The exception
The plain single order. Its settings go to a builder of their own, which is where the payload it may leave out is answered.
What the two roads share
Either hands back a settings object that is already of the right kind, unexamined. Neither looks inside such an object.
Where a mapping arrives instead, the two part: the plain single order's builder tests each value as it reads it, and the table's road fills the record and leaves the testing to the record's own building.
So a reader asking what refuses a bad figure gets a different answer by way of working, though both answers end in a refusal.
The change's side
A change's settings have no exception. Their record is chosen from a table of its own, for every way of working alike, so the plain single order's change goes the same road as a ladder's.
Related
- The payload a mode's settings arrive in
- The five re-keyings
- How a settings field is known to want a tuple
Source evidence
---
row: P49-02
baseline: 49aa659
created: 2026-11-08 17:29 UTC
modified: 2026-11-08 17:29 UTC
evidence:
- claim: "The plain single order's settings have a builder of their own, reached before the table"
source: "`command_translator.py` 378-379, ahead of 381"
- claim: "Every other way of working has its record chosen from the table"
source: "`command_translator.py` 381 into 387-389"
- claim: "A way of working with no entry in that table refuses the command"
source: "`command_translator.py` 383-385"
- claim: "Either road hands back a settings object of the right kind unexamined"
source: "`command_translator.py` 261-262, and 330-331"
- claim: "Where a mapping arrives, the plain single order's builder tests each value itself and the table's road leaves that to the record it fills"
source: "`command_translator.py` 289-294, against 348-352"
- claim: "A change's settings are chosen from a table alone, with no way of working standing apart"
source: "`command_translator.py` 415-424"
---