# scenario: LS query on MC-5SLPM-D capturing the current-vs-requested split
# Captured 2026-04-17 from /dev/ttyUSB0 (Prolific PL2303, 19200 baud,
# unit_id=A, firmware 10v20.0-R24). Device LSS=A (analog setpoint source),
# so serial LS writes are silently ignored — but the LS query still
# reveals the current/requested split.
#
# Wire shape: `<uid> <current> <requested> <unit_code> <unit_label>` (5 fields).
#
# Resolves design §15.6's deferral: the wire already separates
# SetpointState.current (measured value of the controlled variable) from
# SetpointState.requested (setpoint target). The facade populates
# both distinctly; no LV cache dependency.

# Steady state — current == requested.
> ALS
< A +078.95 +078.95 12 SCCM

# Transient — current diverges from requested while the loop re-settles.
# Captured back-to-back with the steady-state query above; the device
# evidently updates the internal target at one cadence and the reported
# current at a slightly different one, exposing the split.
> ALS
< A +078.95 +078.88 12 SCCM
