Ordinal theory and inscriptions

Wonder

1. How can a unit that is currently fungible be made unique without changing the consensus rules?
2. Why does tracking satoshis require software conventions rather than on-chain validation?
3. What happens to a rare satoshi when it passes through a wallet that ignores ordinal numbering?

Explore

Part IOrdinal theory

1. Ordinal theory, proposed by Casey Rodarmor in early 2022 and brought into wide use when his ord software launched inscriptions in January 2023, assigns serial numbers to every satoshi. Since one bitcoin equals 100,000,000 satoshis, the system numbers them sequentially based on the order in which they were mined. This numbering scheme does not alter Bitcoin’s protocol; it operates entirely within existing transaction structures. To a standard node, these are ordinary outputs. Only software that implements the ordinal convention recognizes the specific identities of individual sats.
2. The tracking mechanism relies on first-in, first-out (FIFO) ordering. When a transaction spends inputs to create outputs, the satoshis from the earliest inputs map to the earliest outputs. This deterministic mapping allows participants to trace any specific satoshi through its entire history. The system requires no sidechains or separate tokens; it uses Bitcoin itself as the sole asset, so the identities last as long as the coins themselves.
3. Rarity tiers assign numismatic value based on a satoshi’s position within the blockchain’s structure. An ‘uncommon’ sat is the first one mined in any block. A ‘rare’ sat is the first of a difficulty adjustment period, which retargets every 2016 blocks. An ‘epic’ sat marks the start of a halving epoch. Rarer still are ‘legendary’ sats, the first sat of each cycle in which a halving and a difficulty adjustment coincide (roughly every 24 years), and the single ‘mythic’ sat, the first sat of the genesis block mined in January 2009.
4. This numbering scheme enables inscriptions, where arbitrary data attaches to a specific satoshi. While ordinal theory itself is merely an indexing convention, it provides the substrate for digital artifacts. These artifacts inherit Bitcoin’s immutability and security properties because they reside within valid transactions. Once created, no third party can delete or alter the content associated with that specific serial number.
5. A critical trade-off exists between software compatibility and asset safety. Wallets that do not implement ordinal-aware logic treat all satoshis as identical currency units. If such a wallet spends a ‘rare’ or ‘legendary’ satoshi, it often does so to cover transaction fees or as part of change. To the user, this appears as a loss of value; to the network, it is a routine transfer. The risk of accidental expenditure is inherent in using non-compatible software for ordinal assets.
6. To verify these mechanics, open a block explorer and locate any recent transaction containing multiple inputs. Observe how the satoshi offsets from the first input populate the beginning of the output list. Then, consult an ordinal-specific indexer to view the same transaction. The indexer will display serial numbers and rarity tags that are invisible in standard blockchain data, demonstrating how identity is a layer of interpretation built upon raw ledger entries.

Part IIInscriptions and the 2023 boom

7. Inscriptions attach arbitrary data to a specific satoshi by embedding it within a Taproot script-path witness. The data sits inside an ‘envelope’ composed of OP_FALSE, OP_IF, the content pushes, and OP_ENDIF. Because the condition is false, the code inside the envelope is skipped, while the rest of the script, usually a signature check, runs normally; the data changes nothing about who can spend the coins. This mechanism allows content to be stored directly in the witness section of a standard Bitcoin transaction.
8. Two protocol upgrades made this practical. SegWit, activated in August 2017, discounts the weight of witness data, reducing the cost per byte compared to legacy transactions. Taproot, activated in November 2021, removed previous script size restrictions and simplified script verification. Together, these changes allowed large payloads—images, text, or code—to be embedded economically into the blockchain without requiring a soft fork or sidechain.
9. The first inscription, known as Inscription 0, was created in December 2022. It contained a pixel-art skull image. Casey Rodarmor, who developed ordinal theory and the `ord` software suite, initiated this experiment. The mechanism relies on the immutability of confirmed blocks; once an inscription is mined, it cannot be altered or deleted. This permanence makes digital artifacts as durable as Bitcoin itself, though every byte incurs a permanent cost in block space.
10. A boom followed in early 2023. Collections such as Bitcoin Punks emerged, and individual users began inscribing complex images and interactive web pages. In February 2023, block 774,628 was almost entirely one Taproot Wizards image, making it about 3.96 MB, close to the ceiling of 4 million weight units. This event demonstrated both the potential for high-fidelity storage and the significant impact on network bandwidth and node resource consumption.
11. Within months, millions of inscriptions were created. The surge pushed fees to their highest levels in years as participants competed for block space; much of the May 2023 spike came from BRC-20 tokens, the subject of the next lesson. Supporters view this as a legitimate use of paid resources that generates revenue for miners; critics argue it acts as spam that crowds out standard payments and burdens full nodes with unnecessary bloat. The debate centers on whether Bitcoin’s primary function is electronic cash or a general-purpose ledger.
12. The distinction between inscriptions and other token standards like BRC-20 lies in their structure. Inscriptions bind data to specific satoshis, creating unique digital objects that follow the UTXO model directly. The content itself sits on-chain in the witness, unlike the balances of JSON-based tokens, which exist only in indexers; tracking which satoshi carries an inscription still requires ordinal-aware software.

Key facts

Exercise nine, six, three, one

Fact check

Dates, figures and claims checked against the published sources listed below.

• C. Rodarmor, 'Ordinal Theory Handbook' (docs.ordinals.com)
• Bitcoin Optech newsletter coverage of ordinals (2023)
• ord, the reference implementation (github.com/ordinals/ord)
• Bitcoin Optech, 'Ordinals and Inscriptions' coverage (2023)
• Block 774,628 on-chain transaction data