Changelog¶

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

v0.7.0 (2026-09-19)¶

Feat¶

convert: tiny-polygon accumulator — coarse levels keep the dropped area (#384) (#394)
pyramid: bands in different layers may share a zoom range — multi-layer archives (#385) (#392)
--include-property / --exclude-property / --exclude-all-properties (#386) (#391)
convert: attribute-driven entry zoom — the magnitude ladder (#364) (#375)
pyramid: tile GeoParquet bands directly, one shot (#345) (#368)
export: add --feature-order to pin within-tile draw order (#361) (#366)
convert: add --verbatim, tiling an input exactly as given (#345, #360) (#367)
pyramid: merge per-band archives into one multi-band pyramid
export: default per-tile size cap at 500K (tippecanoe parity)
filter: timestamp column support in --filter
cli,python: expose --representation and --collapse-square (#317, #279)
overview: zoom-band representation selector + tiny-polygon placeholder squares (#317, #279)
convert: attribute filter --filter/--where with stats pushdown (#315)
cli: ergonomic one-step tiles — --keep-overview, controllable temp, visible disk cost (#314)
cli: tiles tuning parity — expose --gsd and --report (#316)
core: warn when remote-input spill dir is on tmpfs/ramfs (#273)
core: AWS_SKIP_SIGNATURE hint on signature errors at custom endpoints (TDD green)
cli: friendly single-file-only rejections on validate/decode/export-pmtiles (TDD green)
core: derive default layer name from multi-partition inputs (TDD green)
python: overview() accepts list[str] input (explicit ordered parts)
cli: --files-from manifest on tiles and overview
core: remote prefix listing, --files-from manifest core, store sharing
core: retune coarse-zoom polygon defaults from the #259 sweep
core: plumb ConvertSource through the streaming pipeline
core: ConvertSource abstraction + local multi-partition resolution
core: configurable input-spill directory + free-space guard (#272)
ci: attach prebuilt CLI binaries and real release notes to GitHub Releases
core: warn on large full-file remote convert (#267)
export: default simple_clip_fastpath on; add --no-simple-clip-fastpath opt-out (#256)
cli: expose overview tuning + memory flags on the tiles one-shot command
export: incremental checkpoints + per-level progress logging (#229)
core: cascade provenance, spec + tuning docs, Python surface (#218)
core: cascade flag, simplify_cascade fold, i_overlay candidate repair (#218)
remote input support — convert directly from s3://, https://, gs:// (#210)
core: opt-in zoom-scaled per-level row-group sizing policy (TDD green)
cli: add decode subcommand (PMTiles -> GeoParquet)
core: PMTiles -> GeoParquet decode module (#112)
core: add PMTiles read-side primitives
warn on antimeridian-suspect features at convert time + user docs (#188)
bench: remote-storage (S3 range-request) access benchmark
core: enforce strict cluster sum invariant (spec §12.1)
core: emit complete coalescing provenance and ranks as object map
core: stamp overview footers with spec version 0.2.0
python: re-point convert() at the overview facade
cli: reimplement tiles as a facade over overview convert -> export
python: expose the overview pipeline (E3)
core: default junction continuation OFF per maintainer sweep review
core: chain density budget, junction continuation, coalescing on by default
