basically announce that via blog posts and some video contents as well. Shortly after that’s been announced, start working our way towards the Shelly test net which thanks to the registration of stake pools will be quite saturated, I thinks we’ve gotten over 2,000 applications for stake pools from people who are interested in helping us beta test and run their own pools, we have to kind of winnow that down so we have a reasonable set of people to work with but throughout the summer and the fall especially the backend of the fall, we’ll be definitely working very closely with those people and getting everything completely smoothed out and cleaned up as we converged to releasing Shelly.

so that’s a bit about incentives and going on to delegation.

We actually have a final scheme for a delegation and with at the point of specification. So if you’re familiar with our wallet backend formal specification that we wrote which is now available at Cardano docs, we’re approaching the problem in a similar aspect so we have a white paper and Philip Kante is now working at translating that white paper into a formal specification which in turn we’re going to be able to use to implement the delegation scheme as this involves lots of cryptography and this involves some discussion about the address structure and potentially changing some of the way our crypto works particularly HD wallets, we have to approach it in a very systematic and rigorous way and we’ve been doing that for the past few months. like incentives paper, eventually we’ll release the delegation paper to e-print and we may submit that to a conference. It’s more interesting from an engineering view point than it is from a novel academic view point so it’s still debatable whether we’ll submit that part for peer review as it’s simply an engineering artifact although we will go through some form of security on it as we get closer to the final system.

So delegation, incentives have been definitely a bear of a problem and they have consumed a huge amount of mental energy over last five months, we certainly learned a lot and we’ve done a lot of simulation and a lot of experiments and we really look forward to turning that over eventually to the community and getting a lot of feedback from you guys and hopefully being able to roll it all into Shelly as soon as possible and get the network completely decentralized.

So one question we do get asked a lot is question about the smart contract side of when will Shelly and Goguen be united. Well there are lot of challenges in that. We also have to finalize our side chain scheme to be able to send assets from Cardano SL to Cardano CL, we have converts to a sidechain solution that we feel pretty comfortable with and we’ve gotten to the point where we’ve begun prototyping that solution and measuring things for example like proof size and transaction latency so how long does it take for a transaction from SL to CL to settle and what are some of the challenges of maintaining the state of multiple blockchains concurrently with Ouroboros, so these are ongoing and interesting questions and they’re ones throughout the summer while we’ll converge the solution for similar to how we’ve converged the solution for both delegation and incentives that among a collection of other small problems are required to be resolved prior to unifying Gogeun and Shelly but the good news is that the harder problem was the result of consensus issues and incentive issues so I don’t imagine that this research stream in going to take an enormous amount of time but its something that will keep you guys updated on throughout the summer and throughout the fall as we get closer to that.

Plutus is our other really interesting language that we’ve been developing and it’s our in-house smart contract language. We believe that smart contracts are intrinsically high assurance artifact and so that means that we have to think very carefully about how do you write good code that doesn’t fail so you don’t have things like the parody hack or the DAO hack and the loss of hundreds of millions of dollars, so that’s everything from how do you work with people who aren’t necessarily technical but they are domain experts like CPAs or lawyers who have lots of business requirement in an intent behind a contractual relationship between parties who are able to model things like the parties, the flows the capital, the events and the conditions and times and so forth one thing should happen and how do you then translate that into actual code and verify that the intent matches the code that has been deployed. This is called semantic gap so we have a partial solution in that we have developed a financial DSL called Marlow. This is the artifact from two years of research that we’ve done in collaboration with professor from Kent University and it’s based on research that Microsoft research did back in the turn of the century from the late 90’s early 2000s with Simon Peyton Jones from a paper called adventures in financial engineering. So Marlow is rapidly converging to a submission and sometime during summer we will finalize the design of Marlow and submit that to APL conference of reasonable prominence and if you’re interested actually, we have a github repository in the IOHK github that has an early version of Marlow within it but basically Marlow will be a financial DSL that people can use the model relationships and these are machine understandable and can be executed in Plutus. our primary smart contract programming language ever that we’d like people to write smart contract in which also is based on the UTXO model instead of the accounts based model is called Plutus and we’ve gone through 12 read signs of language of that since we started and every time we do, we learn little bit more and we try to simplify the language as we go along. We’ve converged to a point where we actually think we have a 1.0 design and throughout this year, we will continue to enhance Plutus and extend our UTXO model to be able to allow Plutus to operate properly so Phil Wadler is the leader of the project and sometime later this year, Phil will start releasing some dedicated content including video content explaining how Plutus works and why we choose the design decision we chose and eventually will allow Plutus to be interoperable with IELE and other frameworks so we’ll build some compilers for that and eventually it’ll run native hopefully with an SL, Plutus actually has a nice role up in that Marlow can be embedded with the Plutus and Plutus can be embedded within Haskell so we have this nice two phase model where some of the code can run on the client side and some of the code can run on the blockchain and hopefully we can even have a meaningful discussion about what formal verification looks like given Plutus.

