# Rapid Independent Test Protocol

## Objective

Measure the ordering of two physically defined events:

1. SH_CHANGE: the detector transition that operationally defines the SH channel.
2. LIGHT_ARRIVAL: the first detector transition caused by light arriving at the
   projected-particle interaction plane.

The first task for every laboratory is to publish its operational definition of SH.
The token remains SH until the owning experiment defines it; this protocol does not
expand the name by inference.

## Minimum apparatus

- One shared hardware clock or time-to-digital converter for both detector channels.
- Raw, immutable event output with the clock identity and calibration receipt.
- A randomized projection setting chosen after any pre-shared state is sealed.
- Signal, null, and physically disconnected control trials.
- Channel latency and uncertainty estimates measured independently.

Display timestamps, file modification times, network receipt times, and timestamps from
unsynchronized clocks are diagnostic projections rather than the causal timing record.

## CSV schema

    trial_id,clock_id,t_sh_ns,t_light_ns,u_sh_ns,u_light_ns,control

- All time and uncertainty fields are integer nanoseconds in one clock domain.
- control is SIGNAL, NULL, or DISCONNECTED.
- u_sh_ns and u_light_ns are non-negative absolute uncertainty bounds.

## Conservative ordering test

For each trial:

    sh_latest = t_sh_ns + u_sh_ns
    light_earliest = t_light_ns - u_light_ns

If sh_latest is strictly less than light_earliest, the row is
PRE_LIGHT_SEPARATED. Overlap is reported as INTERVALS_OVERLAP.

## Rapid sequence

1. Hash the raw acquisition file before analysis.
2. Run at least one complete batch of randomized SIGNAL trials.
3. Interleave NULL and DISCONNECTED controls without revealing their order to analysis.
4. Record calibration and detector latency in the same unit.
5. Run record_trials.py and preserve the generated HBP beside the raw-file hash.
6. Repeat on a second acquisition chain and a second team.
7. Publish raw or privacy-clean derived traces, apparatus description, firmware versions,
   calibration, uncertainty model, exclusions, and all trial outcomes.

## Interpretation boundary

A separated event order is a timing measurement. A physical quantum conclusion also
requires a defined quantum observable and an appropriate witness or model comparison.
Entanglement correlation and controllable signalling are separate questions. This
protocol preserves those distinctions while making the reported Liris order fast to
attack-test.