Casper Community Calls, 24 Nov 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: CasperLabs Community Call

Description

Engineering Status update, Feasibility of the Smart Contract Economy, Mainnet launch timing details.

A

Good morning, good afternoon, good evening, wherever you're coming to us from this is metta parlikar, the cto of casper labs and I'm going to present our weekly engineering and project status, update I'm joined by alex leminov one of our economists on the project and piotr d. uh I just call him po2d, because I can't pronounce his last name so he's just known as p o through d. We have another appeal to the project's pls, so I'm joined by pft who's, our program manager, project manager for the project.

A

So thanks for dialing in uh we'll kick off with our engineering status update and then, after that, we will talk about the feasibility, the smart contract economy, uh more information about progress on mainnet, launch dates and timings. So, on engineering status, we are in the fourth weekly sprint of the 20.12 release cycle. This is thanksgiving week in the united states, so we're actually extending out our cycle for this particular iteration and we're going to do an 8.5 which we don't normally do, but we're going to do an extra sprint.

A

um Just so we're not having to do planning on a holiday. We usually do planning on fridays, so planning will happen next week, we'll be cutting a release uh after you know, probably next week, sometime after you know, after we exit the release cycle, we are doing our security review with trail of bits, so I'll provide a status update on that as well as part of this uh this week, so we are in codename delta, which is our uh final test net phase um with four validators launched.

A

I believe we are up to significantly more peers. Now uh I can take a look here real quickly and see how many peers we have.

A

Let's go up to another pier here. My note actually isn't bonded in here, so we've got a handsome number of peers and I could look at the auction info. But if you go and look at the auction info, I suspect that a good chunk of these are also validators that are bonded in my node isn't running at the moment. So I can't check for a local host, but the network looks good and healthy.

A

um Nice block height there era, id 177 block height, almost 2400, I'm quite certain that the network is moving faster now, yep you'll see just in that time. We've already proposed a block, so the network is moving a lot faster. We did this because we wanted to see how fast it could go. So it's probably one minute walk times.

A

um So we're monitoring that closely want to see how fast the network can go.

A

uh Yeah, so quite a bit more era since then, so beta test net's, not around anymore. The clarity block explorer is now pointing to the new delta test net. So you'll see here it's connected to delta. This is the block height block hash. The explorer looks quite a bit different, quite different because we do not persist the entire blockchain state right. So we, when I say the blockchain state, I mean the consensus state, so you're not going to be able to see all the votes and the attestations that you were able to see before.

A

So those of you that looked at the old scala network, you actually got to see all that information, but that is you know, superfluous information that doesn't really it's very expensive to keep that protocol stated around. So we get rid of the protocol state. We put it away very aggressively in the rust node for efficiency purposes, so current focus and highway is supporting the security audit. We are in our security audit with trail of bits.

A

Initial findings were thin, which is very good news. um They picked on some a piece that was unimplemented or a to do so. We knew we, we know what we have to do there. um So I'm I'm quite optimistic that the audit will, you know, not bear a lot of fruit, so they're working hard to give us. You know, give us some good findings for our money, so we should know we should exit the audit by monday of next week and we should have our final set of findings by that time.

A

The lsc was one of the pieces that they found. This is the fork bomb uh and spam protection uh uh feature.

A

So that's that's a piece that we're building right now and then we also know that we need to disconnect from nodes sending invalid messages on the node rust piece, those of you that are familiar with the delta test net. You know that we ran into some issues when we bounced delta 3, because there was significant or delta 2. Sorry because there was significant crosstalk from delta one. It was on the same port, so we had to kill delta two and launch delta three with a different port number.

A

So we need to build something in where um the there's no crosstalk, if the genesis has doesn't match.

A

Even if it's on the same port right, so we can't expect that there's multiple casper networks that they would necessarily run on different ports, so the genesis hash needs to be sufficient to make sure that the networks don't collide.

