Ethereum Watch

Here are what I find to be the most significant Tritium features described in the whitepaper.

We’ll go one layer at a time.

1) Network Layer

Nexus points out that scalability doesn’t just depend on ledger size (read: the size of messages, and by extension block size) but on how messages are routed among nodes.

Packets are replicated on the bottom Network layer rather than all the way up at the top layer, boosting speed significantly. The Network layer uses Multicast Locking Groups, Nexus’s first step towards “Level 1 locking” in its multidimensional chain structure — which will ultimately have three levels of locks.

These locking groups working in tandem allow for parallelism, while preserving global consensus. Parallelism in various forms is being pursued by blockchain projects seeking scalability, but Nexus is building in parallelism at the very lowest level, and unlike most designs, it will not compromise global consensus or decentralization in order to do so.

Of course, in today’s world, nodes — desktop, laptop, and handheld computers — are often behind router and other firewalls. One of Nexus’s big announcements of the past year was the integration of LISP (Locator/ID Separation Protocol), allowing Nexus nodes to securely have their own identity even if they’re behind a firewall. Otherwise, devices such as your local WiFi router would become bottlenecks, collapsing Nexus’s distributed network into a decentralized network, plagued by choke points.

As the whitepaper puts it, “together, Nexus and LISP form the world’s first truly distributed, secure and scalable application and infrastructure network.”

2a) Ledger Layer: Signature Chains

One major issue with a number of signature schemes, used in every blockchain project, is the need to avoid the reuse of key-pairs. Some cryptocurrencies enforce this behavior, forbidding the reuse of key-pairs or using new addresses by default — such as most Bitcoin wallets today — while others only recommend it be avoided.

Nexus’s Tritium protocol renders the problem moot by bringing Signature Chains to the table, a secure solution that still allows for convenient, flexible use of keys, all the while “protect[ing] account-based transactions and prevent[ing] dust spam attacks.”

With Tritium, Nexus accounts will be accessible with a username, password, and PIN. No more frequent wallet.dat backups necessary. I’ve personally lost coins due to out-of-date wallet backups before, and I’m happy that losing coins to staking heaven forever will no longer be a concern.

Nexus signature chains are built for quantum resistance, like everything else in Nexus’s design. If you use Nexus, your key is updated after every transaction and obscured with hash functions. Your Nexus address is no longer tied to your public key. You can reuse it forever without any hit to security since your public key is never reused.

Signature chains are extremely versatile, too: virtually “any hash function or asymmetric cryptography” can be adopted by the scheme should the need arise.

The adoption of signature chains also allows Nexus to securely move away from Bitcoin’s UTXO model to an Account model. This brings Nexus even farther forward from its roots in Bitcoin and gives it the advantages of the Account model without compromising security, as reusing addresses is no longer a risk.

2b) Ledger Layer: Reputation

Nexus’s ledger layer includes a reputation framework.

As nodes participate in events, they build up a digital “reputation.” “Allowing nodes on the network to verify each other through their reputation” solves a large concern with most Proof-of-Stake (PoS) consensus protocols. “By coupling an economic incentive with greater trust, such as higher returns on verification, there is a (non-trivial) cost incurred by loss of reputation.”

Basically, staking Nexus nodes are economically disincentivized from doing something that would hurt their reputation. As with reputation in the real world, reputation on Nexus takes a good amount of time to build up — by using resources and time to contribute to the network — but can be lost quickly. If any node tries to “witness” events which did not happen, false events contrary to the consensus of the network, that node risks losing trust — and thus losing witnessing power in the future.

Similar reputation systems are found in several other blockchain projects, but usually as the main consensus mechanism. On Nexus, reputation is not a potential central point of failure but, extended into a relationship system, an additional layer of protection against Byzantine attacks.

This relationship system “lays the foundation of what can be considered an immune system for the global network and consensus mechanism, whereby the network can identify an attack by the departure from normal behavior and neutralize it before sustaining damage.”

This makes Nexus not only more scalable but also more secure.

2c) Ledger Layer: Twin Blocks

One issue that hurts the efficiency of legacy blockchains is orphaned blocks.

Tritium supports twin blocks, which reduce the possibility of orphaned blocks and at the same time increase network capacity.

