amoeba commented on code in PR #45401: URL: https://github.com/apache/arrow/pull/45401#discussion_r1938684457
########## CHANGELOG.md: ########## @@ -1,4 +1,25 @@ +# Apache Arrow 19.0.1 (2025-01-30 08:00:00+00:00) + +## Bug Fixes + +* [GH-44513](https://github.com/apache/arrow/issues/44513) - [C++][Python] Pyarrow.Table.join() breaks on large tables v.18.0.0.dev486 +* [GH-45180](https://github.com/apache/arrow/issues/45180) - [C++][Fuzzing] Fix bug discovered by fuzzing +* [GH-45230](https://github.com/apache/arrow/issues/45230) - [Docs] Add LinkedIn social link and fix top nav scaling problems +* [GH-45283](https://github.com/apache/arrow/issues/45283) - [Python][C++][Parquet] "OSError: Repetition level histogram size mismatch" when reading parquet file in pyarrow since 19.0.0 +* [GH-45296](https://github.com/apache/arrow/issues/45296) - [Python] to\_pandas() fails when pandas option 'future.infer\_string' is True +* [GH-45339](https://github.com/apache/arrow/issues/45339) - [Parquet][C++] Reading parquet with an empty list of row group indices fails +* [GH-45357](https://github.com/apache/arrow/issues/45357) - [C++] Disable failing arrow-flight-test when misusing the library + + +## New Features and Improvements + +* [GH-45201](https://github.com/apache/arrow/issues/45201) - [C++][Parquet] Improve performance of writing size statistics +* [GH-45304](https://github.com/apache/arrow/issues/45304) - [C++] Compatibility with newer aws sdk +* [GH-45305](https://github.com/apache/arrow/issues/45305) - [Python] Compatibility with boto 1.36 Review Comment: @kou @raulcd we talked about this on the last arrow community call, would it be good to relabel these three issues or maybe even just rename the header here? They're not technically bug fixes but we don't want new features to be added in patch releases. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
