Engineering, operations & design history
Original-native followup: ten production-pool contexts and one private alternate context
Full repository source: docs/art-direction/canonical-native-followup-20261105/review.md
Technical documentation includes historical decisions, proposals and verification reports. For playing instructions and practical rules, open Gameplay.
No matching entry. Try a shorter phrase or clear the search.
Original-native followup: ten production-pool contexts and one private alternate context
Eleven individually viewed original assets retain their native artwork in 44 normal local desktop/phone Region/Hex views. Ten contexts use the unchanged 39-source/322-permission production pool. The original cold arch uses the separate 40-source/323-permission private pool in a legitimate alternate disposable world; it adds no production-pool map coverage. This package changes no artwork, production renderer, source permissions, saves or geography.
The ten production-pool contexts cover the cohesive city ward in fields, desert arches in desert, tundra longhouse, arctic glyph triptych, crown keep in grasslands, tundra crescent village, tundra barn, cohesive city gate in hills, stone market ward in mountain and stone gate cluster in wetlands. Every source has its own exact-hash authored finding in visual-review.json. Existing native context sheets are referenced without copying them into this package.
All 44 isolated source replays preserve the exact native input under the actual transform. All 88 same-current-state native-reference/restoration pairs are exact. These are pairs of comparisons: 176 individual comparisons. Only 26 of 66 raw initial-return maps are exact; preserve the remaining differences, camera/interaction records and cache evidence. Exact current-state references do not approve raw lifecycle behavior, whole-artwork composition, physical-device behavior or complete memory/performance cost.
Cumulative actual-production-pool fidelity coverage is 23 of 113 source/profile/feature contracts. The remaining 90 default-world contracts are running sequentially in direct-production-all-default-contracts-v1; none of that active run is counted as a terminal result here. Keep the complete private 114-contract plan separate. Availability in a selector or an image paint call alone does not prove final composited world contribution.
The original images retain their complete subjects and native detail. Ground grain scale, foundation/shore/snow contact, full UI and shadow clearance, neighboring composition, seasonal and state suitability remain individual gates. The visible arctic triptych does not resolve the separately retained foreground/cache-owner counterexample. Preserve the full city footprints and arch rubble instead of trimming the artwork to fit a tile. No new game-ready approval is granted.
The two terminal cohorts passed source and protected-file hash checks and closed every owned browser/server. Ten redundant JSON hints totaling 2,696,214 bytes were retired only after complete final results and native witnesses were verified. No raw GPU/RGBA readbacks, catalog snapshots or image duplicates were added. Original artwork, approved references, both Citadel hamlets, source files, persistent saves and the bounded QA-world inputs remain intact.
Local metadata registration can proceed while the preceding canonical-production-native-qualification release builds. Validate its exact eleven evidence leases first; preserve the ten leases for other assets and replace only the cold arch's queued lease with its newer private-context finding. This produces 21 unique queued assets: the prior release's exact-hash closeout can clear ten matching old leases and must preserve all eleven newer findings. Its published commit and asset tree remain unchanged. Publish the followup only after the earlier attempt is terminal and its release state is reconciled; do not claim local findings are live before deployment and live checks. The full library remains 9,215 present catalog images, 9,240 reconciled ledger records and 5,643 pending complete qualifications.
Evidence: analysis.json and visual-review.json bind the terminal result, authored review and retirement receipts. The source cohorts are production third ten ../canonical-native-regression-20261105/direct-production-sources-third10/analysis.json and private alternate cold arch ../canonical-native-regression-20261105/alternate-seed-private-coldarch/analysis.json. The preceding production qualification review remains a separate release.