Bitcoin’s Blockspace War Comes for Taproot: BIP-110

Sometime on or around Saturday, August 9, the Bitcoin network will mine block 961,632. For almost every holder, nothing visible will change. Yet at that exact height a mandatory signaling window opens for BIP-110, a soft fork that would spend the following year restricting how much arbitrary data anyone can pack into a Bitcoin transaction. Barely anyone is signaling for it. That, oddly, is the reason it deserves your attention.

BIP-110 is very unlikely to activate. Miner support has sat in the low single digits for months, nowhere near the 55 percent it needs. But the proposal has become the loudest flashpoint in a dispute that has simmered since Taproot, Bitcoin’s most recent consensus change, went live in November 2021. The argument is about what Bitcoin’s block space is actually for, and Taproot sits at the center of it, because Taproot is what made today’s on-chain data boom cheap and easy. With Bitcoin trading near 64,300 dollars and a total market value around 1.29 trillion dollars (CoinGecko), the outcome of these fights is not academic.

A soft fork almost nobody is signaling

Start with the numbers, because they frame everything else. BIP-110 asks miners to flag support by setting bit 4 in the blocks they produce. To lock in early it needs 1,109 of every 2,016 blocks in a retargeting period, or 55 percent. It has been getting almost none. Independent research from Swiss digital-asset bank AMINA put miner support at roughly 2 percent by July 30 (AMINA Bank), and one late-July tally counted only about 5 exahashes per second signaling out of a network approaching 940 exahashes per second, a fraction of one percent (Bitcoin.com News).

So why does it matter? Because BIP-110 does not simply give up when miners ignore it. If it fails to lock in on support alone, a mandatory signaling window runs from block 961,632 through block 963,647, during which nodes enforcing the rule reject any block that does not signal bit 4. Lock-in would then follow at block 963,648, with activation at block 965,664, roughly September 1 (bips.dev). Those are block heights, not calendar dates, so projections drift by a day or two with hashrate; most trackers put the window opening on or around August 9. The point is that a proposal with almost no support has a built-in mechanism that can still force a confrontation.

What Taproot actually changed

To understand why BIP-110 targets Taproot so precisely, you need to remember what Taproot introduced. Activated at block 709,632 on November 14, 2021 (CoinDesk), it bundled three proposals: BIP-340 brought Schnorr signatures, BIP-341 defined the Pay-to-Taproot output type and Merkelized Alternative Script Trees, and BIP-342 added Tapscript, the scripting language for spending those outputs (bitcoin/bips). Taproot addresses begin with the bc1p prefix.

The design was elegant. A Taproot output can be spent through a simple key path that looks, on-chain, like an ordinary single-signature payment, or through a hidden script path drawn from a tree of possible conditions. Only the branch you actually use is ever revealed, so a complex multisignature or timelock arrangement can settle looking just like everyone else’s spend. Schnorr signatures let multiple keys combine into one, shrinking multisignature transactions and improving privacy. It was, by consensus of almost everyone, a clean upgrade, and it activated smoothly under a process called Speedy Trial with more than 90 percent of miners on board. Nearly five years later it remains the last change to Bitcoin’s consensus rules, the longest quiet stretch in the network’s history.

How Taproot accidentally became a data layer

Here is the twist nobody planned for. Taproot’s flexible witness structure, combined with the witness discount inherited from SegWit, made it unusually cheap to attach large blobs of arbitrary data to a Bitcoin transaction. Developers noticed. In January 2023, Casey Rodarmor launched Ordinals, a scheme for numbering individual satoshis and inscribing content onto them by tucking images, text, and other files inside the Taproot witness, with room for payloads up to roughly four megabytes per transaction. Inscriptions exploded. Rodarmor followed with Runes in April 2024, on halving day, giving Bitcoin a native fungible-token standard, and the BRC-20 experiment layered token accounting on top of inscriptions.

The scale is real. Taproot’s share of transactions peaked around 42 percent in 2024 at the height of the inscription frenzy, then settled back as speculative demand cooled, and tens of millions of inscriptions now sit permanently in the chain (KuCoin research). To one camp this is Bitcoin doing new and useful things. To another it is spam, a diversion of a monetary settlement network into a picture-storage service that every full node must keep forever. BIP-110 is the second camp’s attempt to legislate the first camp out of existence, at least for a year.