A

So that's definitely going to be a piece of what's going to go into the lib p2p network component implementation, we're going to we broke out the rust client into its own library. So we can now build a suite of unit tests for it. uh Dan werner is doing that work and we're finishing the rpc endpoint that returns the endpoint documentation. So this endpoint documentation is really important for people that want to build against the rpc directly without using the client or they want to build their own client right.

A

So this is what this client documentation is for and um we're designing a solution for syncing around blocks being finalized twice. So if I'm synchronizing a new node, I may try to go ahead and re-finalize a block. That's already been finalized, we're also going to be uh gossiping finality signatures as well. This is also to provide more information on the um you know, so that if you wanted to know what the fault tolerance threshold is for a given block, you can get that information.

A

That information is pretty expensive to compute, so we're trying to find an efficient way to do it.

A

Sre and test uh we're focusing on ci cd for clarity, we're looking at network tolerance monitoring, as you may have heard, we have two new sres that join the team they are on boarding. I expect that they are going to be focusing on this in the coming weeks as well. We want them to also provide support to the delta test net and work with validators to help get them onboarded on the ecosystem front, we're creating a javascript sdk this to support this is for supporting our custody provider.

A

They want to use the javascript sdk and then clarity to display the list of validators.

A

We want to finalize the gas pricing model. uh We would need to do this to figure out how much we're going to stuff blocks. So how do we make sure that our blocks are at capacity? So we push through the most uh throughput on the network as possible.

B

Yeah and, strictly speaking, the main problems, there is actually not even so much that it's actually dealing with the details of uh consensus before execution, because it turns out that you know it requires some changes to. uh You know the conventional logic of balance, check-ins that one would expect yup. That's exactly right.

A

B

Likely changes to the ways that our payment code works as well, because I think that, for what we're going to be introducing is actually the concept of a of a full back uh purse. So in other.

A

Words, it says.

B

The originator, the originator of a deploy, will ultimately have to be. You know potentially responsible for the paper right. So uh obviously, if everything is set up correctly on their ends, they're not going to be, uh but you know uh you know you know we will require them effectively to you know backstop uh any problems as it might come up once you actually execute the payment code.

A

Yeah yeah, I see what you're saying, even in the case that it's touring complete payment code. We do eventually want contracts to pay for themselves. I think this is something that's super important. I um I continue to push on this. I don't intend on on walking it back, so uh we will allow for contracts to pay themselves. I think it's important enough in the ux that we support it.

A

So if nothing else then the contract author or the primary primary account associated with the contract would have to provide a fallback burst.

B

uh No well, no, I don't think that's. Actually you know if uv, let's introduce a feature where a contract can you know, maintain a purse.

B

You know once you feel that purse you know that person can be on its own visa file of a fallback purse once it's already in the global state right I mean because the problem that I'm describing the situation where you have no idea where the payment is coming from before you start executing the payment code right right. If you already have a contract that can pay for itself, that's it in the global state that has a purse, and you know it just acts like. I think any other account.

A

Okay, good, that's good to know that is those are features that we will probably get done once we've flushed out all the requirements, we're fully focused right now on the custody provider and their requirements and any exchange requirements.

A

We're talking to some very big exchanges, and we want to make sure that those requirements are met. First and foremost, we also have a requirement for defy contracts, we're leaning towards just for those that want a completely immutable contract for security or safety. There will be a way to specify a contract on deployment to make it completely immutable. We don't recommend this, but we understand that there are folks in the d5 space that say well, wait I don't want to.

A

You know, send my funds to somebody and have them create an upgrade that could potentially, you know, steal the puns away. So I want a completely immutable contract, one where that can never be changed. We don't believe that this is how the world works today, but we understand that there are some concerns around it.

A

Personally, I think that these wallets being locked up and no one being able to ever recover them is a terrible thing so uh being able to provide people. The way to manage their on-chain code is is really critical uh on the contract runtime side, so yeah improve error, handling of transfer functions and there's a small bug, uh unreachable interpreter error hit. I wonder who filed that one? That's a good one!

