Skip to content

Gaze-in-the-Wild first-party readiness handoff

Not required for the official Figshare copy/rights roadmap item

This handoff remains a valid correspondence/candidate-quarantine protocol, but later reviewed evidence directly verified the official Rakshit Kothari-authored Figshare ProcessData + LabelData original-publication bytes and their CC BY 4.0 dataset-record terms. The narrower roadmap item for the official public Figshare copy/current deposit rights is therefore satisfied independently of this handoff. Nothing here or in that later synchronization establishes historical RIT archive byte equivalence, quarantine exit, numeric task mapping, native-60-Hz/GP3 validity, or raw-data retention.

This layer connects two conservative inputs when the correspondence/candidate route is actually needed:

  1. a privacy-safe first-party response record bound to the exact clarification request and a local correspondence SHA-256; and
  2. a live-reverified quarantined candidate ProcessData screen.

Its output is a readiness and blocker report. It is not a quarantine-exit authorization and does not create empirical evidence.

Why this layer still exists

A first-party response may establish facts not resolved by the public Figshare deposits, such as explicit historical-archive identity, a canonical replacement statement, or an authoritative numeric task mapping. Separately, a candidate ProcessData screen can establish that one exact quarantined file is structurally compatible with GazeForge's ProcessData adapter.

Those facts do not automatically establish candidate identity or authorize quarantine exit. The verified current Figshare distribution is also not silently declared byte-identical to the historical RIT-hosted archive.

Strongest possible correspondence/candidate state

The v1 handoff can reach:

ready_for_independent_exact_copy_review

only when all of the following preliminary conditions are satisfied:

  • the first-party response has completed human review;
  • source authority is verified;
  • an authoritative archive/canonical-source location is provided;
  • dataset-file analysis use is explicitly permitted;
  • redistribution terms are reviewed;
  • the redistribution result does not require vocabulary-mapping review; and
  • the candidate kind is eligible for quarantine-exit review.

Even in that strongest state:

  • independent_exact_copy_identity_verified = false;
  • quarantine_exit_ready = false;
  • quarantine_exit_authorized = false; and
  • source_audit_ready = false.

The remaining blocker is independent exact-copy identity verification for the candidate against the reviewed source relevant to that candidate.

Redistribution semantics are not silently rewritten

The first-party response vocabulary can record redistribution_status = "prohibited", while the quarantine-exit authorization vocabulary uses permitted, restricted, or unknown.

The readiness bridge does not silently translate prohibited to restricted. Instead, it adds the blocker redistribution_status_mapping_review_required. Any later vocabulary reconciliation must be explicit and human-reviewed.

This correspondence vocabulary is separate from the later direct Figshare finding that the named dataset records state CC BY 4.0.

Privacy boundary

The handoff serializes:

  • the exact request fingerprint;
  • the exact response fingerprint;
  • the correspondence SHA-256;
  • reviewed response statuses;
  • the exact recovery/tree/screen/file identities; and
  • deterministic readiness/blocker fields.

It does not serialize the raw correspondence body or archive location. Live verification still requires the local correspondence file so its digest can be checked against the reviewed response.

Scientific boundary

The handoff does not establish or create:

  • historical RIT archive equivalence;
  • quarantine-exit authorization;
  • source-audit execution/readiness;
  • participant/trial/task mapping;
  • coordinate semantics;
  • acquisition or corpus sampling cadence;
  • empirical evidence eligibility;
  • task-stratified model validation;
  • cross-dataset performance;
  • native 60 Hz or GP3 validity; or
  • Frozen Evidence performance claims.

For the current roadmap state, see Gaze-in-the-Wild roadmap evidence synchronization. The official Figshare copy/current-rights item is narrow and satisfied; the authoritative numeric task-mapping gate remains open.