Skip to content

Latest commit

 

History

History
126 lines (98 loc) · 3.18 KB

CHANGELOG.md

File metadata and controls

126 lines (98 loc) · 3.18 KB

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.

[0.23.1] - 2024-08-02

Fixed

  • "String length exceeds the maximum length" error when querying large GIS data in Livy (#10)

Changed

  • Updated Rust toolchain and minor dependencies

[0.23.0] - 2024-02-11

Changed

  • Upgraded Spark version to 3.5.0
  • Enabled ANSI mode by default

Added

  • Building a multi-platform image for arm64 architectures

[0.22.1] - 2024-01-11

Fixed

  • Coalesce output of RawQuery to avoid a crash on multiple partitions

[0.22.0] - 2024-01-10

Changed

  • Upgraded to unified ODF schema

[0.21.1] - 2024-01-03

Fixed

  • Escape special symbols in input view names

[0.21.0] - 2024-01-03

Removed

  • Removed deprecated Spark-only ingest

Changed

  • Upgraded to new ODF engine protocol

[0.20.0] - 2024-01-03

Changed

  • Upgraded to new ODF schemas

[0.19.1] - 2023-07-14

Fixed

  • Transform engine will now read the list of input files as specified in the request instead of reading the entire directory of Parquet files

[0.19.0] - 2023-05-29

Changed

  • Configuring Spark to write timestamps in Timestamp(MILLIS,true) logical Parquet format instead of int96 that is considered deprecated and causing compatibility issues

[0.18.0] - 2023-05-05

Changed

  • Removed ingest checkpoints
  • Unique naming of temporary output directories

[0.17.0] - 2022-07-22

Added

  • Support for reading Apache Parquet files directly

[0.16.0] - 2022-04-22

Changed

  • Updated to latest ODF schemas
  • Adapter will now tar/untar checkpoints to have them managed as files

[0.15.1] - 2022-03-16

Fixed

  • Ignore offset column in merge strategies

[0.11.2] - 2021-09-03

Fixed

  • Handle Shapefiles with subdirectories

[0.11.1] - 2021-08-17

Fixed

  • Helpful error messages for invalid event time column upon ingest

[0.11.0] - 2021-08-16

Changed

  • Upgraded to Spark 3.1.2 and latest Sedona

[0.9.1] - 2020-11-14

Fixed

  • Fixed data hashing when dataframe is empty

[0.9.0] - 2020-10-31

Changed

  • Upgraded to Spark 3.0.1 and latest Sedona (GeoSpark)

[0.8.3] - 2020-10-30

Changed

  • Upgraded to ODF manifests

[0.8.0] - 2020-07-12

Changed

  • Upgraded to ODF manifests

[0.7.0] - 2020-06-28

Added

  • Windows support improvements

[0.6.0] - 2020-06-28

Changed

  • Minimizing use of Hadoop FS

[0.5.0] - 2020-06-23

Added

  • Support watermarking in ingest and transform

[0.4.0] - 2020-06-14

Changed

  • Further engine interface improvements
  • Event time column is now required
  • System time column is not considered when computing the hash
  • Result files will be named using system time

[0.3.0] - 2020-05-25

Changed

  • Ensuring data is sorted in Parquet files based on event time
  • Naming output files using system time

[0.2.0] - 2020-05-09

Changed

  • Major refactoring and combined with the ingress

[0.1.0] - 2020-05-03

Changed

  • Separated engine build from kamu-cli