The Starting Point of Internet Interaction: A Comprehensive Explanation of the Mechanism Design of the Decentralized Naming System Handshake

Written by: Ming Ng, Andrew Lee, Cassandra Shi

Compiled by: Perry Wang

Mars Finance Editor's Note: Due to the recent hijacking of DNS on the Binance Smart Chain, many funds were put at risk. This article, originally published in September 2020, is a timely reminder to re-examine the decentralized domain name Handshake.

Names hold a crucial position in human perception of existence and interaction with all things in the world. The core of all interactions lies in the ability of parties to match names with the specific things or people they represent.

Names are an indispensable part of human experience, and one could even question whether something or someone truly exists if it does not have a name.

Correspondingly, names on the internet are vital for our online presence. Users, applications, or machines search for resources on the internet through the names of those resources. This name needs to be understood by humans and must also be uniquely identifiable by machines among billions of potential destinations.

Matching names with the ultimate resources is the starting point for trillions of interactions that occur on the internet every day. Therefore, it is no surprise that naming is the initial starting point of the three core layers of the internet stack: naming (DNS), transport (TCP/IP), and application (HTTP).

Naming requires a unique source of truth, as names within a namespace must be guaranteed to be unique throughout the system. Therefore, an effective naming system must not only be a standard or protocol but must also meet all other requirements for operating a namespace at the scale of the internet—this includes enforcing unique names, managing naming records, scaling for internet access, while still ensuring that anyone, anywhere can have complete access.

The Current State of Naming Spaces

Names are the most valuable assets on the internet, yet we do not truly own our names. All key naming spaces belong to centralized entities that control these naming spaces and take control away from you. This is true for all important naming spaces today—ICANN naming space, Facebook, Twitter, and Google.

Thus, your name on the internet does not belong to you; it belongs to the owners of these centralized naming spaces. With just a keystroke, they can delete anyone. If your name exists within a centralized naming space, your effective right to exist belongs to others.

Centralized naming spaces also determine users' ability to exist. They decide users' ability to search, match, and communicate with others. They unilaterally set frameworks that limit which protocols users can use, what use cases are allowed, and what information can be transmitted.

These centralized naming spaces enforce monopolistic power with little consequence, giving them some of the most valuable properties in today's internet. Verisign, which operates multiple domain registrars, earns billions of dollars annually from the .com domain system it controls, with almost zero innovation, while ICANN can arbitrarily raise the price caps on all top-level domains (TLDs) through its cartel enterprises. Facebook and Twitter completely control how users use their names/accounts, and can delete pages and user identities at will, often without any explanation.

We see a serious threat everywhere that relies on centralized entities for existence and being discovered by others. The internet was supposed to be free of kings, but the ability to eliminate a person's existence and precisely control how names are used has made the owners of these naming spaces the de facto kings/rulers of the internet.

The World Needs Decentralized Naming Spaces

Of course, these centralized naming space owners control digital existence, block access, and enforce monopolistic economic models, which is in stark contrast to the goals of decentralized networks that advocate for the ability to exist, innovate, and create one's own business models without centralized control or systems.

Whether it is decentralized currency, decentralized document systems, or decentralized servers, if these decentralized entities are not incorporated into a widely used naming space, they do not exist for the vast majority of users on the internet.

If there is no decentralized naming space that can be widely read by humans and parsed by machines, users cannot widely adopt the decentralized world.

Technical Standards for Decentralized Naming Systems

Naming systems play a crucial role in discovery, connection, and identification activities. As one of the most fundamental and enduring components of the internet's backbone infrastructure, extremely high standards need to be set for its lifespan, stability, and technical scalability.

For a decentralized naming system to become a legitimate naming space in the decentralized world, the bar is even higher. Without a centralized authority, the whole world must trust that this naming system can maintain stability over a long period, regardless of any potential upheaval or technological advancement, allowing humanity to remain connected.

Thus, the fundamental structure of this naming system needs to meet certain key technical, social, and governance requirements:

1. Truly decentralized: If a decentralized naming system is still controlled by a small group of people, what is the point?

