NEWS HOT-CORE · Editorial Work as an Auditable Workflow · Historical Status
Core First, Add-on Modules After Data Gates, Advanced Statistics Deliberately Disabled.
System reference for merge RC 0.3.4: time and coverage contracts, three distinct processing paths, provenance and quality evidence, and transparent failures rather than fabricated completeness.
Positioning: Workflow, Not a New Truth System
NEWS HOT-CORE is a specified editorial execution contract for time-critical research and a private situational assessment. It uses NIS-related definitions for discourse, representation and market indicators, but it is neither identical to the ROS/NIS core nor demonstrably an integrated runtime merely through this association. Its status is a merge candidate, not an accepted production service.
Path I · Mandatory HOT-CORE
Set time window → broad scan of 13 separately logged topic lanes → thoroughly check substantive candidates against primary sources and independent origins → separate F-candidate sweep (F-Sweep) → cross-lane deduplication and countermodels → deliver core. Batched research limits effort but does not replace proof of coverage for each lane. Optional indicators must not hold back an available core result.
Path II · Active Indicators & Tier 1
Add-on modules run only when readiness, data availability and quality actually suffice. Planned features include qualitative discourse analysis PDI, institutional representation analysis PRI, public read-only market data, verified partial market calculations and a separately labeled speculative layer. Each relevant module row records execution status, quality tier, missingness/blocker and evidence link; reporting status does not oblige inventing values.
Path III · Dormant Tier-2 Registry
Advanced discourse/representation indices and EDDI LITE/FULL remain HARD_LOCKED (BOOTSTRAP_NO_PIPELINE) until codebooks, versioned processing, appropriate history, independent test vectors and reproducible timely validation are available. More data or one passing simple arithmetic test does not automatically activate these models. The registry preserves development goals without creating fictitious ongoing statistics.
Time Contract Instead of Rolling Currency
Editorial cutoff, actual run, event, first publication, material change and observation have distinct timestamps. Late execution does not shift the window. LATE_DISCOVERY is a labeled supplement; POST_CUTOFF remains outside the completed delta and must not be excluded from the next regular run by deduplication. Cosmetic updates are not new factual events; local calendar days and DST are not replaced with generic UTC intervals.
Coverage, Event Ranking & Source Independence
F denotes a documented new systemic/institutional configuration, M a substantial development, T an examined situational update, and N no reportable signal. Media volume or repeated reposts do not strengthen evidence. checked_no_fundamental_change requires actually verified coverage; incomplete_coverage and not_checked are not an all-clear. Contradictory observations, possible false negatives and open candidates remain visible.
Indicators Measure Different Objects
PDI describes accessible discourse samples, not a representative population. PRI distinguishes institutional expectations, observed action and outcome; presence is not proof of impact. Market prices, a single order-book snapshot and a genuine time delta are different observations. Polymarket is restricted under this contract to anonymous public GET market data analysis: no account, wallet, order, trade or individual profiling. Speculative claims do not enter the factual layer.
Market Data: Identity Before Calculation
Contract, delivery month, trading date, unit, currency, price type and demonstrated availability at cutoff must align. Settlement, Last, Spot, Daily Marker, Partial Basket and indicative intraday diagnostics are not interchangeable. A missing leg is filled with neither a proxy nor zero. Numerical derivations require deterministic tool-based verification and input provenance; a valid partial result is not a full index.
Readiness, Provenance & Fail-Soft
INLINE_READY, EXTERNAL_VERIFIED, DORMANT_LOCKED and UNAVAILABLE describe readiness; PASS/PARTIAL/SKIPPED/FAILED describe actual execution. RAW and DERIVED, evidence and interpretation remain separate. Limited legitimate fallbacks end in NULL with a concrete reason instead of fabricated data. Persistent comparisons and owner-only append-only archives require actual write, permission and readback evidence; otherwise status remains unconfirmed.
What the 0.3.4 Merge Consolidates
According to the merge review, the substantive coverage, time, classification and model boundaries of RC3.2 remain intact. The new candidate consolidates technical access/response states, concrete blockers and the separation of partial baskets, intraday data and canonical indicators. An external specification baseline is still UNRESOLVED/NOT_MATERIALIZED; inline invariants are the documented operational reference, not proof that every pipeline is implemented. Editorial safety boundary of this site: technical accessibility does not imply evidentiary validity or blanket permission for use/publication.
Public Description, Private Results
The site documents architecture and validation logic but runs no daily news job, market API or archiving service. Private results remain PRIVATE_PERSONAL_ONLY / OWNER_ONLY; the system overview is not a reason to publish briefings, raw threads or daily archives. The 76 reported candidate checks replace neither live acceptance nor target runtime, persistence or production evidence.