cli: add --coalesce-lines, --coalesce-snap, --coalesce-max-level-rows
core: wire line coalescing through both overview pipelines
core: add line network coalescing engine (Q3)
cli: default point-thinning 16.0 when clustering
core: validate clustering metadata (point_count column + values)
cli: --cluster and --accumulate-attribute flags on overview
core: per-level point clustering with point_count + numeric aggregation
cli: expose --no-streaming and --read-batch-size overview flags
core: two-pass streaming overview convert (H3)
overview: per-level row-group sizing + string/geometry stats suppression (H1)
overview: add export-pmtiles (E0)
overview: add Q2 density-based per-level budget
overview: retune line-thinning default 2.0 -> 1.0
overview: expose --gsd-base knob, document tuning, add true-scale renders
overview: class-aware cell-winner ranking (Q1)
cli: add overview and validate subcommands
core: add overview convert pipeline (P5)
overview: add overview reader and conformance validator (P4)
overview: add level metadata model and GeoParquet overview writer (P3)
overview: add world-space GSD-driven simplification (P2)
overview: add pure level-assignment engine (P1)
simplify: add coalesced-linestring simplification and noop removal
python: add simplify_factor to Python bindings
cli: add --simplify and --simplify-factor flags
pipeline: integrate simplify_geometry_for_tile into encoding
simplify: add unified simplify_geometry_for_tile helper
simplify: add boundary-preserving polygon ring simplification
simplify: add boundary-preserving linestring simplification
simplify: add tile boundary detection
core: add simplify_factor to TilerConfig
core: Wave 3 - adaptive retry loop and zoom propagation
core,cli: Wave 2 - adaptive threshold integration
core: Wave 1 - adaptive threshold infrastructure
coalesce: implement property passthrough for keep-first/strict modes
coalesce: wire predictive architecture into pipeline
pipeline: implement tile-level geometry coalescing
pipeline: wire coalesce_config through tile encoding
add CoalesceConfig and CLI flags for geometry coalescing
core: add SpatialGrid and CoalesceTargets for coalescing
core: implement coalesce_geometries for geometry merging
core: add covering_tiles and percentile utilities
cli: add --memory-budget flag to control sort segment size
core: implement disk-spilling external sort for memory-bounded processing
core: implement spatial bucketing for memory-bounded tile generation
core: add row group filtering and memory-bounded processing infrastructure
core: improve error messages for directory path failures
support partitioned GeoParquet directories
add zoom_range_for_bbox() unified min/max zoom function
add min_zoom_for_bbox() for visibility-based zoom filtering
add per-feature max zoom optimization (--auto-max-zoom)
core: skip clipping for features fully contained within tile bounds
core: add drop-smallest filtering to legacy TileIterator path
core: add drop-smallest filtering to streaming path
core: implement drop-smallest filtering in encode_tile_from_raw
config,cli,python: add drop-smallest-as-needed configuration
core: add geometry_pixel_area_world dispatcher for all types
core: add Sortable trait to TileRef for external sorting
core: add TileRef for lightweight sorting
core: add disk-backed GeometryStore for bounded memory

