Exactly one size
An order's intent carries a size in coin and a size in money, and must carry exactly one. The test compares their presence rather than their values, so giving both and giving neither are refused on the same line.
A new name that differs
A request to amend a protective pair names the list it is replacing and the list replacing it, and the two must differ. A request naming one list twice is refused.
Figures that must disagree
A reported disagreement over a balance carries what was expected and what was found, and the two must differ. One whose figures agree is refused — the record exists to say a balance was wrong, and a record saying it was right is turned away.
Those two figures are separately held to being finite figures of the exact kind, and to nothing about their sign, so either may stand below nought.
Three runs of one length
A planned ladder carries a run of places, a run of prices and a run of sizes, and the three must be of one length. The lengths are compared in a single chain, so a run shorter than the first and one longer than the last are refused by the same line.
The fifth borrows its rule: the record of a bracket beginning builds another record out of five of its own fields purely to be checked.
The rest
These five are the whole of it. Every other test in the module reads one field.
Related
Source evidence
---
row: P61-03
baseline: 49aa659
created: 2026-11-08 19:05 UTC
modified: 2026-11-08 19:05 UTC
evidence:
- claim: "An order's intent must carry exactly one of a size in coin and a size in money, the two tested against each other by their presence"
source: "`events.py` 193-195"
- claim: "A request to amend a protective pair must name a new list differing from the old"
source: "`events.py` 949-951"
- claim: "A reported disagreement over a balance must have its two figures differ, so one whose figures agree is refused"
source: "`events.py` 1721-1723, the two held to kind and finiteness alone at 1717-1719"
- claim: "A planned ladder must have its three runs of one length"
source: "`events.py` 1421-1423"
- claim: "These five are the only places in the forty-two where one of a record's own fields is weighed against another"
source: "`events.py` 193, 883, 949, 1421 and 1721, with every other test reading one field"
---