Neutral Technical Overview · July 2026

BIP‑110 &
Bitcoin Storage

OP_RETURN, Ordinals, Bitcoin Core v30, and the temporary soft‑fork proposal — history, both sides, and where it stands.
Ben Sigman
The distinction that decides everything

Policy is not consensus.

Policy / Standardness

What a node chooses to relay and accept into its mempool by default. Reversible. A single operator can override it.

Consensus

What makes a block valid. A transaction can be consensus‑valid yet non‑standard — miners may still include it.

Most of the 2025 controversy was about policy defaults — not the rules that define a valid block.
2010 · Foundational context

A temporary limit that stayed.

Satoshi added the 1 MB block‑size cap as an anti‑DoS measure — slipped in under a commit labeled “fix openssl linkage problems.”
“We can phase in a change later if we get closer to needing it.”
— Satoshi Nakamoto, October 2010
2013–2015 · The controlled nudge

OP_RETURN: a less harmful place to put data.

  • A provably‑unspendable output — data never enters the UTXO set
  • Limited by policy to ~80 bytes, one output per transaction
  • Meant to steer people away from fake outputs that bloat storage forever
2023 → · The witness discount
4×
Ordinals put arbitrary data in the Taproot witness — which gets a 4× fee discount, making big payloads far cheaper than OP_RETURN. Inscriptions became a major driver of block‑space demand.
~2023–2025 · Why the default stopped matching reality

When mempools diverge, blocks propagate slower.

Miners and alt relays (MARA Slipstream, Libre Relay) were already accepting oversized OP_RETURNs. When node mempools don’t match miner templates, BIP‑152 compact‑block reconstruction fails more often — extra round‑trips, higher latency, and a real centralization pressure favoring larger, better‑connected miners.
2025 · The Core v30 change

The limit came off.

  • Apr 2025 — Poinsot proposes relaxing OP_RETURN standardness
  • May–Jun — Sanders’ PR merged: default raised to 100,000 bytes, multiple outputs allowed
  • Oct 2025 — Core v30 ships. Policy only — consensus unchanged
The case FOR the change · Poinsot, Todd, Sanders

“The old limit was fiction.”

  • It was already ineffective — relays bypassed it for months
  • Harm reduction — stop pushing data into UTXO‑bloating fake outputs
  • Big data is still cheaper via witness anyway — this mostly affects small use cases
The case AGAINST · Dashjr, Mow, Knots users

“Don’t legitimize spam.”

  • Raising the default encourages non‑monetary use — drifting from Bitcoin as money
  • Node burden & legal risk — archival nodes store whatever is inscribed
  • Operators can push back by running Bitcoin Knots
Late 2025–2026 · The response

BIP‑110: a temporary soft fork.

What it is

A ~1‑year consensus soft fork to restrict several methods of embedding large contiguous data. Assigned Dec 2025 — author Dathon Ohm, draft input credited to Luke Dashjr.

Where it stands

Miner signaling remains under 1% of hashrate. A mandatory signaling window is projected for early August 2026. Core has not implemented or endorsed it.

The through‑line

Temporary limits
never stay temporary.

From the 2010 block‑size cap to BIP‑110 — the same tension recurs: protective defaults vs. permanent scarcity. The question isn’t settled; it’s the debate that defines what Bitcoin is for.
Source: BIP 110 and Bitcoin Storage · Compiled July 2026
Made with xavior.ai