- Rarity stream picks are property-tested against the VolmLib rarity selection
- Pregen capacity growth scales the bake by the same factor as engine start
- Rarity selection streams pick from precomputed cumulative tables
- Implode and surface biome fixes take primitive coordinates
- Child selection reuses one derived child seed instead of a new RNG per level
- Base biome selection reads bridge and inferred streams directly
- Planes fill lazily per column with ready bits and batch only missing columns on full-chunk requests
- Natural fills batch only the base heights Terrain3D still needs and reuse the sparse memo
- Bounds lattice and hydrology planner reads stay out of the bake
- Bake capacity is eight times the noise cache, covering the measured working set
- Generator evaluation counters and a bake census print from the probe
- Parity test compares batched chunk base heights and evaluation counts with the scalar path
- Compiled generator programs evaluate layer trees in batches through NoiseProgram
- TerrainProgram accumulates interpolator group means over chunk columns
- ChunkBakeCache holds compute-once per-chunk planes with provisional results never published
- Height, biome, region, rock, fluid and cave streams read the bake as thin facades
- Sparse hydrology and bounds reads stay scalar with a direct-mapped base height memo
- ColumnBake assembles id planes, heights, fluid surface, columns and the apron on demand
- Fell back to per-group grid sampling when a height sample nests inside another on the same thread
- Covered nested sampling with a bit-exact parity test
- Read interpolator corner bounds through a per-thread chunk block and skip groups whose cell corners are all zero
- Prefilled chunk context caches inline on generation threads with heights ahead of biomes
- Kept the burst-pool prefill only for the platform main thread, with heights and biomes in one task
- Painted surface fluid from the chunk context cache instead of re-sampling the fluid stream
- Removed the unused decoration, derivative and identity streams and Engine.getBiomeID
- Skipped the shore-height noise when a column sits outside the proven shore band
- Dropped interpolator groups whose bounds are zero in every base-capable biome
- Stop writing cave ceiling markers, which nothing reads
- Write cave floor markers only when the dimension has cave spawners
- Reuse the zone floor subterrain cell for zone decoration
- Place ellipsoid deposit clumps from the footprint cells instead of rebuilding the clump object
- Skip ore cave-biome lookups when no pack biome scales ore deposits
- Reject terrain-limited clumps before biome work when scaling cannot apply
- Compute the chunk terrain limit once per chunk
- Draw the exposure discard roll only when the discard chance is positive
- Mod targets run through bounded Gradle workers with separate caches and daemon slots
- Target logs retain verification failures and bounded output summaries
- Nested tests use bounded forks and JVM thread counts
- Lifecycle tests control deadlines through injected clocks
- Timeout overrides retain original values for restoration
- Timeout restoration preserves operator changes and symbolic links
- Tests cover timeout persistence and failed configuration writes
- Runtime targets use separate Gradle caches and shared module outputs
- Modded build inputs include test fixtures and runtime cache settings
- Compact connected roots extend bounded gaps into supporting terrain
- Unsupported cliff placements leave no partial trees
- Root and canopy roles follow object scaling and rotation
- Canopies preserve solid trunk cells across chunk replay and saved continuations
- Branch wood replaces leaf clusters within each tree
- Regression tests cover slopes and overlapping tree placements
- Gameplay checks compare grounded spindle geometry across both packs
- Added DeferredPlacementRecords to gate tracking on Identifier and TileWrapper slices
- Cleared deferred custom and tile records at positions native structure placement changed
- Implemented block change tracking in the Bukkit and modded structure policies
- Covered invalidation through the custom block materialization pass
- Covered modded wrapper change tracking and protected cells
- Passed disabled change sets in existing wrapper tests
- Skip deferred custom block records whose live block no longer matches the provider carrier type
- Leave air-carrier furniture and unresolvable carriers to the provider unchanged
- Resolve each carrier once per chunk pass and log skipped records once per chunk at debug level
- Skip tile records quietly when the live block no longer matches the tile material
- Apply the same carrier check to the modded custom block pass
- Add Bukkit, CraftEngine and modded regression tests for flattened, replaced, intact and furniture records
- Truncated or malformed generation manifests and epoch metadata report as invalid history again after the strict reader change
- The datapack metadata test expects the range and supported_formats to start at the 1.21.1 format
- Merge feature/modded-versions into the modded fix integration
- Take loader pins from gradle/mod-targets.json for every target
- Pin the 26.3 NeoForge target to 26.3.0.83
- Build Fabric, Forge and NeoForge per target listed in gradle/mod-targets.json (1.21.1, 26.3)
- Add 1.21.1 adapter source overlays under adapters/*/src/versions/1.21.1
- Add verifyModdedMatrix and per-target Java runtime verification of loader jars
- Re-inject persistent dimensions from the levels-ready hook before the watchdog starts
- Wait for queued chunks and the final save before pregen completes
- Add the 1.21.1 data version and datapack supported_formats range
- Cascade regions whose sea or shore pool loses every biome
- Use Gson calls available on the 1.21.1 runtime
