nonoise₿itcoin

Since its introduction in 2014, OP_RETURN was limited to a single output per transaction. This helped ensure that it was used sparingly and primarily for small proofs or certificates.

In Bitcoin Core v30, this restriction has been removed. Transactions with multiple OP_RETURN outputs are now considered standard and will be relayed by default.

This change significantly increases the amount of arbitrary data that can be carried in a single transaction. It also weakens the historical reasoning behind OP_RETURN as a small and controlled data outlet.

My questions:

1. Why was the one-per-transaction limit removed after being enforced for more than a decade?
2. Was the impact on spam and chain bloat carefully studied before merging this change?
3. How does allowing multiple OP_RETURNs align with the original compromise that OP_RETURN was meant to represent?
4. What guidance would Core give to node operators who believe this change undermines Bitcoin’s efficiency as money?

The limit was removed because it was an arbitrary policy, not a consensus rule. It was a "relay policy" that various wallets had begun to work around anyway, leading to inconsistencies. The change in v30 simply aligns Bitcoin Core's default policy with what was already happening in practice.

The impact was debated, but the view of the developers who merged it is that node operators who are concerned about chain bloat have always had the ultimate tool: the ability to set their own, stricter policies. The change doesn't force anyone to accept these transactions; it just changes the default.

The original compromise was from a time when the threat of spam was viewed differently. This update reflects a pragmatic shift, acknowledging that data use cases exist and that trying to stifle them with one policy is less effective than letting individual node operators choose their own limits.