The coins, counted cashier by cashier

This table is a count and not a survey. Each coin on it was read on a deposit page on a named day, and the five cashiers that were not opened from here contribute nothing to it — neither a zero nor a guess. A coin that appears in two cashiers is counted twice, because the question the table answers is how many cashiers take it, not how many coins exist. The newest entries, dated by certificate.

Three of the coins in the count hold a stable value, and they do a different job from the rest. A balance sitting in one of them does not change size between two evenings, and a transfer made in one arrives at the size it left. On the route from Turkey that second property is the important one: it is the difference between knowing what will land and finding out afterwards. Roulette, counted pocket by pocket.

A coin on a deposit page is only half an answer. The same stable coin runs on several networks at once, and the cashier issues an address for exactly one of them at a time. Which network the deposit page offers decides what the transfer costs, how long it takes to confirm, and — on a small deposit — whether the whole exercise is worth doing at all. A fee that is trivial on a large transfer can be a noticeable share of a small one, and the route from Turkey is full of small ones. Metaspins: what is written down.

• The address belongs to one network. A transfer sent on another does not arrive, and nobody can reverse it — not the cashier, not the exchange.
• The choice is made twice. Once at the desk that sends the coin and once at the cashier that receives it, and the two have to agree.
• The fee is paid again on the way out. A withdrawal is a transfer in the other direction on the same rails, priced the same way.

This register does not print fee figures anywhere, for the same reason it prints no payout times: a fee moves with network traffic rather than with anything an operator decides, and a number captured on one afternoon describes that afternoon. What is scored instead is whether the cashier offers a network cheap enough to be usable at all, which is a property of the deposit page and can be read.

The count below is honest about being a count. Three things it does not answer, and each of them is a separate reading: whether the minimum deposit in that coin is small enough to be useful, whether the same coin can be withdrawn as well as deposited, and whether a balance is held in the coin or converted into something else the moment it arrives.

That last one matters more than it sounds. A cashier that converts an incoming stable coin into a house balance has taken the stability away at the door: the protection a dollar-pegged coin offers on the route stops at the cashier's front page. Where a cashier says which way it works, it is worth reading; where it says nothing, nothing is assumed here.

Where a cashier is missing from this table, the deposit page was not opened from here. That is a gap in the reading and it is marked as one.

Breadth earns points in this register, but a wide cashier is not the same thing as a good one. What it buys is optionality: more ways for a coin bought on a Turkish desk to be delivered, and a better chance that one of them runs on a network whose fee does not swallow a small deposit. Beyond that, the fifteenth coin in a cashier does nothing for anybody who only ever uses the first. How the breadth block turns this count into points is on the method page, and the records it feeds are in the register.