Practitioner's Perspective: Public Network Architecture

Loading...

Del curso dictado por University at Buffalo

Decentralized Applications (Dapps)

174 calificaciones

University at Buffalo

Decentralized Applications (Dapps)

174 calificaciones

Curso 3 de 4 en SpecializationBlockchain

This third course of the Blockchain specialization prepares you to design and develop end-to-end decentralized applications (Dapps) – which provide anyone with access to the blockchain’s features and services. You will use Truffle IDE, smart contracts, a simple web client and a MetaMask client. You will learn about the architecture of a Dapp: the front-end client interface, backed by the blockchain and smart contracts. The course covers the basic design of a Dapp, Truffle development process and commands (init, develop, test and migrate), test-driven development of Dapp, Dapp application models and emerging standards that are essential for predictable Dapp behavior. Main concepts are delivered through videos, demos and hands-on exercises.

De la lección

Decentralized Applications (Dapps)

We will explore the notion of the blockchain server as the foundation for a Decentralized Application. We will demonstrate how to install the blockchain server and establish a peer-to-peer network of nodes. It is a common practice to develop and test a Dapp on a local test network before deploying it on a public network.

Ethereum APIs10:30

Practitioner's Perspective: Public Network Architecture5:26

Conoce a los instructores

Bina Ramamurthy
Teaching Professor
Computer Science and Engineering Department

[MUSIC]

The problem is those private networks are invisible.

You don't know that they exists.

There is no telling how many hyperledger fabric networks are out there today.

We don't think that there is millions of them or thousands of them, but

it's an honorable number because it's not any kind of subscripts like the public

network, where you can actually do any kind of analysis on that.

So what's needed is a registar.

And we think that the public network is where that's going to live,

just like you have DNS today and you can go and register a name.

There is something on Ethereum called the ENX, right?

The naming service for Ethereum.

And that allows you to do some really cool things.

In fact, students of this course should really look at the naming service on

Ethreum, because it's a structure that allows you to some really interesting

things around, not only identity, but around organization and

things like access control and user management.

The naming services is vastly underutilized compared to

where it's going.

It's very important piece of technology, the Ethereum main net.

And with that and some other things, you can start to imagine

having a registrar of both private networks and public ones.

So from ConsenSys's point of view, I think and

from Ethereum's point of view, there's a vision that some of us share

that I think is right where the public main net becomes sort of the referee

of millions of what we call en lateral peer-to-peer connections.

You and me, us this room, bigger group, huge group, public group, right?

And you want all those to be able to coordinate so that our smart contract and

they book other smart contracts, they can book other smart contracts,

ultimately with a public network, but without invoking race conditions and

all the other problems that the classic distributed systems have.

So here you have this sort of a kind of time keeper, you have a timestamp on

the block and you have this public massive public network

that's pretty hard to corrupt, vanishingly small probability of corrupting, right?

Not zero probability, but limiting the zero.

And that public network, especially Ethereum,

because it's got the attributes of the turn and complete system and

it's more than a key value per storage system.

It's a little more complex.

It allows for more attributes.

We think it's a pretty god referee for all these millions

of this factorial of peer-to-peer connections,

peerwise connections between individual nodes, right?

And once you have that, then you can say, I want to register that.

I'm going to register my node, my node and your node, our connection.

I can register it, or I can say this is a network, or

this is another bigger network.

This is the supply chain for the automotive industry network, but

here is the rubber providence supply chain or providence network for

making sure that we don't have conflict materials in our rubber supply.

All right, those are two different networks that created at different times

without even knowing about each other.

But if they're all coordinated by the public main net,

then they can start to as long as we don't get the data models too wrong.

We can start to bring those together and say, all right, we started here.

But now we're going to start doing transactions with this larger thing.

All right, I'm good here.

I'm a tire manufacturer.

I've got this rubber providence supply chain, but

I'm part of the automotive supply chain.

I need to be part of both networks.

And I don't want to have two separate nodes that do one and

the other separately.

I want one set of nodes that does everything.

If I'm setting up that consortium,

I want to be able to register that as a thing that I can join on the main net.

What tools are out there to build interesting solutions or

to tackle these problems, right?

The interesting thing is that a lot of these tools are just merging.

There's a lot of opportunity to build new tools, so

if you're a tool builder, now's your time.

Get out there, tell us about it.

My job is the seeker of awesomeness, so if you've got an awesome tool,

come to me, something about it.

Maybe we can turn you into a new venture.

But there are tools and then there are tools.

So there are tools like plasma and protocol level tools,

and then there are tools on top of that, just like computer science has

always been a layer, a set of abstraction layers all the way down and up, right?

So here we are at this abstraction layer.

We've got distributed systems.

We have ways of managing those distributed systems.

We have consensus layers and all these other things.

We have the data layer and on top of that, we have

this layer of enabling tools that help you build interesting solutions.

Two of my favorites are uPort and OpenLaw.

Explora nuestro catálogo

Inscríbete de manera gratuita y obtén recomendaciones personalizadas, actualizaciones y ofertas.

Coursera brinda acceso universal a la mejor educación del mundo, al asociarse con las mejores universidades y organizaciones, para ofrecer cursos en línea.

© 2018 Coursera Inc. Todos los derechos reservados.

Descargar en la App Store

Consíguelo en Google Play