Units, Not Percentages
Before a token has a price, every obligation you settle in it forces a choice most people treat as clerical: do you owe a fixed *number* of tokens, or a fixed *share* of supply? The choice looks cosmetic. It decides who bears supply risk — and the unpriced window is the one moment you are forced to make it in the open.
Two ways to denominate
An obligation payable in a token can be written two ways. As a fraction — "X% of supply" — it floats: mint more tokens and the nominal count owed rises with them, but the holder's *relative weight* stays fixed. As a quantity — "N tokens" — it pins the count, and lets the holder's relative weight move as supply changes. Neither form is neutral. You are choosing which variable to hold constant, and therefore which party absorbs the consequence when the denominator moves. A fraction protects the holder's share; a quantity protects the protocol's cap table. Pretending the two are interchangeable is how counterparties end up owning something other than what they thought they agreed to.
The proportional-adjustment refinement
There is a third form, stricter than either: hard-code the *quantity*, then attach a clause that re-bases it if total supply changes. In ordinary conditions the counterparty reads a fixed number — legible, no modeling required. But if the protocol alters the denominator, the number honors the ratio it originally stood for. It says both things at once: the count is real, and so is the fraction it represented. That is the version you write when you want the obligation readable by a non-specialist and fair to a specialist at the same time. It costs a sentence of drafting and removes a decade of ambiguity.
Why the unpriced window is the right time
This is the argument for building before there is a price. When a token is minted but not priced — as spirit's parent token has been on base since June — you cannot paper over a denomination choice with a valuation. You have to state the quantity out loud, in units, with no market to argue about. Minting creates the ledger entry; pricing is a separate, later act. The gap between them is not a delay to endure. It is the only window in which the terms are auditable before they are tradeable — the window where the numbers earn the trust they will need when a price finally arrives.