Fix¶

core: raise the glob floor to 0.3.4 and convert with into
export: declare the requested min zoom even when coarse levels generalized to nothing (#380) (#390)
mvt: clean polygons in tile space after quantization (#383) (#393)
export: withhold coalesced_count from tiles when it is 1 everywhere (#379) (#389)
pmtiles: apply contiguous-offset rule to leaf directory entries (#377) (#378)
decode: name the section or tile that failed to decompress, with byte offset (#381) (#388)
export: publish renamed source columns under their source name (#359) (#365)
core: release the producer before joining it in the streaming engines (#362) (#363)
declare the real MSRV (1.95, was 1.75) and verify it in CI
input: import ObjectStoreExt for object_store 0.14
pyramid: rebuild the band merge around review findings
cli: teach the bare-form rewrite about pyramid
pmtiles: never write leaf_dirs_offset = 0
overview: box LevelSink::Spill for arrow 59
pmtiles: split oversized directories into leaves in PmtilesWriter
export: encode UTC-stamped timestamps instead of advertising nulls
export: encode DATE, TIMESTAMP and DECIMAL instead of dropping them
export: stop out-of-domain bboxes claiming every tile column
export: --tile-buffer is tile pixels, not MVT extent units
export: derive tile membership from the buffer-expanded bbox
deps: pin aws-smithy-types below 1.7.0, unbreaking Check
ci: pin rust-toolchain at the v1 tag, unbreaking CI Lint
python: bump pip past PYSEC-2026-3721, unbreaking Python Quality
input: replace deprecated GlobError::into_error, unbreaking CI on main
python: sync tylertoo.pyi tile_size_limit default to 512000
convert: assume OGC:CRS84 for explicit null GeoParquet crs
python: add filter kwarg to overview type stub
overview: auto-rename reserved-column collisions instead of rejecting (#288)
core: review fixes — remote URL classification, CRS/metadata diagnostics, cached single footer
python: add spill_dir to type stub
core: stop re-fetching oversized remote column chunks per page (#261)
core: enable allow_http for http:// remote inputs
clip: keep MultiPolygon parts that split into pieces at tile bounds (#244)
core: address CI gates for #211 clamp (clippy complexity, must-use, xenon)
core: auto-clamp empty overview levels instead of failing (#211)
green the mypy and security-audit gates for remote input
python: add bbox parameter to overview() type stub
core: clippy/MSRV cleanups in bbox filter code
python: wire bbox regional extract through the Python bindings
core: emit spec-compliant MVT polygon ring winding
core: migrate to arrow/parquet 58 APIs for geoarrow 0.8
bench: migrate deprecated criterion::black_box to std::hint::black_box
core: validate geo:overviews footer against actual row groups on open
core: reject nonsensical overview conversion knobs up front
core: row-align overview winner tables around skipped hostile geometries
core: use Web Mercator Y for latitude in tile-local coordinate transform
clippy: last 1.96 lint in debug_antarctica test
clippy: appease Rust 1.96 lints (unnecessary_sort_by, manual_checked_ops)
clippy: resolve -D warnings failures across overview + simplify
overview: reject case-insensitive level column collisions (F2)
test: increase memory budget to 500MB for tarpaulin overhead
test: increase memory budget for CI environments
core: fix clippy len_zero warning in test
core: improve CannotReduceFurther error with actionable suggestions
core: improve CannotReduceFurther error with actionable suggestions
core: complete adaptive threshold algorithm integration
core: track actual RSS instead of throughput for memory reporting
coalesce: address adversarial review findings
address clippy warnings
core: add gap-based dropping to clustering path + tests
core: sort records by original Hilbert index for gap-based dropping
core: make temp file names unique with pid/tid for parallel safety
core: remove unused segment_idx field in external sort
core: prioritize exact "geometry" column match over partial
core: apply PR #116's multi-file process_geometries_parallel
use single pipeline for all files, not separate pipelines per file
core: apply PR #116 directory support to get_row_group_count
core: update extract_crs to handle directory paths
update test calls after rebase on PR #133
test: use array instead of vec for fixed-size test data
bench: remove unnecessary mut and casts in tile_ref benchmark

Refactor¶

core: split the four conversion drivers into named phases
core: fold away the near-identical writer and codec pairs
core: sweep dead code and tighten the public surface
core: remove legacy per-tile pipeline
tiles: complete excision — drop CLI flags + lib re-export
tiles: remove unproven zoom-dependent simplification (#158)
overview: unify Crs and METERS_PER_DEGREE in level.rs
cli: replace --auto-max-zoom with --zoom-by-area flags
core: replace auto_max_zoom with zoom_by_area in TilerConfig