If two blocks at the same block height (the same place in the blockchain) are constructed by different channels — Nexus runs on three channels, namely prime, hash, and holding/staking — the two blocks will both be accepted, as long as there are no locking conflicts, such as attempted double spends.

As Nexus builds out its multidimensional chain, twin blocks help scale in the horizontal dimension. They also provide cross-channel accountability. In other words, the three block verification channels can verify blocks produced by one another, across channels. I find this encouraging: Some theorize that multiple channels offer greater attack surface for bad actors, but cross-channel verification mitigates that risk.

So Nexus’s twin blocks are yet another innovation that boost both scalability and security.

The core work in Tritium, and the most detail in the Tritium Whitepaper, is concerned with the Network and Ledger layers. But the paper continues to quickly survey Nexus’s planned higher layers, as well, so we’ll do that here.

3) Register Layer

In short, registers enable “higher levels of abstraction of the ledger,” offloading strain from processing nodes in the Nexus network. Translation: more speed, more efficiency.

One example use of the register layer the whitepaper gives, critically relevant to the current state of blockchain, is for the registers to “reference hashes of specific checkpoints in their private databases, to allow the easy verification of data states within internal systems without compromising privacy.”

Most blockchains do not have good systems for confidentiality in place, but reference hashes will provide organizations running projects on Nexus a useful way to implement confidentiality when dealing with things like medical records, intellectual property, and compliance with privacy regulations.

4) Operation Layer

The next level of abstraction will be where operations live. Operations are “byte-level instructions that can be used to perform basic tasks on registers.”

Opcodes can be compiled from domain-specific languages, if desired, and sample opcodes described in the whitepaper include READ, WRITE, DEBIT, CREDIT, AUTHORIZE, TRANSFER, and GETHASH.

This is an important layer for developers, providing a basic language with which to easily communicate with registers, but there’s still too much heavy lifting here for advanced applications.

We still need more abstraction.

5) API Layer

Let me interject a quote straight from the whitepaper here that comments both on current issues in smart contract platforms and on Nexus’s ambitions to solve them:

Contracts hold the key to the future of distributed systems, although current implementations have issues with accuracy, performance, and ease-of-use.

These limitations make the implementation process needlessly complicated and expensive, which raises the barriers to entry and leads inevitably to obsolescence.

Considering this, [Nexus’s] advanced contract engine will be accessible through an intuitive API set designed specifically for ease-of-use and seamless integration with existing software.

As with all of the layers so far, as we move up we’re finding the technology easier and easier for developers to communicate with.

This adds an API to the stack, allowing developers to retrieve or push data from and to the Nexus ledger, establish terms for new contracts, execute transactions, and so on. Unlike some smart contract platforms such as Ethereum, where blockchain operations are primarily done by code written in a particular language, Nexus’s API framework means that developers will be able to code applications in whichever language they like.

6) Logical Layer

As these layers become more abstract, so does the whitepaper’s discussion of them, since development of the upper layers is farther in the future than the development of the foundational lower levels.

The Logical layer will be “the core application space of the software stack,” using the API layer to “interact with the network, apply new states to registers, execute register operation methods, or authorize to signature chains.”

7) Interface Layer

Finally, there’s the Interface layer. The Tritium Whitepaper says this “will improve the accessibility of the Ledger layer by interacting through API calls and operation codes … Developers can focus on providing an intuitive and responsive UI instead of the intricacies of blockchain design.”

The Interface layer will have a modular design for the easy construction of applications. “This will engender the birth of a marketplace where developers can share and sell custom modules built for the Nexus interface.”

Other projects are figuring out modular app construction tools on top of their current products, but Nexus is building with this in mind from the start to make development easier on Nexus than it is on many other contract platforms today.

Development will also be faster and more secure, thanks to the work being done via the Tritium upgrade. Building the lowest layers with the top levels in mind allows Nexus to construct everything for maximum security and efficiency, from the Lower Level Database and Lower Level Library at the Network Application level up through signature chains, twin blocks, and more. Each improvement in the foundation means a more secure, more flexible building on top.

So while there is still a ways to go in Nexus’s construction, we’ve seen impressive work on the lower levels. Tritium is implementing revolutionary new technologies into these levels, and Amine and Obsidian will ramp creative, efficient, secure solutions all the way up to the top floors.

Keep your eye on Nexus. If the Tritium Whitepaper is anything to go by, it’ll be an exciting ride.