In January 2023, a developer named Casey Rodarmor released a program that did something seemingly harmless: it numbered each satoshi in the order it was mined and allowed a file to be attached to one of them. Within weeks, images, texts, and even entire games were recorded within Bitcoin blocks, fees skyrocketed, and the network became embroiled in a debate that continues to this day.
The mechanism is ingenious and worth understanding before passing judgment. Taproot, from the previous lesson, allows a branch of a script to be revealed when spending. This revealed content acts as a witness, the area that SegWit created, which occupies a cheaper space in the block. An inscription hides the file within a script segment that is never executed — and the network carries, replicates, and stores it forever.
The effect was immediate and measurable. In May 2023, the mempool accumulated hundreds of thousands of transactions, and the fees for a regular payment rose to tens of dollars. In block 840,000, the halving block of April 2024, miners collected more than 37 bitcoins in fees, almost double the issuance reward. For miners, it was good news. For those wanting to send money that afternoon, it was not.
Thus, two camps emerged, each with arguments worth hearing.
On one side: the rules do not distinguish types of data, nor should they. Whoever pays the fee buys the space, period. Filtering content would require someone deciding what is legitimate use, and it is precisely that someone the network was built not to have. Moreover, high fees are the revenue that will replace issuance when it ends — and it's good to discover early if there is demand willing to pay for security.
On the other: the network was made to transfer value, and block space is a scarce common good. Each recorded image is carried forever by thousands of nodes worldwide, making it more expensive to run one and pushing in the direction the block size war tried to avoid. Some also note that the cheap witness was created for signatures, not files: using it this way exploits a price not intended for this purpose.
The dispute also appeared in the code. Some node operators began using filters that refuse to relay such transactions; others argue that filters don't work because miners can receive the transaction directly and include it anyway — which indeed happens. By 2025, the discussion reached the limits of the OP_RETURN free data field, with proposals and reversals. No side won cleanly.
It's worth noting that this debate is older than ordinals. In 2010, there was already discussion about recording messages in transactions; OP_RETURN itself was created in 2014 precisely to offer a cheap and disposable place for those who wanted to record data, preventing them from doing worse things. The underlying question — is Bitcoin money, or is it a public space for those who pay for it? — is practically as old as the genesis block.
What history shows is how the network resolves this type of issue: poorly, slowly, and without authority. No one can prohibit, no one can mandate, and the result emerges from what thousands of independent operators decide to run. It's slow and uncomfortable, and it's the same property that prevents someone from censoring your payment.
You have reached the end of the intermediate track. You know what a key is, how a wallet organizes it, how the network validates, and what exists on top of it. What comes next is the mathematics underlying all of this, with calculations on the table: finite fields, the secp256k1 curve, the complete calculation of a signature, and the exact point where it breaks. It's the beginning of the advanced track.