Hey everyone, Unfortunally, I found another issue where we produce incorrect metadata <https://github.com/apache/iceberg-python/pull/2408> when merging manifests. This was the result of a big refactor.
Let's get this PR in, and I'll cut another RC right away. Kind regards, Fokko Op za 30 aug 2025 om 19:21 schreef Kevin Liu <kevinjq...@apache.org>: > +1 (non-binding) > > [x] Download links are valid. > [x] Checksums and signatures. > [x] LICENSE/NOTICE files exist > [x] No unexpected binary files > [x] All source files have ASF headers > [x] Can compile from source > [x] Built and tested > > Had an issue running `make test-coverage` due to version conflict with the > `datafusion` dependency. If you see `*Fatal Python error: Bus error*`, > this is because the resolver installed the latest datafusion version 49 > which is incompatible with older versions of pyiceberg-core. > A workaround is to manually upgrade to datafusion 48 and rerun tests. > ``` > poetry run pip install datafusion==48 > ``` > I've added a PR to include `poetry.`lock in the source distribution so > that dependencies are pined. > https://github.com/apache/iceberg-python/pull/2405 > The underlying issue is due to a breaking change in the datafusion FFI, > see https://github.com/apache/datafusion-python/issues/1217 > I don't think this is blocking for the RC and release. > > Thanks (again) for running the release, Fokko! Very excited to see this > finally come through. > > Best, > Kevin Liu > > On Fri, Aug 29, 2025 at 11:46 AM Fokko Driesprong <fo...@apache.org> > wrote: > >> Hi Everyone, >> >> With great excitement I propose that we release the following RC as the >> official PyIceberg 0.10.0 release. >> >> A summary of the high level features: >> >> - Initial V3 support (read-only): Initial-defaults, UnknownType, >> Reading Deletion Vectors >> - Introduction of the AuthManager with various implementations >> - First TableMaintenance feature: Expire Snapshots >> - Replace Arrow partition transforms with Iceberg-Rust >> - Support for ADLS >> - Fixed a lot of bugs and improvements: UUID support, Imported Hive >> tables, Avro Gzip compression, reduced upsert memory pressure, several >> performance optimizations around metadata tables, parallel add-files >> operation, and much more! >> >> Big shout out to the community for working on this, and joining the >> discussion on Slack, GitHub and the dev-list! >> >> The commit ID is 86b0fe8766245fcc0132da82475901aae9f542e8 >> >> >> - This corresponds to the tag: pyiceberg-0.10.0rc2 >> (b8b292b9648fed1cdd5722f356c139652497825b) >> - >> https://github.com/apache/iceberg-python/releases/tag/pyiceberg-0.10.0rc2 >> - >> >> https://github.com/apache/iceberg-python/tree/86b0fe8766245fcc0132da82475901aae9f542e8 >> >> The release tarball, signature, and checksums are here: >> >> - https://dist.apache.org/repos/dist/dev/iceberg/pyiceberg-0.10.0rc2/ >> >> You can find the KEYS file here: >> >> - https://downloads.apache.org/iceberg/KEYS >> >> Convenience binary artifacts are staged on pypi: >> >> https://pypi.org/project/pyiceberg/0.10.0rc2/ >> >> And can be installed using: pip3 install pyiceberg==0.10.0rc2 >> >> Instructions for verifying a release can be found here: >> >> - https://py.iceberg.apache.org/verify-release/ >> >> Please download, verify, and test. >> >> Please vote in the next 72 hours. >> >> [ ] +1 Release this as PyIceberg 0.10.0 >> [ ] +0 >> [ ] -1 Do not release this because... >> >> Kind regards, >> Fokko >> >
