What is a hard fork, and how is it different from a soft fork?
A fork changes a blockchain's rules. How soft forks and hard forks differ, why some split a chain in two and what replay risk means for holders.
Every so often a headline announces that a blockchain is about to fork, and it can sound as if the network is about to break in two. Usually it is not. Most forks are scheduled software upgrades that the whole network adopts on the same day. A few are something else: a disagreement that ends with two chains, two coins, and two communities that each claim the original name.
What is a fork in a blockchain?
A fork is a change to the rules that a blockchain's nodes use to decide which blocks are valid. Every node runs software that checks each new block against those rules. When the software changes the rules, nodes running the new version, and nodes running the old one may no longer agree on which blocks count. How they disagree separates a soft fork from a hard fork. A brief tie between two competing blocks under unchanged rules is a different thing, closer to the reorganisations covered in our note on blockchain finality (opens in a new tab).
What is a soft fork?
A soft fork tightens the rules so that blocks valid under the new rules are still valid under the old ones. Nodes that never upgrade keep accepting the new blocks, so the chain stays in one piece. The catch is that the upgraded side must control most of the block production. If it does not, old-rule blocks that break the new rules can still win, and the change does not stick.
Bitcoin's Segregated Witness (SegWit) upgrade is the best-known soft fork. It changed how transaction signature data is stored and activated on 24 August 2017 at block 481,824.
What is a hard fork?
A hard fork loosens, or changes the rules so that some blocks valid under the new rules are invalid under the old ones. A node that does not upgrade, rejects those blocks and stops following the chain everyone else is on. A hard fork therefore needs every node operator to upgrade, or it leaves the ones who did not on a separate chain.
Most hard forks are planned upgrades with near-universal support. Ethereum has introduced most of its major changes this way. Client teams agree an activation point in advance, and test the change on public networks first. When almost everyone upgrades, the old chain is simply abandoned and no lasting second chain appears.
What happens when a hard fork splits the community?
A contentious hard fork happens when a meaningful share of users, miners, or validators refuses the new rules and keeps the old chain running. Both chains share the entire history up to the fork block. From that point they diverge, and anyone who held the coin before the split holds a balance on both.
Ethereum's split of July 2016 is the standard example. An attacker had drained ether from a large investment contract called The DAO. In response, a hard fork at block 1,920,000 on 20 July 2016, moved about 12 million ether from the affected contracts into a recovery contract for return to investors. Most of the network adopted the fork. A minority rejected the idea of rewriting a contract's outcome, and kept the unforked chain running. That chain continues today as Ethereum Classic.
Bitcoin split in a similar way on 1 August 2017. A group that wanted larger blocks to fit more transactions launched Bitcoin Cash, as a hard fork with a bigger block size limit. Bitcoin itself went the other way and activated SegWit as a soft fork a few weeks later.
What risks come with a chain split?
The main technical risk in a chain split is replay. If both chains accept the same transaction format, a transaction signed on one chain can be copied and broadcast on the other, moving coins the sender meant to leave alone. Node operators who stayed on the original Ethereum chain in 2016 were warned about exactly this. Newer forks usually add replay protection: a change that makes a signature valid on one chain only.
The practical risks sit with the services around the chain. Exchanges commonly pause deposits and withdrawals around a fork, and decide for themselves whether to list or credit any new coin. Smart contracts, oracles and bridges built for one chain may behave unpredictably on a copy of it. A split can also leave each chain with fewer miners or validators than the original had, which may make each one cheaper to attack.
Does a fork create free coins?
A contentious hard fork credits pre-split holders with a balance on both chains, but a second balance is not the same as new value. The two coins trade separately, and how the market prices either one after a split cannot be known in advance. Whether a holder can reach the new coin at all also depends on the wallet, or exchange holding it. A planned upgrade that the whole network adopts creates no second coin.
Where Northtape fits
Northtape runs no node and takes no side in any fork. It is a news desk. Stories about hard forks, network upgrades, and outages are detected by the Risk Radar protocol lens.
FAQs
Is a hard fork bad for a blockchain? Not necessarily. Most hard forks are routine scheduled upgrades that the whole network adopts, and they leave a single chain behind.
Does a soft fork require everyone to upgrade? No. Nodes that do not upgrade keep accepting the new blocks, although they cannot fully check the new rules. Most block producers do need to enforce those rules for the change to hold.
What is replay protection? Replay protection is a change made at a chain split that stops a transaction signed on one chain from being valid on the other.
Who decides whether a fork happens? No single party. Developers write and release the software, but a fork takes effect only as far as node operators, block producers, exchanges and users choose to run it.
None of this is investment advice. It describes how blockchain forks work, not whether any coin produced by one is worth holding, buying or selling.