Perf¶

export: halve member-store fill budget to 128 MiB
export: single-read fan-out pass 2 for partitioning mode (#235)
convert: bound pass-1 winner-grid memory with budgeted level waves (#306)
export: size partition wave from densest partition, not flat constant (#311)
writer: parallelize GeoParquet WKB encode (#304)
export: memory-budget preflight for auto partition wave (#303)
convert: size auto RAM-vs-spill from measured geometry (#305)
export: auto-scale partition wave to available cores (#293)
convert: recalibrate auto buffered-output estimate to measured cost (#294)
convert: add per-phase RSS instrumentation (#295)
convert: make auto profile pick backing from workload, not mode (#294)
core: parallelize parquet row-group encoding (#296)
core: fuse pass-1 geometry traversals in streaming convert (#274)
convert: --in-flight-batches auto — core-aware pass-2 concurrency
remote: stage selected row groups up front to coalesce and parallelize fetches (#286, #287)
core: parallelize explicit-list remote connects (TDD green)
core: spill remote input to disk to bound re-reads to ~1× (#219)
core: parallelize convert's dominant serial stages (#264)
bench: refresh big-file profile, add sweep harness + convert guard + docs
export: opt-in simple-clip fast path skipping i_overlay fallback (#239)
clip: replace O(V²) self-intersection scan with sweepline (#241)
convert: cap O(V²) validity check on oversized RDP candidates (#242)
export: fix O(V²) clip stall on large-extent polygons (#237)
export: single-read fan-out scan, kill cross-level prefix re-read (#233)
export: wave-batched shared read, kill per-partition re-reads (#228)
export: parallelize scan_level and gzip compression (#227)
export: top-down recursive tile splitting for large polygons (#226)
core: cascading simplification in all three pass-2 paths (#218)
core: single-read pipelined pass-2 engine + --profile presets (#213, #212)
bench: read AWS bucket/region/profile from env vars
bench: four-way overview layout benchmark + column-per-zoom prototype
bench: row-group sizing sweep against real S3 (#202)
core: add --bbox row-group filtering for regional extracts (#102)
bench: demonstrate the remote latency floor with a parallel range-request reader
core: stream finished tile partitions to PMTiles writer (H3b)
core: parallelize export clip and MVT encode with rayon (H3c lever 2)
core: export bbox-containment clip fast path (H3c lever 4)
core: parallelize pass-2 simplification with rayon (H3c lever 1)
core: cut validation waste in overview simplify (H3c lever 3)
core: H3(c) wall-time profile of overview pipeline
core: add row-level predicate pushdown for spatial filtering
core: add column projection for geometry-only reads
core: revert to extsort crate to fix 7x performance regression

v0.6.0 (2026-03-11)¶

Feat¶

implement point clustering with position averaging (#25)
implement accumulator system for attribute aggregation (#23)
implement gap-based density detection (#24)
support WKT geometry encoding (#35)
implement tiny polygon accumulation (#85)
profiling: add fine-grained spans to read_parquet phase
add time profiling with tracing
add memory profiling with dhat

Fix¶

remove needless borrow in benchmark
resolve clippy warnings and unused imports
remove WKT fixture from repo, tests skip when missing
gracefully skip WKT tests when fixture is missing
skip profiling integration tests when dhat-heap feature enabled
use tempfile crate for proper temp directory isolation in tests
use cross-platform temp directories in integration tests

Refactor¶

replace wagyu-rs with i_overlay for polygon clipping

Perf¶

parallel row group I/O for ~24% speedup
reuse file handle across row groups (#41)
parallelize tile encoding in Phase 3 (#90)

v0.5.0 (2026-03-10)¶

Feat¶

core: wire up pipeline to use WorldCoord throughout (Phase 2)
core: add WorldCoord-based hierarchical clipping (Phase 2)
core: add WorldCoord-based feature drop functions (Phase 2)
core: add WorldCoord support to MVT encoding and validation (Phase 1)
core: add WorldCoord support to clipping modules (Phase 1)
core: add WorldCoord-based simplification functions (Phase 1)
core: add WorldCoord type for 32-bit integer coordinates (Phase 0)
clip: integrate wagyu-rs for robust polygon clipping
add --deterministic flag and fix PR #63 review feedback

Fix¶

clip: enable U-shape split test with wagyu-rs v0.2.1
clip: add wagyu fallback for edge case geometry handling (#94)
change default compression to gzip and add CRS validation
change default compression to gzip for compatibility
implement leaf directory support for large PMTiles archives
core: clamp tile coordinates and bounds to valid ranges
core: fix issue #83 - geometry coordinates collapsing to zeros
resolve clippy warnings in tests
core: align feature_drop coordinate precision with MVT encoding
ci: download fixtures from release instead of LFS
ci: remove 1.8GB fixture from LFS to fix bandwidth quota
Remove unused MIN_EXPECTED_TILES constant
Remove #[ignore] from regression tests - fixture is in LFS
use is_empty() instead of len() > 0 for Clippy
Fix clippy warning and add clipping benchmarks
tile: clamp latitude to Web Mercator bounds
use wagyu-rs from crates.io instead of path dependency
ci: update benchmark group names after consolidation