The fee dynamics are what make the argument visceral. During the hottest inscription and token-minting waves, ordinary users watched the cost of a simple payment jump as data transactions bid up limited block space, and the mempool, the queue of unconfirmed transactions, stayed swollen for days. Miners, by contrast, banked the extra fees, which is why the economic incentive to keep data flowing runs directly against the ideological push to shut it off. That divide, users against miners, purists against builders, is the fault line every one of these proposals runs along.

The OP_RETURN detour: Bitcoin Core v30 and the Knots revolt

The current standoff did not start with BIP-110. It started with a relay-policy change. In October 2025, Bitcoin Core v30 raised the default limit on OP_RETURN data, the field designed for small amounts of provably unspendable metadata, from 83 bytes to 100,000 bytes, and began allowing multiple such outputs per transaction (CoinDesk, Unchained). The reasoning was pragmatic: if people are going to embed data anyway, better to channel it into an OP_RETURN output that stays out of the UTXO set than to bloat the witness or, worse, create fake spendable outputs that nodes must track forever.

Not everyone agreed. Longtime developer Luke Dashjr, who maintains the rival Bitcoin Knots client and co-founded the OCEAN mining pool, called the change malicious code that he warned would kill Bitcoin almost immediately, and critics including Nick Szabo cautioned that looser limits invite spam, node-cost inflation, and permanent storage of illicit material. The backlash showed up in the node count. Bitcoin Knots, which lets operators set their own stricter data limits, climbed to more than a fifth of listening nodes, around 23 percent, up from roughly 4 percent and only a few hundred nodes at the start of 2025 (OAK Research). This is the Core-versus-Knots split, and BIP-110 is the version of that argument that tries to reach the consensus layer, where policy stops being a setting each operator chooses and becomes a rule the whole network enforces.

Inside BIP-110: what Reduced Data would ban

BIP-110, formally the Reduced Data Temporary Softfork, was authored by Dathon Ohm, assigned in December 2025, and reached Complete status in June 2026 (bips.dev). It is not a vague statement of principle. It is a set of seven concrete consensus rules, and several of them cut directly into the machinery Taproot introduced.

Read that list again and the irony becomes hard to miss. A proposal pitched as saving Bitcoin from data spam does much of its work by hobbling Taproot’s own scripting surface: the annex, the control block, conditional logic inside Tapscript. The grandfather clause softens the blow by exempting coins that already exist, so nobody’s current holdings become unspendable. And the whole thing is temporary by design. The restrictions would run for 52,416 blocks, about one year, after which the rule set expires and enforcement stops. It is less a permanent constitutional amendment than a one-year emergency measure, which is part of why supporters frame it as low-risk and opponents call it a dangerous precedent dressed up as a trial.

The signaling gambit and the chain-split risk

The activation method is where BIP-110 gets genuinely controversial. It uses a modified version of the BIP9 signaling process, but with the lock-in threshold set at 55 percent rather than the roughly 95 percent that governed past soft forks. A lower bar sounds easier, and that is exactly the concern. Taproot activated with the comfortable backing of well over 90 percent of hashrate, which meant essentially no risk that a meaningful minority would keep mining under the old rules. A rule that can lock in on 55 percent, backed by a mandatory signaling window that rejects non-signaling blocks, invites the opposite scenario.

Picture the current state: support near 2 percent, and a window that tells enforcing nodes to reject any block that fails to signal. If a small minority of nodes enforces BIP-110 while the overwhelming majority of hashrate keeps producing perfectly valid, non-signaling blocks under the existing rules, you get two chains running at once, at least for a while. That means replay risk, where a transaction on one chain can be rebroadcast on the other, paused deposits at exchanges, and general operational chaos regardless of which chain ultimately wins. Casa co-founder Jameson Lopp called the proposal reckless and warned that the low 55 percent threshold greatly increases the chances of a chain split (Bitcoin.com News). The mechanism designed to make BIP-110 hard to ignore is the same mechanism that makes it hazardous.

Bitcoin has been here before, which is part of why the warnings carry weight. The scaling wars of 2017 ended with a user-activated soft fork campaign that pressured miners into activating SegWit, and with a faction breaking away to create Bitcoin Cash rather than accept the result. Those events proved two things at once: that a determined minority can force a protocol question onto the table, and that when the opposing camps are both large, the outcome is a permanent split rather than a clean resolution. BIP-110 borrows the coercive spirit of that era, a rule that punishes non-signaling blocks, without the broad backing that ultimately made SegWit stick. High coercion paired with low support is precisely the combination that unsettles engineers.

