{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp04_retrieval_first_paired20_20260802","cell_id":"layer_decay_and_expiration_management__film_media_production","arm":"RETRIEVAL_FIRST","candidate_id":"C1","hypothesis_id":"H1","version":0,"title":"Revision-Triggered Expiry for Editorial Notes","problem":"Obsolete notes, review links, and cut instructions remain visible after a sequence changes, allowing feedback tied to an earlier cut to appear actionable on the current cut.","actors":["Picture editors","Assistant editors","Director","Producer","Review coordinator or post-production supervisor","Authorized reviewers"],"observable_state":"For each versioned editorial note, the system exposes the referenced sequence and cut, lifecycle state, creation and transition times, renewal or supersession relationship, current-cut applicability, and preserved decision lineage. When a new cut is promoted, prior-cut notes no longer remain current by default.","consequence":"A stale instruction can be mistaken for current creative intent, producing contradictory edits, avoidable reconciliation work, and rework during director and producer review through picture lock.","affected_objective":"Reduce edits attributable to superseded feedback and the time picture editors and assistant editors spend reconciling contradictory notes without erasing the history of creative decisions.","intervention":"When an authorized user promotes a new cut, automatically transition every still-active note attached to prior cuts to expired or superseded. Require a director, producer, or delegated review coordinator to renew a still-applicable note onto the new cut; give each renewed note a link to its predecessor. Demote expired notes from current-action views while keeping them available in version history. The transition changes visibility and applicability only; it does not delete notes or decide creative intent.","structural_mapping":[{"archetype_element":"Accumulated temporal layers","domain_realization":"Versioned notes, review links, annotations, and cut instructions deposited across successive sequence revisions."},{"archetype_element":"Layer identity and age index","domain_realization":"Each note is identified by note ID, referenced sequence, referenced cut version, author, creation time, and lifecycle-transition time."},{"archetype_element":"Decay or expiration trigger","domain_realization":"Promotion of a successor cut makes active notes attached to earlier cuts stale candidates and triggers their automatic expired or superseded state."},{"archetype_element":"Preservation exception and renewal","domain_realization":"A still-valid instruction is explicitly renewed onto the successor cut by an authorized reviewer instead of remaining active by default."},{"archetype_element":"Supersession marker and lineage","domain_realization":"The renewed or replacement note records its predecessor, while the prior note records its successor and remains inspectable in history."},{"archetype_element":"Active versus archived access state","domain_realization":"Current-cut views show active and renewed notes by default; expired notes are demoted but remain accessible through version history and audit views."},{"archetype_element":"Review cadence and revalidation","domain_realization":"Every promoted cut creates a bounded revalidation point before review continues."}],"mechanism_mapping":[{"mechanism_slug":"stale_layer_detection_dashboard","role":"Provides an identity-resolved view of notes whose referenced cut is no longer current, allowing editors and reviewers to inspect state transitions and unresolved renewal decisions.","counterfactual_removal":"Without this inventory and stale-state surface, expired notes could remain distributed across review links or be overlooked, making the intervention difficult to audit and stale-feedback incidents difficult to attribute."},{"mechanism_slug":"time_to_live_ttl_policy","role":"Adapts expiry from elapsed clock time to a revision-relative lifetime: a note's active lease ends when its referenced cut is superseded, unless an authorized actor renews it for the successor cut.","counterfactual_removal":"Without the revision-triggered expiry rule, prior-cut notes would remain active until someone manually resolves them, collapsing the proposal back to existing version scoping and manual completion."},{"mechanism_slug":"tombstone_or_deletion_marker","role":"Leaves an expired or superseded marker at the prior note's identity and records any renewed or replacement successor so that demotion cannot be mistaken for absence or erasure.","counterfactual_removal":"Without the marker and successor link, users could not reliably reconstruct why a note disappeared from the active view or distinguish an intentionally superseded instruction from a missing record."}],"causal_chain":["A new cut changes the context against which earlier editorial notes were authored.","Promotion of that cut triggers lifecycle transitions for notes attached to prior cuts.","Prior notes become expired or superseded rather than remaining actionable by default.","Authorized reviewers explicitly renew only instructions that still apply, creating predecessor-successor links.","Current-action views exclude unrenewed stale notes while version history preserves them.","Editors encounter fewer obsolete instructions as apparent current intent.","Edits attributable to superseded feedback and subsequent reconciliation work decline."],"baseline":"Professional review systems already provide substantial overlapping functionality: version-scoped comments, ordered version histories, latest-version designation, manual Open/Done or resolved states, state filtering, comparison, exports, and some approval-triggered resolution. The comparison baseline is this version-scoped, history-preserving workflow with manual note resolution, but without automatic expiry of every prior-cut note at cut promotion, explicit renewal onto the successor cut, and preserved per-note successor lineage.","nearest_rivals":["VideoReview version-scoped comments and resolution states","Autodesk Flow Production Tracking note/version/status model","Frame.io version stacks plus Mark As Complete","Clipback revision tracking with approval-triggered resolution"],"remaining_contrastive_claim":"In a film editorial review workflow, promoting a