A

That's a good bug! uh This week, a few folks are out so we're out thursday friday. um The company isn't shut down, so the european developers will continue to work with the us. Folks will not be around.

A

um We have new timings for our weekly workshops, thursday and at 7 00 a.m, pacific and 11 p.m, pacific, which is 4 p.m. Japan, time uh you're still holding those this week.

B

uh Yes, I think we do this. Even even the usher is on holiday. I can take over.

A

Okay, fantastic yeah, and just so everyone knows, those weekly workshops are open forum. You do not need a special invitation if you are a smart contract developer and you want help or the specific use case. You're working on these are open forums that you can join both of them, the ones that uh the one in the afternoon uh I mean pacific japan, time 11 pm on thursday has been used by broadleaf pretty extensively.

A

So if you want to see a real world use case, you're welcome to join in and uh learn about what we're building for broadleaf and, of course, the 7am one that one's been very scantily attended. But it's uh it's a good one. It's a good one! It's good, it's good for folks to hop in and learn about what they can do with casper all right.

A

So the question from the community this week was the smart contract economy and I'll have alex talk about that, because alex wrote a paper on that some time back well on paper, it's too strong a word, but I believe it was a medium post about.

B

The smart contacts yeah a blog post.

A

About the smart contract economy.

B

uh Yes- uh and uh you know I mean my my views for was still uh correspond to what was stated there and the status there is essentially is that uh we actually have a very poor understanding of what the potential uses of the blockchain are right. I mean, I think, there's a uh there's like sort of a, I feel, a lot of misplaced skepticism.

B

uh uh You know in this general universe about you know the feasibility of uh you know these decentralized.

B

You know consensus platforms uh right, you know their doubts, for their proof of stake is going to work and they know so and so forth, but other kinds of also like this is combined with unreasonable enthusiasm about things like d5 right, and uh you know just my belief is that both of these are wrong right I mean you should not be overly ecstatic about d5, but you should be. uh You know fairly optimistic about uh you know, blockchains just being generally functional.

B

uh The problem is that nobody knows how to turn that you know basic functionality into something that the business can use uh and specifically into something that you cannot replicate with other existing systems right. uh The uh analogy that always go to is, you know the development of the commercial internet right where you know we are essentially, uh let's say somewhere around probably 1994, or you know 1995 right now in the sense that you know the basic building blocks. The basic protocols underline at right.

B

Tcpap, you know, like uh you know, http are already there, but the thousands being built on top of those protocols right. The actual application layer is still extremely underdeveloped right. So we are still in the world where uh you know most websites are basically you know their html pages right. I mean there's nothing super exciting, going on about them, they're, not really implementing any. uh You know uh interactive functionalities, it isn't just about you know uh in taking information or you know connecting maybe one. You know piece of information to another piece of information right.

B

So in particular what we don't have you know? We don't have. uh You know, search engines right. We don't have true search engines. We have directories right, we have alta vista, we don't have google, uh we don't really have uh you know any e-commerce right because, like all the little you know uh back-end stuff that you don't really think about when you shop online that doesn't exist. Yet right I mean there's, you know no way to process. You know credit card payments uh to buy. You know toilet paper on amazon.

B

You know there is no amazon, in fact, so that's kind of where we are at uh you know. Yeah.

A

I tend to agree, I would say, there's not even the browser, I would say the browser.

B

I would say the browser doesn't.

A

Even exist yet I mean honestly like for me, the killer app for blockchain really is going to be a self-sovereign wallet, self-sovereign identity and an integrated wallet right and where I can retain control of my private keys and it's easy, secure recoverable all those really really important things um where you know losing your keys and all of your funds associated with it is not commonplace right.

A

So yeah. From my perspective, I would say that the browser barely even exists at this point like we're in, like darpa net mode still.

B

Yeah fair enough right! Yes, so, uh but of course, once you have the browser, you still need something to actually.

A

Yeah absolutely yeah. That's.