What the builders are saying

The technical establishment has been blunt. Adam Back, the Blockstream chief executive whose Hashcash work Satoshi cited in the white paper, dismissed BIP-110 in plain terms: “It really doesn’t work, breaks multiple things, doesn’t have tech nor ecosystem consensus. This is not the way” (Bitcoin.com News). Lopp’s chain-split warning lands in the same place from a security angle. Developer Jimmy Song, by contrast, has said he does not know enough about the downstream consequences to take a firm side, a reminder that even seasoned Bitcoiners find the tradeoffs genuinely hard.

What makes the debate confusing is that the people most worried about data spam are not the ones pushing BIP-110. Dashjr and the Knots camp want data curbed, but they have generally pursued that through client policy and node choice, not a 55 percent consensus fork. Most of Core and most of the industry oppose the idea of consensus-enforced content rules on principle, preferring to let fee markets and relay policy handle the problem. Supporters of BIP-110 exist, and have gathered grassroots backing, but they are a small and loosely organized group up against a near-unanimous wall of protocol engineers. That is the real story of the signaling numbers: this is not a close vote that might tip, it is a proposal the technical community has already rejected, now testing whether a minority can force the issue anyway.

The messier cousin: Paul Sztorc’s eCash hard fork

BIP-110 is not the only fork event on the August calendar, and the other one is arguably more disruptive. Longtime developer Paul Sztorc, founder of LayerTwo Labs, has scheduled a hard fork he calls eCash for around block 964,000, roughly August 21. It would copy Bitcoin’s code, apply a one-time difficulty reset, airdrop eCash one-for-one to every Bitcoin holder, and switch on seven Drivechain-style sidechains for decentralized exchange, privacy, prediction markets, non-fungible tokens, identity, and quantum resistance (CoinDesk).

The controversy is not the airdrop. It is that eCash would, on the new chain, reassign hundreds of thousands of dormant Satoshi-linked coins, an act CoinDesk reported many in the community are simply calling theft. Fidelity Digital Assets has objected to the precedent of rewriting balances at addresses their owners never touched. Crucially, this is an opt-in hard fork that creates a separate asset, not a change to Bitcoin itself, so it does not split the main chain the way a contested soft fork could. Its danger is different: it forces custodians, auditors, corporate treasuries, and fund sponsors to decide, on a deadline, how to handle an airdropped coin they never asked for.

What a chain split means for US holders, ETFs, and the SEC

Here is a point that trips up newcomers: no regulator approves or blocks a Bitcoin fork. Consensus rules are decided by the people who run nodes, mine blocks, and hold coins, not by Washington. The Securities and Exchange Commission has no button to press on BIP-110, and neither does any other agency. That is by design, and it is one of the things that makes Bitcoin what it is. If you are tracking the year’s supervisory milestones in our rundown of the deadlines that decide 2026, note that none of them touch the protocol layer.

Regulation bites afterward, at the edges where Bitcoin meets the traditional system. A split or a large airdrop forces exchanges and custodians to decide whether to support a new chain, to implement replay protection, and possibly to halt deposits while they sort it out. Sponsors of the spot Bitcoin ETFs that the SEC cleared generally retain discretion over forked assets and are under no obligation to distribute them to shareholders, and some mandates prohibit holding forked coins at all. On the tax side, the Internal Revenue Service has treated coins received through a hard fork and airdrop as ordinary income at the moment of receipt, valued at fair market price, since its 2019 guidance. Self-custody holders keep both sides of any split automatically, but they also inherit the job of managing replay risk themselves. In other words, the consensus fight is permissionless, but its consequences run straight into custody agreements, fund prospectuses, and tax returns.

Covenants: the upgrades that cannot get consensus either

BIP-110 is the loud negative case, a change most people do not want. The mirror image is the set of changes many people do want but still cannot ship. Covenants, which let a coin carry rules about how it can be spent next, are the clearest example. Two candidates dominate the discussion, and neither has a path to activation.