new cut automatically marks each prior-cut note expired or superseded, carries forward only explicitly renewed notes, and retains visible note lineage, thereby reducing stale-feedback edits beyond version scoping and manual resolution.","authority_safety":{"decision_authority":"The director and producer retain authority over creative intent and note applicability. A designated review coordinator or post-production supervisor may administer lifecycle transitions and renewals under their delegation. Editors may flag suspected errors and restore visibility but may not reinterpret or discard creative instructions unilaterally.","authorized_first_step":"Run a read-only retrospective replay on one completed, picture-locked production: inventory notes and cut promotions, simulate the proposed state transitions, and have authorized reviewers adjudicate whether each simulated expiry, renewal, and lineage link would have been correct.","excluded_actions":["Deleting or irreversibly altering notes, media, review links, or audit history","Automatically deciding that an instruction lacks creative value","Applying simulated state changes to a live production","Hiding expired notes from authorized historical or dispute review","Overriding legal holds, contractual retention duties, evidentiary needs, or studio records policy","Renewing a note without director, producer, or delegated-review authority","Treating expiry as approval of the successor cut"],"halt_rollback":"Halt the pilot if a cut promotion cannot be reliably identified, note-to-cut identity is ambiguous, an authorized reviewer disputes a transition, or a retention restriction is encountered. Because the first step is read-only, rollback consists of discarding the simulated labels and derived views while retaining the untouched source records and an audit of the test."},"negative_tests":{"strongest_counterevidence":"The closest products already cover most of the workflow: VideoReview binds comments to cuts and supplies Open/Done states; Autodesk links notes to Versions and supports no-longer-relevant closure; Frame.io combines version stacks with manual completion; and Clipback reports approval-triggered automatic resolution. These collisions sharply limit the proposal to the specific combination of cut-promotion-triggered per-note expiry, explicit renewal, and successor lineage.","problem_falsifier":"The problem is falsified for the target workflow if an audit across the three intended productions finds no edits or material reconciliation time attributable to notes from superseded cuts, or shows that current version scoping and manual resolution already prevent such notes from appearing actionable.","intervention_falsifier":"The intervention is falsified if an existing reproducible professional workflow already performs revision-triggered per-note expiry, explicit renewal or carry-forward, and preserved successor lineage, or if controlled use across three productions fails to reduce edits attributable to superseded feedback by at least 25% relative to version-scoped manual resolution.","risks":["Valid cross-version instructions may be demoted if reviewers fail to renew them.","Mandatory renewal may add review labor or create a cut-promotion bottleneck.","Ambiguous cut promotion or duplicated note identities may cause incorrect transitions.","Lifecycle labels may create false confidence when the underlying creative intent remains disputed.","Users may recreate notes instead of renewing them, breaking lineage.","Expired-note demotion could conceal context needed for audit, contractual review, or creative dispute resolution.","Products or private studio automations not covered by the bounded search may already implement the remaining behavior."]},"next_evidence_step":"On one completed production, select at most three sequences with at least three promoted cuts each and review no more than 150 version-linked notes. Replay each cut promotion in a read-only prototype, generate proposed expired, renewed, and superseded states, and ask two authorized reviewers to independently adjudicate transition correctness. Measure false expiry, missed stale-note detection, renewal workload, lineage completeness, and retrospectively identifiable edits caused by superseded feedback. Proceed to a prospective controlled comparison only if at least 95% of note-to-cut links are resolvable, no protected record would be hidden, and reviewer agreement on the simulated transitions is at least 90%.","prior_art_status":"SEARCHED_BOUNDED","revision_record":{"parent_version":null,"progress_targets_addressed":["Specified the complete actor, state, consequence, intervention, authority, safety, causal, falsification, and evidence structure for H1.","Constrained the proposal to the residual distinction left after substantial prior-art collision.","Separated lifecycle visibility changes from deletion and creative decision authority."],"conceptual_changes":["Converted the terse lifecycle idea into a revision-triggered active lease for each editorial note.","Defined explicit renewal and predecessor-successor lineage as the narrow residual mechanism.","Treated expiry as demotion from current applicability, not destruction or a judgment that the note lacks historical value."],"operational_changes":["Defined cut promotion as the transition event and specified per-note states and required metadata.","Limited the first test to a read-only retrospective replay on one completed production.","Added authorization, halt conditions, rollback, and quantitative readiness thresholds."],"evidence_changes":["Incorporated the bounded prior-art record's substantial collisions with VideoReview, Autodesk Flow Production Tracking, Frame.io, and Clipback.","Retained the original three-production 25% effect threshold as the intervention falsifier.","Added bounded precursor measures for transition accuracy, reviewer agreement, renewal workload, and lineage completeness."],"claim_changes":["Narrowed the candidate from general lifecycle labeling to automatic cut-promotion-triggered per-note expiry, explicit renewal, and visible successor lineage.","Made no world-novelty claim and acknowledged inaccessible products, private automations, patents, and undocumented configurations.","Preserved the exact remaining contrastive claim from the independent prior-art record."]}}