Did BIP-110 Change Bitcoin’s Inscription Rules—or Split Off a Different Chain?

Did BIP-110 change Bitcoin’s inscription rules on the main network? No. On August 8, 2026, nodes enforcing the proposal’s mandatory-signaling rule rejected a block accepted by the wider network and followed a minority branch. That split did not put BIP-110’s proposed transaction limits into effect on Bitcoin’s main chain. The distinction is between the rules a particular node enforces, the branch it follows, and the rules adopted by the network’s main chain.1 4


What the community was arguing about

BIP-110, formally titled “Reduced Data Temporary Softfork,” came out of a dispute over whether Bitcoin should carry arbitrary data as well as monetary transactions. Its author argued that consensus rules should temporarily limit data storage and refocus development on Bitcoin as money. In the proposal’s original public discussion, Jameson Lopp objected that the technical questions were contentious and that the proposed moral or legal framework was unacceptable to him. That exchange shows a real disagreement among participants; it is not a survey of Bitcoin users or developers.2

After the split, the argument moved from principle to the meaning of “activation.” In an August 9 thread about closing the BIP, Greg Tonoski, who said he operated a BIP-110 node, argued that the proposal had activated because its branch produced signaling blocks. Other participants replied that blocks on a minority branch did not amount to adoption by the Bitcoin network. The same word was being used for two events: a node’s own software enforcing a branch, and the proposed rules becoming the rules followed by Bitcoin’s main chain. The thread documents that disagreement; it does not establish a representative view of Bitcoin users or developers.3


What the proposal would have changed

BIP-110 was not a single switch labeled “ban Ordinals.” The specification proposed seven temporary consensus rules that applied to particular transaction constructions. For example, it allowed OP_RETURN output scripts up to 83 bytes and invalidated certain pushed or witness data above 256 bytes; it also restricted several witness-version and Taproot-script uses. Those details matter because the word “inscription” covers many kinds of content and transaction patterns. A label alone does not establish whether a specific transaction would have violated the proposed rules; the construction and the validating node’s rules matter.1

The proposal used a modified signaling process. It set a threshold of 55%—1,109 of 2,016 blocks—and included a mandatory-signaling window. If its deployment had progressed as written, the additional rules would have taken effect later and remained active for about a year before expiring. The window that triggered the split began at block 961,632: BIP-110 nodes rejected a block without the required signal, while the broader network accepted it. The two groups then followed different histories.1

A contemporaneous report from The Block said the BIP-110 branch had produced two blocks by 6 p.m. Eastern on August 8, while the main chain was at block 961,640. It also reported that just 2.53% of blocks had signaled support in the preceding two-week period, far below the proposal’s threshold. These figures describe the conditions during the initial split; they do not mean the alternative branch could never produce another block.4

A later archive from bitcoin-forks.org records the SHA-256d BIP-110 branch through block 961,639—eight blocks after the last shared block, 961,631. The archive explains its vantage points: a Bitcoin Knots node configured to enforce BIP-110 followed that branch, while its ordinary Bitcoin Core node followed the majority chain. It also says its record ends at 961,639 because the observer did not upgrade to follow the later BLAKE2b proof-of-work rules. That qualification is important. The archive documents what its nodes observed; it does not establish that every later network using BIP-110-related rules disappeared.5

The BIP repository changed the proposal’s status to Closed on August 9, citing a chain split and stalled mining.1 That is a status in the BIP process, not a vote by every Bitcoin user and not a rule change applied retroactively to the main chain. Bitcoin Knots’ September 2 release 29.4.1 included BLAKE2b proof of work and minor changes to BIP-110’s activation and expiry rules.6 A related pull request documents the BLAKE2b transition at block 961,640, after block 961,639, the last SHA-256d block; its review records a check of the first BLAKE2b block’s hash against the release checkpoint.7 This documents a later continuation on an incompatible minority branch, not Bitcoin mainnet adopting BIP-110.


