What is Soft Fork?
How it works
A soft fork tightens the protocol's validation rules, making previously valid blocks or transactions invalid under the new rules. Because the new rules are a subset of the old rules, any block that satisfies the new rules also satisfies the old ones. This means nodes that have not upgraded can still accept and relay blocks created by upgraded nodes, as long as those blocks do not violate the old rules. However, non-upgraded nodes may not enforce the new rules, so they could temporarily accept invalid blocks if a majority of miners do not follow the fork. To activate a soft fork, a supermajority of miners (often 95%) must signal readiness via block version bits, after which a lock-in period begins and the new rules take effect.
A well-known example is Bitcoin's Segregated Witness (SegWit) upgrade, activated in 2017. SegWit changed the transaction format to separate signature data (witness) from the core transaction data. Old nodes see a SegWit transaction as a valid, but empty, part of the block (since the witness data is placed in a new field they ignore). This allowed SegWit to increase block capacity and fix transaction malleability without requiring every node to upgrade. The activation used a BIP 9 signaling mechanism: miners included a specific bit in their blocks to vote, and after reaching 95% threshold during a retarget period, the soft fork locked in and later activated.
Other Bitcoin soft forks include BIP 66 (strict DER signature encoding) and BIP 65 (CHECKLOCKTIMEVERIFY). These upgrades demonstrate how soft forks can introduce new features or security improvements while preserving backward compatibility. Because old nodes can still participate, soft forks reduce the risk of a chain split, but they rely on economic majority (miners/validators) to actually enforce the new rules. If a majority adopts the soft fork, non-upgraded nodes eventually see only blocks that obey the new rules, making them effectively obsolete without forcing a hard split.
Why it matters
Soft forks enable blockchain networks to upgrade securely without forcing every participant to update immediately. They maintain a single chain, avoiding the community and economic splits that hard forks can cause. This makes soft forks the preferred method for introducing non-controversial changes, such as bug fixes, efficiency improvements, or adding new opcodes, because they allow gradual adoption while keeping the network unified.
Real-world examples
Two canonical Bitcoin soft forks are BIP 66 (2015) which enforced strict DER signature format, and Segregated Witness (SegWit, activated 2017) which separated signature data to increase block capacity and fix transaction malleability. Both were activated via miner signaling and retained backward compatibility, meaning non-upgraded nodes could still process blocks.
FAQ
What is the difference between a soft fork and a hard fork?
A soft fork tightens rules (backward-compatible), so old nodes can still accept new blocks; a hard fork changes rules so old nodes reject new blocks, creating a permanent chain split unless all nodes upgrade.
How is a soft fork activated in Bitcoin?
Typically through BIP 9 version bits signaling: miners set a specific bit in block headers to vote. After a threshold (e.g. 95% in a retarget period) is reached, the fork locks in and activates on a predetermined block height.
Can a soft fork be rejected by the network?
Yes, if miners do not reach the activation threshold, the soft fork does not activate. Also, even after activation, if a majority of hash power does not follow the new rules, the network may temporarily diverge until a clearer majority emerges.
Related terms
Go deeper
Browse the complete crypto glossary to explore related terms and concepts.
Browse Glossary