OP_CTV is the conservative option, framed as a safe building block for vaults; its activation client is out, with a minimum activation height near May 2027, but no lock-in in sight (AMINA Bank). OP_CAT is broader and more powerful, and has seen far more experimentation on test networks, but has no activation mechanism at all. The reason both stall is the same reason BIP-110 will fail and Taproot succeeded: Bitcoin changes only on overwhelming agreement, and there is no CEO to break a tie. Blockstream researcher Andrew Poelstra has pushed back on what he calls ossificationists, arguing Bitcoin is already evolving and that developers need to argue passionately and correctly for beneficial changes (Bitcoin Magazine). Ethereum, by contrast, ships this kind of programmability through a Foundation-led roadmap; its EIP-7702 upgrade turned ordinary accounts into smart accounts in one coordinated hard fork, a governance model Bitcoin deliberately does not have.

The quantum clock is the real deadline

Step back from the data war and there is a larger reason people keep trying to touch Bitcoin’s consensus rules: the clock on its cryptography. In late March 2026, Google Quantum AI published research suggesting that breaking the elliptic-curve signatures Bitcoin relies on might take fewer than 500,000 physical qubits and somewhere near 1,200 to 1,450 logical qubits, with an attack measured in minutes rather than the millions of qubits once assumed (CoinDesk, Google Research). Roughly 6.9 million BTC, close to a third of the supply, already sit at addresses whose public key is exposed on-chain.

Taproot has an uncomfortable role here too. A Pay-to-Taproot output reveals its full public key by default, so every Taproot coin is quantum-exposed from the moment it is created, not just when it is spent. The same output type sitting at the heart of the data war is also the most quantum-visible. The proposed fix is BIP-360, or Pay-to-Merkle-Root, authored by Hunter Beast, Ethan Heilman, and Isabel Foxen Duke, which restructures Taproot outputs to remove the vulnerable key-path spend while preserving the scripting that powers Lightning and other layers (bip360.org); a companion proposal, BIP-361, sketches a phased sunset of legacy signatures. Work is moving from paper to code, with firms like BTQ Technologies demonstrating a BIP-360 implementation on a dedicated quantum testnet (The Quantum Insider). The proof-of-work that secures Bitcoin today, the subject of our look at hashrate growth and the price of security, does nothing to protect an exposed public key from a future quantum attacker; only a signature change can. Views on urgency vary. Ethereum Foundation researcher Justin Drake said his confidence in a quantum breakthrough this decade has shot up significantly, while Binance founder Changpeng Zhao struck a calmer note, arguing that all crypto has to do is upgrade to quantum-resistant algorithms, so no need to panic (CoinDesk). Both agree the migration itself, coordinating a signature swap across a decentralized network, is the hard part.

Taproot’s quieter wins: multisig, Lightning, and stablecoins

It would be easy to read all this as Taproot being a mistake. It is not. Away from the inscription noise, Taproot is doing exactly the work it was designed for. Schnorr key aggregation, standardized as MuSig2 in BIP-327, lets several signers combine into a single on-chain signature. A field report from custodian BitGo measured a key-path MuSig2 input at 57.5 virtual bytes against 104.5 for a native SegWit multisig, roughly 45 percent smaller, while making a multisignature wallet look identical to a single-signature one on-chain (Bitcoin Optech). That privacy and efficiency is why Taproot features increasingly show up in modern multisig best practices for serious self-custody.

The applications keep coming. Simple Taproot Channels are now in production on Lightning. Taproot Assets, the protocol from Lightning Labs, carried a wrapped version of Tether’s USDT onto Lightning in March 2026, giving Bitcoin’s payment layer a dollar rail (btc.network), with the RGB protocol pursuing a competing design. And Citrea, the first zero-knowledge rollup built directly on Bitcoin, reached mainnet in January 2026 leaning on Taproot and BitVM to settle back to the base chain (The Block). The uncomfortable fact for BIP-110 is that the same Taproot machinery it would restrict, the annexes, the control blocks, the rich Tapscript, is the machinery these layers are built on. Blunt tools cut both ways.

There is also a quieter, organic adoption story that has nothing to do with inscriptions. A growing share of wallets now generate Taproot addresses by default, major exchanges have added bc1p withdrawal support, and every Lightning channel built on the newer Taproot format inherits the privacy and fee benefits without its users ever thinking about it. Strip those capabilities to fight spam and you also slow the parts of Bitcoin that are working exactly as designed. That is the objection to BIP-110 compressed into a single line: it aims at a symptom with a tool broad enough to wound the patient.

