Note: This article was originally published on Gavin Wood's blog "Insights into a Modern World" on April 17, 2014, when Gavin was still the co-founder and CTO of Ethereum and had not yet founded Polkadot. In this blog, Gavin comprehensively explains what the Web3.0 era in his mind should be like and the four components that make up Web3.0.

Wu talks about blockchain, translated by Ping Xiong

As we move into the future, we find that there is a growing need for a zero-trust system of interactions.

Even before Snowden, we were aware that entrusting our information to arbitrary entities on the internet was fraught with danger. However, in the aftermath of Snowden, this awareness fell mainly on those who believed that large organizations and governments often try to overreach. As a result, we realize that entrusting our information to corporate entities is a fundamentally wrong model. The chances of a corporate entity keeping its hands off our data are simply the cost of the effort required minus its expected benefit. Given that companies’ revenue models often require them to know as much as possible about their users, realists will realize that the potential for surreptitious abuse is unlikely to be overestimated.

The protocols and technologies on the web, and indeed the entire Internet, are a great technology preview. SMTP, FTP, HTTP(S), PHP, HTML, and Javascript are just a few of the major tools that have contributed to the cloud-based applications we see today, such as Google Drive, Facebook, and Twitter, not to mention countless other applications such as gaming, shopping, banking, and dating. However, in the future, these protocols and technologies will have to be redesigned based on our new understanding of the interaction between society and technology.

Web 3.0, or what might be called the “post-Snowden” web, is a reimagining of the things we already use the web for, but with a fundamentally different model for interactions between parties. What we consider public, we publish. What we assume is consistent, we put on a consensus ledger. What we consider private, we keep secret and never reveal. Communications always take place over encrypted channels and only with anonymous identities as endpoints; nothing traceable (like IP addresses) is ever used.

In short, we designed a system that mathematically enforces our prior assumptions because no government or organization can reasonably be trusted.

The post-Snowden web has four components: static content publishing, dynamic information, trustless transactions, and an integrated user interface.

release

First, we already have a lot of decentralized, encrypted information publishing systems. What these systems do is take a short intrinsic address (i.e., a hash value) of some information, and if needed in the future, you can get the information itself based on the hash value. New information can be submitted to it. Once downloaded, we can guarantee that it is the correct information because the hash value is inherent to it. This static publishing system takes on most of the work of HTTP(S) and all of the work of FTP. There are many implementations of this technology, but the easiest to cite is BitTorrent. Every time you click on a BitTorrent magnet link, what you are doing is telling your client to download the data corresponding to that hash value.

In Web 3.0, this part of the technology is used to publish and download any (potentially large) static information we are willing to share. We can, just like BitTorrent, incentivize others to maintain and share this information; however, combined with other parts of Web 3.0, we can make it more efficient and accurate. Because the incentive framework is inherent to the protocol, it is DDoS-proof by design. How about this bonus?

Communications

The second part of Web 3.0 is an anonymous low-level messaging system based on identities. This is used for communication between people on the web. It uses strong cryptography so that some guarantees can be made about messages; they can be encrypted with the public key of an identity to guarantee that only that identity can decode it. They can be signed by the sender's private key to guarantee that it really came from the sender and provide the receiver with a secure receipt of the communication. A shared secret can provide the opportunity for secure communication, including between groups, without the need for proof of receipt.

Since each of these provides the ultimate message logistics, it becomes unnecessary to use transport protocol-level addresses; what was once an address consisting of a user or port and an IP address is now just a hash value.

Messages will have a time to live, distinguishing between a publish message, which might want to live as long as possible to ensure that as many identities as possible see it, and an immediate signaling message, which might want to be transmitted across the network as quickly as possible. Thus, the binary opposition of latency and lifespan is interchangeable.

The actual physical routing will be done via a game-theoretic adaptive networking system. Each node tries to maximize their value to other nodes, asserting that other nodes have valuable information about them. A node whose information is not valuable will be disconnected, and their location will be connected to some other, possibly unknown (or possibly secondary) node. In order for a node to be useful, information with some specific properties will be required (e.g., the sender address or subject, both unencrypted, starting with a specific bit string).