2. Focused on the responsibilities of the naming system: The responsibilities of the naming system need to be extremely focused and swift. If the DNS system is designed to also handle the transmission of 4K video, imagine how reliably it could operate?

3. Achieve trustless access as much as possible: Anyone should be able to access the naming space directly in a completely trustless manner without consuming a lot of resources.

4. Compatible with other parts of the internet: Allow seamless integration and use with other applications, users, and technology stacks.

5. Stability and upgradability at the protocol level: Allow continuous innovation without disrupting regular operations.

The Mechanism Design of Handshake

Considering these goals, as well as the overall objectives of decentralized root domains and authorities, Handshake is the only naming system fundamentally suited to serve as a decentralized network naming space.

Focusing on the Responsibilities of the Naming System

Let's first think about the inherent complexity of a large-scale naming system like the internet. Unlike other layers of the internet stack, the naming layer DNS is the only one that is a system rather than a protocol. The main difference between a system and a protocol is that a protocol cannot enforce the uniqueness of names, which is crucial for the normal operation of the naming space. It can be said that the naming layer of the internet is also the most complex layer to date, facing many competing technical, political, and economic demands.

As an independent blockchain, Handshake has the space to develop and manage itself without interfering with other projects and does not have to compete for priority with other use cases (such as gaming or DeFi) operating in parallel on the same network. Additionally, other blockchains have some fundamental limitations, such as the Bitcoin network limiting the size of OPs, while Ethereum is notoriously difficult to synchronize.

If Handshake were to depend on another blockchain, the instability caused by competition for priority among use cases and political interests would undermine the core requirement of decentralized naming infrastructure—stability. The naming infrastructure must be highly stable; remember that users, hosting providers, and developers must be confident that names will exist in the same format for a long time. For example, the skyrocketing gas prices on Ethereum due to the DeFi boom and the complex migration to ETH2 have created significant uncertainty about how applications will operate in the future and whether retail users will have the same access as whales.

Finally, creating a native auction system is very complex and requires highly specific native elements, such as making tokens unusable for certain periods. If the tokens of the Handshake protocol, HNS, are not native tokens, it would greatly increase the complexity of the system.

Decentralization

Another key consideration is decentralization. Remember, the goal here is to achieve a truly decentralized, uncensorable naming space free from centralized control and strategies. Anything below this goal would be completely redundant.

Ethereum is the most decentralized smart contract platform to date, but it is still insufficient as the underlying blockchain for a truly decentralized naming system. Ethereum-based systems must either be strictly immutable or design governance mechanisms with single or multiple signers. For example, the ENS system on Ethereum has a multisig composed of 7 parts, making it auditable, or hindering any future innovation or upgrades. These mechanisms either obstruct future innovation or fail to meet decentralization requirements.

What about sidechain solutions? Sidechains primarily rely on the security of the main chain, and since they share task prioritization with the main chain, the concerns about multi-task competition for priority remain unresolved in this solution. Additionally, there are currently no decentralized sidechains in the Bitcoin network. Counterparty is a one-way system, Liquid requires a small consortium multisig, and Rootstock is currently pushing for a consortium awaiting Bitcoin as the main chain to support Drivechain.

For all the issues with proof of work (PoW), its foundation is a decentralized competitive relationship due to the limited number of miners, and there are clear differences in the concerns of developers, users, and miners. This contrasts sharply with proof of stake (PoS), which encourages stakeholders to collude and centralize management, creating an environment that is largely a form of plutocracy.

Therefore, truly decentralized naming with upgradeability is most likely to be realized on an independent PoW blockchain with strong hash rates, a robust ecosystem, and miners confident in the blockchain's value.

Outstanding Trustless Solutions

Compared to other naming blockchains, the entire Handshake technology stack is designed to create a human-readable, truly decentralized, fully accessible, and secure naming space.

The naming data in Handshake is stored in a new data structure called Urkel Tree, specifically designed for this purpose. The proofs are small and can be quickly verified, allowing for name resolution with minimal computation.