Perf¶

Add pre-clip bounding box filter for large geometries
Implement hierarchical clipping across zoom levels
Replace Wagyu with Sutherland-Hodgman for tile clipping

v0.4.0 (2026-02-25)¶

Feat¶

python: add progress callback support
python: add streaming mode and parallel control parameters
python: add property filtering and layer name parameters

Fix¶

add version sync safeguards and fix pyproject.toml version mismatch
add clippy allow for too_many_arguments and add clippy to pre-commit
use workspace dependency for gpq-tiles-core version

v0.2.0 (2026-02-25)¶

Fix¶

release: complete v0.2.0 release setup

v0.1.0 (2026-02-24)¶

Feat¶

set up commitizen for automated versioning and releases
quality: warn about pathologically small row groups
default to zstd compression, expose parallel options in CLI
add progress bars for cleaner output
parallelize geometry processing within row groups (#37)
parallelize tile processing for large geometries (#33)
cli: add --streaming-mode flag with progress reporting
pipeline: implement ExternalSort streaming mode
pipeline: add StreamingMode::ExternalSort variant
core: add external sort and WKB serialization modules
streaming: add StreamingPmtilesWriter with LowMemory mode
streaming: add memory budget configuration and tracking
streaming: add row-group-based streaming tile generation
quality: add GeoParquet file quality detection for streaming
add tile deduplication with XXH3 hashing and run_length encoding
add compression options (gzip, brotli, zstd, none) for PMTiles output
add property filtering with --include/-y, --exclude/-x, --exclude-all/-X flags
add 17K feature fixture for parallelization benchmarks
add tilestats metadata to PMTiles output
auto-extract field metadata from GeoParquet schema
add field metadata support to PMTiles writer
derive layer name from input filename, add --layer-name CLI flag
complete Phase 5 Python bindings with uv/ruff tooling
add benchmark suite with generate_tiles_from_geometries API
pipeline: add Rayon parallel tile generation
pipeline: wire spatial indexing into tile generation
spatial-index: add space-filling curve sorting for efficient tile generation
complete Phase 3 with density-based dropping
integrate feature dropping into pipeline (Phase 3)
add point thinning (1/2.5 drop rate per zoom)
add line dropping (coordinate quantization algorithm)
implement tiny polygon dropping with diffuse probability
implement PMTiles v3 writer (Tasks 7-9)
implement tiler pipeline wiring clip → simplify → MVT
implement MVT encoding for vector tiles
add golden comparison tests against tippecanoe output
implement geometry clipping with correct BooleanOps
implement zoom-based simplification
implement Arrow-native geometry batch processing (TDD green)

Fix¶

release: add version to core dep, add READMEs, fix benchmark filter
release: install protoc inside manylinux container
python: copy README into crate for sdist builds
ci: use tag triggers for release, skip slow benchmarks
consolidate release workflows and fix version bump detection
guard against degenerate linestrings in simplify and fix flaky test
PMTiles now compatible with pmtiles.io and standard viewers
golden: update stale Z8 test to use full pipeline
resolve three medium/low priority issues
simplify geometry in tile-local pixel coordinates
handle antimeridian crossing in tiles_for_bbox
clip: preserve all polygon parts when clipping produces MultiPolygon
use real bbox calculation instead of world bounds
resolve CI timeouts and coverage linker errors
upgrade pyo3 0.24 → 0.28 for Python 3.14 support
resolve CI failures for benchmark, check, test, and security audit

Perf¶

streaming: add memory benchmarks for streaming pipeline