In Web 3.0, the communication part allows nodes to communicate, update, and self-organize in real time, publishing information that does not require internal trust or priority for later reference. In the traditional web, this is most of the information passed over HTTP in an AJAX-style implementation.

consensus

The third part of Web 3.0 is the consensus engine. Bitcoin brought many of us the idea of ​​consensus-based applications. However, this was only a tentative first step. A consensus engine is a way to agree on certain rules of interaction, knowing that future interactions (or non-interactions) will automatically and irreversibly be executed exactly as specified. It is effectively an all-encompassing social contract and draws its power from the network effects of consensus.

The fact that a breach of one agreement can affect all others is key to creating a strong social contract, thereby reducing the chances of breach or willful neglect. For example, the more a reputation system is isolated from an individual’s social interaction system, the less effective it will be. A reputation system combined with features similar to Facebook or Twitter will be more effective than one without, because a user’s intrinsic value depends on how they are viewed by their friends, partners, or colleagues. A particularly poignant example is the dilemma of whether and when to friend an employer or a dating partner on Facebook.

The consensus engine will be used for all trusted information publishing and modification. This will happen through a fully universal global transaction processing system. The first viable example of this is the Ethereum project.

Traditional networks do not fundamentally solve the consensus problem, but instead fall back on centralized trust in authoritative organizations such as ICANN, Verisign, and Facebook.

front end

The fourth and final component of the Web 3.0 experience is the technology that brings it all together; the "browser" and user interface. Interestingly, this will look fairly similar to the browser interface we already know and love. There will be a URI bar, a back button, and of course, the lion's share will be used to display Dapps (i.e. web pages/websites).

Using this consensus-based name resolution system (different from Namecoin in the app), the URI can be simplified to a unique address (i.e. hash) for the frontend of the app. Through the information publishing system, this can be expanded to the collection of files required for the frontend (e.g. an archive containing .html, .js, .css, and .jpg files). This is the static part of the dapp (-let).

It does not contain dynamic content, but is served over other communication channels. For collecting and submitting dynamic but public content whose origin must be absolutely certain and must remain unchanged forever ("fixed"), such as reputation, balance, etc., there is a javascript-based API for interacting with the consensus engine. For collecting and submitting dynamic, potentially private content (which is necessarily volatile and may be destroyed or lack availability), a P2P messaging engine can be used.

There will be some cosmetic differences, we will see a shift away from traditional client-server URL patterns like "https://address/path" to new forms of addresses like "goldcoin" and "uk.gov". Name resolution will be done via contracts based on a consensus engine, which can be easily redirected or augmented by users. Periods will allow for multiple levels of name resolution, for example, "uk.gov" might pass the "gov" subname to the name resolver given by "uk".

Due to the permanent transient nature of information being provided to browsers automatically or incidentally through updates to the consensus backend and maintenance of the peer-to-peer network, we will see backend DApps or dapplets play a large role in our Web 3.0 experience.

After the initial sync process, page load times will be reduced to zero, as static data is pre-downloaded and guaranteed to be up-to-date, while dynamic data (delivered via the consensus engine P2P messaging engine) is also maintained up-to-date. While syncing, the user experience will be very reliable, although the actual information displayed may be out-of-date.

For users of Web 3.0, all interactions will be conducted in an anonymous, secure, and, for many services, trustless manner. For those services that require a third party, these tools will give users and application developers the ability to spread trust across multiple different, potentially competing entities, significantly reducing the amount of trust people must place in the hands of any particular single entity.

With the separation of front-end and back-end APIs, we will see additional capabilities to leverage different front-end solutions to provide superior user experiences. For example, Qt's QtQuick and QML technologies can replace the HTML/CSS combination of traditional web technologies and provide native interfaces and rich accelerated graphics with minimal syntactic overhead and an efficient reactive programming paradigm.

Migration to Web3.0

This transition will be gradual.

On Web 2.0, we will increasingly see sites that leverage similar Web 3.0 components on the backend, such as Bitcoin, BitTorrent, and Namecoin. This trend will continue, and Ethereum, the true Web 3.0 platform, will likely be used by sites that want to provide transaction proofs for their content, such as voting sites and exchanges. Of course, a system is only as secure as its weakest link, so these sites will eventually move themselves to Web 3.0 browsers, which can provide end-to-end security and trustless interactions.

Say hello to Web 3.0, the secure social operating system.