So look for some updates on that throughout the summer that it’s really exciting work stream and it’s something we think is going to add a lot of value to the space, so we kind of have a nice blend of principles and pragmatism. Yeah if you want legacy interoperability, you can use the K-EVM and the languages you’ve come to know and love like solidity and future languages like Viper will be backward compatible with that, if want to see the best and greatest on the VM side, IELE is really an amazing piece of engineering and there should be a paper submitted sometime in July formalizing IELE and that will probably go up on e-print, so you’ll you can kind of see what we’ve been able to come up with partnership with RV and it’s really explore the power of K framework and if you’re very principled and you want to write high assurance code, you’ll have the Plutus option, so as these things start going from the lab into implementation, eventually they’ll be pulled into the Goguen project and eventually we’ll be able to pull Goguen into Shelly via side chains and the whole system will come together and at that point we should be completely decentralized, really wonderful system with smart contract capability that’s on par and better than what an Ethereum must offer, faster safer and cheaper to operate.

So looking beyond that there are really two final work stream that we are going to be investing and have already begun investing a lot of resources into. One of them is the governance and sustainability side of Cardano, so you may have noticed recently that we just published our treasury paper to e-print. It’s major accomplishment that comes from over a year of research that we’ve done and we’ve already gotten expressed interest from ZenCash and are in discussions with other cryptocurrencies to implement or assist in the implementation of our treasury scheme or a light version of treasury scheme for their own use, so our hope is that this treasury system can be used in many cryptocurrencies as it does add a lot of self sustainability into the ecosystem. So if you’re really curious on how that system works, we have a video from Bingsheng the project leader there on Yutube and IOHK Yutube channel and we also have published a paper to e-print although the paper is very technical, it does capture most of the reasons why we proceed that, so that research line needs another four papers we think, and we will be pursuing those over the next year and at the end of that research stream, we should be able to begin implementation of that treasury system into Cardano and at a very deep level.

The other side is scalability research and we’ve begun work on Ouroboros Hydra and we should now that Ouroboros Genesis is done and it just need a little bit of polishing begin that research stream and that’s where we’re going to get true sharding and true scalability in the system,

so the other side of this is the network side of things and we have really started getting into the meat and potatoes of RINA and we’ve been thinking deeply about what’s the best way of pursuing RINA, we estimate that RINA the way that we’d like to construct it contains about six hundred man months of engineering time, it’s equivalent to basically building a second internet in terms of complexity of engineering , so we’re trying to parse that out and how large the team needs to be, how to runs some things in parallel, which things require researching, which things can we borrow from the seven years of research that’s occurred in the RINA community towards actual production implementations. So throughout the summer and throughout the fall, we’ll be making some specialized announcements about what we’re planning on dealing with RINA and when RINA should come into play but that’s one of the last piecies that we anticipate being put into Cardano.

So Ouroboros Hydra, RINA and Trasury are kind of the next major components of our system, the way we constructed our treasury mechanics is that we’ve kind of abstracted voting to a point where they can be reused for forking the protocol as well, so at just different parameterization and different threshold for decision making, so whenever we end up implementing the treasury into the system in tandem will also implement some form of a forking protocol, so that Cardano improvement proposal can be voted on, so that’s a 2019 topic and as we get closer to 2019, information will continue to get richer and we will publish more paper, but I think the longer the short is that we’ve really gotten a good handle on some of the most difficult problems in space and we’ve been able to put together a pretty tremendous team of people who now fully get how to build the best cryptocurrency and we’ve seen just tremendous returns in our engineering philosophy, for example with the formal wallet specification, we just didn’t stop at writing a specification, Polina, one of our mathematicians is now actually writing a calcus verification of it and that’ll be done very soon, so we are starting to see a lot of people who generally don’t work together like PL experts, cryptographers, game theorists, distributed systems experts, mathematicians who normally live in their own silos all coming together towards a common goal and understanding the common collection of problems that we wish to solve and we’re actually able to translate than research fairly quickly into code, there’s been a lot of hiccups along the way, you know Daedalus has been certainly a very challenging and there were some mistakes in the beginning with how we rolled out and how we design things that we certainly learned it and cause usability issues, so we apology for that but it looks like there’s a light at the end of the tunnel.

Another thing is we’re working on alternative clients we’ve been exploring a browser based client for Cardano and we’re going to see if that’s feasible, we’ve already done three proofs of concepts and we have another presentation internally next week and our hope is to see if that can bear fruit, we’ve also been in discussions with people who are constructing mobile clients and discussions with how do we build a lighter version of Daedalus, some of the challenges are that we have kind of a redesign on the way our wallet works and we have to wait for that redesign to be finished in terms of implementation into design prior to moving on to things like ledger and lighter weight wallets and mobile wallet and so forth.

The good news is that we’re nearly done with that process, the bad news is we kind of have to wait for that process to finish before we can move on to reasonable experiences for mobile wallets, and this includes major improvements and things like for example much faster restoration, it shouldn’t take hours to restore a wallet, should take seconds to minutes and there’s a variety of reasons why it currently takes hours, but our hope is to convert that seconds to minutes, similarly we need to redo things like the data layer of the system, so we can get much more efficient storage, much more efficient download of blockchain and so forth and these things will come and we know how to solve them, it’s just matter of engineering way in a writing code to get through it.

So anyway thank you for your time, thank you for listening, we’ve been working real hard on everything and it’s been hell of a last six months and we’ve learned a lot and we really appreciate the patience and I look forward to releasing Cardano 1.2, I think you guys are really gonna like that, I think you kind of like 1.3 and 1.4 and we’re just going to keep chipping away at things and till we get it done, thank you for being part of the ecosystem, thanks for the faith and the patience, and I’ll talk to you guys soon. [End]

5 Likes

작업 고생하셨습니다 잘 읽겠습니다!

1 Like

카페에다가 인용 좀 할게요 IamMintingYou 님~!

1 Like