B

Nobody what nobody really knows, uh but uh you know the inflow of investment uh into you know. Crypto platforms suggests that you know you're. Probably in this, like you know, bubbly, you know 1999 mode right, although still without that, you know insane variety of possible applications, but uh you know nevertheless right there a lot of people in the spheres a lot of money floating around, and so what's you know, one should expect that what was going to happen is you know, probably a series of uh you know, bubbles and every time a bubble bursts.

B

There are some survivors and then eventually that evolves. You know, basically by selection process into an actual. uh You know viable ecosystem yeah.

A

I would tend to agree, I mean the reason that the uh that we saw the bubble we saw in 2017 is because the internet in and of itself did not create wealth. You know early on in its very nasa days right. It didn't create this market right. There was no crypto economics associated inherent in the product itself. I mean when bitcoin went on its rally, it generated it created a lot of wealth right.

A

It's there's a huge wealth transfer and same with the ico boom right, so people were able to it basically created a capital market that didn't exist before, which is why we saw such a premature boom right, like the boom of 1999 with the internet, was all traditional markets like these? Were it was literally an ipo boom that happened right and all of the venture capital that flowed into it right um was is very different than what we saw with crypto right. There was. This is not even necessarily a venture capital right.

A

There was venture capital was literally like these. A lot of these venture firms were created as a result of the 2017 boom. Is it was it's almost like it took the entire model and flipped it on its head, um which is what I think crypto is going to do. Ultimately to a lot of these. You know fintech markets, but um yeah. I tend to agree that that we'll see a series of bubbles and bursts- and you know casper- is very different in that our approach is very different.

A

We're taking very much this approach that um we believe that the underlying protocol needs to be decentralized and we will enable a lot of decentralized applications, but I believe that even these, ultimately, these decentralized protocols need to be managed and governed in some mechanism right, and so what we've built is a way for owners to be it a decentralized governance protocol, a community or an individual manage their own chain code, um and so uh you get immutable transactions with the ability to upgrade your on-chain contracts and manage your on-chain contracts through a governance process, and it can be a single signer or it could be a multi-signer up to n number of individuals right.

A

It's unlimited in that regard, so.

B

Yeah, certainly, I think that you're, providing this kind of you know, essentially what amounts to market completeness is probably important uh to you know, spurn the growth, but you know again, the question is you know how they actually direct the growth into something, that's productive, because you know part of my skepticism. uh You know about d5.

B

Fundamentally, you know it's even more about just moving money around than traditional financial markets right, because exactly traditional financial markets do have the uh you know, property of uh you know basically managing the risk for something that ultimately has a physical, productive expression. Right I mean, ultimately, you know one is not too far off on things that it's really about.

B

You know like a literal car factory right I mean it's a risk associated with it, and these risks are, you know, being managed through the financial markets right, but it's not clear what the correspondent physical factory is for defy right. So that's right.

A

B

A financial markets that are disconnected completely from actual physical capital markets and that's.

A

B

A problem, uh so uh you know the question is, then you know how does one make you know? How does what makes the equivalent of a factory for you know for blockchain right? What does a factory look like? What does it make right and there it has to be? You know we have to very carefully identify what it is that is different about. uh You know blockchain as a business technology. Right I mean so. As you often say right I mean we are aiming to be part of the middleware right. I mean what.

A

Was the middleware.

B

Layer, not necessarily you know something that you know just you know street level users interact with, uh and so there you you have to okay. Well, uh you know, basically what uh you know, what are we providing? We are providing uh publicly verifiable computations and.

A

That's basically.

B

What it does, and so the question is okay, well, for a business, why do you need a publicly verifiable computation? So far? I have not seen like uh you know anything. You don't think particularly you know great, uh but you know I like. I have some speculation since the medium article actually about what kind of markets uh can you know um can actually benefit from this.

A

Right right and yeah from my perspective, I think if a business uh can benefit from additional layer of trust, then they should go ahead and use blockchain. But the thing is, you need to make it as easy as possible for them to integrate blockchain and yes, we have. I have strong opinions that you know the blockchain should remain at the protocol level at the protocol layer right and for the user for the end user. They shouldn't really care the blockchain really doesn't matter to them right.