- /locate biome on non-stacked Iris dimensions searches the visible biomes chunks are filled with instead of falling back to the vanilla structure-biome search
- BiomeSearchFilter rejects columns whose region, land, sea, shore, cave, carving, subterrain or image-map biomes cannot hold a target, so terrain height is computed only for columns that can match
- Hydrology planning state is consulted only when a target comes from a river policy, so a locate no longer queues hydrology tile plans across its whole radius
- Admitted samples resolve inside their generation-history coordinate scope, so saved-biome reads for generated chunks no longer fail the command
- HintedLocator surface hints and IrisBiome visible key enumeration are shared with the locate filter
- Bukkit and modded biome policies share the same filtered search
- Saved Iris dimensions are prepared in parallel during world load (server-about-to-start on NeoForge and Forge, server-starting on Fabric), before the server watchdog runs
- Server start only creates the restored levels, builds their engines in parallel, and publishes them, so its wait no longer grows one world at a time
- Each restored dimension opens its generation history once instead of twice and resolves its active pack once instead of three times
- An epoch snapshot whose content matches its validated source pack reuses that validation instead of running the full pack validator for every world
- Level creation reads the configured pack and structure biome keys from data warmed during preparation
- Immutable generation packs feed the prepared hydrology cache identity from the data's memoized fingerprint instead of hashing the pack twice per engine build
- Restore logging reports each dimension's preparation and engine time and how long the server thread waited
- Themed regions, those at least five times rarer than the dimension's median region, no longer receive adopted biomes while a common region accepts the biome's role
- Themed regions still take a mod biome when no common region has a pool for its role or every common template is full
- Category affinity now chooses among common regions, so a rare biome is no longer locked inside a region that can be absent for thousands of blocks
- Each pack biome now serves as the template for at most five adopted land biomes and three adopted sea, shore, or cave biomes
- A mod biome whose best template is full falls back to the next best template under the cap
- A mod biome that finds every compatible template full is skipped with a reason naming the cap instead of being dropped
- Converted patch decorators also lose the share of surface columns that earlier surface patches in the declaring biomes already occupy, estimated from those patches' expected coverage in the catalog
- Canopy and surface occupancy are combined into one vegetation occlusion estimate per decorated feature
- For land biomes with known categories, the region climate tie-breaker averages only the region's templates that share a category with the mod biome, so a mixed region no longer looks temperate on average
- A small region affinity term favors regions whose templates share the mod biome's categories
- A mod biome that brings its own vegetation pays a template penalty scaled by the share of the template's top surface layer that cannot hold plants, so snow, ice, sand, and stone templates no longer swallow its trees
- The top layer palette is resolved through biome-palette snippets and weighted by palette entry weight
- Mod biomes without their own vegetation keep the closest climate template
- Ores declared only by adopted cave biomes are written to the deposits of the region each cave biome joined, scoped to those cave biomes, instead of to the cave biome itself where the deposit stage never reads them
- Ores of adopted surface biomes still go to the biome
- The worldgen catalog (format 2) records the members of every block tag referenced by a captured placed feature, so ore targets using mod block tags convert instead of passing through
- Ore conversion supports noise_based_count and noise_threshold_count placement counts
- Converted ores that produce another converted ore's host block are written ahead of it, so an ore whose host comes from a sibling ore finds that host during the deposit stage
- Patch decorators converted from features that snap to a leaf-blocked heightmap lose the share of placements that land on tree canopy in the declaring biomes
- Canopy is estimated from the tree-like features that run before the decoration in each declaring biome's vegetal step, using their expected per-chunk count and a typical crown footprint
- Patches that snap below leaves or run before the trees keep their full density
- importedFeatures gains biomeRules, which deny placed feature families or exact keys only in registry biomes matched by key, namespace, or tag; Iris custom biomes match through their vanilla derivative
- Mod ingest classifies tree-like passthrough features from the catalog feature JSON, including trees nested in random selectors, fallen trees, and huge mushrooms
- Tree-like mod features that a mod injects into vanilla biomes the pack already uses are denied in minecraft biomes, so they no longer double the pack's own trees while adopted mod biomes keep them
- Studio schema digest updated for the new rule shape
- Mod biomes pick their template across all regions by climate distance, biome category agreement from minecraft and c tags, a reuse penalty that spreads biomes over equally good templates, and the region's overall climate
- Regions much rarer than the dimension's median region are treated as themed and only receive a mod biome when nothing else fits
- Templates derived from void biomes no longer act as climate references
- /iris ingest mods accepts excludeRegions=<a,b> to keep mod biomes out of named regions; unknown region names are refused
