Honest Pricing

A v1 policy for Bitcoin block space Bitcoin does not lack block space rules. It lacks a consistent price, and its our duty to always be there for our beloved 'Digital Capital'. Since August 2017, a byte in the transaction body and a byte in the witness have not paid the same. Under BIP 141, non-witness data counts as 4 weight units per byte and witness data as 1. In virtual bytes that is a 75 percent discount. The discount was justified principally by the different resource profile of witness data, most notably signatures: witness bytes do not contribute to the legacy txid and are weighted at one quarter of non-witness bytes. SegWit also fixed transaction malleability, changed how the block limit is accounted for, and made room for later script upgrades. None of that required a file to inherit a signature’s price. From 2023 onward, inscriptions put arbitrary payloads in a Taproot envelope inside the witness. The protocol meter treats that payload like a signature. A payment and a JPEG compete for the same block under different weights(not necessary against innovation in indexers/protocols base on Bitcoin). That is the block space debate in one line. Not whether data may exist on Bitcoin(it has, since the genesis coinbase). The question is whether extra payload should clear at signature rates. Honest Pricing is a node policy that answers no, without touching consensus, without filtering content, and without another fork. What this is not It is not BIP 110. It does not invalidate large OP_RETURN, inscriptions, annexes, or data pushes. It does not change chain selection. A valid most-work block remains a valid most-work block. It is not a moral list. The policy does not ask whether the bytes are Runes, a hash, a poem, or a PNG. It asks whether they are required to move value, and whether payload above a small franchise has been quoted at body weight. It is not a new fee market. Miners already sell space in satoshis per virtual byte. Honest Pricing only restates the economic size of some transactions before comparing them to a chain-derived reference rate. Three pieces The mechanism closes as a v1 policy, not consensus. Three pieces fit naturally: a 2,016-block EMA provides a deterministic, chain-derived reference rate; the historic OP_RETURN allowance provides a small payload franchise; and restoring witness payload from one weight unit per byte to four removes the SegWit subsidy without penalizing ordinary spending. v1 deliberately accepts two simplifications. The reference rate is a robust historical proxy, not a perfect market-clearing price. Payload classification must remain structural rather than semantic. Neither requires a consensus change. The reference rate Every full node can deterministically derive two quantities from the chain: the fees paid by confirmed transactions and their total virtual size. No external oracle, committee, or price feed is required. v1 uses an exponential moving average calibrated to a 2,016-block span — matching Bitcoin’s difficulty-adjustment interval. This average is a deliberate approximation. total fees / total vbytes is not the marginal feerate of the last transaction that entered the block, and it is not Bitcoin Core’s confirmation-based estimator. A miner can also circulate fees through self-dealing transactions and move the ratio. Those are mechanism facts, not implementation footnotes. For v1, a simple, auditable number every node can recompute from the tip is the point. A later version can graduate to a better estimator without changing the rest of the policy. What gets counted Split the transaction on structure, not on content. Base is monetary structure: version, value-bearing inputs and outputs, spending scripts, real signatures, and control data required to satisfy the spend. Base is never multiplied. Sending bitcoin does not pay a surcharge because someone inscribed a file last week. Payload franchise is the legacy 83-byte datacarrier allowance: the total scriptPubKey size fixed as standard in Bitcoin Core 0.12.0, typically about 80 bytes of payload plus script overhead. A hash, a short protocol tag, a compact Runes field stays inside that line. Bitcoin Core 30 raised the default -datacarriersize to 100,000 bytes and allowed multiple data-carrier outputs; a node can still set -datacarriersize=83. Honest Pricing does not re-litigate that relay default. It prices what exceeds the old franchise. Excess witness payload is witness bytes not required to spend( the inscription envelope, bulk pushes, anything that is using the witness as disk). Ambiguous bulk fails closed: treat it as payload, not as a signature. The coefficient is not a penalty Under BIP 141 a witness byte already contributes 1 weight unit to protocol weight. Adding another 3 brings it to 4, the same as a body byte. Multiplying raw payload by 4 on top of protocol weight would overcount. A ~50 KB witness payload moves from about 12,500 protocol vB to about 50,000 economic vB. At reference rate= 10 sat/vB that is on the order of 500,000 satoshis, not 125,000. A simple payment of ~140 vB remains ~1,400 satoshis. Same block. Same reference rate. Different claim on space. Why policy is the right layer Relay and mining policy is how Bitcoin has always nudged behavior without splitting the chain: dust limits, standard scripts, the old datacarrier cap itself. Core v30 widened default relay of OP_RETURN because users had already routed around the 83-byte cork, into the witness, or across a miner’s private desk. Opening a prunable channel does not require leaving the witness discount attached to files. If another miner includes underpriced payload, Honest Pricing nodes follow the chain. They simply do not help that payload jump their mempool. Adoption is voluntary. A rule that needs a flag day and a hashrate threshold is a different project. 2026 already ran that experiment. Economic weight is a local quote. Consensus weight stays BIP 141. What v1 leaves out v1 does not prune. Witness data does not enter the UTXO set and can be discarded as part of historical block pruning after validation; OP_RETURN outputs likewise create no spendable UTXO. A later, explicitly discardable datacarrier can be argued on its own. Pricing first. v1 does not pretend classification is free. The test is structural: is this byte required to satisfy the spend? It is not a MIME type and not a content list. v1 does not claim the EMA is the true clearing price of the next block. The closed loop 2014: OP_RETURN exists so data does not live forever in the UTXO set. 2017: witness exists so signatures leave the txid, with a weight that matches their resource profile. 2023: that weight is reused as a file market. 2025: default datacarrier relay widens because the old cork had already failed. 2026: a consensus ban fails to gather work. Honest Pricing derives a reference rate using a 2,016-block EMA span, leaves the legacy datacarrier allowance unadjusted, adds 3 weight units only to excess witness payload, and never multiplies money. Bitcoin remains permissionless. Block space stops being on sale at signature rates. Thanks for your time, the only thing more valuable than Bitcoin. AT.