Disclaimer: This article is for informational purposes only and does not constitute financial or technical advice. Always conduct your own research before making decisions based on blockchain technology.
The Core Concept
On July 21, 2017, Bitcoin achieved one of the most significant technical milestones in its eight-year history: the lock-in of BIP 91, a miner-activated soft fork proposal that paved the way for Segregated Witness activation. To understand why this matters, one must first grasp the fundamental challenge that Bitcoin faced. The network had been stuck in a scaling debate for nearly three years, with transaction fees rising and confirmation times lengthening as adoption outpaced block capacity. SegWit, originally proposed in late 2015 by Bitcoin Core developer Pieter Wuille, offered a solution that would effectively increase block capacity by approximately 1.8 times while also fixing transaction malleability — a critical bug that had plagued Bitcoin since its inception.
The problem was not technical but political. SegWit required 95 percent of miners to signal support via BIP 9 version bits signaling, and that threshold had proven unreachable. Miners, many of whom operated under the banner of the large-block movement, wanted a block size increase instead of or in addition to SegWit. The deadlock threatened to split the network into competing chains. BIP 91, authored by James Hilliard, offered an elegant compromise: it lowered the activation threshold to 80 percent of hash power over a 336-block window and forced SegWit activation by rejecting non-SegWit-signaling blocks.
How It Works Under the Hood
BIP 91 operates through a mechanism known as a miner-activated soft fork, or MASF. Unlike BIP 9, which relied on miners voluntarily embedding version bits in their block headers, BIP 91 introduced a direct enforcement layer. Here is how the signaling worked in practice. Each Bitcoin block contains a version field in its 80-byte header. Miners supporting BIP 91 set specific bits in this field to indicate their intent. The proposal required that 269 out of 336 consecutive blocks — representing 80 percent of approximately 2.3 days of mining — signal support for the mechanism.
Once the 269-block threshold was reached, BIP 91 entered a 144-block grace period — roughly one day — before enforcement began. During enforcement, any miner producing a block that did not also signal for SegWit via BIP 141 would have their block rejected by the BIP 91-compliant majority. This economic incentive structure was crucial: non-compliant miners would waste electricity and computational resources producing blocks that the network would simply ignore.
The technical elegance of BIP 91 lay in its indirect approach. Rather than modifying SegWit itself, it created a parallel enforcement mechanism that leveraged Bitcoin’s existing consensus rules. By forcing miners to signal for BIP 141, BIP 91 essentially shortcut the stalled BIP 9 process. The result was that SegWit would activate within the original BIP 141 timeline — approximately two weeks after 95 percent of blocks signaled — but the path to reaching that 95 percent was now guaranteed by the 80 percent BIP 91 majority.
Real-World Applications
The activation of SegWit via BIP 91 had immediate practical implications for Bitcoin users and businesses. Transaction malleability, the bug that allowed third parties to modify transaction IDs without changing the actual payment, had been a major obstacle for Lightning Network development. With SegWit’s fix for malleability, payment channel implementation became far more reliable, clearing the path for Lightning Network’s deployment on the mainnet in the months that followed.
SegWit also introduced native segwit addresses — the now-familiar addresses starting with “bc1” — which offered lower transaction fees due to their smaller data footprint. Exchanges and wallet providers began preparing for SegWit integration almost immediately. Bitfinex, one of the largest exchanges at the time, announced plans to support SegWit withdrawals. Hardware wallet makers like Ledger and Trezor also rushed to add SegWit compatibility to their firmware.
Perhaps most significantly for everyday users, SegWit’s effective block size increase from 1 MB to approximately 1.8 MB meant that more transactions could fit in each block. With Bitcoin processing around 250,000 transactions per day in mid-2017 and fees regularly exceeding $2, any capacity increase was welcome. The fee reduction would not be immediate — it required widespread SegWit adoption by wallets and services — but the foundation was now in place.
Scalability and Limitations
Despite its significance, BIP 91 was not a silver bullet for Bitcoin’s scalability challenges. The 1.8x effective block size increase that SegWit provided was modest compared to the exponential growth in demand that 2017 would bring. By December 2017, Bitcoin transaction fees would regularly exceed $30, demonstrating that SegWit alone could not keep pace with the flood of new users entering the ecosystem.
The BIP 91 compromise also came with strings attached. The New York Agreement that underpinned the miner consensus included a commitment to a subsequent 2x hard fork that would double Bitcoin’s base block size to 2 MB. This “SegWit2x” proposal was highly controversial and would ultimately be cancelled in November 2017 after failing to gain sufficient community support. The cancellation led to the Bitcoin Cash hard fork on August 1, 2017, when a group of miners and developers who were dissatisfied with the SegWit-only approach created an alternative chain with 8 MB blocks.
There were also technical limitations to the BIP 91 approach itself. By relying on miner enforcement rather than full node consensus, BIP 91 concentrated power in the hands of mining pools. Critics argued that this undermined Bitcoin’s decentralization principles. If miners could force protocol changes through 80 percent hash power alone, what prevented them from making changes that benefited themselves at the expense of users? The debate highlighted the ongoing tension between miner influence and user sovereignty in Bitcoin governance.
The Future Horizon
Looking beyond July 2017, the BIP 91 lock-in represented a watershed moment for Bitcoin’s technical evolution. The activation of SegWit opened the door to a new generation of Layer 2 scaling solutions, with Lightning Network being the most prominent. Lightning would eventually enable near-instant, low-cost Bitcoin transactions by routing payments through off-chain payment channels, leveraging SegWit’s malleability fix to ensure channel security.
The success of BIP 91 also demonstrated that Bitcoin’s governance, while messy and contentious, could produce results. The network had faced an existential threat — a potential chain split that could have destroyed billions in value — and navigated it through a combination of technical ingenuity, economic incentives, and community negotiation. This precedent would be tested repeatedly in the years to come as Bitcoin continued to scale.
For the developers, miners, and users who had lived through the scaling debate, July 21, 2017, was a day of cautious celebration. SegWit was finally within reach, the network had avoided a catastrophic split, and the path forward was clear. But the challenges ahead — Lightning Network implementation, the SegWit2x debate, and the relentless pressure of growing demand — promised that Bitcoin’s technical evolution was far from over. The blockchain had proven its resilience. Now it had to prove it could scale.
transaction malleability was the real killer bug. segwit fixing that enabled lightning, which changed everything
fixing malleability also enabled atomic swaps and better cross chain bridges. segwit was over 100 lines of code that unlocked thousands
without malleability fixes there is no lightning, no atomic swaps, no taproot. segwit was the foundation for everything that came after
Three years of debate over a capacity increase that turned out to be less than 2x. Meanwhile fees were hitting $50.
2x capacity increase sounds small but it enabled lightning which handles orders of magnitude more. the real upgrade was second layer
fees hit $50 and the debate was about whether 1.8x was enough. looking back the whole thing seems absurd when lightning handles millions of txs for fractions of a cent
the 95% BIP 9 threshold was unreachable because a few pools held the network hostage. hilliard dropping it to 80% was the move
bitmain controlled so much hashrate in 2017 that they could single handedly block 95% signaling. ASIC centralization was the real story behind the scaling war
hilliards BIP 91 was the pragmatic compromise that nobody wanted to admit was necessary. dropping from 95% to 80% signaling was what actually broke the deadlock