What is a blockchain?

A blockchain is a shared, append-only ledger that many independent computers keep copies of and agree on, without any single party being in charge. Records are grouped into blocks, and each block is cryptographically linked to the one before it — forming a chain whose history is expensive to rewrite and easy to audit. The idea was first combined into a working system by the pseudonymous Satoshi Nakamoto in the 2008 Bitcoin whitepaper, which solved a long-standing problem: how to let strangers agree on a single record of who owns what, with no trusted middleman.

The core idea: an append-only, shared ledger

Most databases let an administrator insert, update, and delete rows. A blockchain deliberately gives up that flexibility. New data can only be added — never silently changed or removed — and the right to add it is governed by rules that every participant enforces independently.

Because many parties each hold a full copy and check every new block against the same rules, no single operator can quietly alter a balance or reorder history. That is the trade the design makes: it sacrifices the speed and convenience of a central database in exchange for not having to trust any single operator.

How the “chain” stays honest

Three primitives do most of the work:

• Hashing. A cryptographic hash function turns any data into a fixed-length fingerprint. Change one character and the fingerprint changes completely. Each block stores the hash of the previous block, so the blocks form a tamper-evident sequence — edit an old block and every later block’s link breaks.
• Digital signatures. Transactions are signed with a private key, proving the sender authorised them without revealing the key. Anyone can verify the signature with the matching public key.
• Consensus. Because copies are spread across the network, participants need a way to agree on which new block is the next valid one. The mechanism that coordinates this agreement — and makes cheating expensive — is the heart of every blockchain. We cover the main families in the consensus guide.

Public, private, and permissioned

“Blockchain” is not one thing. Networks differ mainly in who is allowed to participate:

• Public / permissionless (Bitcoin, Ethereum): anyone can read the ledger, submit transactions, and help secure the network. Openness is the point, and security comes from making attacks economically irrational.
• Permissioned / private: a known set of organisations runs the network. These are faster and cheaper but reintroduce trusted parties, which often means a conventional database with good access controls would serve just as well.

Choosing between them is mostly a question of who needs to trust whom. If one organisation controls the data and everyone already trusts it, a blockchain adds cost without adding much.

What blockchains are good — and bad — at

Blockchains are well suited to situations where multiple parties who don’t fully trust each other need to share one authoritative record, and where tamper-evidence and auditability matter more than raw speed. They are poorly suited to high-throughput, low-latency workloads, private data, or anything where a single trusted operator is acceptable.

Every design also runs into the blockchain trilemma — the observation that decentralisation, security, and scalability pull against each other, so improving one often costs another. Much of modern blockchain engineering, from new consensus mechanisms to layer-2 scaling, is an attempt to ease that tension. The cluster articles below go one layer deeper into each piece.