The strongest counterargument

Supporters can reasonably say that the split was not imaginary. Nodes enforcing BIP-110 did reject blocks that their software considered invalid and follow a different branch. Their mandatory-signaling rule was in force locally; that did not mean the proposal’s transaction restrictions had reached the later ACTIVE stage in its deployment plan.1 A minority branch can also become economically important if enough miners, wallets, exchanges, and users choose to coordinate around it.4 The proposal’s time limit and its exemption for UTXOs created before activation were design choices intended to constrain its effects, not freeze existing coins.1

But local enforcement does not prove broad adoption. The low signaling rate and the split into a separate branch show that this deployment did not replace the rules followed by Bitcoin’s main chain. Those observations describe this deployment; they do not measure every later BIP-110-related network or define what counts as economically meaningful adoption.4 5 6 7 Nor does the Closed label settle the policy debate over whether block space should be neutral toward the data a valid transaction carries. It records the result of this proposal’s deployment attempt; it does not answer the broader argument about data, fees, or Bitcoin’s purpose.3


What it means for inscription users

The careful answer is narrower than either slogan. BIP-110 did not impose its proposed inscription restrictions on Bitcoin’s main chain. It did cause enforcing nodes to reject the main-chain branch and follow an alternative one; that minority branch later continued under BLAKE2b proof of work with related rule changes. Calling this a Bitcoin-wide Ordinals ban is inaccurate. Saying nothing happened is inaccurate too: nodes disagreed about which blocks were valid, and a minority branch produced blocks.

For readers following Ordinals and Ordinals Inscriptions, the distinction is between a protocol proposal and the rules a particular node actually validates. The debate also touches how Bitcoin Script and Taproot handle transaction data, but the word “inscription” by itself does not determine validity. Software and chain selection do. The episode exposed a continuing dispute over block space; it did not resolve it.

Last updated: 2026-11-17


Related Articles


Related Glossary Terms


See Also

Sources

  1. Dathon Ohm, “BIP 110: Reduced Data Temporary Softfork,” Bitcoin Improvement Proposals repository, version 1.0.1, updated August 9, 2026, abstract, specification, deployment, and changelog. Supports the proposal’s rules, threshold, time limit, grandfathering, and Closed status; accessed November 17, 2026.
  2. Dathon Ohm et al., “BIP Proposal: Reduced Data Temporary Softfork,” Bitcoin Development Mailing List, October 26, 2025. Supports the stated rationale and contemporaneous objection; the thread is a small discussion sample, not a representative survey.
  3. Jon Atack et al., “BIP 110: update status to closed,” Bitcoin Development Mailing List, August 9, 2026. Supports the documented disagreement over activation, branch adoption, and closing the proposal.
  4. Zack Abrams, “Bitcoin’s BIP-110 supporters split onto minority chain as main network pulls ahead,” The Block, August 8, 2026. Supports the block-961,632 split, contemporaneous two-block snapshot, and reported signaling rate.
  5. Bitcoin-forks.org, “Methods: How These Numbers Are Produced,” undated archival methodology page. Supports the later eight-block SHA-256d record and explains the observer’s vantage points and why its record ends before the BLAKE2b branch; accessed November 17, 2026.
  6. Luke Dashjr, “Bitcoin Knots v29.4.1.knots20260508,” GitHub release, September 2, 2026. Supports the release date, BLAKE2b proof of work, and minor BIP-110 activation/expiry changes.
  7. Luke Dashjr et al., “HARDFORK: BLAKE2b Parameters (Mainnet & Testnet4),” Bitcoin Knots pull request #385, August 30–September 1, 2026. Documents the mainnet BLAKE2b activation parameters and checkpoints for blocks 961,639 and 961,640; a review comment records verification of block 961,640’s hash against the checkpoint. This is evidence about the Knots branch and its implementation, not a record of Bitcoin mainnet adoption.