FastChunkGen: Project Infinity 0.1 config options

Config options (how Project Infinity 0.1 sets them)

44 settings this pack ships for the mod. We don't have the mod's own defaults to compare them against yet.

Show settings
SettingProject Infinity 0.1 value
fixes.chunkIoThreadingIssues
(Default: true) Fix thread-safety issues in vanilla chunk IO.
config/fastchunkgen.toml
default
fixes.disableLoggingShutdownHook
(Default: true) Remove the log4j2 shutdown hook so shutdown logs survive, and force exit if a dedicated server hangs.
config/fastchunkgen.toml
default
fixes.enforceSafeWorldRandomAccess
(Default: false) Crash instead of warn when something uses the world random off-thread. For debugging.
config/fastchunkgen.toml
default
fixes.generalThreadingIssues
(Default: true) Fix thread-safety issues in the vanilla chunk system.
config/fastchunkgen.toml
default
fixes.vanillaWorldGenBugs
(Default: true) Fix vanilla world generation bugs.
config/fastchunkgen.toml
default
fixes.worldGenThreadingIssues
(Default: true) Fix thread-safety issues in vanilla world generation. Required by threadedWorldGen.
config/fastchunkgen.toml
default
generalOptimizations.autoSave.delay
(Default: 20000) Milliseconds before a dirty chunk is auto-saved. ENHANCED only.
config/fastchunkgen.toml
default
generalOptimizations.autoSave.mode
(Default: ENHANCED) VANILLA, ENHANCED (save during idle time) or PERIODIC (every 6000 ticks). Keep the quotes.
config/fastchunkgen.toml
default
generalOptimizations.midTickChunkTasksInterval
(Default: 100000) Nanoseconds between chunk tasks run inside the tick loop. -1 disables.
config/fastchunkgen.toml
default
generalOptimizations.optimizeAsyncChunkRequest
(Default: true) Serve off-thread chunk requests without blocking the server thread.
config/fastchunkgen.toml
default
generalOptimizations.optimizeNoiseSampling
(Default: true) Faster Perlin noise sampling. Produces identical terrain.
config/fastchunkgen.toml
default
generalOptimizations.optimizeScheduling
(Default: true) Optimize chunk task scheduling.
config/fastchunkgen.toml
default
generalOptimizations.overrideModernFixWorldGenAllocations
(Default: true) Apply our world generation allocation patches ahead of ModernFix's equivalent ones. Disable to let ModernFix win instead.
config/fastchunkgen.toml
default
generalOptimizations.reduceAllocations
(Default: true) Reduce allocations in hot paths.
config/fastchunkgen.toml
default
globalExecutorParallelism
(Default: 19) Worker threads used for chunk generation and IO.
config/fastchunkgen.toml
default
ioSystem.async
(Default: true) Load and unload chunks off the server thread.
config/fastchunkgen.toml
default
ioSystem.chunkDataCacheLimit
(Default: 32768) Hard cap on cached chunk writes.
config/fastchunkgen.toml
default
ioSystem.chunkDataCacheSoftLimit
(Default: 8192) Cached chunk writes before flushing to disk starts.
config/fastchunkgen.toml
default
ioSystem.chunkStreamVersion
(Default: -1) Compression for newly saved chunks. -1 vanilla, 1 GZip, 2 Zlib, 3 none. Existing chunks stay readable.
config/fastchunkgen.toml
default
ioSystem.fireForgeChunkDataEvents
(Default: true) Fire Forge ChunkDataEvent.Save on async saves. Off means mods lose per-chunk data.
config/fastchunkgen.toml
default
ioSystem.forgeChunkDataEventsOnMainThread
(Default: true) Fire Forge chunk data events and deserialize chunk capabilities on the server thread. Off is unsafe for most mods.
config/fastchunkgen.toml
default
ioSystem.gcFreeChunkSerializer
(Default: false) Allocation-free chunk writer. Cannot fire Forge ChunkDataEvent.Save, so mods lose their per-chunk data. Back up first. Set to false for the following reasons: Incompatible with architectury@9.2.14 (*) (defined in fastchunkgen)
config/fastchunkgen.toml
default
ioSystem.optimizations
(Default: true) Enable chunk IO optimizations.
config/fastchunkgen.toml
default
ioSystem.recoverFromErrors
(Default: true) Regenerate chunks that fail to load. Destroys their contents.
config/fastchunkgen.toml
default
ioSystem.replaceImpl
(Default: true) Use the rewritten IO implementation.
config/fastchunkgen.toml
default
ioSystem.serializeBlockEntitiesOnMainThread
(Default: true) Snapshot block entity NBT on the server thread when a chunk save starts. Disabling serializes them on IO workers instead, which is faster but races modded block entities and Forge capabilities.
config/fastchunkgen.toml
default
ioSystem.skipMissingExternalChunkDelete
(Default: true) Check that the oversized chunk file exists before deleting it after each chunk save. Avoids an internal exception per save on Windows.
config/fastchunkgen.toml
default
maxViewDistance
(Default: 32) Highest view/render distance allowed (vanilla caps at 32, max 125). Higher values need more memory.
config/fastchunkgen.toml
default
threadedLighting.enabled
(Default: true) Run light updates on a dedicated thread.
config/fastchunkgen.toml
default
threadedWorldGen.allowGenerationCancellation
(Default: true) Abort an in-progress generation step when the chunk is no longer requested. Set to false if a pregenerator reports skipped chunks.
config/fastchunkgen.toml
default
threadedWorldGen.allowThreadedFeatures
(Default: true) Generate features (trees, ores, structures) in parallel.
config/fastchunkgen.toml
default
threadedWorldGen.asyncScheduling
(Default: true) Schedule generation tasks off the server thread.
config/fastchunkgen.toml
default
threadedWorldGen.enabled
(Default: true) Generate chunks on multiple threads.
config/fastchunkgen.toml
default
threadedWorldGen.reduceLockRadius
(Default: true) Lock fewer neighbouring chunks per generation step. More parallelism.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.cacheDensityFunctionHashes
(Default: true) Cache the hashCode of density function and spline records. They are immutable, so the value never changes.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.enabled
(Default: true) Optimize vanilla world generation.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.identityPreservingDensityFunctionMapAll
(Default: true) Reuse density function nodes whose inputs did not change when a chunk maps the noise router instead of rebuilding every node per chunk, and skip the throwaway holder pair NoiseChunk unwraps immediately.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.optimizeAquifer
(Default: true) Faster aquifer sampling.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.optimizeRandomInstances
(Default: true) Cheaper random instances during world generation.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.reuseSurfaceRules
(Default: true) Build the surface rule tree once per worker and point it at each new chunk and carved block instead of rebuilding every rule object each time.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.shareNoiseBetweenChunks
(Default: true) Compute noise that neighbouring chunks both need once instead of in each chunk.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.skipRedundantDensityFunctionWrapping
(Default: true) Skip the NoiseChunk wrapper map for density functions it would hand straight back. Only marker nodes own a cache that has to be shared.
config/fastchunkgen.toml
default
vanillaWorldGenOptimizations.useEndBiomeCache
(Default: true) Cache End biome lookups.
config/fastchunkgen.toml
default
version
config/fastchunkgen.toml
3