Secondly, HNSD, a highly unique application written in C, only handles DNS functions within Handshake (avoiding any currency-related elements). Its design mechanism ensures speed and lightweight operation as much as possible. It is written in C and can be compiled on every computer system in use today. HNSD verifies compact proofs, cryptographically secures DNS records on the Handshake blockchain, and writes them into the chain with the maximum proof of work.

The lightweight resolver means anyone can use Handshake with minimal effort. The compact Urkel proofs bring high efficiency, allowing Handshake solutions to be executed on the most rudimentary devices.

Thus, Handshake is the only naming system that allows anyone at any level of resource to access the naming space without any centralized components and without any level of security trade-offs. Users or applications can choose to resolve names either by running a full node (HNSD client) or through third-party resolvers. In many cases, this spectrum of accessibility is the litmus test for a trustless naming resolution system.

Compatibility with Other Parts of the Internet

HNS can work in collaboration with the entire DNS stack across all key pathways (technical, solutions, and new systems) without being burdened by the political, economic, and technical constraints of DNS.

Handshake will not replace the ICANN root domain system but will extend it. When querying a name, the Handshake resolver first checks the TLD's blockchain. If it is not found there, it "falls back" to the ICANN DNS system. Users running Handshake can browse the entire internet as usual, but now they can also resolve names rooted in the blockchain.

After resolving the root domain, DNS resolution on Handshake is indistinguishable from the old internet systems. Resolvers follow the domain and subdomain chain indicated by the URL and connect with centralized, authoritative name servers around the world, which still run the exact same software as before.

Stability and Upgradability at the Protocol Level

Handshake allows any content of up to 512 bytes to be inserted into the Urkel Tree as a name. Currently, all Handshake software is configured to only allow read/write of properly formatted DNS records, but writing software that uses data for other purposes is straightforward.

Moreover, PoW is the most powerful mechanism known to us that can provide infrastructure-level stability while having a clear upgrade path.

Regarding governance, philosophically, the most interesting aspect of HNS is the experiment with a no-king structure. There are no leaders, no foundations, and no core teams for product releases. Handshake follows the same upgrade path as Bitcoin through soft forks and off-chain social coordination, proving that Handshake has the most flexible governance mechanism.

Alternative methods like PoS governance are still in their infancy and have not stood the test of time. The approach of the Ethereum Name Service (ENS) is more fragile, as a group of multisig key holders (regardless of their reputation) will control the governance and upgrades of the backbone infrastructure of the decentralized network.

Handshake and Bitcoin are two truly decentralized projects without kings, and the cryptocurrency space should explore new coordination mechanisms. We will discuss this topic in future articles.

Handshake: Expanding the Internet to the Stars and Beyond

The idea of controlling names through centralized systems is fundamentally incompatible with decentralized networks, which need to be free from any centralized elements.

Without a widely resolvable and recognized naming system, decentralized networks and everything within them (decentralized servers, decentralized document spaces) will be invisible and non-existent in the eyes of the mainstream world.

HNS is the only naming space designed to withstand any length of time, infinitely expand use cases, and maintain a balance between convenience and decentralization.

Of course, for Handshake to achieve the widespread adoption it needs, it will take some time. However, the winding road to Handshake gaining legitimacy is not the same as the validation curve that Bitcoin had to traverse before being recognized as a legitimate store of value and becoming a major competitor to gold and the dollar.

By making decentralized networks as easy to access as centralized networks, HNS aims to extend the internet beyond its current boundaries. If you are interested in participating in this historic process, we strongly encourage you to become part of this ecosystem—join the community, acquire HNS tokens, and stake your claim in this field.

Most importantly, participate in the community by providing code, ideas, or simply a good vibe. Handshake and decentralized networks are still in their early stages of development, and it is crucial that everyone’s sense of existence, collaboration, and coordination within the internet is still in its infancy.

We believe that every internet user should truly own a decentralized name in a world where they can use it freely and feel secure knowing that the name truly belongs to them. We believe Handshake will be the system that carries this vital naming space. Safe travels!