A

It shouldn't matter to them in at least in so far as it should be completely invisible to them right. They see they reap the benefits of blockchain. Just like http. We all reap the benefits of http, but nobody cares right. Nobody thinks about http and tcp right. It's all behind the scenes exactly.

B

Yes, I think that once we launch we'll be able to, uh you know, think a little bit more clearly about you know where to uh you know where to take this, so that you know actually novel users of the blockchain are encouraged right, because I mean they are like you know: they're, even infrastructural problems with like chasing d5, because uh with the amount of money slashing there, you end up flooding.

B

uh You know, you know you know essentially as a channel, so is that you know nobody else can use as a blockchain anymore, because so much money is bidding up. Yeah.

A

B

Know the transfer capacity effectively.

A

Yeah, it's hard to argue with that, though, from a token perspective, but yeah.

B

I totally agree with you like from the long run perspective. That is not good right because, again, as they say, it's uh ultimately defined the stage is almost entirely speculative by the virtue of being. uh You know divorced from the you know, let's say the actual productive economy right I mean it's basically I mean at the end of the day I mean what we are working on is essentially a you know, a very advanced account in technology right but like at the end of the day for an accounting technology to be useful.

B

It has to be actually you know, accounting for something that you know ultimately can be consumed right. So.

A

Yeah yeah, I tend to I tend to agree like from my perspective.

A

uh I would like to see you know if you kind of step back and like what is the charter and what's the mission of casper labs, our mission really is to you know, make the benefit of blockchain more accessible to everybody. Right like we want to scale opportunity for everyone. I maintain that that is still one of our core tenants, one of our core pillars of our mission right um and you know we want to scale opportunity for developers.

A

We want to scale access to find new financial instruments for the individual and we need to enable blockchain to kind of cross into traditional markets and into like real real use cases right and whether we like it or not. You know financial markets are very heavily regulated today. Right and I just don't. I don't see I'm very much a pragmatist. I do not see governments catching up to the regulation, particularly united states. Other countries will be able to smaller countries, specifically like smaller countries will be able to catch up with the regulation very quickly right.

A

So you you'll see liechtenstein. Switzerland right switzerland historically has been a leader in financial fintech estonia. These small governments are able to rapidly adopt new regulations because it gives them an edge right. The united states is like it's a behemoth in terms of financial markets. It's going to move so slowly right and I've said this over and over again, when I worked at avalera, I worked at avalaire from 2011 to 2016.

A

back then, and that's almost 10 years ago, when I started there, governments in the united states, state state and local governments, united states student did not know how to tax sas software, and literally the problem is well. uh If my server is in one state and my customer is another state, what where does the tr? What's the transaction? What's the point of origin where's nexus, they are so far behind they're still stuck on nexus laws right like there's.

A

This whole thing about amazon, paying taxes, amazon cut a special deal with the state of washington right because their warehouses are all over the country. Their customers are all over the world. Where do they?

A

How does the state of washington get their piece and for the longest time they were only paying taxes on goods and services, goods that were delivered in washington state from amazon like you could buy stuff from amazon for many many years I did for almost a decade and I never paid sales tax right because there was no nexus rule right in washington. I mean it only. I only did business in the state of washington didn't do any business state of california.

A

I didn't have to pay sales tax and anything I purchased, and that just think about how long the internet's been around and they still haven't, figured out just how to tax services right. This is even taxing the goods, the transfer of goods from one state to another. They didn't have any rules for that, and so, uh if you ask me, it's going to take a very long time, the sec is moving as fast as they can.

A

Their first rule is always is to outlaw it because then they just wait and watch and see where the thing goes right, but I suspect it's going to take a very long time to see any movement in the regulatory space around crypto, you know and so being able to work. Within that framework, I think is going to be critical for this blockchain success.