- The mod ingest summary and manifest namespace list only include namespaces that produced at least one ingest item, so mods without worldgen content no longer print empty summary lines
- Ores converted by mod ingest use the FULL_HEIGHT deposit placement scope, so they replace their host blocks anywhere in their height band like vanilla ore features instead of being culled above the column surface
- Pack validation rejects borrowedDimensionType on platforms that generate their own Iris dimension type, with an error naming the supported loaders
- Dimension type keys and registry contracts honor borrowedDimensionType only when the platform generation registry supports borrowed types, so Bukkit always derives its generated iris type
- The Bukkit runtime contract always expects the generated Iris type; hotload still refuses a borrowedDimensionType change
- Mod ingest never selects the irisworldgen namespace or a namespace equal to a dimension key of the source pack or any installed Iris pack, so Iris's own scoped and legacy alias biomes are not adopted
- Explicitly requesting one of those namespaces is refused
- A namespace that belongs to a loaded mod is never treated as an Iris dimension alias namespace
- above_bottom and below_top ore anchors resolve against the pack's own dimension height, or the borrowed dimension type's height when the pack borrows one, instead of the vanilla dimension type
- Converted deposit heights now match the pack-relative rebase, so a 0..256 pack no longer receives out-of-world ore bands
- importedFeatures gains disabledExact, which denies normalized complete placed feature keys without matching key families
- Mod ingest puts converted feature keys in disabledExact, so converting mod:ore_tin no longer suppresses mod:ore_tin_large
- Mod ingest keeps vanilla and unselected mod features generating when the source pack already ran importedFeatures, and otherwise disables vanilla plus every unselected mod namespace that has placed features
- Studio schema digest updated for the new field
- Mod ingest compares source and target by real path and file identity, so a differently cased target on a case-insensitive filesystem can no longer resolve to and delete the source pack
- replace=true only replaces a target that contains ingest/manifest.json
- Ingest builds into a hidden staging directory and swaps it in atomically, deleting the previous output only after the new pack is complete
- IrisDimension borrowedDimensionType reuses an existing registered dimension type instead of generating one
- Pack validation rejects borrowedDimensionType values that are not namespaced registry keys
- Generation contracts record the borrowed type key and capture it as an external registry definition
- Borrowed types bind when the pack min equals the type min_y and the pack top fits, with both ranges in the refusal
- Studio hotload rejects borrowed type changes and the world check validates borrowed ranges
- Forced datapack skips generated types and presets for borrowed dimensions and accepts retained borrowed epochs
- modded.json dimensionTakeovers writes a LevelStem per level key using the irisworldgen:iris generator
- Takeover changes quarantine the level chunk, entity, POI, and Iris data before load and track applied state per world
- Fresh levels adopt generation history with content-addressed biomes instead of legacy aliases
- /iris world takeover <dimension> <pack[:dimension]|off> with help and translations in every locale
- Captured the worldgen catalog after server start, scanning registries on the server thread and encoding and writing on a worker into config/irisworldgen/catalog/worldgen-catalog.json
- Added WorldgenCatalogFingerprint over namespace versions and registry keys
- Added /iris catalog [namespace|refresh] and /iris ingest mods <source> <target> with keyed dimension, namespaces and replace options
- Added catalog messages and help entries for every shipped locale
- Added tests for catalog assembly, namespace summaries, ingest arguments, the registry scan and command registration
- ModIngester copies the source pack with IrisProjectCopier into a new target pack whose folder name becomes the dimension key, honouring replaceTarget and rejecting overlapping or already-ingested sources
- Mod biomes are classified by minecraft/c tags into land, sea, shore, and cave pools, skipping rivers, other-dimension biomes, and biomes missing the dimension tag
- Each adopted biome joins the region with the nearest land climate and copies the nearest-climate template biome with the mod biome as its derivative and no custom-biome overrides
- Template trees and procedural trees are dropped when the mod biome brings its own non-decorator vegetation
- Ore and scattered ore features become Iris deposits per target with resolved height anchors, counts, rarity, vein size, air-exposure discard, and host blocks
- Vanilla-declared ores become dimension deposits filtered to the declaring biomes when they do not cover the whole dimension; mod-only ores become biome deposits
- Surface simple_block and legacy random_patch features become Iris decorators on adopted mod biomes and on pack biomes whose derivative declares them
- Remaining features pass through native generation, with importedFeatures enabled and minecraft plus every converted key disabled
- Passthrough features shadowed by a disabled prefix or a disabled step are reported as skipped
- Mod structures are reported as adopted unless importedStructures disables them or none of their biomes generate
- The target pack records ingest/manifest.json with the catalog fingerprint, namespaces, generated and modified files, and every item
- Tests cover climate placement, biome classification, ore and patch conversion against current 26.3 formats, propagation, passthrough, manifest, idempotent replace, and pack validation
