Viewpoint: Decentralized "domain names" may become the identity system of the Web 3.0 era

Author: Yang Min, Developer of Decentralized Account System DAS

Source: Orange Paper

Undoubtedly, decentralized identity is a core piece of the Web 3.0 puzzle. From the era of altcoins, to the ICO era, and then to the DEFI/NFT era, there has never been a shortage of projects telling stories about decentralized identity. Investors and speculators are also willing to pay for such stories. However, today we still do not have a practical decentralized identity system. In other words, the products that currently claim to offer decentralized identity are largely dispensable.

Why is this the case?

Where does the problem lie? Is decentralized identity a false proposition, or is the technical difficulty of implementing decentralized identity very high? Neither of these is the case. The problem may lie in:

1. The very goal of decentralized identity
2. And the paths chosen to achieve this goal.

Problem 1: The Goal of Decentralized Identity Itself

This issue refers to the functional boundaries of decentralized identity, what responsibilities it should undertake, and what form the product should take. This is a challenging question for us at present. This state of difficulty is not unique to us. When a person from the web1.0 era describes what applications might exist in the web2.0 era, or when someone from the first decade of the 21st century tries to describe what life in the mobile internet era would be like, they face the same difficulties.

Problem 2: Path Selection for Achieving the Goal

This issue refers to the fact that, in the absence of a clear goal, most teams tend to adopt a "create first, compare later" approach. They first create a product, add various features, and then go to the market to validate whether this is what people will need in the future. However, these features often come with strong assumptions. For example, it is assumed that users are willing to install applications and encrypt and upload identity data; it is assumed that developers will access this data in some way to provide services to users. These product features may be correct at some future point, but they fail to spark any interest from current users, as they do not address present problems.

Some Certain Judgments on Decentralized Identity

Although we cannot depict the full picture of decentralized identity in advance, we can still make some very certain judgments.

Decentralized Identity Needs to Be Human-Readable

Regardless of what the future decentralized identity system looks like, no matter how complex the underlying business is, it must have a globally unique identifier to represent each object, making it easy for us to recognize. Just as we do not start describing a person from all their details, despite each of us having different cultural backgrounds, personalities, preferences, intelligence, height, etc., we first mention their name or their ID number.

Human-readable means that these globally unique identifiers should be easy to read and spell for humans.

Decentralized Identity Evolves, It Will Not Appear Suddenly

We will not suddenly have a brilliant idea one day and figure out the final form of decentralized identity. It should evolve gradually from a product form that current users are willing to continue using. Just as the mobile device we use most frequently today—the smartphone—did not suddenly design itself into its current form. We still refer to smartphones as phones, which is a clear trace of evolution.

The willingness of current users to continue using is very important. In fact, the responsibilities that today's smartphones undertake are more in line with what personal digital assistants (PDAs) in the 1990s aimed to achieve. Yet today's smartphones are called phones, not PDAs.

A key point here is that if you can only take one device when you go out (in fact, you don't even need to say "if"; when you can only take one device, no one wants to take two), you will definitely choose to take your phone rather than a PDA. Because at that time, stable calls anytime and anywhere were more important than texting anytime and anywhere. The phone is a rigid demand, and thus people are more willing to carry their phones long-term, leading to a continuous increase in the retention of phones.

When a new technology emerges, it will naturally be chosen to be layered onto the phone. The technology itself also needs to survive, and it will choose the host that is most conducive to its survival. Over time, as more and more new technologies are layered onto the phone, the phone can do everything that a PDA can do, and people have no reason to use a PDA separately. Therefore, usefulness is the premise for existence, and existence can continue to evolve.

Pragmatic Decentralized Identity

So based on the above logic, what kind of products are likely to evolve into the future decentralized identity?

The answer is decentralized "domain names."

Decentralized "Domain Names" Provide Globally Unique Identifiers That Are Human-Readable

Just as internet domain names solve the problem of hard-to-remember IP addresses, blockchain-based decentralized "domain names" solve the problem of hard-to-remember blockchain addresses. Previously, to make a transfer, you needed to copy and paste the other party's address and carefully verify it each time. With decentralized "domain names," you can directly input easy-to-read and easy-to-spell names like alice.bit or bob.eth for transfers. At the same time, these names are globally unique and can only be managed by the user's private key.

Decentralized "Domain Names" Have the Basis for Evolution

There are at least two aspects that make users willing to continuously hold decentralized "domain names": on one hand, the practicality brought by the ability to make transfers; on the other hand, their uniqueness makes users willing to use them as a personalized identity marker.

Moreover, unlike traditional domain names that can only be associated with limited data types like IP addresses, decentralized "domain names" support associating with any data type, allowing different applications to read the data they care about and provide different services to users. For example, in a transfer scenario, a wallet will read the blockchain address under the "domain name"; in a personal homepage scenario, a personal homepage application will read various profile information of the user; in an end-to-end encrypted communication scenario, the application will read the user's public key. The support for any data type gives decentralized "domain names" great flexibility, making it easy to layer applications on top of them. Therefore, the practicality of decentralized "domain names" will be continuously reinforced.

Thus, we have reason to believe that decentralized "domain names," which can solve current practical problems, represent an early form of decentralized identity. Compared to directly building decentralized identity products, continuously layering decentralized identity-related features on top of decentralized "domain names" is a more pragmatic approach.

Others

For decentralized "domain name" projects, it is recommended to pay attention to ENS and DAS.

The decentralized "domain names" mentioned in the text are in quotes because we should not let this single use case limit our imagination; we should refer to them as decentralized accounts.