A

Right and again, it's all about central control um and being able to centrally control everything associated with your account is going to set us up really well, even for centrally central finance extremely well. It's just a repeating theme. Over and over again.

A

B

I I do have you know like I also like uh on top of being skeptical about d5, I'm also skeptical, about integrating traditional finance. uh You know into blockchain applications just because that you know it is always stamped in face, because it's so highly politicized that you know. Essentially, you know every four years, you're just waiting for the sword to fall on you. You know, if you do anything that is outside of traditional bounds of finance right. So it's always. uh I.

A

Mean hey if you've been running a d5 protocol and you've been the central point of upgrading it and, if you're concerned about you, know regulatory coming after you, it's pretty easy right. You have a decentralized governance process that rotates who can sign upgrades on your on your protocol and you just release it into the wild and even with casper. You can make it completely decentralized right. So there isn't a single head to cut off right.

A

um You could have people all over the world with multi-signature keys on the protocol upgrades it could be as your protocols on casper can still be managed, as decentralized as bitcoin. If you so choose to do that right, um it's absolutely possible to do that with casper or you can just say it's an immutable contract. Now it's out there in the wild, the keys are burned. It can never be upgraded right, that's also possible to do so.

A

The key from my perspective at the protocol level, is providing the flexibility so that the market can take it where it wants to go right. We just build a decentralized protocol. Then we step away from it and speaking of stepping away from it, so you know people are really curious about our mainnet launch our token generation event and I'm just going to talk really quickly. We said that we were going to be. You know, ready to ship by q4.

A

I maintain from a feature completeness perspective. We are over 90 of the way 95 of the way there. I would actually say what we're focusing on now is performance um details around how blocks are proposed right in terms of the uh you know how many transactions we can fit into a block so we're bearing down now on the performance metrics around the blockchain, but in terms of feature completeness, we're pretty much there right. We have some small security enhancements around liveness issues that we discovered in delta test net.

A

We have refinements to do to the joiner. We have bug, fixing we're doing security review and we're integrating our custody providers. So from our perspective, engineering is going to now go into the mode of where we start upgrading the blockchain right. We start we get ready used to do how to do protocol hard forks. We will aggressively upgrade this blockchain. That is our intention that we will be in a pattern of performing upgrades. We will continue to do that because we believe that we will continue to get feedback from the community and new feature requests.

A

We want to implement side chains for horizontal scalability, horizontal scaling of the blockchain. We don't know about composite sharding, that's research that needs to kick off, but you know, andreas is very excited to start the sidechain research right away.

A

We are working on setting up our foundation so that we can be you know from a regulatory perspective, we're compliant for the public uh token sale. We have signed agreements with custody providers and uh the public sale.

A

uh You know people that are going to run the public sale for us we're also in talks with you know, late stage talks with multiple exchanges for listing the token. So those of you that have questions about that. These are all the pieces that we're lining up and getting ready for our mainnet launch.

A

This is the reason why we're looking at q1 for mainnet launch it's not about code complete! It's really about regulatory custody. Solutions um exchange listings and um being ready to launch on uh with it with the public sale right. So these are important things that we need to get done correctly, and you know the regulatory landscape changes very rapidly, and while we wanted to do the sale in early january, it will probably go to march simply because our custody provider, we want to make sure that's all buttoned up.

A

That's going to take us through the end of january, and then chinese new year hits, and so we will wait until after chinese new year to execute the launch and the sale. So for those of you that are curious about timing and the reason for it. This is the reason for it. We believe that everybody that is, you know is a vft holder will want us to do the right thing.

A

We want to make sure the protocol is hardened, secure and performant, and so we're going to take that extra time to make sure that's all tied up with a pretty big bow on it. So any questions from attendees on the.

A

Call cool so thanks for dialing in uh everyone in the u.s have a wonderful and safe thanksgiving, and I will speak with you on december 1st holy toledo. I believe that's december 1st november's got 30 days in it right tuesday is december 1st I will talk to you next week. Cheers everyone thanks for.

B