MODULE 1
Decentralization 101
What a distributed ledger is and why it removes the need to trust an intermediary.
DLT fundamentals: blocks and chains of information
A ledger is a list of facts in order: who paid whom, who signed what, who has access to where. Your bank keeps one. Your phone company keeps another. In both cases the list lives on their computers, and you have to take its word for it.
A distributed ledger — DLT — is that same list held at once on many independent computers. Nobody holds the original, because everybody does. Changing the past no longer means convincing one company: it means convincing the majority of a network whose members do not even know each other.
Facts are not stored loose but grouped into blocks. And every block carries inside it the fingerprint — the *hash* — of the block before it: a short number that sums up its entire contents and changes completely if a single letter is touched.
That is where the chain comes from, and where its useful property comes from. If someone alters an old record, that block's fingerprint stops adding up, and with it the next one's, and the next one's. Tampering cannot be hidden. That is the whole idea, and there is no more to it.
One frequent confusion is worth dismantling: a distributed ledger is not a cryptocurrency. The chain is the structure; the currency was merely its first famous use. At Xatoxi, every channel and every vault keeps a replicated ledger of its own, and what it records is not coins but messages, permissions and files.
Nodes: the backbone of every decentralized network
A node is a device that keeps a copy of the ledger and checks whatever reaches it. It is not a company's server: it is a computer, a phone or any machine somebody contributes to the network.
What is gained by having no centre is exactly what is lost in convenience. There is no place to switch off, no directory to seize, no company to demand a user's records from. Nor is there anybody who can give you back a key you lost.
The practical problem is how they find one another with no central directory. The answer is a distributed hash table (DHT): each node knows a few others and knows who to ask about the rest, much as you reach an address in an unfamiliar city by asking at one corner after another.
The network's strength is proportional to the number of nodes, and that has a consequence worth saying out loud: a young decentralized network is fragile. Not by design, but by size. Every device that joins makes attacking it more expensive.
At Xatoxi every node is a verified person, and that is not an identity ornament. It is what stops somebody from standing up ten thousand fake nodes and taking the network's majority — the attack known in the literature as *Sybil*.
Cryptography: symmetric vs. asymmetric, RSA/ECC and their limits
Symmetric cryptography uses a single key to lock and to unlock, like a house key. It is fast, and it is still what encrypts the bulk of the data; AES is its most widespread example. Its problem is old and simple: how do you get the key to the other person without anyone copying it on the way.
Asymmetric cryptography solved that in the seventies with a pair of keys. A public one, handed out without fear, which can only lock. And a private one, which never leaves your device and is the only thing that unlocks. Anyone can write to you; only you can read.
The two systems holding up today's internet are RSA and ECC. RSA rests on multiplying two enormous primes being easy and undoing that multiplication being ruinously expensive. ECC rests on a related problem over elliptic curves, and reaches the same security with far shorter keys — which is why it is the one phones carry.
And here is the limit to understand: neither of them is impossible to break. Both are expensive to break. Their security is not a law of mathematics, it is a sum of time and electricity — and that sum is done by a classical computer. Change the kind of computer and the sum changes: that is what module 2 is about.
One last detail, which separates a serious channel from a decorative one: continuous key renewal, or *forward secrecy*. If keys are renewed often, stealing today's does not open last month's conversations. Xatoxi's secure channel works this way.
Peer consensus, without mining or wasted energy
If everybody keeps a copy, who decides the order of the facts? That is the consensus problem, and it is the heart of any decentralized network. Without an answer, two honest nodes can end up with two different versions of the truth and no way to tell which one counts.
The best-known answer is mining, or *proof of work*: nodes compete at solving a deliberately useless problem, and whoever solves it earns the right to write the next block. It works, and it has been working for over a decade. Its price is literal: security is paid for in electricity, and rises with it.
There is another family of answers, older than cryptocurrencies: peer consensus. Nodes do not compete, they agree. They announce what they have seen, they vote, and they accept an order once a sufficient majority coincides. These protocols tolerate a share of the nodes lying or failing, which is why they are called *Byzantine fault tolerant*.
The difference in cost is what decides which devices can take part. Peer consensus spends messages, not watts: talking is the expensive part, not computing. That is what lets a node fit in a phone rather than in an industrial unit.
Xatoxi works this way: every channel or vault keeps a replicated ledger of its own, validated by peer consensus, with no mining and no wasted energy.
Questions, or want to go deeper?
This module covers the fundamentals. If you work at an organization asking these questions in earnest — what to keep, where, and with what guarantees — we are listening.
Talk to the team