Blockchain types and how networks agree

Why it matters

The word blockchain conceals who may write, who validates, how failure is resolved and what makes a history expensive to replace.

Public, private and permissioned

A public network normally lets anyone read state and submit transactions, while participation in validation depends on protocol rules. A permissioned ledger restricts writers or validators to identified organizations.

Permissioning can improve privacy, accountability and throughput for institutional workflows. It also concentrates admission and governance, so it should not borrow the trust claims of an open network without qualification.

Layers, sidechains and application chains

A layer-one network provides its own consensus and settlement. A layer-two system executes or batches activity while relying on another network for some combination of data, dispute resolution or final settlement.

Sidechains and application-specific chains may use separate validator sets. Their bridge and exit assumptions matter as much as their headline speed.

Proof of work and proof of stake

Proof of work makes proposing history costly through computation and energy. Proof of stake selects validators using committed assets and can penalize protocol-detectable misbehavior.

Work is externally expended; stake is internally accounted. Their concentration pressures, recovery paths, finality and attack economics differ, and implementations within each category differ too.

Compare the full security model

Ask who can validate, what clients enforce, how finality is defined, where users obtain data, how upgrades occur and what happens during a split.

Throughput alone is not security. A fast network may move coordination into a smaller validator set or more demanding hardware; a slower one may impose other costs on users and operators.