Why this fight keeps coming back

Zoom out and the recurring pattern is really about economics. Block space is scarce and, as Bitcoin’s value climbs, increasingly valuable. Every block is an auction, and data users bid for the same room as payment users. When inscriptions run hot, fees spike and payment users complain; when they cool, miners lose a revenue stream they had come to rely on. Miners are also the ones who signal on soft forks, so their incentives, some ideological, most economic, shape whether any of these proposals moves. Restriction-minded hashpower is concentrated in pools like Dashjr’s OCEAN, but even there the share is small, which is precisely why BIP-110’s signaling numbers are what they are. Fees, hashrate, and ideology all feed into how miners vote, and right now none of the three points toward BIP-110.

Underneath the economics is governance. Bitcoin has no authority that can mandate a change, which is a feature: it is why the network resisted the block-size push of 2017 and why it has stayed stable since Taproot. The flip side is that genuinely good ideas, covenants and quantum resistance among them, move at the same glacial pace as bad ones. BIP-110 will almost certainly expire without activating. But the question it poses, what belongs in Bitcoin’s blocks and who gets to decide, is structural, and it will be back the next time inscriptions run hot or fees run cold.

What to watch next

• Block 961,632, on or around August 9: does any pool actually signal bit 4 once the mandatory window opens, or does support stay near zero?
• The Bitcoin Knots node share: whether the data-restriction camp keeps growing past its current level near 23 percent, or plateaus after the BIP-110 push fizzles.
• Block 964,000, around August 21: how custodians, auditors, and ETF sponsors respond to Paul Sztorc’s eCash airdrop and its disputed reassignment of dormant coins.
• OP_CTV’s activation window into May 2027, and whether any covenant proposal builds the consensus that BIP-110 lacks.
• The pace of BIP-360 and BIP-361, the real long-term deadline, as quantum research keeps advancing.

The likeliest outcome this month is anticlimax: BIP-110 fails to lock in, the mandatory window passes without a lasting split, and eCash launches as a small breakaway chain that most of the industry ignores. But even an anticlimax is informative. It would confirm, once more, that Bitcoin’s rules do not change without something close to unanimity, that Taproot’s flexibility is both its gift and its lightning rod, and that the quiet stretch since November 2021 is less a truce than a standoff nobody has found the votes to break.

Frequently Asked Questions

What is BIP-110 and when could it activate?

BIP-110, the Reduced Data Temporary Softfork proposed by Dathon Ohm, is a one-year soft fork that would limit arbitrary data in Bitcoin transactions. Its mandatory signaling window opens around August 9, 2026 at block 961,632; it would lock in at block 963,648 and activate near September 1 at block 965,664, but only if it clears 55 percent miner signaling. Support has hovered near 2 percent, so activation is very unlikely.

How is BIP-110 connected to Taproot?

Taproot’s flexible scripting and the witness discount made it cheap to embed images, text, and tokens in Bitcoin transactions, which fueled Ordinals, inscriptions, and Runes. BIP-110 targets exactly those features, invalidating Taproot annexes, capping control blocks, and banning OP_SUCCESS, OP_IF, and OP_NOTIF inside Tapscript, alongside tighter limits on OP_RETURN and witness data.

Could BIP-110 cause a Bitcoin chain split?

Temporarily, yes. The mandatory signaling window means enforcing nodes reject blocks that do not signal, so if a small minority enforces the rule against a large non-signaling majority, two chains can coexist for a period, bringing replay risk and paused exchange deposits. Casa co-founder Jameson Lopp has called the 55 percent threshold reckless for precisely this reason.

What is the eCash hard fork planned for August 2026?

It is a separate, opt-in hard fork led by developer Paul Sztorc, targeted near block 964,000 around August 21, that would airdrop eCash one-for-one to Bitcoin holders and enable Drivechain-style sidechains. It is controversial because it would reassign dormant Satoshi-linked coins on the new chain; Fidelity Digital Assets has objected to rewriting balances at addresses their owners never controlled. It is distinct from BIP-110.

Does the SEC have to approve a Bitcoin soft fork?

No. Consensus changes are decided by node operators, miners, and users, not by any regulator, so the SEC neither approves nor blocks BIP-110. Regulation bites afterward: a split or airdrop forces exchanges, custodians, and spot Bitcoin ETF sponsors to decide what to support, and the IRS has treated forked or airdropped coins as ordinary income at receipt.

By the HOGE Wire Bitcoin desk, covering Bitcoin protocol development, mining, and market structure.